From 7e329f1d8598f0a792f111db22e9d5a48a2e3f99 Mon Sep 17 00:00:00 2001 From: admin <572701190@qq.com> Date: Tue, 26 May 2026 08:51:10 +0800 Subject: [PATCH] Import sanitized HIS processing tools --- .gitignore | 35 + README.md | 19 + 患者列表处理/.env.example | 19 + 患者列表处理/.gitignore | 22 + 患者列表处理/README.md | 125 + 患者列表处理/人工复核网页端/.dockerignore | 4 + 患者列表处理/人工复核网页端/.env.example | 15 + 患者列表处理/人工复核网页端/Dockerfile | 17 + 患者列表处理/人工复核网页端/README.md | 62 + 患者列表处理/人工复核网页端/app.py | 2127 +++++++++++++ .../人工复核网页端/docker-compose.yml | 19 + 患者列表处理/人工复核网页端/requirements.txt | 5 + 患者列表处理/人工复核网页端/static/app.css | 1339 ++++++++ 患者列表处理/人工复核网页端/static/app.js | 2042 ++++++++++++ .../人工复核网页端/templates/index.html | 382 +++ .../人工复核网页端/templates/login.html | 29 + 患者列表处理/工作流_Gitea版.md | 202 ++ .../数据处理工作区/01_科室分类规则.json | 258 ++ .../数据处理工作区/02_患者列表OCR归档.py | 1346 ++++++++ .../数据处理工作区/03_人工复核修正.template.json | 1 + .../数据处理工作区/04_合并批次结果.py | 247 ++ .../数据处理工作区/05_同步PostgreSQL单表.py | 233 ++ .../数据处理工作区/06_PostgreSQL建表结构.sql | 84 + .../数据处理工作区/07_处理程序说明.md | 72 + .../08_PostgreSQL调整Patient_Lists列顺序.sql | 168 + .../09_PostgreSQL住院号非空唯一约束.sql | 102 + 患者首页处理/.gitignore | 22 + 患者首页处理/README.md | 259 ++ 患者首页处理/工作流_Gitea版.md | 306 ++ 患者首页处理/数据可视化网页端/.dockerignore | 4 + 患者首页处理/数据可视化网页端/.env.example | 17 + 患者首页处理/数据可视化网页端/Dockerfile | 19 + 患者首页处理/数据可视化网页端/README.md | 58 + 患者首页处理/数据可视化网页端/app/__init__.py | 1 + 患者首页处理/数据可视化网页端/app/main.py | 2779 +++++++++++++++++ .../数据可视化网页端/app/static/app.js | 1393 +++++++++ .../数据可视化网页端/app/static/index.html | 273 ++ .../数据可视化网页端/app/static/styles.css | 1193 +++++++ .../数据可视化网页端/docker-compose.yml | 20 + .../数据处理工作区/01_配置规则/01_科室分类规则.json | 245 ++ .../02_解析入库/02_患者首页PDF解析与入库.py | 2181 +++++++++++++ .../03_人工复核/03_人工复核导出与回写.py | 224 ++ .../04_质量体检/04_字段核验与数据库体检.py | 175 ++ .../05_备用读取/05_备用PDF转Markdown_Mineru.py | 130 + .../06_图片对照核验/06_PDF转图片与对照核验.py | 697 +++++ .../07_Kimi视觉兜底/07_Kimi图片识别辅助.py | 205 ++ 46 files changed, 19175 insertions(+) create mode 100644 .gitignore create mode 100644 README.md create mode 100644 患者列表处理/.env.example create mode 100644 患者列表处理/.gitignore create mode 100644 患者列表处理/README.md create mode 100644 患者列表处理/人工复核网页端/.dockerignore create mode 100644 患者列表处理/人工复核网页端/.env.example create mode 100644 患者列表处理/人工复核网页端/Dockerfile create mode 100644 患者列表处理/人工复核网页端/README.md create mode 100644 患者列表处理/人工复核网页端/app.py create mode 100644 患者列表处理/人工复核网页端/docker-compose.yml create mode 100644 患者列表处理/人工复核网页端/requirements.txt create mode 100644 患者列表处理/人工复核网页端/static/app.css create mode 100644 患者列表处理/人工复核网页端/static/app.js create mode 100644 患者列表处理/人工复核网页端/templates/index.html create mode 100644 患者列表处理/人工复核网页端/templates/login.html create mode 100644 患者列表处理/工作流_Gitea版.md create mode 100644 患者列表处理/数据处理工作区/01_科室分类规则.json create mode 100644 患者列表处理/数据处理工作区/02_患者列表OCR归档.py create mode 100644 患者列表处理/数据处理工作区/03_人工复核修正.template.json create mode 100644 患者列表处理/数据处理工作区/04_合并批次结果.py create mode 100644 患者列表处理/数据处理工作区/05_同步PostgreSQL单表.py create mode 100644 患者列表处理/数据处理工作区/06_PostgreSQL建表结构.sql create mode 100644 患者列表处理/数据处理工作区/07_处理程序说明.md create mode 100644 患者列表处理/数据处理工作区/08_PostgreSQL调整Patient_Lists列顺序.sql create mode 100644 患者列表处理/数据处理工作区/09_PostgreSQL住院号非空唯一约束.sql create mode 100644 患者首页处理/.gitignore create mode 100644 患者首页处理/README.md create mode 100644 患者首页处理/工作流_Gitea版.md create mode 100644 患者首页处理/数据可视化网页端/.dockerignore create mode 100644 患者首页处理/数据可视化网页端/.env.example create mode 100644 患者首页处理/数据可视化网页端/Dockerfile create mode 100644 患者首页处理/数据可视化网页端/README.md create mode 100644 患者首页处理/数据可视化网页端/app/__init__.py create mode 100644 患者首页处理/数据可视化网页端/app/main.py create mode 100644 患者首页处理/数据可视化网页端/app/static/app.js create mode 100644 患者首页处理/数据可视化网页端/app/static/index.html create mode 100644 患者首页处理/数据可视化网页端/app/static/styles.css create mode 100644 患者首页处理/数据可视化网页端/docker-compose.yml create mode 100644 患者首页处理/数据处理工作区/01_配置规则/01_科室分类规则.json create mode 100755 患者首页处理/数据处理工作区/02_解析入库/02_患者首页PDF解析与入库.py create mode 100755 患者首页处理/数据处理工作区/03_人工复核/03_人工复核导出与回写.py create mode 100755 患者首页处理/数据处理工作区/04_质量体检/04_字段核验与数据库体检.py create mode 100644 患者首页处理/数据处理工作区/05_备用读取/05_备用PDF转Markdown_Mineru.py create mode 100644 患者首页处理/数据处理工作区/06_图片对照核验/06_PDF转图片与对照核验.py create mode 100644 患者首页处理/数据处理工作区/07_Kimi视觉兜底/07_Kimi图片识别辅助.py diff --git a/.gitignore b/.gitignore new file mode 100644 index 0000000..a0c0412 --- /dev/null +++ b/.gitignore @@ -0,0 +1,35 @@ +# Local secrets +.env +.env.* +!.env.example +.pgpass + +# Patient data and generated outputs +数据处理结果区/ +待处理-患者目录图片集群/ +已处理-患者目录图片集群/ +待处理-患者首页PDF/ +已处理-患者首页PDF/ +*.csv +*.jsonl +*.pdf +*.png +*.jpg +*.jpeg +*.xlsx +*.xls + +# Runtime/cache +__pycache__/ +*.py[cod] +*.log +*.tmp +.DS_Store +.idea/ +.vscode/ + +# Local review settings +review_settings.local.json +数据可视化网页端/review_settings.local.json +人工复核网页端配置.json +03_人工复核修正.json diff --git a/README.md b/README.md new file mode 100644 index 0000000..a702b1a --- /dev/null +++ b/README.md @@ -0,0 +1,19 @@ +# HIS 数据处理工具集 + +本仓库合并保存 HIS 相关处理程序、网页端、配置模板和流程文档。 + +## 目录 + +- `患者列表处理/`：患者列表 OCR、合并、PostgreSQL 同步、人工复核网页端。 +- `患者首页处理/`：患者首页 PDF 解析入库、人工复核、质量体检、可视化网页端、Kimi 视觉兜底工具。 + +## 入库边界 + +仓库只保存程序、文档、结构脚本和配置模板，不保存以下内容： + +- 原始 PDF、图片、截图、OCR 裁剪图等患者数据。 +- `数据处理结果区/` 中的结构化结果、日志、质检报告和复核导出。 +- `.env`、`.pgpass`、本地网页端配置、明文数据库连接信息、API Key 或登录密码。 +- 两个旧工程的 `.git` 历史。 + +需要运行程序时，请在本地根据各子目录的 `.env.example` 创建 `.env`，并从安全渠道填入数据库和 API 配置。 diff --git a/患者列表处理/.env.example b/患者列表处理/.env.example new file mode 100644 index 0000000..6a97306 --- /dev/null +++ b/患者列表处理/.env.example @@ -0,0 +1,19 @@ +HIS_DB_HOST=DB_HOST +HIS_DB_PORT=5432 +HIS_DB_NAME=DB_NAME +HIS_DB_USER=DB_USER +HIS_DB_PASSWORD=请填写数据库密码 + +REVIEW_APP_USERNAME=admin +REVIEW_APP_PASSWORD=请填写网页端登录密码 +REVIEW_APP_SECRET_KEY=请填写随机字符串 +REVIEW_APP_PORT=8090 +REVIEW_POSTGRES_SYNC=1 + +KIMI_API_KEY=请填写Kimi或Moonshot API Key +KIMI_MODEL=kimi-k2.6 +KIMI_TIMEOUT_SECONDS=80 +KIMI_IMAGE_MAX_WIDTH=1200 + +TENCENTCLOUD_SECRET_ID=请填写腾讯云SecretId +TENCENTCLOUD_SECRET_KEY=请填写腾讯云SecretKey diff --git a/患者列表处理/.gitignore b/患者列表处理/.gitignore new file mode 100644 index 0000000..3d7389d --- /dev/null +++ b/患者列表处理/.gitignore @@ -0,0 +1,22 @@ +# Patient images and generated results +待处理-患者目录图片集群/ +已处理-患者目录图片集群/ +数据处理结果区/ + +# Local correction data can contain patient information +数据处理工作区/03_人工复核修正.json +数据处理工作区/人工复核网页端配置.json + +# Local secrets and runtime files +.env +.pgpass +工作流_本地使用版.md +*.log +2026_5_23_HIS患者目录图片集群处理*.zip +__pycache__/ +*.py[cod] + +# OS/editor noise +.DS_Store +.idea/ +.vscode/ diff --git a/患者列表处理/README.md b/患者列表处理/README.md new file mode 100644 index 0000000..763fd91 --- /dev/null +++ b/患者列表处理/README.md @@ -0,0 +1,125 @@ +# HIS 患者列表图片归档 + +这个项目用于把 HIS 患者列表截图批量 OCR 为结构化记录，并同步到 PostgreSQL 单表。 + +## 目录约定 + +- `待处理-患者目录图片集群/`：后续新下载、尚未归档的图片集群。 +- `已处理-患者目录图片集群/`：已经完成归档的原始图片集群。 +- `数据处理工作区/`：处理脚本、科室分类规则、数据库建表脚本和说明文档。 +- `人工复核网页端/`：Docker 化的人工复核网页端。 +- `数据处理结果区/已处理-患者目录图片集群/`：各批次本地归档结果。 +- `数据处理结果区/信息记录/`：全局信息记录和批次信息记录。 + +数据、图片、OCR 缓存、处理结果不应提交到 Git。 + +## 工作区脚本 + +1. `数据处理工作区/01_科室分类规则.json` +2. `数据处理工作区/02_患者列表OCR归档.py` +3. `数据处理工作区/03_人工复核修正.json` +4. `数据处理工作区/04_合并批次结果.py` +5. `数据处理工作区/05_同步PostgreSQL单表.py` +6. `数据处理工作区/06_PostgreSQL建表结构.sql` +7. `数据处理工作区/07_处理程序说明.md` +8. `数据处理工作区/08_PostgreSQL调整Patient_Lists列顺序.sql` +9. `数据处理工作区/09_PostgreSQL住院号非空唯一约束.sql` + +`03_人工复核修正.json` 可能包含患者信息，仓库中只保留模板文件。 + +## 处理新批次 + +```bash +export TENCENTCLOUD_SECRET_ID='...' +export TENCENTCLOUD_SECRET_KEY='...' + +python3 数据处理工作区/02_患者列表OCR归档.py \ + --input "待处理-患者目录图片集群/批次文件夹名" \ + --output "数据处理结果区/已处理-患者目录图片集群/批次文件夹名-列表归档结果" \ + --ocr-engine table-v3 \ + --batch-size 6 \ + --image-padding-y 24 \ + --workers 1 \ + --folder-workers 2 \ + --timeout 90 \ + --max-retries 1 +``` + +处理完成后优先查看： + +- `患者列表_结构化.json` +- `患者列表_记录.csv` +- `复核报告.json` +- `重复住院号报告.json` + +如果只想用已有 OCR 缓存重建结果： + +```bash +python3 数据处理工作区/02_患者列表OCR归档.py \ + --input "待处理-患者目录图片集群/批次文件夹名" \ + --output "数据处理结果区/已处理-患者目录图片集群/批次文件夹名-列表归档结果" \ + --ocr-engine table-v3 \ + --batch-size 6 \ + --image-padding-y 24 \ + --workers 1 \ + --folder-workers 2 \ + --rebuild-from-cache +``` + +确认无误后，把原始图片目录移动到 `已处理-患者目录图片集群/`。 + +## 合并与入库 + +```bash +python3 数据处理工作区/04_合并批次结果.py + +export HIS_DB_HOST='DB_HOST' +export HIS_DB_PORT='5432' +export HIS_DB_NAME='DB_NAME' +export HIS_DB_USER='DB_USER' +export HIS_DB_PASSWORD='...' + +python3 数据处理工作区/05_同步PostgreSQL单表.py +``` + +PostgreSQL 只使用 `"Patient_Lists"` 一个正式表。科室映射、OCR 缓存路径、拼接图路径、OCR 请求号等处理过程信息保留在本地工作区和结果目录，不进入数据库正式表。 + +## 人工复核网页端 + +在根目录 `.env` 中配置 `REVIEW_APP_USERNAME`、`REVIEW_APP_PASSWORD`、`REVIEW_APP_SECRET_KEY`、`REVIEW_APP_PORT`；如需抽查功能，再配置 `KIMI_API_KEY`、`KIMI_MODEL`。可参考 `.env.example`。启动： + +```bash +cd 人工复核网页端 +docker compose up -d --build +``` + +默认访问 `http://127.0.0.1:8090`。网页端包含概览、复核、抽查、抽查查看、设置区域，会读取各批次 `复核报告.json`，按图片内行号裁剪原图，人工修订会保存到 `数据处理工作区/03_人工复核修正.json` 并尽量同步 PostgreSQL。出院时间允许为空；若填写，则会校验时间格式以及是否早于入院时间。 + +入库后建议核对： + +```sql +SELECT count(*) FROM "Patient_Lists"; +SELECT count(*) FROM "Patient_Lists" WHERE review_status = '需人工复核'; +SELECT inpatient_no FROM "Patient_Lists" GROUP BY inpatient_no HAVING count(*) > 1; +``` + +## 关键规则 + +- 默认使用 `RecognizeTableAccurateOCR` 表格识别 V3。 +- 患者字段顺序固定为：姓名、性别、年龄、住院号、诊断、入院时间、最后书写时间、住院天数、出院时间、手术后天数。 +- 拼接图默认给每张原图上下各加 `24px` 白边，减少贴边表格行被 OCR 忽略的概率。 +- 处理时优先 6 张拼接；如果行数偏少、接口超时或识别失败，程序会动态降到 4/3/2/单张。 +- `住院号` 是唯一强制校验条件：不能为空且全库唯一。重复住院号按后出现记录覆盖先出现记录；格式异常但非空的住院号保留，后续在网页端人工核验。 +- `出院时间` 允许为空；若入院时间晚于出院时间，程序和 PostgreSQL 都会把它作为需复核问题处理。 +- `自动复核通过` 表示 OCR 结果经过规则校验后无明显异常；`人工复核通过` 表示该行命中了人工修正配置，修正后校验通过。 +- 抽查结果直接写入 PostgreSQL 的 `audit_*` 字段，不再使用本地 `抽查记录.json`。 +- `AI修改-待确认` 只作为网页端辅助修正草稿，不会被“提交待同步”当作人工确认结果入库。 +- 旧批次若用缓存重建，要沿用当时的拼接和白边参数，避免新图片名和旧 OCR 缓存不匹配。 + +## 迁移注意 + +- 需要 Python 3、`Pillow` 和 PostgreSQL 客户端 `psql`。 +- 腾讯云密钥和数据库密码只通过环境变量传入，不要写入脚本或提交到文件。 +- `raw_ocr/` 是 OCR 响应缓存，迁移时建议保留，可减少重复调用。 +- `merged_images/` 是拼接图，便于复核。 +- PostgreSQL 结构以 `数据处理工作区/06_PostgreSQL建表结构.sql` 为准。 diff --git a/患者列表处理/人工复核网页端/.dockerignore b/患者列表处理/人工复核网页端/.dockerignore new file mode 100644 index 0000000..f86ec77 --- /dev/null +++ b/患者列表处理/人工复核网页端/.dockerignore @@ -0,0 +1,4 @@ +.env +__pycache__/ +*.py[cod] +*.log diff --git a/患者列表处理/人工复核网页端/.env.example b/患者列表处理/人工复核网页端/.env.example new file mode 100644 index 0000000..a5aaf90 --- /dev/null +++ b/患者列表处理/人工复核网页端/.env.example @@ -0,0 +1,15 @@ +REVIEW_APP_USERNAME=admin +REVIEW_APP_PASSWORD=change-me +REVIEW_APP_SECRET_KEY=change-me-to-a-random-string +REVIEW_APP_PORT=8090 +REVIEW_POSTGRES_SYNC=1 +KIMI_API_KEY=change-me +KIMI_API_ENABLED=1 +KIMI_MODEL=kimi-k2.6 +KIMI_TIMEOUT_SECONDS=120 +KIMI_IMAGE_MAX_WIDTH=1200 +KIMI_AUDIT_MAX_TOKENS=768 +KIMI_CORRECTION_MAX_TOKENS=768 + +# Docker 内默认挂载项目根目录到 /workspace；本机直接运行时可改为项目根目录。 +WORKSPACE_ROOT=/workspace diff --git a/患者列表处理/人工复核网页端/Dockerfile b/患者列表处理/人工复核网页端/Dockerfile new file mode 100644 index 0000000..1316b54 --- /dev/null +++ b/患者列表处理/人工复核网页端/Dockerfile @@ -0,0 +1,17 @@ +FROM python:3.12-slim + +ENV PYTHONDONTWRITEBYTECODE=1 +ENV PYTHONUNBUFFERED=1 + +WORKDIR /app + +COPY requirements.txt . +RUN pip install --no-cache-dir -r requirements.txt + +COPY app.py ./app.py +COPY templates ./templates +COPY static ./static + +EXPOSE 8000 + +CMD ["gunicorn", "-b", "0.0.0.0:8000", "app:app", "--workers", "2", "--threads", "4", "--timeout", "180"] diff --git a/患者列表处理/人工复核网页端/README.md b/患者列表处理/人工复核网页端/README.md new file mode 100644 index 0000000..f8a44f2 --- /dev/null +++ b/患者列表处理/人工复核网页端/README.md @@ -0,0 +1,62 @@ +# HIS 患者列表人工复核网页端 + +用于查看 `复核报告.json` 中的待处理记录，按原图行号生成局部截图，保存人工修订到 `数据处理工作区/03_人工复核修正.json`。 + +## 配置 + +在项目根目录 `.env` 中增加： + +```bash +REVIEW_APP_USERNAME=admin +REVIEW_APP_PASSWORD=请填写本地登录密码 +REVIEW_APP_SECRET_KEY=请填写随机字符串 +REVIEW_APP_PORT=8090 +KIMI_API_KEY=请填写Kimi或Moonshot API Key +KIMI_MODEL=kimi-k2.6 +``` + +如果未设置 `REVIEW_APP_PASSWORD`，程序会回退使用 `HIS_DB_PASSWORD` 作为登录密码。 + +默认启用 PostgreSQL 联动，保存或删除人工修订时会先按 `image_path + image_row_no` 更新 `"Patient_Lists"` 对应记录；如果合并去重后正式表只保留了同住院号的另一条记录，会继续按 `inpatient_no` 匹配更新。PostgreSQL 暂不可用时，本地修订不会丢失，会写入待同步状态，恢复连接后点击“提交待同步”即可补交。需要临时关闭时可设置： + +```bash +REVIEW_POSTGRES_SYNC=0 +``` + +## Docker 启动 + +```bash +cd 人工复核网页端 +docker compose up -d --build +``` + +访问： + +```text +http://127.0.0.1:8090 +``` + +## 复核后重建 + +网页保存后，人工修订会写入： + +```text +数据处理工作区/03_人工复核修正.json +``` + +之后对相应批次运行 `02_患者列表OCR归档.py --rebuild-from-cache`，再执行合并和入库脚本即可。 + +出院时间默认允许为空；确有出院时间时，网页端会校验时间格式以及是否早于入院时间。 +PostgreSQL 中也已加入对应约束：空出院时间放行；入院时间晚于出院时间时，该记录必须保持 `需人工复核` 状态。 + +时间字段支持输入 `2021-11-24 9:46:59`、`2021-12-8 9:38:00` 这类单数字月、日、小时格式；网页端和后端保存时会统一补全为 `YYYY-MM-DD HH:MM:SS`。 + +保存前如果仍有校验提示，网页会弹窗确认是否继续保存；无提示保存成功后会自动切到下一条待处理记录。 + +页面右上角会显示 PostgreSQL 连接状态，也会每 20 秒自动检测一次；如果本地修订保存成功但数据库同步失败，会标记为待同步，之后点击“提交待同步”可重新提交。提交后会汇总成功、未匹配、失败数量，未成功项可选择保留或从待同步列表删除。 + +字段顺序固定为：入院时间、最后书写时间、住院天数、出院时间。字段校验提示统一显示在截图下方，右侧编辑区不再重复显示。截图区支持鼠标滚轮缩放，“恢复尺寸”会回到图片原始大小。 + +抽查页会随机抽取患者行，调用 Kimi 多模态模型核对裁剪截图与结构化字段；抽查结论、AI 反馈、人工反馈直接写入 PostgreSQL 的 `audit_*` 字段，可在“抽查查看”页回看。复核页也可以调用“Kimi自动修改当前项/5项”，生成的内容会标记为 `AI修改-待确认`，不会进入待同步；人工点击“保存确认”后才会作为正式修订同步数据库。 + +设置页可新增网页端用户，为每个配置用户单独设置是否能看到概览、复核、抽查、抽查查看、设置，也可修改抽查模型和 API Key。设置内容保存在本地 `数据处理工作区/人工复核网页端配置.json`，该文件不提交到 Git。 diff --git a/患者列表处理/人工复核网页端/app.py b/患者列表处理/人工复核网页端/app.py new file mode 100644 index 0000000..1c290b5 --- /dev/null +++ b/患者列表处理/人工复核网页端/app.py @@ -0,0 +1,2127 @@ +#!/usr/bin/env python3 +# -*- coding: utf-8 -*- +"""Manual review web app for HIS patient-list OCR results.""" + +from __future__ import annotations + +import datetime as dt +import base64 +import csv +import hashlib +import io +import json +import os +import random +import re +import secrets +import socket +import urllib.error +import urllib.request +from dataclasses import dataclass +from pathlib import Path +from typing import Any + +import psycopg +from psycopg.rows import dict_row +from dotenv import load_dotenv +from flask import Flask, Response, abort, jsonify, redirect, render_template, request, send_file, session, url_for +from PIL import Image + + +load_dotenv() + +COLUMNS = [ + "姓名", + "性别", + "年龄", + "住院号", + "诊断", + "入院时间", + "最后书写时间", + "住院天数", + "出院时间", + "手术后天数", +] + +DATETIME_PATTERN = re.compile(r"\d{4}-\d{2}-\d{2} \d{2}:\d{2}:\d{2}") +FLEX_DATETIME_PATTERN = re.compile(r"^(\d{4})-(\d{1,2})-(\d{1,2})\s+(\d{1,2}):(\d{1,2}):(\d{1,2})$") +INPATIENT_NO_PATTERN = re.compile(r".*") +IMAGE_EXTENSIONS = {".png", ".jpg", ".jpeg", ".bmp", ".webp"} +IGNORED_REVIEW_TIPS = {"缺少出院时间"} +IGNORED_AI_ISSUE_PATTERNS = ( + "最后书写时间与出院时间顺序异常", + "最后书写时间晚于出院时间", + "出院后仍有书写记录", + "已出院后仍有书写记录", +) +FORMAT_ONLY_AI_ISSUE_PATTERNS = ( + "实际值一致", + "时间值一致", + "格式差异", + "补零差异", + "小时格式", + "日期格式补零", + "仅格式", + "无需修正", +) +AI_PENDING_STATE = "AI修改-待确认" +STILL_CONFIRM_STATE = "修订后仍需确认" +MANUAL_PASSED_STATE = "人工复核通过" +DATETIME_COLUMNS = {"入院时间", "最后书写时间", "出院时间"} + +WORKSPACE_ROOT = Path(os.getenv("WORKSPACE_ROOT", ".")).resolve() +RESULT_ROOT = WORKSPACE_ROOT / os.getenv("REVIEW_RESULT_ROOT", "数据处理结果区/已处理-患者目录图片集群") +CORRECTIONS_PATH = WORKSPACE_ROOT / os.getenv("REVIEW_CORRECTIONS_PATH", "数据处理工作区/03_人工复核修正.json") +CONFIG_PATH = WORKSPACE_ROOT / os.getenv("REVIEW_CONFIG_PATH", "数据处理工作区/人工复核网页端配置.json") +POSTGRES_SYNC_ENABLED = os.getenv("REVIEW_POSTGRES_SYNC", "1").lower() not in {"0", "false", "no", "off"} +MERGED_RESULT_PATH = WORKSPACE_ROOT / os.getenv("REVIEW_MERGED_RESULT_PATH", "数据处理结果区/合并_患者列表_结构化.json") +MERGED_CSV_PATH = WORKSPACE_ROOT / os.getenv("REVIEW_MERGED_CSV_PATH", "数据处理结果区/合并_患者列表_记录.csv") +RESULT_INFO_DIR = WORKSPACE_ROOT / os.getenv("REVIEW_RESULT_INFO_DIR", "数据处理结果区/信息记录") +PERMISSION_KEYS = ("overview", "review", "audit", "audit_history", "settings") +PERMISSION_LABELS = { + "overview": "概览", + "review": "复核", + "audit": "抽查", + "audit_history": "抽查一览", + "settings": "设置", +} +FULL_PERMISSIONS = {key: True for key in PERMISSION_KEYS} +KIMI_TIMEOUT_SECONDS = int(os.getenv("KIMI_TIMEOUT_SECONDS", "120")) +KIMI_IMAGE_MAX_WIDTH = int(os.getenv("KIMI_IMAGE_MAX_WIDTH", "1200")) +KIMI_AUDIT_MAX_TOKENS = int(os.getenv("KIMI_AUDIT_MAX_TOKENS", "768")) +KIMI_CORRECTION_MAX_TOKENS = int(os.getenv("KIMI_CORRECTION_MAX_TOKENS", "768")) + +app = Flask(__name__) +app.secret_key = ( + os.getenv("REVIEW_APP_SECRET_KEY") + or os.getenv("FLASK_SECRET_KEY") + or hashlib.sha256((os.getenv("HIS_DB_PASSWORD", "") + str(WORKSPACE_ROOT)).encode("utf-8")).hexdigest() + or secrets.token_hex(32) +) + + +@dataclass(frozen=True) +class ReviewItem: + key: str + batch_name: str + source: str + record: dict[str, Any] + origin: str + + +def json_response(data: Any, status: int = 200) -> Response: + return app.response_class( + json.dumps(data, ensure_ascii=False), + status=status, + mimetype="application/json; charset=utf-8", + ) + + +def error_response(message: str, status: int = 400) -> Response: + return json_response({"error": message}, status=status) + + +def normalize_text(value: Any) -> str: + if value is None: + return "" + return re.sub(r"\s+", " ", str(value)).strip() + + +def normalize_datetime_text(value: Any) -> str: + text = normalize_text(value) + if not text: + return "" + match = FLEX_DATETIME_PATTERN.fullmatch(text) + if not match: + return text + year, month, day, hour, minute, second = (int(part) for part in match.groups()) + try: + parsed = dt.datetime(year, month, day, hour, minute, second) + except ValueError: + return text + return parsed.strftime("%Y-%m-%d %H:%M:%S") + + +def item_key(image_path: str, row_no: int) -> str: + return hashlib.sha1(f"{image_path}|{row_no}".encode("utf-8")).hexdigest()[:20] + + +def require_login() -> None: + if not session.get("authenticated"): + abort(401) + + +@app.before_request +def auth_gate() -> None: + if request.path in {"/login", "/favicon.ico"} or request.path.startswith("/static/"): + return + if not session.get("authenticated"): + if request.path.startswith("/api/"): + abort(401) + return redirect(url_for("login")) + + +def load_json(path: Path, default: Any) -> Any: + if not path.exists(): + return default + return json.loads(path.read_text(encoding="utf-8")) + + +def load_config() -> dict[str, Any]: + return load_json(CONFIG_PATH, {"users": [], "kimi": {}}) + + +def normalized_permissions(value: Any) -> dict[str, bool]: + if not isinstance(value, dict): + return dict(FULL_PERMISSIONS) + return {key: bool(value.get(key, False)) for key in PERMISSION_KEYS} + + +def current_permissions() -> dict[str, bool]: + return normalized_permissions(session.get("permissions") or FULL_PERMISSIONS) + + +def require_permission(permission: str) -> None: + require_login() + if not current_permissions().get(permission): + abort(403) + + +def atomic_write_json(path: Path, data: Any) -> None: + path.parent.mkdir(parents=True, exist_ok=True) + temp_path = path.with_suffix(path.suffix + ".tmp") + temp_path.write_text(json.dumps(data, ensure_ascii=False, indent=2), encoding="utf-8") + temp_path.replace(path) + + +def save_config(config: dict[str, Any]) -> None: + atomic_write_json(CONFIG_PATH, config) + + +def password_hash(password: str, salt: str | None = None) -> dict[str, str]: + salt = salt or secrets.token_hex(16) + digest = hashlib.sha256((salt + password).encode("utf-8")).hexdigest() + return {"salt": salt, "hash": digest} + + +def password_matches(password: str, user: dict[str, Any]) -> bool: + salt = str(user.get("salt", "")) + expected = str(user.get("password_hash", "")) + digest = hashlib.sha256((salt + password).encode("utf-8")).hexdigest() + return bool(expected) and secrets.compare_digest(digest, expected) + + +def correction_index() -> dict[str, dict[str, Any]]: + items = load_json(CORRECTIONS_PATH, []) + index: dict[str, dict[str, Any]] = {} + for item in items: + if item.get("已提交人工通过"): + continue + image_path = normalize_text(item.get("图片路径")) + row_no = int(item.get("图片内行号") or 0) + if image_path and row_no: + index[item_key(image_path, row_no)] = item + return index + + +def archived_correction_keys() -> set[str]: + keys: set[str] = set() + for item in load_json(CORRECTIONS_PATH, []): + if not item.get("已提交人工通过"): + continue + image_path = normalize_text(item.get("图片路径")) + row_no = int(item.get("图片内行号") or 0) + if image_path and row_no: + keys.add(item_key(image_path, row_no)) + return keys + + +def save_correction(item: dict[str, Any]) -> None: + data = load_json(CORRECTIONS_PATH, []) + key = item_key(item["图片路径"], int(item["图片内行号"])) + replaced = False + for index, existing in enumerate(data): + existing_key = item_key(normalize_text(existing.get("图片路径")), int(existing.get("图片内行号") or 0)) + if existing_key == key: + data[index] = item + replaced = True + break + if not replaced: + data.append(item) + atomic_write_json(CORRECTIONS_PATH, data) + + +def set_correction_sync_status(key: str, sync_status: dict[str, Any]) -> None: + data = load_json(CORRECTIONS_PATH, []) + changed = False + for item in data: + existing_key = item_key(normalize_text(item.get("图片路径")), int(item.get("图片内行号") or 0)) + if existing_key == key: + item["PostgreSQL同步"] = sync_status + changed = True + break + if changed: + atomic_write_json(CORRECTIONS_PATH, data) + + +def delete_correction(key: str) -> bool: + data = load_json(CORRECTIONS_PATH, []) + kept = [] + removed = False + for item in data: + existing_key = item_key(normalize_text(item.get("图片路径")), int(item.get("图片内行号") or 0)) + if existing_key == key: + removed = True + else: + kept.append(item) + if removed: + atomic_write_json(CORRECTIONS_PATH, kept) + return removed + + +def result_dirs() -> list[Path]: + if not RESULT_ROOT.exists(): + return [] + return sorted(path for path in RESULT_ROOT.glob("*-列表归档结果") if path.is_dir()) + + +def build_review_items() -> list[ReviewItem]: + items: dict[str, ReviewItem] = {} + for result_dir in result_dirs(): + batch_name = result_dir.name.removesuffix("-列表归档结果") + report = load_json(result_dir / "复核报告.json", {}) + for source, records in ( + ("需人工复核记录", report.get("需人工复核记录", [])), + ("人工修正记录", report.get("人工修正记录", [])), + ): + for record in records: + patient = normalize_patient(record.get("患者信息", {})) + tips = effective_review_tips(record.get("复核", {}).get("提示", [])) + if source == "需人工复核记录" and not tips and not validate_patient(patient, {}): + continue + image = record.get("图片信息", {}) + image_path = normalize_text(image.get("图片路径")) + row_no = int(image.get("图片内行号") or 0) + if not image_path or not row_no: + continue + key = item_key(image_path, row_no) + items[key] = ReviewItem(key, batch_name, source, record, "review_report") + + return sorted(items.values(), key=lambda item: (item.batch_name, item.record.get("来源文件夹", ""), item.record.get("图片信息", {}).get("图片名", ""), item.record.get("图片信息", {}).get("图片内行号", 0))) + + +def parse_datetime(text: str) -> dt.datetime | None: + text = normalize_datetime_text(text) + if not text: + return None + if not DATETIME_PATTERN.fullmatch(text): + return None + try: + return dt.datetime.strptime(text, "%Y-%m-%d %H:%M:%S") + except ValueError: + return None + + +def validate_patient(patient: dict[str, Any], options: dict[str, Any] | None = None) -> list[str]: + options = options or {} + warnings: list[str] = [] + if not normalize_text(patient.get("姓名")): + warnings.append("缺少姓名") + if normalize_text(patient.get("性别")) not in {"男", "女"}: + warnings.append("性别异常") + age = normalize_text(patient.get("年龄")) + if age and not re.fullmatch(r"\d{1,3}岁", age): + warnings.append("年龄格式异常") + if not normalize_text(patient.get("住院号")): + warnings.append("缺少住院号") + admission = normalize_datetime_text(patient.get("入院时间")) + discharge = normalize_datetime_text(patient.get("出院时间")) + last_write = normalize_datetime_text(patient.get("最后书写时间")) + + admission_dt = parse_datetime(admission) + discharge_dt = parse_datetime(discharge) + last_write_dt = parse_datetime(last_write) + + if not admission: + warnings.append("缺少入院时间") + elif not admission_dt: + warnings.append("入院时间格式异常") + + if discharge and not discharge_dt: + warnings.append("出院时间格式异常") + + if last_write and not last_write_dt: + warnings.append("最后书写时间格式异常") + if admission_dt and discharge_dt and discharge_dt < admission_dt: + warnings.append("出院时间早于入院时间") + if admission_dt and last_write_dt and last_write_dt < admission_dt: + warnings.append("最后书写时间早于入院时间") + + hospital_days = normalize_text(patient.get("住院天数")) + if hospital_days and not hospital_days.isdigit(): + warnings.append("住院天数格式异常") + postoperative_days = normalize_text(patient.get("手术后天数")) + if postoperative_days and not re.fullmatch(r"后\d+天", postoperative_days): + warnings.append("手术后天数格式异常") + return warnings + + +def patient_change_list(before: dict[str, Any], after: dict[str, Any]) -> list[dict[str, str]]: + changes: list[dict[str, str]] = [] + normalized_before = normalize_patient(before or {}) + normalized_after = normalize_patient(after or {}) + for column in COLUMNS: + old = normalize_text(normalized_before.get(column, "")) + new = normalize_text(normalized_after.get(column, "")) + if old != new: + changes.append({"字段": column, "修改前": old, "修改后": new}) + return changes + + +def change_log_entries(before: dict[str, Any], after: dict[str, Any], source: str) -> list[dict[str, str]]: + return [{**item, "修改者": source} for item in patient_change_list(before, after)] + + +def normalized_change_log(value: Any) -> list[dict[str, str]]: + if not isinstance(value, list): + return [] + normalized: list[dict[str, str]] = [] + for item in value: + if not isinstance(item, dict): + continue + field = normalize_text(item.get("字段") or item.get("column")) + if field not in COLUMNS: + continue + normalized.append( + { + "字段": field, + "修改前": normalize_text(item.get("修改前") if "修改前" in item else item.get("before", "")), + "修改后": normalize_text(item.get("修改后") if "修改后" in item else item.get("after", "")), + "修改者": normalize_text(item.get("修改者") or item.get("source") or "人工"), + } + ) + return normalized + + +def fallback_change_log(record: dict[str, Any], correction: dict[str, Any] | None) -> list[dict[str, str]]: + if not correction: + return [] + existing_log = normalized_change_log(correction.get("修改记录")) + if existing_log: + return existing_log + source = "AI" if correction.get("AI修改") or correction.get("修改来源") == "AI修改" else "人工" + return change_log_entries(record.get("患者信息", {}), merged_patient(record, correction), source) + + +def patient_change_summary(before: dict[str, Any], after: dict[str, Any]) -> str: + changes = patient_change_list(before, after) + if not changes: + return "" + return "；".join(f"{item['字段']}：{item['修改前'] or '空'} -> {item['修改后'] or '空'}" for item in changes) + + +def effective_review_tips(tips: Any) -> list[str]: + if not isinstance(tips, list): + tips = [tips] if normalize_text(tips) else [] + return [normalize_text(tip) for tip in tips if normalize_text(tip) and normalize_text(tip) not in IGNORED_REVIEW_TIPS] + + +def merged_patient(record: dict[str, Any], correction: dict[str, Any] | None) -> dict[str, Any]: + patient = dict(record.get("患者信息", {})) + if correction: + patient.update(correction.get("患者信息", {})) + for column in COLUMNS: + patient.setdefault(column, "") + for column in ("入院时间", "出院时间", "最后书写时间"): + patient[column] = normalize_datetime_text(patient.get(column, "")) + return patient + + +def normalize_patient(patient: dict[str, Any]) -> dict[str, str]: + normalized = {column: normalize_text(patient.get(column, "")) for column in COLUMNS} + normalized["住院号"] = normalized["住院号"].upper() + for column in ("入院时间", "出院时间", "最后书写时间"): + normalized[column] = normalize_datetime_text(normalized[column]) + return normalized + + +def has_valid_inpatient_no(patient: dict[str, Any]) -> bool: + return bool(normalize_text(patient.get("住院号"))) + + +def record_has_valid_inpatient_no(record: dict[str, Any]) -> bool: + return has_valid_inpatient_no(record.get("患者信息", {})) + + +def hospital_days_value(value: Any) -> int | None: + text = normalize_text(value) + return int(text) if text.isdigit() else None + + +def review_notes(warnings: list[str], options: dict[str, Any], manual_note: str) -> str: + notes = list(warnings) + if manual_note: + notes.append(f"人工备注: {manual_note}") + return "；".join(notes) + + +def db_configured() -> bool: + return all(os.getenv(name) for name in ("HIS_DB_HOST", "HIS_DB_NAME", "HIS_DB_USER", "HIS_DB_PASSWORD")) + + +def postgres_connection() -> psycopg.Connection: + return psycopg.connect( + host=os.getenv("HIS_DB_HOST"), + port=int(os.getenv("HIS_DB_PORT", "5432")), + dbname=os.getenv("HIS_DB_NAME"), + user=os.getenv("HIS_DB_USER"), + password=os.getenv("HIS_DB_PASSWORD"), + connect_timeout=5, + ) + + +def update_postgres_record( + record: dict[str, Any], + patient: dict[str, str], + options: dict[str, Any], + warnings: list[str], + manual_note: str, + manual_corrected: bool, +) -> dict[str, Any]: + if not POSTGRES_SYNC_ENABLED: + return {"enabled": False, "updated": 0, "message": "PostgreSQL同步已关闭"} + if not db_configured(): + return {"enabled": True, "updated": 0, "error": "数据库环境变量未完整配置"} + + image = record.get("图片信息", {}) + image_path = normalize_text(image.get("图片路径")) + row_no = int(image.get("图片内行号") or 0) + if not image_path or not row_no: + return {"enabled": True, "updated": 0, "error": "缺少图片路径或图片内行号，无法定位数据库记录"} + if not has_valid_inpatient_no(patient): + return {"enabled": True, "updated": 0, "error": "缺少住院号，未同步到正式表", "retryable": False} + if warnings: + review_status = "需人工复核" + elif manual_corrected: + review_status = "人工复核通过" + else: + review_status = "自动复核通过" + notes = review_notes(warnings, options, manual_note) + params = { + "image_path": image_path, + "image_row_no": row_no, + "patient_name": patient["姓名"], + "gender": patient["性别"], + "age": patient["年龄"], + "inpatient_no": patient["住院号"], + "diagnosis": patient["诊断"], + "admission_time": patient["入院时间"], + "last_write_time": patient["最后书写时间"], + "hospital_days": hospital_days_value(patient["住院天数"]), + "discharge_time": patient["出院时间"], + "postoperative_days": patient["手术后天数"], + "review_status": review_status, + "review_notes": notes, + "manual_corrected": manual_corrected, + } + set_sql = """ + UPDATE "Patient_Lists" + SET patient_name = %(patient_name)s, + gender = %(gender)s, + age = %(age)s, + inpatient_no = %(inpatient_no)s, + diagnosis = %(diagnosis)s, + admission_time = %(admission_time)s, + last_write_time = %(last_write_time)s, + hospital_days = %(hospital_days)s, + discharge_time = %(discharge_time)s, + postoperative_days = %(postoperative_days)s, + review_status = %(review_status)s, + review_notes = %(review_notes)s, + manual_corrected = %(manual_corrected)s, + imported_at = now() + """ + try: + with postgres_connection() as conn: + with conn.cursor() as cur: + cur.execute( + 'SELECT record_id FROM "Patient_Lists" WHERE image_path = %(image_path)s AND image_row_no = %(image_row_no)s', + params, + ) + image_match = cur.fetchone() + inpatient_matches = [] + if patient["住院号"]: + cur.execute('SELECT record_id FROM "Patient_Lists" WHERE inpatient_no = %(inpatient_no)s LIMIT 2', params) + inpatient_matches = cur.fetchall() + if image_match and inpatient_matches and inpatient_matches[0][0] != image_match[0]: + cur.execute(set_sql + " WHERE record_id = %(record_id)s", {**params, "record_id": inpatient_matches[0][0]}) + updated = cur.rowcount + if updated: + cur.execute('DELETE FROM "Patient_Lists" WHERE record_id = %(record_id)s', {"record_id": image_match[0]}) + match_method = "inpatient_no覆盖重复图片记录" if updated else "" + elif image_match: + cur.execute(set_sql + " WHERE record_id = %(record_id)s", {**params, "record_id": image_match[0]}) + updated = cur.rowcount + match_method = "image_path+image_row_no" if updated else "" + elif len(inpatient_matches) == 1: + cur.execute(set_sql + " WHERE record_id = %(record_id)s", {**params, "record_id": inpatient_matches[0][0]}) + updated = cur.rowcount + match_method = "inpatient_no" if updated else "" + else: + updated = 0 + match_method = "" + result = {"enabled": True, "updated": updated, "match_method": match_method} + if updated == 0: + result["message"] = "PostgreSQL未找到对应记录，已保留本地修订，后续可重建/入库后再同步" + result["retryable"] = True + return result + except Exception as exc: + return {"enabled": True, "updated": 0, "error": str(exc), "retryable": True} + + +def sync_status_from_result(result: dict[str, Any]) -> dict[str, Any]: + if not result.get("enabled"): + state = "未启用" + elif result.get("error"): + state = "失败" + elif result.get("updated", 0) > 0: + state = "成功" + else: + state = "待同步" + return { + "状态": state, + "更新时间": dt.datetime.now().isoformat(timespec="seconds"), + "更新行数": result.get("updated", 0), + "匹配方式": result.get("match_method", ""), + "提示": result.get("error") or result.get("message", ""), + } + + +def pending_postgres_sync_count() -> int: + count = 0 + for correction in load_json(CORRECTIONS_PATH, []): + if correction.get("AI修改") or correction.get("已提交人工通过"): + continue + if not has_valid_inpatient_no(correction.get("患者信息", {})): + continue + status = correction.get("PostgreSQL同步", {}) + if status.get("状态") != "成功": + count += 1 + return count + + +def db_status() -> dict[str, Any]: + status = { + "enabled": POSTGRES_SYNC_ENABLED, + "configured": db_configured(), + "ok": False, + "pending_sync_count": pending_postgres_sync_count(), + } + if not POSTGRES_SYNC_ENABLED or not db_configured(): + return status + try: + with postgres_connection() as conn: + with conn.cursor() as cur: + cur.execute('SELECT 1 FROM "Patient_Lists" LIMIT 1') + status["ok"] = True + except Exception as exc: + status["error"] = str(exc) + return status + + +def ensure_audit_columns() -> None: + if not db_configured(): + return + statements = [ + 'ALTER TABLE "Patient_Lists" ADD COLUMN IF NOT EXISTS audit_result text', + 'ALTER TABLE "Patient_Lists" ADD COLUMN IF NOT EXISTS audit_ai_feedback text', + 'ALTER TABLE "Patient_Lists" ADD COLUMN IF NOT EXISTS audit_ai_raw_output text', + 'ALTER TABLE "Patient_Lists" ADD COLUMN IF NOT EXISTS audit_manual_feedback text', + 'ALTER TABLE "Patient_Lists" ADD COLUMN IF NOT EXISTS audit_machine_verdict text', + 'ALTER TABLE "Patient_Lists" ADD COLUMN IF NOT EXISTS audit_source text', + 'ALTER TABLE "Patient_Lists" ADD COLUMN IF NOT EXISTS audit_checked_by text', + 'ALTER TABLE "Patient_Lists" ADD COLUMN IF NOT EXISTS audit_checked_at timestamptz', + 'ALTER TABLE "Patient_Lists" ADD COLUMN IF NOT EXISTS audit_patient_before jsonb', + 'ALTER TABLE "Patient_Lists" ADD COLUMN IF NOT EXISTS audit_patient_after jsonb', + 'ALTER TABLE "Patient_Lists" ADD COLUMN IF NOT EXISTS audit_change_summary text', + 'CREATE INDEX IF NOT EXISTS idx_patient_lists_audit_result ON "Patient_Lists"(audit_result)', + ] + with postgres_connection() as conn: + with conn.cursor() as cur: + for statement in statements: + cur.execute(statement) + + +def valid_inpatient_sql(alias: str = "") -> str: + prefix = f"{alias}." if alias else "" + return f"{prefix}inpatient_no IS NOT NULL AND btrim({prefix}inpatient_no) <> ''" + + +def sample_db_records(source: str, count: int) -> list[dict[str, Any]]: + if not db_configured(): + return [] + ensure_audit_columns() + review_where = "review_status = '自动复核通过'" if source == "db_auto_passed" else "review_status = '人工复核通过'" + where = f"{review_where} AND {valid_inpatient_sql()}" + with postgres_connection() as conn: + with conn.cursor(row_factory=dict_row) as cur: + cur.execute(f'SELECT * FROM "Patient_Lists" WHERE {where} ORDER BY random() LIMIT %s', (count,)) + return [db_row_to_public(row) for row in cur.fetchall()] + + +def update_audit_postgres(public: dict[str, Any], payload: dict[str, Any]) -> dict[str, Any]: + if not POSTGRES_SYNC_ENABLED: + return {"enabled": False, "updated": 0, "message": "PostgreSQL同步已关闭"} + if not db_configured(): + return {"enabled": True, "updated": 0, "error": "数据库环境变量未完整配置"} + ensure_audit_columns() + image_path = normalize_text(public.get("image_path")) + row_no = int(public.get("image_row_no") or 0) + item_payload = payload.get("item") if isinstance(payload.get("item"), dict) else {} + original_patient = normalize_patient(payload.get("original_patient") or item_payload.get("audit_original_patient") or public.get("patient", {})) + patient = normalize_patient(item_payload.get("patient") or public.get("patient", {})) + inpatient_no = normalize_text(patient.get("住院号") or public.get("patient", {}).get("住院号")).upper() + if not inpatient_no: + return {"enabled": True, "updated": 0, "error": "缺少住院号，未同步到正式表", "retryable": False} + change_summary = patient_change_summary(original_patient, patient) + params = { + "image_path": image_path, + "image_row_no": row_no, + "inpatient_no": inpatient_no, + "patient_name": patient["姓名"], + "gender": patient["性别"], + "age": patient["年龄"], + "diagnosis": patient["诊断"], + "admission_time": patient["入院时间"], + "last_write_time": patient["最后书写时间"], + "hospital_days": hospital_days_value(patient["住院天数"]), + "discharge_time": patient["出院时间"], + "postoperative_days": patient["手术后天数"], + "audit_result": normalize_text(payload.get("audit_result") or payload.get("status")), + "audit_ai_feedback": normalize_text(payload.get("audit_ai_feedback") or payload.get("model_result")), + "audit_ai_raw_output": normalize_text(payload.get("audit_ai_raw_output")), + "audit_manual_feedback": normalize_text(payload.get("audit_manual_feedback")), + "audit_machine_verdict": normalize_text(payload.get("audit_machine_verdict")), + "audit_source": normalize_text(payload.get("audit_source")), + "audit_checked_by": session.get("username", ""), + "audit_patient_before": json.dumps(original_patient, ensure_ascii=False), + "audit_patient_after": json.dumps(patient, ensure_ascii=False), + "audit_change_summary": change_summary, + } + set_sql = """ + UPDATE "Patient_Lists" + SET patient_name = %(patient_name)s, + gender = %(gender)s, + age = %(age)s, + inpatient_no = %(inpatient_no)s, + diagnosis = %(diagnosis)s, + admission_time = %(admission_time)s, + last_write_time = %(last_write_time)s, + hospital_days = %(hospital_days)s, + discharge_time = %(discharge_time)s, + postoperative_days = %(postoperative_days)s, + audit_result = %(audit_result)s, + audit_ai_feedback = %(audit_ai_feedback)s, + audit_ai_raw_output = %(audit_ai_raw_output)s, + audit_manual_feedback = %(audit_manual_feedback)s, + audit_machine_verdict = %(audit_machine_verdict)s, + audit_source = %(audit_source)s, + audit_checked_by = %(audit_checked_by)s, + audit_patient_before = %(audit_patient_before)s::jsonb, + audit_patient_after = %(audit_patient_after)s::jsonb, + audit_change_summary = %(audit_change_summary)s, + audit_checked_at = now() + """ + try: + with postgres_connection() as conn: + with conn.cursor() as cur: + updated = 0 + match_method = "" + image_match = None + inpatient_matches = [] + if image_path and row_no: + cur.execute( + 'SELECT record_id FROM "Patient_Lists" WHERE image_path = %(image_path)s AND image_row_no = %(image_row_no)s', + params, + ) + image_match = cur.fetchone() + if inpatient_no: + cur.execute('SELECT record_id FROM "Patient_Lists" WHERE inpatient_no = %(inpatient_no)s LIMIT 2', params) + inpatient_matches = cur.fetchall() + if image_match and inpatient_matches and inpatient_matches[0][0] != image_match[0]: + cur.execute(set_sql + " WHERE record_id = %(record_id)s", {**params, "record_id": inpatient_matches[0][0]}) + updated = cur.rowcount + if updated: + cur.execute('DELETE FROM "Patient_Lists" WHERE record_id = %(record_id)s', {"record_id": image_match[0]}) + match_method = "inpatient_no覆盖重复图片记录" if updated else "" + elif image_match: + cur.execute(set_sql + " WHERE record_id = %(record_id)s", {**params, "record_id": image_match[0]}) + updated = cur.rowcount + match_method = "image_path+image_row_no" if updated else "" + elif len(inpatient_matches) == 1: + cur.execute(set_sql + " WHERE record_id = %(record_id)s", {**params, "record_id": inpatient_matches[0][0]}) + updated = cur.rowcount + match_method = "inpatient_no" if updated else "" + result = {"enabled": True, "updated": updated, "match_method": match_method} + if updated == 0: + result["message"] = "PostgreSQL未找到对应抽查记录" + return result + except Exception as exc: + return {"enabled": True, "updated": 0, "error": str(exc)} + + +def audit_history_rows( + page: int, + page_size: int, + source: str = "", + sort: str = "checked_desc", + query: str = "", + status: str = "", +) -> dict[str, Any]: + if not db_configured(): + return {"items": [], "total": 0, "page": page, "page_size": page_size} + ensure_audit_columns() + clauses = [""" + COALESCE(audit_result, '') <> '' + OR COALESCE(audit_ai_feedback, '') <> '' + OR COALESCE(audit_ai_raw_output, '') <> '' + OR COALESCE(audit_manual_feedback, '') <> '' + OR COALESCE(audit_machine_verdict, '') <> '' + OR COALESCE(audit_change_summary, '') <> '' + """] + clauses.append(valid_inpatient_sql()) + params: list[Any] = [] + if source: + clauses.append("audit_source = %s") + params.append(source) + status_clauses = { + "manual_pending": "COALESCE(audit_result, '') = ''", + "manual_passed": "audit_result = '人工核验通过'", + "manual_failed": "audit_result = '人工核验异常'", + "manual_unsure": "audit_result = '暂不确定'", + "ai_failed": "audit_machine_verdict = '异常'", + } + if status in status_clauses: + clauses.append(status_clauses[status]) + if query: + like = f"%{query}%" + clauses.append(""" + patient_name ILIKE %s + OR inpatient_no ILIKE %s + OR major_department ILIKE %s + OR sub_department ILIKE %s + OR source_folder ILIKE %s + """) + params.extend([like, like, like, like, like]) + where = " AND ".join(f"({clause})" for clause in clauses) + order_sql = "audit_checked_at ASC NULLS LAST, record_id ASC" if sort == "checked_asc" else "audit_checked_at DESC NULLS LAST, record_id DESC" + with postgres_connection() as conn: + with conn.cursor(row_factory=dict_row) as cur: + cur.execute(f'SELECT count(*) AS total FROM "Patient_Lists" WHERE {where}', params) + total = int(cur.fetchone()["total"]) + cur.execute( + f'SELECT * FROM "Patient_Lists" WHERE {where} ORDER BY {order_sql} LIMIT %s OFFSET %s', + (*params, page_size, (page - 1) * page_size), + ) + rows = [db_row_to_public(row) for row in cur.fetchall()] + return {"items": rows, "total": total, "page": page, "page_size": page_size} + + +def reset_audit_history_records() -> dict[str, Any]: + if not POSTGRES_SYNC_ENABLED: + return {"enabled": False, "updated": 0, "message": "PostgreSQL同步已关闭"} + if not db_configured(): + return {"enabled": True, "updated": 0, "error": "数据库环境变量未完整配置"} + ensure_audit_columns() + where = """ + COALESCE(audit_result, '') <> '' + OR COALESCE(audit_ai_feedback, '') <> '' + OR COALESCE(audit_ai_raw_output, '') <> '' + OR COALESCE(audit_manual_feedback, '') <> '' + OR COALESCE(audit_machine_verdict, '') <> '' + OR COALESCE(audit_source, '') <> '' + OR COALESCE(audit_change_summary, '') <> '' + OR audit_checked_at IS NOT NULL + """ + with postgres_connection() as conn: + with conn.cursor() as cur: + cur.execute( + f""" + UPDATE "Patient_Lists" + SET audit_result = NULL, + audit_ai_feedback = NULL, + audit_ai_raw_output = NULL, + audit_manual_feedback = NULL, + audit_machine_verdict = NULL, + audit_source = NULL, + audit_checked_by = NULL, + audit_patient_before = NULL, + audit_patient_after = NULL, + audit_change_summary = NULL, + audit_checked_at = NULL + WHERE {where} + """ + ) + updated = cur.rowcount + return {"enabled": True, "updated": updated} + + +def audit_overview_summary() -> dict[str, Any]: + summary = { + "total": 0, + "unreviewed": 0, + "passed": 0, + "failed": 0, + "unsure": 0, + "ai_only": 0, + "manual_review_passed_source": 0, + "db_auto_passed_source": 0, + } + if not db_configured(): + return summary + try: + ensure_audit_columns() + with postgres_connection() as conn: + with conn.cursor(row_factory=dict_row) as cur: + cur.execute( + """ + SELECT + count(*) FILTER ( + WHERE COALESCE(audit_result, '') <> '' + OR COALESCE(audit_ai_feedback, '') <> '' + OR COALESCE(audit_ai_raw_output, '') <> '' + OR COALESCE(audit_machine_verdict, '') <> '' + ) AS total, + count(*) FILTER ( + WHERE COALESCE(audit_result, '') = '' + AND ( + COALESCE(audit_ai_feedback, '') <> '' + OR COALESCE(audit_ai_raw_output, '') <> '' + OR COALESCE(audit_machine_verdict, '') <> '' + ) + ) AS unreviewed, + count(*) FILTER (WHERE audit_result = '人工核验通过') AS passed, + count(*) FILTER (WHERE audit_result = '人工核验异常') AS failed, + count(*) FILTER (WHERE audit_result = '暂不确定') AS unsure, + count(*) FILTER ( + WHERE COALESCE(audit_result, '') = '' + AND COALESCE(audit_machine_verdict, '') <> '' + ) AS ai_only, + count(*) FILTER (WHERE audit_source = 'manual_review_passed') AS manual_review_passed_source, + count(*) FILTER (WHERE audit_source = 'db_auto_passed') AS db_auto_passed_source + FROM "Patient_Lists" + """ + ) + row = cur.fetchone() or {} + summary.update({key: int(row.get(key) or 0) for key in summary}) + except Exception as exc: + summary["error"] = str(exc) + return summary + + +def processing_items_overview() -> dict[str, Any]: + corrections = correction_index() + archived = archived_correction_keys() + items = [ + public + for item in build_review_items() + if item.key not in archived + for public in [public_item(item, corrections.get(item.key))] + if has_valid_inpatient_no(public["patient"]) + ] + state_counts = {"待处理": 0, AI_PENDING_STATE: 0, STILL_CONFIRM_STATE: 0, MANUAL_PASSED_STATE: 0} + for item in items: + state_counts[item["manual_state"]] = state_counts.get(item["manual_state"], 0) + 1 + audit_summary = audit_overview_summary() + confirm_review = state_counts[AI_PENDING_STATE] + state_counts[STILL_CONFIRM_STATE] + return { + "review_total": len(items), + "pending_review": state_counts["待处理"], + "ai_pending": state_counts[AI_PENDING_STATE], + "still_issue": confirm_review, + "manual_passed": state_counts[MANUAL_PASSED_STATE], + "postgres_pending_sync": pending_postgres_sync_count(), + "audit_unreviewed": audit_summary.get("unreviewed", 0), + "audit_total": audit_summary.get("total", 0), + "audit_failed": audit_summary.get("failed", 0), + "audit_unsure": audit_summary.get("unsure", 0), + "audit_summary": audit_summary, + } + + +def public_item(item: ReviewItem, correction: dict[str, Any] | None = None) -> dict[str, Any]: + record = item.record + image = record.get("图片信息", {}) + review = record.get("复核", {}) + correction = correction or correction_index().get(item.key) + options = (correction or {}).get("复核选项", {}) + patient = merged_patient(record, correction) + warnings = validate_patient(patient, options) + if correction and correction.get("AI修改"): + state = AI_PENDING_STATE + elif correction and not warnings: + state = MANUAL_PASSED_STATE + elif correction: + state = STILL_CONFIRM_STATE + else: + state = "待处理" + return { + "key": item.key, + "batch_name": item.batch_name, + "source": item.source, + "origin": item.origin, + "major_department": record.get("大科室", ""), + "sub_department": record.get("子科室", ""), + "source_folder": record.get("来源文件夹", ""), + "image_path": image.get("图片路径", ""), + "image_name": image.get("图片名", ""), + "image_row_no": image.get("图片内行号", ""), + "patient": patient, + "original_patient": record.get("患者信息", {}), + "review_status": review.get("状态", ""), + "review_tips": effective_review_tips(review.get("提示", [])), + "manual_state": state, + "manual_note": (correction or {}).get("复核备注", ""), + "change_source": (correction or {}).get("修改来源") or ("AI修改" if (correction or {}).get("AI修改") else ("人工修改" if correction else "")), + "change_log": fallback_change_log(record, correction), + "ai_corrected": bool((correction or {}).get("AI修改")), + "ai_feedback": (correction or {}).get("AI反馈", ""), + "options": options, + "validation_warnings": warnings, + "updated_at": (correction or {}).get("更新时间", ""), + "postgres_sync": (correction or {}).get("PostgreSQL同步", {}), + "audit_original_patient": patient, + "audit_ai_feedback": record.get("audit_ai_feedback", ""), + "audit_ai_raw_output": record.get("audit_ai_raw_output", ""), + "audit_manual_feedback": record.get("audit_manual_feedback", ""), + "audit_machine_verdict": record.get("audit_machine_verdict", ""), + "audit_source": record.get("audit_source", ""), + "audit_checked_by": record.get("audit_checked_by", ""), + "audit_checked_at": str(record.get("audit_checked_at") or ""), + "audit_change_summary": record.get("audit_change_summary", ""), + } + + +def public_record(record: dict[str, Any]) -> dict[str, Any]: + image = record.get("图片信息", {}) + patient = record.get("患者信息", {}) + image_path = normalize_text(image.get("图片路径")) + row_no = int(image.get("图片内行号") or 0) + return { + "key": item_key(image_path, row_no), + "batch_name": record.get("处理批次", ""), + "major_department": record.get("大科室", ""), + "sub_department": record.get("子科室", ""), + "source_folder": record.get("来源文件夹", ""), + "image_path": image_path, + "image_name": image.get("图片名", ""), + "image_row_no": row_no, + "patient": {column: patient.get(column, "") for column in COLUMNS}, + "audit_original_patient": {column: patient.get(column, "") for column in COLUMNS}, + "audit_result": record.get("audit_result", ""), + "audit_ai_feedback": record.get("audit_ai_feedback", ""), + "audit_ai_raw_output": record.get("audit_ai_raw_output", ""), + "audit_manual_feedback": record.get("audit_manual_feedback", ""), + "audit_machine_verdict": record.get("audit_machine_verdict", ""), + "audit_source": record.get("audit_source", ""), + "audit_checked_by": record.get("audit_checked_by", ""), + "audit_checked_at": str(record.get("audit_checked_at") or ""), + "audit_change_summary": record.get("audit_change_summary", ""), + } + + +def db_row_to_public(row: dict[str, Any]) -> dict[str, Any]: + patient = { + "姓名": row.get("patient_name", ""), + "性别": row.get("gender", ""), + "年龄": row.get("age", ""), + "住院号": row.get("inpatient_no", ""), + "诊断": row.get("diagnosis", ""), + "入院时间": row.get("admission_time", ""), + "最后书写时间": row.get("last_write_time", ""), + "住院天数": "" if row.get("hospital_days") is None else str(row.get("hospital_days")), + "出院时间": row.get("discharge_time", ""), + "手术后天数": row.get("postoperative_days", ""), + } + image_path = normalize_text(row.get("image_path")) + row_no = int(row.get("image_row_no") or 0) + return { + "key": item_key(image_path, row_no), + "record_id": row.get("record_id"), + "batch_name": row.get("batch_name", ""), + "major_department": row.get("major_department", ""), + "sub_department": row.get("sub_department", ""), + "source_folder": row.get("source_folder", ""), + "image_path": image_path, + "image_name": row.get("image_name", ""), + "image_row_no": row_no, + "patient": patient, + "audit_original_patient": row.get("audit_patient_before") or patient, + "review_status": row.get("review_status", ""), + "audit_result": row.get("audit_result", ""), + "audit_ai_feedback": row.get("audit_ai_feedback", ""), + "audit_ai_raw_output": row.get("audit_ai_raw_output", ""), + "audit_manual_feedback": row.get("audit_manual_feedback", ""), + "audit_machine_verdict": row.get("audit_machine_verdict", ""), + "audit_source": row.get("audit_source", ""), + "audit_checked_by": row.get("audit_checked_by", ""), + "audit_checked_at": str(row.get("audit_checked_at") or ""), + "audit_change_summary": row.get("audit_change_summary", ""), + } + + +def all_patient_records() -> list[dict[str, Any]]: + if not MERGED_RESULT_PATH.exists(): + return [] + merged = load_json(MERGED_RESULT_PATH, {}) + return [ + record + for record in merged.get("患者记录", []) + if record.get("图片信息", {}).get("图片路径") and record_has_valid_inpatient_no(record) + ] + + +def record_lookup_key(record: dict[str, Any]) -> str: + image = record.get("图片信息", {}) + image_path = normalize_text(image.get("图片路径")) + row_no = int(image.get("图片内行号") or 0) + return item_key(image_path, row_no) if image_path and row_no else "" + + +def storage_patient(patient: dict[str, Any]) -> dict[str, Any]: + normalized = normalize_patient(patient) + stored: dict[str, Any] = {column: normalized.get(column, "") for column in COLUMNS} + if stored["住院天数"].isdigit(): + stored["住院天数"] = int(stored["住院天数"]) + return stored + + +def merged_csv_row(record: dict[str, Any]) -> dict[str, Any]: + image = record.get("图片信息", {}) + patient = normalize_patient(record.get("患者信息", {})) + review = record.get("复核", {}) + tips = review.get("提示", []) + if not isinstance(tips, list): + tips = [tips] if normalize_text(tips) else [] + return { + "处理批次": record.get("处理批次", ""), + "大科室": record.get("大科室", ""), + "子科室": record.get("子科室", ""), + "来源文件夹": record.get("来源文件夹", ""), + "图片路径": image.get("图片路径", ""), + "图片名": image.get("图片名", ""), + "图片内行号": image.get("图片内行号", ""), + "拼接组序号": image.get("拼接组序号", ""), + "OCR请求ID": image.get("OCR请求ID", ""), + "姓名": patient["姓名"], + "性别": patient["性别"], + "年龄": patient["年龄"], + "住院号": patient["住院号"], + "诊断": patient["诊断"], + "入院时间": patient["入院时间"], + "最后书写时间": patient["最后书写时间"], + "住院天数": patient["住院天数"], + "出院时间": patient["出院时间"], + "手术后天数": patient["手术后天数"], + "复核状态": review.get("状态", ""), + "复核提示": "；".join(normalize_text(tip) for tip in tips if normalize_text(tip)), + "人工修正": bool(review.get("人工修正")), + } + + +def write_merged_csv(records: list[dict[str, Any]]) -> None: + fields = [ + "处理批次", + "大科室", + "子科室", + "来源文件夹", + "图片路径", + "图片名", + "图片内行号", + "拼接组序号", + "OCR请求ID", + "姓名", + "性别", + "年龄", + "住院号", + "诊断", + "入院时间", + "最后书写时间", + "住院天数", + "出院时间", + "手术后天数", + "复核状态", + "复核提示", + "人工修正", + ] + MERGED_CSV_PATH.parent.mkdir(parents=True, exist_ok=True) + temp_path = MERGED_CSV_PATH.with_suffix(MERGED_CSV_PATH.suffix + ".tmp") + with temp_path.open("w", encoding="utf-8-sig", newline="") as handle: + writer = csv.DictWriter(handle, fieldnames=fields) + writer.writeheader() + for record in records: + writer.writerow(merged_csv_row(record)) + temp_path.replace(MERGED_CSV_PATH) + + +def commit_manual_passed_records() -> dict[str, Any]: + corrections = load_json(CORRECTIONS_PATH, []) + review_items = {item.key: item for item in build_review_items()} + merged = load_json(MERGED_RESULT_PATH, {"汇总": {}, "患者记录": []}) + records = merged.get("患者记录", []) + record_index = {record_lookup_key(record): record for record in records if record_lookup_key(record)} + summary = { + "eligible": 0, + "merged_updated": 0, + "merged_already_current": 0, + "merged_missing": 0, + "postgres_success": 0, + "postgres_already_success": 0, + "postgres_not_found": 0, + "postgres_failed": 0, + "skipped": 0, + "skipped_missing_key": 0, + "skipped_ai": 0, + "skipped_archived": 0, + "skipped_issue": 0, + "archived_corrections": 0, + "items": [], + } + changed_merged = False + changed_corrections = False + archived_now: set[str] = set() + + for correction in corrections: + key = item_key(normalize_text(correction.get("图片路径")), int(correction.get("图片内行号") or 0)) + if not key: + summary["skipped"] += 1 + summary["skipped_missing_key"] += 1 + continue + if correction.get("已提交人工通过"): + summary["skipped"] += 1 + summary["skipped_archived"] += 1 + continue + if correction.get("AI修改"): + summary["skipped"] += 1 + summary["skipped_ai"] += 1 + continue + patient = normalize_patient(correction.get("患者信息", {})) + warnings = validate_patient(patient, correction.get("复核选项", {})) + if warnings: + summary["skipped"] += 1 + summary["skipped_issue"] += 1 + continue + summary["eligible"] += 1 + + merged_record = record_index.get(key) + if merged_record: + stored_patient = storage_patient(patient) + review = merged_record.setdefault("复核", {}) + manual_note = normalize_text(correction.get("复核备注", "")) + desired_review = { + "状态": "人工复核通过", + "提示": [], + "人工修正": True, + } + if manual_note: + desired_review["人工备注"] = manual_note + needs_update = ( + merged_record.get("患者信息") != stored_patient + or any(review.get(key) != value for key, value in desired_review.items()) + ) + if needs_update: + merged_record["患者信息"] = stored_patient + review.update(desired_review) + review["人工复核提交时间"] = dt.datetime.now().isoformat(timespec="seconds") + summary["merged_updated"] += 1 + changed_merged = True + else: + summary["merged_already_current"] += 1 + else: + summary["merged_missing"] += 1 + + sync_status = correction.get("PostgreSQL同步", {}) + if sync_status.get("状态") == "成功": + summary["postgres_already_success"] += 1 + summary["items"].append({"key": key, "status": "already_success"}) + correction["已提交人工通过"] = True + correction["提交归档时间"] = dt.datetime.now().isoformat(timespec="seconds") + archived_now.add(key) + changed_corrections = True + continue + + target = review_items.get(key) + if target is None: + summary["postgres_failed"] += 1 + summary["items"].append({"key": key, "status": "failed", "message": "未找到本地复核条目"}) + correction["已提交人工通过"] = True + correction["提交归档时间"] = dt.datetime.now().isoformat(timespec="seconds") + archived_now.add(key) + changed_corrections = True + continue + result = update_postgres_record( + target.record, + patient, + correction.get("复核选项", {}), + [], + normalize_text(correction.get("复核备注", "")), + True, + ) + correction["PostgreSQL同步"] = sync_status_from_result(result) + changed_corrections = True + if result.get("error"): + summary["postgres_failed"] += 1 + item_status = "failed" + elif result.get("updated", 0) > 0: + summary["postgres_success"] += 1 + item_status = "success" + else: + summary["postgres_not_found"] += 1 + item_status = "not_found" + summary["items"].append({"key": key, "status": item_status, "postgres": result}) + correction["已提交人工通过"] = True + correction["提交归档时间"] = dt.datetime.now().isoformat(timespec="seconds") + archived_now.add(key) + changed_corrections = True + + if changed_merged: + merged.setdefault("汇总", {})["人工复核提交时间"] = dt.datetime.now().isoformat(timespec="seconds") + atomic_write_json(MERGED_RESULT_PATH, merged) + write_merged_csv(records) + summary["archived_corrections"] = len(archived_now) + if changed_corrections: + atomic_write_json(CORRECTIONS_PATH, corrections) + RESULT_INFO_DIR.mkdir(parents=True, exist_ok=True) + report_path = RESULT_INFO_DIR / "提交已人工通过数据报告.json" + atomic_write_json(report_path, {**summary, "生成时间": dt.datetime.now().isoformat(timespec="seconds")}) + summary["report_path"] = str(report_path.relative_to(WORKSPACE_ROOT)) + summary["postgres"] = db_status() + return summary + + +def audit_public_by_key(key: str) -> dict[str, Any] | None: + if key in archived_correction_keys(): + return None + for item in build_review_items(): + public = public_item(item, correction_index().get(item.key)) + if public["key"] == key and has_valid_inpatient_no(public["patient"]): + return public + for record in all_patient_records(): + public = public_record(record) + if public["key"] == key: + return public + if db_configured(): + try: + with postgres_connection() as conn: + with conn.cursor(row_factory=dict_row) as cur: + cur.execute( + """ + SELECT * FROM "Patient_Lists" + WHERE encode(sha1((image_path || '|' || image_row_no::text)::bytea), 'hex') LIKE %(prefix)s + LIMIT 1 + """, + {"prefix": key + "%"}, + ) + row = cur.fetchone() + return db_row_to_public(row) if row else None + except Exception: + return None + return None + + +def kimi_config() -> dict[str, Any]: + config = load_config().get("kimi", {}) + enabled_value = config.get("enabled", os.getenv("KIMI_API_ENABLED", "1")) + enabled = str(enabled_value).lower() not in {"0", "false", "no", "off"} + return { + "api_key": config.get("api_key") or os.getenv("KIMI_API_KEY") or os.getenv("MOONSHOT_API_KEY") or "", + "model": config.get("model") or os.getenv("KIMI_MODEL") or "kimi-k2.6", + "enabled": enabled, + } + + +def kimi_audit_prompt(public: dict[str, Any]) -> str: + return ( + "你是HIS患者列表OCR抽查助手。请核对截图中目标患者所在行与给定结构化字段是否一致；目标行通常位于裁剪区域中部附近。" + "重点核对姓名、性别、年龄、住院号、诊断、入院时间、最后书写时间、住院天数、出院时间、手术后天数。" + "患者出院后医生可能继续补充病历，所以最后书写时间晚于出院时间不算异常，不要把它列为问题。" + "日期时间只比较真实时间值，不比较补零格式；例如2021-11-12 7:39:05与2021-11-12 07:39:05完全一致，必须判定通过，不要列为异常，也不要建议改成未补零格式。" + "只返回一个JSON对象，不要解释、不要Markdown表格、不要代码块外文本。" + "JSON格式为：{\"结论\":\"通过/异常/不确定\",\"问题\":[...],\"建议修正\":{...}}。" + f"结构化字段：{json.dumps(public['patient'], ensure_ascii=False)}" + ) + + +def equivalent_field_value(field: str, left: Any, right: Any) -> bool: + left_text = normalize_text(left) + right_text = normalize_text(right) + if left_text == right_text: + return True + if field in DATETIME_COLUMNS: + return normalize_datetime_text(left_text) == normalize_datetime_text(right_text) + return False + + +def clean_audit_model_json(data: dict[str, Any], public: dict[str, Any] | None = None) -> dict[str, Any]: + cleaned = dict(data) + patient = normalize_patient((public or {}).get("patient", {})) + suggestions = cleaned.get("建议修正") + if isinstance(suggestions, dict): + kept_suggestions = {} + for field, value in suggestions.items(): + normalized_field = normalize_text(field) + if normalized_field in COLUMNS and equivalent_field_value(normalized_field, value, patient.get(normalized_field, "")): + continue + kept_suggestions[field] = value + cleaned["建议修正"] = kept_suggestions + issues = cleaned.get("问题") + if isinstance(issues, list): + kept = [] + removed = [] + for issue in issues: + text = normalize_text(issue) + if any(pattern in text for pattern in IGNORED_AI_ISSUE_PATTERNS) or any(pattern in text for pattern in FORMAT_ONLY_AI_ISSUE_PATTERNS): + removed.append(text) + else: + kept.append(issue) + cleaned["问题"] = kept + suggestions = cleaned.get("建议修正") + if not cleaned.get("问题") and (not isinstance(suggestions, dict) or not suggestions): + cleaned["结论"] = "通过" + return cleaned + + +def parse_model_verdict(text: str) -> str: + cleaned = normalize_text(text) + if cleaned.startswith("```"): + cleaned = re.sub(r"^```(?:json)?\s*", "", cleaned) + cleaned = re.sub(r"\s*```$", "", cleaned) + try: + data = json.loads(cleaned) + except Exception: + match = re.search(r"\{.*\}", cleaned, flags=re.S) + if not match: + return "" + try: + data = json.loads(match.group(0)) + except Exception: + return "" + verdict = normalize_text(data.get("结论") or data.get("verdict") or data.get("result")) + return verdict if verdict in {"通过", "异常", "不确定"} else verdict + + +def call_kimi_vision(public: dict[str, Any]) -> dict[str, str]: + config = kimi_config() + if not config["enabled"]: + raise RuntimeError("目前无AI功能") + if not config["api_key"]: + raise RuntimeError("未设置 AI API Key，请在设置页配置") + path = safe_workspace_path(public["image_path"]) + cropped = crop_image(path, int(public["image_row_no"] or 1), 2) + image_base64 = image_to_base64(cropped) + prompt = kimi_audit_prompt(public) + payload = { + "model": config["model"], + "temperature": 0.6, + "max_tokens": KIMI_AUDIT_MAX_TOKENS, + "thinking": {"type": "disabled"}, + "messages": [ + { + "role": "user", + "content": [ + {"type": "image_url", "image_url": {"url": f"data:image/png;base64,{image_base64}"}}, + {"type": "text", "text": prompt}, + ], + } + ], + } + request = urllib.request.Request( + "https://api.moonshot.cn/v1/chat/completions", + data=json.dumps(payload, ensure_ascii=False).encode("utf-8"), + headers={"Authorization": f"Bearer {config['api_key']}", "Content-Type": "application/json"}, + method="POST", + ) + try: + with urllib.request.urlopen(request, timeout=KIMI_TIMEOUT_SECONDS) as response: + data = json.loads(response.read().decode("utf-8")) + except urllib.error.HTTPError as exc: + detail = exc.read().decode("utf-8", errors="replace") + raise RuntimeError(detail or str(exc)) from exc + except (TimeoutError, socket.timeout) as exc: + raise RuntimeError(f"AI接口超时（{KIMI_TIMEOUT_SECONDS}秒），请稍后重试或减少批量数量") from exc + except urllib.error.URLError as exc: + raise RuntimeError(f"AI接口连接失败：{exc.reason}") from exc + result = data.get("choices", [{}])[0].get("message", {}).get("content", "") + parsed = parse_json_object(result) + if parsed: + parsed = clean_audit_model_json(parsed, public) + clean_result = json.dumps(parsed, ensure_ascii=False, indent=2) if parsed else "" + verdict_source = clean_result or result + return { + "prompt": prompt, + "result": clean_result, + "raw_result": result, + "machine_verdict": parse_model_verdict(verdict_source), + } + + +def parse_json_object(text: str) -> dict[str, Any]: + cleaned = normalize_text(text) + fenced = re.search(r"```(?:json)?\s*(\{.*?\})\s*```", text or "", flags=re.S | re.I) + candidates = [fenced.group(1)] if fenced else [cleaned] + decoder = json.JSONDecoder() + for start, char in enumerate(cleaned): + if char == "{": + candidates.append(cleaned[start:]) + for candidate in candidates: + try: + data = json.loads(candidate) + return data if isinstance(data, dict) else {} + except Exception: + try: + data, _ = decoder.raw_decode(candidate) + return data if isinstance(data, dict) else {} + except Exception: + continue + return {} + + +def call_kimi_correction(public: dict[str, Any]) -> dict[str, Any]: + config = kimi_config() + if not config["enabled"]: + raise RuntimeError("目前无AI功能") + if not config["api_key"]: + raise RuntimeError("未设置 AI API Key，请在设置页配置") + path = safe_workspace_path(public["image_path"]) + cropped = crop_image(path, int(public["image_row_no"] or 1), 2) + image_base64 = image_to_base64(cropped) + prompt = ( + "你是HIS患者列表OCR修正助手。请根据截图中目标患者所在行修正给定结构化字段；目标行通常位于裁剪区域中部附近。" + "字段只能包含：姓名、性别、年龄、住院号、诊断、入院时间、最后书写时间、住院天数、出院时间、手术后天数。" + "时间格式尽量输出YYYY-MM-DD HH:MM:SS；出院时间可以为空；手术后天数为空或形如后X天。" + "请只返回JSON：{\"患者信息\":{...},\"说明\":\"...\"}。" + f"当前字段：{json.dumps(public['patient'], ensure_ascii=False)}" + ) + payload = { + "model": config["model"], + "temperature": 0.6, + "max_tokens": KIMI_CORRECTION_MAX_TOKENS, + "thinking": {"type": "disabled"}, + "messages": [ + { + "role": "user", + "content": [ + {"type": "image_url", "image_url": {"url": f"data:image/png;base64,{image_base64}"}}, + {"type": "text", "text": prompt}, + ], + } + ], + } + request = urllib.request.Request( + "https://api.moonshot.cn/v1/chat/completions", + data=json.dumps(payload, ensure_ascii=False).encode("utf-8"), + headers={"Authorization": f"Bearer {config['api_key']}", "Content-Type": "application/json"}, + method="POST", + ) + try: + with urllib.request.urlopen(request, timeout=KIMI_TIMEOUT_SECONDS) as response: + data = json.loads(response.read().decode("utf-8")) + except urllib.error.HTTPError as exc: + detail = exc.read().decode("utf-8", errors="replace") + raise RuntimeError(detail or str(exc)) from exc + except (TimeoutError, socket.timeout) as exc: + raise RuntimeError(f"AI接口超时（{KIMI_TIMEOUT_SECONDS}秒），请稍后重试或减少批量数量") from exc + except urllib.error.URLError as exc: + raise RuntimeError(f"AI接口连接失败：{exc.reason}") from exc + result = data.get("choices", [{}])[0].get("message", {}).get("content", "") + parsed = parse_json_object(result) + patient = parsed.get("患者信息", parsed) + if not isinstance(patient, dict): + patient = {} + corrected = dict(public.get("patient", {})) + corrected.update({column: patient.get(column, corrected.get(column, "")) for column in COLUMNS}) + return {"prompt": prompt, "result": result, "patient": normalize_patient(corrected), "note": normalize_text(parsed.get("说明", ""))} + + +def filtered_items() -> list[dict[str, Any]]: + corrections = correction_index() + archived = archived_correction_keys() + batch = normalize_text(request.args.get("batch")) + status = normalize_text(request.args.get("status") or "pending") + sort = normalize_text(request.args.get("sort") or "source") + query = normalize_text(request.args.get("q")).lower() + items = [] + for item in build_review_items(): + if item.key in archived: + continue + public = public_item(item, corrections.get(item.key)) + if not has_valid_inpatient_no(public["patient"]): + continue + if batch and public["batch_name"] != batch: + continue + if status == "pending" and public["manual_state"] != "待处理": + continue + if status == "done" and public["manual_state"] != MANUAL_PASSED_STATE: + continue + if status == "still_issue" and public["manual_state"] != STILL_CONFIRM_STATE: + continue + if status == "confirming" and public["manual_state"] not in {STILL_CONFIRM_STATE, AI_PENDING_STATE}: + continue + if status == "ai_pending" and public["manual_state"] != AI_PENDING_STATE: + continue + if query: + haystack = json.dumps(public, ensure_ascii=False).lower() + if query not in haystack: + continue + items.append(public) + if sort == "updated_desc": + items.sort(key=lambda item: (bool(item.get("updated_at")), item.get("updated_at") or ""), reverse=True) + return items + + +def safe_workspace_path(path_text: str) -> Path: + raw = Path(path_text) + path = raw if raw.is_absolute() else WORKSPACE_ROOT / raw + resolved = path.resolve() + if WORKSPACE_ROOT not in resolved.parents and resolved != WORKSPACE_ROOT: + abort(403) + if not resolved.exists(): + abort(404) + return resolved + + +def crop_image(path: Path, row_no: int, context_rows: int = 2) -> Image.Image: + if path.suffix.lower() not in IMAGE_EXTENSIONS: + abort(404) + with Image.open(path) as image: + source = image.convert("RGB") + width, height = source.size + inferred_rows = max(1, int(row_no or 1), round(height / 39.0)) + row_height = max(1.0, height / inferred_rows) + row_index = min(inferred_rows - 1, max(0, int(row_no or 1) - 1)) + target_top = int(row_index * row_height) + target_bottom = int(min(height, (row_index + 1) * row_height)) + top = int(max(0, target_top - row_height * context_rows)) + bottom = int(min(height, target_bottom + row_height * context_rows)) + if bottom <= top: + top = max(0, min(height - 1, target_top)) + bottom = min(height, top + max(1, int(row_height * max(1, context_rows + 1)))) + return source.crop((0, top, width, bottom)) + + +def image_to_base64(image: Image.Image) -> str: + source = image.convert("RGB") + if source.width > KIMI_IMAGE_MAX_WIDTH: + ratio = KIMI_IMAGE_MAX_WIDTH / source.width + next_size = (KIMI_IMAGE_MAX_WIDTH, max(1, round(source.height * ratio))) + source = source.resize(next_size, Image.Resampling.LANCZOS) + buffer = io.BytesIO() + source.save(buffer, format="PNG", optimize=True) + return base64.b64encode(buffer.getvalue()).decode("ascii") + + +def image_crop_response(path: Path, row_no: int, context_rows: int = 2) -> Response: + cropped = crop_image(path, row_no, context_rows) + buffer = io.BytesIO() + cropped.save(buffer, format="PNG") + buffer.seek(0) + return send_file(buffer, mimetype="image/png", max_age=0) + + +@app.route("/") +def index() -> str: + return render_template("index.html") + + +@app.route("/favicon.ico") +def favicon() -> Response: + return Response(status=204) + + +@app.route("/login", methods=["GET", "POST"]) +def login() -> str | Response: + if request.method == "POST": + username = request.form.get("username", "") + password = request.form.get("password", "") + expected_user = os.getenv("REVIEW_APP_USERNAME", "admin") + expected_password = os.getenv("REVIEW_APP_PASSWORD") or os.getenv("HIS_DB_PASSWORD") + if expected_password and secrets.compare_digest(username, expected_user) and secrets.compare_digest(password, expected_password): + session["authenticated"] = True + session["username"] = username + session["permissions"] = dict(FULL_PERMISSIONS) + return redirect(url_for("index")) + for user in load_config().get("users", []): + if secrets.compare_digest(username, str(user.get("username", ""))) and password_matches(password, user): + session["authenticated"] = True + session["username"] = username + session["permissions"] = normalized_permissions(user.get("permissions")) + return redirect(url_for("index")) + return render_template("login.html", error="登录失败"), 401 + return render_template("login.html", error="") + + +@app.route("/logout", methods=["POST"]) +def logout() -> Response: + session.clear() + return redirect(url_for("login")) + + +@app.route("/api/session") +def api_session() -> Response: + require_login() + return json_response({ + "username": session.get("username", ""), + "permissions": current_permissions(), + "permission_labels": PERMISSION_LABELS, + "kimi_enabled": kimi_config()["enabled"], + }) + + +@app.route("/api/summary") +def api_summary() -> Response: + require_login() + corrections = correction_index() + archived = archived_correction_keys() + items = [ + public + for item in build_review_items() + if item.key not in archived + for public in [public_item(item, corrections.get(item.key))] + if has_valid_inpatient_no(public["patient"]) + ] + batches: dict[str, dict[str, Any]] = {} + state_counts = {"待处理": 0, STILL_CONFIRM_STATE: 0, MANUAL_PASSED_STATE: 0, AI_PENDING_STATE: 0} + for item in items: + state_counts[item["manual_state"]] = state_counts.get(item["manual_state"], 0) + 1 + batch = batches.setdefault( + item["batch_name"], + {"batch_name": item["batch_name"], "total": 0, "pending": 0, "ai_pending": 0, "still_issue": 0, "done": 0}, + ) + batch["total"] += 1 + if item["manual_state"] == MANUAL_PASSED_STATE: + batch["done"] += 1 + elif item["manual_state"] == AI_PENDING_STATE: + batch["ai_pending"] += 1 + batch["still_issue"] += 1 + elif item["manual_state"] == STILL_CONFIRM_STATE: + batch["still_issue"] += 1 + else: + batch["pending"] += 1 + state_counts["待确认"] = state_counts[AI_PENDING_STATE] + state_counts[STILL_CONFIRM_STATE] + return json_response( + { + "total": len(items), + "state_counts": state_counts, + "batches": sorted(batches.values(), key=lambda value: value["batch_name"]), + "corrections_path": str(CORRECTIONS_PATH), + "postgres": db_status(), + "audit_summary": audit_overview_summary(), + "permissions": current_permissions(), + } + ) + + +@app.route("/api/processing-items") +def api_processing_items() -> Response: + require_login() + return json_response(processing_items_overview()) + + +@app.route("/api/items") +def api_items() -> Response: + require_permission("review") + page = max(1, int(request.args.get("page", 1))) + page_size = min(100, max(10, int(request.args.get("page_size", 40)))) + items = filtered_items() + start = (page - 1) * page_size + return json_response({"items": items[start : start + page_size], "total": len(items), "page": page, "page_size": page_size}) + + +@app.route("/api/items/") +def api_item(key: str) -> Response: + require_permission("review") + if key in archived_correction_keys(): + abort(404) + corrections = correction_index() + for item in build_review_items(): + if item.key == key: + public = public_item(item, corrections.get(key)) + if not has_valid_inpatient_no(public["patient"]): + abort(404) + return json_response(public) + abort(404) + + +@app.route("/api/items//validate", methods=["POST"]) +def api_validate(key: str) -> Response: + require_permission("review") + payload = request.get_json(force=True) + patient = normalize_patient(payload.get("patient", {})) + options = payload.get("options", {}) + return json_response({"warnings": validate_patient(patient, options), "patient": patient}) + + +@app.route("/api/items//correction", methods=["POST", "DELETE"]) +def api_correction(key: str) -> Response: + require_permission("review") + corrections = correction_index() + target = None + for item in build_review_items(): + if item.key == key: + target = item + break + if target is None: + abort(404) + if request.method == "DELETE": + original_patient = normalize_patient(target.record.get("患者信息", {})) + original_warnings = validate_patient(original_patient, {}) + pg_sync = update_postgres_record(target.record, original_patient, {}, original_warnings, "", False) + delete_correction(key) + return json_response({"ok": True, "postgres": pg_sync}) + payload = request.get_json(force=True) + patient = normalize_patient(payload.get("patient", {})) + options = {} + warnings = validate_patient(patient, options) + if not has_valid_inpatient_no(patient): + return error_response("缺少住院号，不能保存到复核表格。", status=400) + previous_correction = corrections.get(key) + previous_patient = merged_patient(target.record, previous_correction) + change_log = fallback_change_log(target.record, previous_correction) + change_log.extend(change_log_entries(previous_patient, patient, "人工")) + image = target.record.get("图片信息", {}) + correction = { + "图片路径": image.get("图片路径", ""), + "图片内行号": int(image.get("图片内行号") or 0), + "患者信息": patient, + "复核选项": options, + "复核备注": normalize_text(payload.get("manual_note", "")), + "修改来源": "人工修改", + "修改记录": change_log, + "更新时间": dt.datetime.now().isoformat(timespec="seconds"), + } + pg_sync = update_postgres_record(target.record, patient, options, warnings, correction["复核备注"], True) + correction["PostgreSQL同步"] = sync_status_from_result(pg_sync) + save_correction(correction) + return json_response({"ok": True, "warnings": warnings, "item": public_item(target, correction), "postgres": pg_sync}) + + +@app.route("/api/items//kimi-correction", methods=["POST"]) +def api_item_kimi_correction(key: str) -> Response: + require_permission("review") + corrections = correction_index() + target = None + for item in build_review_items(): + if item.key == key: + target = item + break + if target is None: + abort(404) + public = public_item(target, corrections.get(key)) + try: + result = call_kimi_correction(public) + except RuntimeError as exc: + return error_response(str(exc), status=502) + except Exception as exc: + return error_response(f"AI修改失败：{exc}", status=500) + if not has_valid_inpatient_no(result["patient"]): + return error_response("AI结果缺少有效住院号，已拒绝写入复核表格。", status=422) + previous_correction = corrections.get(key) + previous_patient = public.get("patient", {}) + change_log = fallback_change_log(target.record, previous_correction) + change_log.extend(change_log_entries(previous_patient, result["patient"], "AI")) + image = target.record.get("图片信息", {}) + correction = { + "图片路径": image.get("图片路径", ""), + "图片内行号": int(image.get("图片内行号") or 0), + "患者信息": result["patient"], + "复核选项": {}, + "复核备注": result["note"] or "AI修改，待人工确认", + "修改来源": "AI修改", + "修改记录": change_log, + "更新时间": dt.datetime.now().isoformat(timespec="seconds"), + "AI修改": True, + "AI反馈": result["result"], + "PostgreSQL同步": { + "状态": "AI待确认", + "更新时间": dt.datetime.now().isoformat(timespec="seconds"), + "提示": "AI修改结果需人工保存确认后再同步数据库", + }, + } + save_correction(correction) + return json_response({"ok": True, "item": public_item(target, correction), "result": result}) + + +@app.route("/api/postgres/status") +def api_postgres_status() -> Response: + require_login() + return json_response(db_status()) + + +@app.route("/api/postgres/sync", methods=["POST"]) +def api_postgres_sync() -> Response: + require_permission("review") + corrections = load_json(CORRECTIONS_PATH, []) + review_items = {item.key: item for item in build_review_items()} + summary = {"total": 0, "success": 0, "failed": 0, "not_found": 0, "skipped": 0, "items": []} + changed = False + for correction in corrections: + if correction.get("AI修改") or correction.get("已提交人工通过"): + summary["skipped"] += 1 + continue + sync_status = correction.get("PostgreSQL同步", {}) + if sync_status.get("状态") == "成功": + summary["skipped"] += 1 + continue + key = item_key(normalize_text(correction.get("图片路径")), int(correction.get("图片内行号") or 0)) + target = review_items.get(key) + if target is None: + summary["failed"] += 1 + summary["items"].append({"key": key, "status": "failed", "message": "未找到本地复核条目"}) + continue + patient = normalize_patient(correction.get("患者信息", {})) + if not has_valid_inpatient_no(patient): + summary["skipped"] += 1 + continue + options = correction.get("复核选项", {}) + manual_note = normalize_text(correction.get("复核备注", "")) + warnings = validate_patient(patient, options) + result = update_postgres_record(target.record, patient, options, warnings, manual_note, True) + correction["PostgreSQL同步"] = sync_status_from_result(result) + changed = True + summary["total"] += 1 + if result.get("error"): + summary["failed"] += 1 + item_status = "failed" + elif result.get("updated", 0) > 0: + summary["success"] += 1 + item_status = "success" + else: + summary["not_found"] += 1 + item_status = "not_found" + summary["items"].append({"key": key, "status": item_status, "postgres": result}) + if changed: + atomic_write_json(CORRECTIONS_PATH, corrections) + summary["postgres"] = db_status() + return json_response(summary) + + +@app.route("/api/postgres/sync/drop_failed", methods=["POST"]) +def api_postgres_sync_drop_failed() -> Response: + require_permission("review") + payload = request.get_json(force=True) + keys = set(str(key) for key in payload.get("keys", []) if str(key)) + if not keys: + return json_response({"removed": 0}) + corrections = load_json(CORRECTIONS_PATH, []) + kept = [] + removed = 0 + for correction in corrections: + key = item_key(normalize_text(correction.get("图片路径")), int(correction.get("图片内行号") or 0)) + if key in keys: + removed += 1 + continue + kept.append(correction) + if removed: + atomic_write_json(CORRECTIONS_PATH, kept) + return json_response({"removed": removed, "postgres": db_status()}) + + +@app.route("/api/audit/sample", methods=["POST"]) +def api_audit_sample() -> Response: + require_permission("audit") + payload = request.get_json(force=True) + count = min(20, max(1, int(payload.get("count", 5)))) + source = normalize_text(payload.get("source") or "manual_review_passed") + if source == "manual_review_passed": + corrections = correction_index() + archived = archived_correction_keys() + candidates = [ + public + for item in build_review_items() + if item.key not in archived + for public in [public_item(item, corrections.get(item.key))] + if public["manual_state"] == MANUAL_PASSED_STATE and has_valid_inpatient_no(public["patient"]) + ] + sampled = random.sample(candidates, min(count, len(candidates))) if candidates else [] + else: + sampled = sample_db_records("db_auto_passed", count) + for item in sampled: + item["audit_source"] = source + return json_response({"items": sampled}) + + +@app.route("/api/audit/kimi", methods=["POST"]) +def api_audit_kimi() -> Response: + require_permission("audit") + payload = request.get_json(force=True) + key = normalize_text(payload.get("key")) + public = payload.get("item") if isinstance(payload.get("item"), dict) else None + if public is None: + public = audit_public_by_key(key) + if public is None: + abort(404) + try: + result = call_kimi_vision(public) + except RuntimeError as exc: + return error_response(str(exc), status=502) + except Exception as exc: + return error_response(f"AI抽查失败：{exc}", status=500) + return json_response(result) + + +@app.route("/api/audit/result", methods=["POST"]) +def api_audit_result() -> Response: + require_permission("audit") + payload = request.get_json(force=True) + key = normalize_text(payload.get("key")) + public = payload.get("item") if isinstance(payload.get("item"), dict) else None + if public is None: + public = audit_public_by_key(key) + if public is None: + abort(404) + result = update_audit_postgres(public, payload) + return json_response({"ok": result.get("updated", 0) > 0, "postgres": result}) + + +@app.route("/api/audit/history") +def api_audit_history() -> Response: + require_permission("audit_history") + page = max(1, int(request.args.get("page", 1))) + page_size = min(100, max(10, int(request.args.get("page_size", 40)))) + source = normalize_text(request.args.get("source")) + status = normalize_text(request.args.get("status")) + sort = normalize_text(request.args.get("sort") or "checked_desc") + query = normalize_text(request.args.get("q")) + return json_response(audit_history_rows(page, page_size, source, sort, query, status)) + + +@app.route("/api/settings") +def api_settings() -> Response: + require_permission("settings") + config = load_config() + kimi = kimi_config() + env_user = os.getenv("REVIEW_APP_USERNAME", "admin") + return json_response( + { + "permissions": current_permissions(), + "permission_labels": PERMISSION_LABELS, + "users": [ + {"username": env_user, "created_at": "环境变量用户", "permissions": dict(FULL_PERMISSIONS), "builtin": True}, + *[ + { + "username": user.get("username"), + "created_at": user.get("created_at", ""), + "permissions": normalized_permissions(user.get("permissions")), + "builtin": False, + } + for user in config.get("users", []) + ], + ], + "kimi_model": kimi["model"], + "kimi_api_key_set": bool(kimi["api_key"]), + "kimi_enabled": bool(kimi["enabled"]), + } + ) + + +@app.route("/api/settings/users", methods=["POST"]) +def api_settings_users() -> Response: + require_permission("settings") + payload = request.get_json(force=True) + username = normalize_text(payload.get("username")) + password = str(payload.get("password") or "") + permissions = normalized_permissions(payload.get("permissions")) + if not username or not password: + return json_response({"error": "用户名和密码不能为空"}, status=400) + config = load_config() + users = config.setdefault("users", []) + if any(user.get("username") == username for user in users): + return json_response({"error": "用户已存在"}, status=400) + hashed = password_hash(password) + users.append({ + "username": username, + "password_hash": hashed["hash"], + "salt": hashed["salt"], + "permissions": permissions, + "created_at": dt.datetime.now().isoformat(timespec="seconds"), + }) + save_config(config) + return json_response({"ok": True}) + + +@app.route("/api/settings/users//permissions", methods=["POST"]) +def api_settings_user_permissions(username: str) -> Response: + require_permission("settings") + payload = request.get_json(force=True) + config = load_config() + for user in config.get("users", []): + if user.get("username") == username: + user["permissions"] = normalized_permissions(payload.get("permissions")) + save_config(config) + return json_response({"ok": True}) + abort(404) + + +@app.route("/api/settings/kimi", methods=["POST"]) +def api_settings_kimi() -> Response: + require_permission("settings") + payload = request.get_json(force=True) + config = load_config() + kimi = config.setdefault("kimi", {}) + model = normalize_text(payload.get("model")) or "kimi-k2.6" + api_key = normalize_text(payload.get("api_key")) + kimi["model"] = model + if api_key: + kimi["api_key"] = api_key + save_config(config) + return json_response({"ok": True}) + + +@app.route("/api/settings/kimi/enabled", methods=["POST"]) +def api_settings_kimi_enabled() -> Response: + require_permission("settings") + payload = request.get_json(force=True) + config = load_config() + kimi = config.setdefault("kimi", {}) + kimi["enabled"] = bool(payload.get("enabled")) + save_config(config) + return json_response({"ok": True, "enabled": kimi["enabled"]}) + + +@app.route("/api/settings/commit-manual-passed", methods=["POST"]) +def api_settings_commit_manual_passed() -> Response: + require_permission("settings") + return json_response(commit_manual_passed_records()) + + +@app.route("/api/settings/reset-audit-history", methods=["POST"]) +def api_settings_reset_audit_history() -> Response: + require_permission("settings") + result = reset_audit_history_records() + if result.get("error"): + return error_response(result["error"], status=500) + return json_response(result) + + +@app.route("/api/crop") +def api_crop() -> Response: + require_login() + path = safe_workspace_path(request.args.get("path", "")) + row_no = max(1, int(request.args.get("row", 1))) + context_rows = min(5, max(0, int(request.args.get("context", 2)))) + return image_crop_response(path, row_no, context_rows) + + +@app.route("/api/image") +def api_image() -> Response: + require_login() + path = safe_workspace_path(request.args.get("path", "")) + if path.suffix.lower() not in IMAGE_EXTENSIONS: + abort(404) + return send_file(path, max_age=0) + + +if __name__ == "__main__": + host = os.getenv("REVIEW_APP_HOST", "127.0.0.1") + port = int(os.getenv("REVIEW_APP_PORT", "8090")) + app.run(host=host, port=port) diff --git a/患者列表处理/人工复核网页端/docker-compose.yml b/患者列表处理/人工复核网页端/docker-compose.yml new file mode 100644 index 0000000..472f8d1 --- /dev/null +++ b/患者列表处理/人工复核网页端/docker-compose.yml @@ -0,0 +1,19 @@ +name: his-list-review + +services: + manual-review: + build: + context: . + image: his-list-manual-review:latest + container_name: his-list-manual-review + env_file: + - ../.env + environment: + WORKSPACE_ROOT: /workspace + REVIEW_APP_HOST: 0.0.0.0 + REVIEW_APP_PORT: 8000 + ports: + - "${REVIEW_APP_PORT:-8090}:8000" + volumes: + - ..:/workspace + restart: unless-stopped diff --git a/患者列表处理/人工复核网页端/requirements.txt b/患者列表处理/人工复核网页端/requirements.txt new file mode 100644 index 0000000..63e644e --- /dev/null +++ b/患者列表处理/人工复核网页端/requirements.txt @@ -0,0 +1,5 @@ +Flask==3.0.3 +gunicorn==23.0.0 +Pillow==10.4.0 +psycopg[binary]==3.2.9 +python-dotenv==1.0.1 diff --git a/患者列表处理/人工复核网页端/static/app.css b/患者列表处理/人工复核网页端/static/app.css new file mode 100644 index 0000000..2fea719 --- /dev/null +++ b/患者列表处理/人工复核网页端/static/app.css @@ -0,0 +1,1339 @@ +:root { + --ink: #1f2933; + --muted: #68717d; + --line: #d8dee4; + --paper: #f4f6f8; + --panel: #ffffff; + --field: #f9fafb; + --teal: #147c72; + --teal-dark: #0d5d55; + --cinnabar: #c94a36; + --amber: #b7822f; + --ok: #497c43; + --shadow: 0 18px 50px rgba(31, 41, 51, 0.12); + color-scheme: light; +} + +* { + box-sizing: border-box; +} + +body { + margin: 0; + background: var(--paper); + color: var(--ink); + font-family: "Noto Sans CJK SC", "Source Han Sans SC", "Microsoft YaHei", sans-serif; + letter-spacing: 0; +} + +button, +input, +select, +textarea { + font: inherit; +} + +.topbar { + min-height: 76px; + display: grid; + grid-template-columns: minmax(280px, max-content) minmax(260px, 1fr) max-content; + align-items: center; + gap: 14px; + padding: 14px 22px; + border-bottom: 1px solid var(--line); + background: rgba(255, 255, 255, 0.94); + position: sticky; + top: 0; + z-index: 10; +} + +#saveState { + max-width: min(760px, 100%); + justify-self: center; + color: var(--ink); + font-size: 16px; + font-weight: 800; + line-height: 1.35; + text-align: center; + text-decoration: underline; + text-underline-offset: 4px; + white-space: normal; + overflow-wrap: anywhere; + overflow: hidden; + display: -webkit-box; + -webkit-box-orient: vertical; + -webkit-line-clamp: 2; + pointer-events: none; +} + +.eyebrow { + color: var(--teal); + font-size: 12px; + font-weight: 800; + letter-spacing: 0; +} + +h1 { + margin: 2px 0 0; + font-size: 22px; + font-weight: 800; +} + +.top-actions { + display: flex; + align-items: center; + justify-content: flex-end; + gap: 12px; +} + +.top-nav { + display: flex; + gap: 6px; + padding: 4px; + border: 1px solid var(--line); + border-radius: 8px; + background: var(--field); +} + +.nav-button { + min-height: 32px; + border: 0; + padding: 6px 10px; + background: transparent; + color: var(--muted); +} + +.nav-button.active { + background: var(--teal); + color: #fff; +} + +.save-state { + color: var(--muted); + font-size: 13px; +} + +.pg-state { + color: var(--ok); + font-size: 13px; + max-width: 360px; + overflow: hidden; + text-overflow: ellipsis; + white-space: nowrap; +} + +.pg-state.warn { + color: var(--cinnabar); +} + +.workspace { + height: calc(100vh - 76px); + display: grid; + grid-template-columns: minmax(320px, 26vw) minmax(420px, 1fr) minmax(390px, 28vw); + min-height: 640px; +} + +.page-view.hidden { + display: none; +} + +.hidden { + display: none !important; +} + +.overview-view, +.audit-view, +.settings-view { + min-height: calc(100vh - 76px); + padding: 18px 22px; +} + +.audit-view { + height: calc(100vh - 76px); + min-height: 0; + display: grid; + grid-template-rows: auto minmax(0, 1fr); + gap: 14px; + overflow: hidden; +} + +.audit-history-view { + height: calc(100vh - 76px); + min-height: 0; +} + +.overview-grid, +.batch-grid, +.settings-layout { + display: grid; + gap: 12px; +} + +.overview-grid { + grid-template-columns: repeat(auto-fit, minmax(150px, 1fr)); + margin-bottom: 16px; +} + +.overview-panel, +.settings-panel, +.audit-toolbar, +.audit-detail { + border: 1px solid var(--line); + border-radius: 8px; + background: var(--panel); + padding: 16px; +} + +.overview-panel h2 { + margin: 0 0 12px; + font-size: 18px; +} + +.batch-grid { + grid-template-columns: repeat(auto-fill, minmax(300px, 1fr)); +} + +.settings-layout { + grid-template-columns: minmax(560px, 1fr) minmax(340px, 420px); + align-items: start; +} + +.settings-panel { + display: grid; + gap: 12px; +} + +.settings-heading { + display: flex; + align-items: center; + justify-content: space-between; + gap: 12px; +} + +.settings-heading h2 { + margin: 0; + font-size: 18px; +} + +.settings-side, +.new-user-form { + display: grid; + gap: 12px; +} + +.settings-form-grid { + display: grid; + grid-template-columns: 1fr 1fr; + gap: 10px; +} + +.settings-label { + color: var(--muted); + font-size: 12px; + font-weight: 700; + margin-bottom: 7px; +} + +.settings-note { + margin: 0; + color: var(--muted); + font-size: 13px; + line-height: 1.6; +} + +.settings-panel label, +.audit-toolbar label { + display: grid; + gap: 6px; + color: var(--muted); + font-size: 12px; + font-weight: 700; +} + +.settings-panel input, +.audit-toolbar input, +.audit-toolbar select, +.audit-history-filters input, +.audit-history-filters select { + width: 100%; + border: 1px solid var(--line); + border-radius: 8px; + background: var(--field); + padding: 10px 11px; + color: var(--ink); + outline: none; +} + +.audit-toolbar { + display: grid; + grid-template-columns: minmax(180px, 220px) 150px auto minmax(420px, auto) minmax(150px, 1fr); + gap: 10px; + align-items: end; +} + +.ai-disabled .audit-toolbar { + grid-template-columns: minmax(180px, 220px) 150px auto minmax(150px, 1fr); +} + +.inline-field { + display: grid; + grid-template-columns: auto minmax(0, 1fr); + align-items: center; + gap: 8px; +} + +.audit-toolbar label.inline-field, +.audit-history-filters .inline-field { + display: grid; + grid-template-columns: auto minmax(0, 1fr); + gap: 8px; + align-items: center; +} + +.field-prefix { + color: var(--muted); + font-size: 12px; + font-weight: 900; + white-space: nowrap; +} + +.settings-panel input:focus, +.audit-toolbar input:focus, +.audit-toolbar select:focus, +.audit-history-filters input:focus, +.audit-history-filters select:focus { + border-color: var(--teal); + box-shadow: 0 0 0 3px rgba(20, 124, 114, 0.12); +} + +.audit-layout { + display: grid; + grid-template-columns: minmax(280px, 24vw) minmax(420px, 1fr) minmax(320px, 24vw); + gap: 14px; + min-height: 0; + height: 100%; + overflow: hidden; +} + +.audit-history-view .audit-toolbar { + grid-template-columns: auto minmax(140px, 1fr); + margin-bottom: 0; +} + +.audit-history-layout { + height: 100%; + min-height: 0; + overflow: hidden; + grid-template-columns: minmax(300px, 25vw) minmax(440px, 1fr) minmax(340px, 26vw); + gap: 16px; +} + +.audit-history-layout > .audit-history-left, +.audit-history-layout > .audit-detail, +.audit-history-layout > .audit-side, +.audit-layout > .queue-list, +.audit-layout > .audit-detail, +.audit-layout > .audit-side { + min-height: 0; + max-height: 100%; + overflow: auto; + overscroll-behavior: contain; +} + +.audit-layout > .audit-detail { + border: 0; + background: transparent; + padding: 2px 4px 14px; +} + +.audit-history-left { + display: grid; + grid-template-rows: auto minmax(0, 1fr); + gap: 12px; + overflow: hidden; +} + +.audit-history-layout > .audit-history-left { + overflow: hidden; +} + +.audit-history-filters { + display: grid; + gap: 8px; + padding: 12px; + border: 1px solid var(--line); + border-radius: 8px; + background: var(--panel); +} + +.audit-history-left > .queue-list { + overflow: auto; + align-content: start; + padding-right: 4px; + gap: 10px; +} + +.audit-history-layout .queue-item { + padding: 11px 12px; + min-height: 94px; +} + +.audit-history-layout .queue-title { + line-height: 1.25; +} + +.audit-history-layout .queue-meta { + display: block; + margin-top: 3px; + line-height: 1.45; + white-space: normal; + overflow-wrap: anywhere; +} + +.finder-grid { + display: grid; + grid-template-columns: repeat(2, minmax(0, 1fr)); + gap: 8px; +} + +.finder-card { + border: 1px solid var(--line); + border-radius: 8px; + background: var(--field); + padding: 10px; +} + +.finder-card strong { + display: block; + color: var(--ink); + font-size: 22px; + line-height: 1.1; +} + +.finder-card span { + color: var(--muted); + font-size: 12px; +} + +.queue-panel, +.image-panel, +.edit-panel { + min-width: 0; + min-height: 0; + overflow: auto; +} + +.queue-panel { + border-right: 1px solid var(--line); + padding: 16px; +} + +.queue-panel:focus { + outline: none; +} + +.image-panel { + padding: 16px 18px; +} + +.edit-panel { + border-left: 1px solid var(--line); + padding: 16px; + background: var(--panel); +} + +.stats-grid { + display: grid; + grid-template-columns: repeat(auto-fit, minmax(86px, 1fr)); + gap: 8px; + margin-bottom: 12px; +} + +.stat { + background: var(--panel); + border: 1px solid var(--line); + border-radius: 8px; + padding: 10px; +} + +.stat strong { + display: block; + font-size: 24px; + line-height: 1; +} + +.stat span { + color: var(--muted); + font-size: 12px; +} + +.batch-card { + border: 1px solid var(--line); + border-radius: 8px; + background: var(--panel); + padding: 12px; +} + +.batch-card h3 { + margin: 0 0 10px; + font-size: 14px; +} + +.batch-metrics { + display: grid; + grid-template-columns: repeat(4, minmax(0, 1fr)); + gap: 8px; +} + +.batch-metrics span { + min-width: 0; + color: var(--muted); + font-size: 12px; +} + +.batch-metrics strong { + display: block; + color: var(--ink); + font-size: 18px; + line-height: 1.1; +} + +.filters { + display: grid; + gap: 8px; + margin-bottom: 12px; +} + +.filter-row { + display: grid; + grid-template-columns: minmax(0, 1fr) minmax(0, 1fr); + gap: 8px; +} + +.filter-row #sortFilter { + grid-column: 1 / -1; +} + +.filters .review-filter-field { + display: grid; + grid-template-columns: auto minmax(0, 1fr); + align-items: center; + gap: 8px; +} + +.review-status-row { + display: grid; + grid-template-columns: minmax(0, 1.35fr) minmax(130px, 0.65fr); + gap: 8px; + align-items: stretch; +} + +.status-segment { + display: grid; + grid-template-columns: 1fr; + padding: 3px; + border: 1px solid var(--line); + border-radius: 8px; + background: #eef3f4; +} + +.status-pill { + min-height: 36px; + border: 0; + border-radius: 6px; + padding: 7px 9px; + background: transparent; + color: var(--muted); + font-weight: 800; +} + +.status-pill.active { + background: var(--teal); + color: #fff; + box-shadow: 0 5px 14px rgba(20, 124, 114, 0.18); +} + +.ai-action-panel { + border: 1px solid rgba(20, 124, 114, 0.28); + border-radius: 8px; + background: linear-gradient(180deg, #f5fbf8, #eef6f4); + padding: 9px; + display: grid; + gap: 8px; +} + +.review-ai-panel { + margin-top: 2px; +} + +.audit-ai-panel { + min-width: 420px; + grid-template-columns: auto 1fr; + align-items: center; + gap: 10px; +} + +.ai-action-title { + display: flex; + align-items: center; + gap: 7px; + color: var(--teal-dark); + font-size: 12px; + font-weight: 900; +} + +.ai-badge { + width: 28px; + height: 22px; + display: inline-grid; + place-items: center; + border-radius: 7px; + background: var(--teal); + color: #fff; + font-size: 12px; + font-weight: 900; +} + +.ai-action-buttons { + display: grid; + grid-template-columns: 1fr 1fr; + gap: 8px; +} + +.review-ai-panel .ai-action-buttons { + grid-template-columns: repeat(3, minmax(0, 1fr)); +} + +.audit-ai-panel .ai-action-buttons { + min-width: 0; +} + +.ai-button { + background: #fff; + border-color: rgba(20, 124, 114, 0.34); + font-weight: 800; +} + +.filters select, +.filters input, +.edit-form input, +.edit-form textarea { + width: 100%; + border: 1px solid var(--line); + border-radius: 8px; + background: var(--field); + padding: 10px 11px; + color: var(--ink); + outline: none; +} + +.filters select:focus, +.filters input:focus, +.edit-form input:focus, +.edit-form textarea:focus { + border-color: var(--teal); + box-shadow: 0 0 0 3px rgba(20, 124, 114, 0.12); +} + +.queue-list { + display: grid; + gap: 8px; +} + +.queue-item { + width: 100%; + text-align: left; + background: var(--panel); + border: 1px solid var(--line); + border-radius: 8px; + padding: 10px 11px; + cursor: pointer; + transition: border-color 0.15s ease, transform 0.15s ease, box-shadow 0.15s ease; +} + +.queue-item:hover, +.queue-item.active { + border-color: var(--teal); + box-shadow: var(--shadow); +} + +.queue-item:hover { + transform: translateY(-1px); +} + +.queue-title { + font-weight: 800; + display: flex; + justify-content: space-between; + gap: 8px; + align-items: center; + min-width: 0; +} + +.queue-title > span:first-child { + min-width: 0; + overflow: hidden; + text-overflow: ellipsis; + white-space: nowrap; +} + +.queue-badges { + display: flex; + justify-content: flex-end; + gap: 6px; + flex: 0 1 auto; + flex-wrap: wrap; + min-width: 0; +} + +.queue-badges .badge { + white-space: nowrap; +} + +.queue-badges .validation-badge { + max-width: 100%; +} + +.queue-meta, +.record-meta { + color: var(--muted); + font-size: 12px; + line-height: 1.55; +} + +.queue-load-state { + color: var(--muted); + font-size: 12px; + line-height: 1.55; + text-align: center; + padding: 8px 0 2px; +} + +.badge { + display: inline-flex; + align-items: center; + height: 22px; + padding: 0 8px; + border-radius: 999px; + font-size: 12px; + white-space: nowrap; + background: #eef1f4; + color: var(--muted); +} + +.badge.pending { + background: #f3e1dc; + color: var(--cinnabar); +} + +.badge.done { + background: #dde9dc; + color: var(--ok); +} + +.badge.warn { + background: #f0e4cb; + color: var(--amber); +} + +.badge.ai { + background: #dbe7f5; + color: #2f5f91; +} + +.record-strip { + display: flex; + justify-content: space-between; + gap: 12px; + align-items: flex-start; + margin-bottom: 12px; +} + +.index-note { + margin-bottom: 10px; + border: 1px solid rgba(20, 124, 114, 0.28); + border-radius: 8px; + background: #f3fbf8; + color: var(--teal-dark); + padding: 10px 12px; + font-size: 15px; + font-weight: 800; + text-align: center; + text-decoration: underline; + text-underline-offset: 4px; +} + +.record-title { + font-size: 18px; + font-weight: 800; +} + +.record-title .title-badges { + display: inline-flex; + vertical-align: middle; + margin-left: 8px; +} + +.crop-frame { + background: #e5e9ee; + border: 1px solid var(--line); + border-radius: 8px; + min-height: 280px; + display: grid; + place-items: center; + overflow: auto; + cursor: grab; + user-select: none; + overscroll-behavior: contain; + touch-action: none; +} + +.crop-frame.is-dragging { + cursor: grabbing; +} + +.crop-frame img { + max-width: none; + display: block; + height: auto; + transition: width 0.12s ease; + pointer-events: none; +} + +.zoom-actions { + display: flex; + align-items: center; + gap: 8px; + margin-top: 8px; +} + +.tips-box, +.validation-box { + margin-top: 12px; + display: grid; + gap: 8px; +} + +.tips-box .validation { + font-size: 15px; + font-weight: 800; + text-decoration: underline; + text-underline-offset: 4px; +} + +.tip, +.validation { + border-radius: 8px; + border: 1px solid var(--line); + background: var(--panel); + padding: 9px 10px; + font-size: 13px; +} + +.validation.error { + border-color: rgba(201, 74, 54, 0.42); + color: var(--cinnabar); + background: #fff7f4; +} + +.validation.ok { + border-color: rgba(73, 124, 67, 0.38); + color: var(--ok); + background: #f5fbf2; +} + +.validation.info { + border-color: rgba(104, 113, 125, 0.28); + color: var(--muted); + background: #f9fafb; +} + +.model-result { + min-height: 150px; + border: 1px solid var(--line); + border-radius: 8px; + padding: 12px; + background: var(--field); + color: var(--ink); + overflow: auto; +} + +.model-result table { + width: 100%; + border-collapse: collapse; + background: var(--panel); + border-radius: 8px; + overflow: hidden; +} + +.model-result th, +.model-result td { + border: 1px solid var(--line); + padding: 9px 10px; + vertical-align: top; + text-align: left; + font-size: 13px; +} + +.model-result th { + width: 120px; + color: var(--muted); + background: #eef3f4; + font-weight: 900; +} + +.model-result pre { + margin: 0; + white-space: pre-wrap; + overflow-wrap: anywhere; + font: inherit; +} + +.ai-full-output { + border: 1px solid var(--line); + border-radius: 8px; + background: var(--field); + padding: 10px 12px; +} + +.ai-full-output summary { + cursor: pointer; + color: var(--muted); + font-size: 12px; + font-weight: 900; +} + +.ai-full-output pre { + margin: 10px 0 0; + max-height: 220px; + overflow: auto; + white-space: pre-wrap; + overflow-wrap: anywhere; + color: var(--ink); + font: inherit; + font-size: 13px; +} + +.change-log { + display: grid; + gap: 8px; +} + +.change-log:empty { + display: none; +} + +.change-log details { + border: 1px solid rgba(20, 124, 114, 0.25); + border-radius: 8px; + background: #f5fbf8; + padding: 10px; +} + +.change-log summary { + cursor: pointer; + color: var(--teal-dark); + font-size: 12px; + font-weight: 900; +} + +.change-log-source { + margin: 8px 0 0; + color: var(--muted); + font-size: 12px; + font-weight: 800; +} + +.change-log table { + width: 100%; + border-collapse: collapse; + margin-top: 8px; + background: var(--panel); +} + +.change-log th, +.change-log td { + border: 1px solid var(--line); + padding: 7px 8px; + text-align: left; + vertical-align: top; + font-size: 12px; +} + +.audit-frame { + height: 360px; + min-height: 360px; + margin: 12px 0; + display: block; + padding: 0; +} + +.audit-image-stage { + width: max-content; + height: max-content; + min-width: 100%; + min-height: 100%; + display: flex; + align-items: center; + justify-content: center; +} + +.audit-image-stage img { + flex: 0 0 auto; + margin: 0; +} + +.ai-result-section, +.ai-prompt-output, +.ai-full-output { + width: 100%; +} + +.audit-side, +.history-card { + border: 1px solid var(--line); + border-radius: 8px; + background: var(--panel); + padding: 14px; +} + +.audit-side { + display: grid; + gap: 12px; + align-content: start; +} + +.audit-side h2 { + margin: 0; + font-size: 17px; +} + +.audit-side label, +.model-label { + display: grid; + gap: 6px; + color: var(--muted); + font-size: 12px; + font-weight: 700; +} + +.audit-side select, +.audit-side input, +.audit-side textarea, +.model-label textarea { + width: 100%; + border: 1px solid var(--line); + border-radius: 8px; + background: var(--field); + padding: 10px 11px; + color: var(--ink); +} + +.patient-kv, +.audit-verdict-grid { + display: grid; + grid-template-columns: auto 1fr; + gap: 6px 10px; + font-size: 13px; +} + +.audit-edit-grid { + display: grid; + grid-template-columns: 1fr 1fr; + gap: 10px; +} + +.audit-edit-grid label { + display: grid; + gap: 5px; + color: var(--muted); + font-size: 12px; + font-weight: 700; +} + +.audit-edit-grid label.wide { + grid-column: 1 / -1; +} + +.audit-edit-grid textarea { + min-height: 72px; + resize: vertical; +} + +.patient-kv span:nth-child(odd), +.audit-verdict-grid span { + color: var(--muted); +} + +.audit-verdict-grid { + padding: 10px; + border: 1px solid var(--line); + border-radius: 8px; + background: var(--field); +} + +.audit-verdict-grid strong { + min-width: 0; + overflow-wrap: anywhere; +} + +.history-list { + display: grid; + gap: 10px; +} + +.history-card { + display: grid; + gap: 8px; +} + +.edit-form { + display: grid; + gap: 12px; +} + +.edit-form label { + display: grid; + gap: 6px; + color: var(--muted); + font-size: 12px; + font-weight: 700; +} + +.edit-form input, +.edit-form textarea { + color: var(--ink); + font-size: 14px; + font-weight: 500; +} + +.form-grid { + display: grid; + grid-template-columns: 1fr 1fr; + gap: 10px; +} + +.check-row { + grid-template-columns: auto 1fr !important; + align-items: center; +} + +.check-row input { + width: 18px; + height: 18px; +} + +.button-row { + display: grid; + grid-template-columns: 1fr 1fr 1fr; + gap: 8px; +} + +.review-actions { + grid-template-columns: repeat(3, 1fr); +} + +.permission-grid, +.settings-list { + display: grid; + gap: 8px; +} + +.permission-grid { + grid-template-columns: repeat(auto-fit, minmax(118px, 1fr)); +} + +.permission-grid label, +.user-permissions label { + display: inline-flex; + align-items: center; + gap: 6px; + min-height: 38px; + padding: 7px 9px; + border: 1px solid var(--line); + border-radius: 8px; + background: var(--field); + color: var(--ink); + font-size: 13px; + font-weight: 500; +} + +.permission-grid input, +.user-permissions input { + width: auto; +} + +.settings-user { + border: 1px solid var(--line); + border-radius: 8px; + padding: 12px; + display: grid; + gap: 10px; +} + +.user-permissions { + display: grid; + grid-template-columns: repeat(auto-fit, minmax(104px, 1fr)); + gap: 8px; +} + +button, +.ghost-button, +.primary-button, +.secondary-button, +.danger-button { + border: 1px solid var(--line); + border-radius: 8px; + min-height: 40px; + padding: 9px 12px; + cursor: pointer; + text-decoration: none; + display: inline-grid; + place-items: center; + background: var(--panel); + color: var(--ink); +} + +button:disabled { + cursor: not-allowed; + opacity: 0.58; +} + +.compact-button { + min-height: 32px; + padding: 6px 10px; + font-size: 13px; +} + +.primary-button { + background: var(--teal); + border-color: var(--teal); + color: #fff; + font-weight: 800; +} + +.primary-button:hover { + background: var(--teal-dark); +} + +.secondary-button:hover, +.ghost-button:hover { + border-color: var(--teal); +} + +.danger-button { + color: var(--cinnabar); +} + +.danger-button:hover { + border-color: var(--cinnabar); +} + +.login-shell { + min-height: 100vh; + display: grid; + place-items: center; + background: var(--paper); +} + +.login-panel { + width: min(420px, calc(100vw - 32px)); + background: var(--panel); + border: 1px solid var(--line); + border-radius: 8px; + padding: 28px; + box-shadow: var(--shadow); +} + +.login-mark { + width: 54px; + height: 54px; + display: grid; + place-items: center; + border-radius: 8px; + background: var(--teal); + color: #fff; + font-weight: 900; + margin-bottom: 16px; +} + +.login-form { + display: grid; + gap: 14px; +} + +.login-form label { + display: grid; + gap: 7px; + color: var(--muted); + font-size: 13px; + font-weight: 700; +} + +.login-form input { + width: 100%; + border: 1px solid var(--line); + border-radius: 8px; + padding: 11px 12px; + background: var(--field); +} + +.login-form button { + background: var(--teal); + color: white; + font-weight: 800; + border-color: var(--teal); +} + +.form-error { + color: var(--cinnabar); + font-size: 13px; +} + +@media (max-width: 1120px) { + .workspace { + grid-template-columns: 330px minmax(420px, 1fr); + height: auto; + } + + .edit-panel { + grid-column: 1 / -1; + border-left: 0; + border-top: 1px solid var(--line); + } +} + +@media (max-width: 760px) { + .topbar { + height: auto; + grid-template-columns: 1fr; + align-items: flex-start; + } + + #saveState { + justify-self: stretch; + text-align: left; + } + + .workspace { + display: block; + } + + .queue-panel, + .image-panel, + .edit-panel { + border: 0; + border-bottom: 1px solid var(--line); + } + + .form-grid, + .button-row, + .stats-grid, + .overview-grid, + .settings-layout, + .settings-form-grid, + .audit-toolbar, + .audit-layout { + grid-template-columns: 1fr; + } +} diff --git a/患者列表处理/人工复核网页端/static/app.js b/患者列表处理/人工复核网页端/static/app.js new file mode 100644 index 0000000..413e677 --- /dev/null +++ b/患者列表处理/人工复核网页端/static/app.js @@ -0,0 +1,2042 @@ +const columns = [ + "姓名", + "性别", + "年龄", + "住院号", + "诊断", + "入院时间", + "最后书写时间", + "住院天数", + "出院时间", + "手术后天数", +]; + +const state = { + items: [], + current: null, + auditItems: [], + auditCurrent: null, + auditHistory: [], + auditHistoryCurrent: null, + processingOverview: null, + permissions: {}, + permissionLabels: {}, + kimiEnabled: true, + activeView: "overview", + aiBusy: false, + aiTaskLabel: "", + imageZoom: 1, + imagePan: null, + auditImageZoom: 1, + auditImagePan: null, + auditHistoryImageZoom: 1, + auditHistoryImagePan: null, + page: 1, + pageSize: 60, + totalItems: 0, + loadingItems: false, + pgTimer: null, + debounce: null, +}; + +const datetimeColumns = new Set(["入院时间", "出院时间", "最后书写时间"]); + +const $ = (selector) => document.querySelector(selector); + +function permissionKeyForView(view) { + return view === "auditHistory" ? "audit_history" : view; +} + +function permissionEntries() { + const labels = state.permissionLabels || {}; + const fallback = { + overview: "概览", + review: "复核", + audit: "抽查", + audit_history: "抽查一览", + settings: "设置", + }; + return Object.entries({ ...fallback, ...labels }); +} + +function badgeClass(item) { + if (item.manual_state === "AI修改-待确认") return "ai"; + if (item.manual_state === "人工复核通过") return "done"; + if (item.manual_state === "修订后仍需确认") return "warn"; + return "pending"; +} + +function auditBadgeClass(item) { + if (item.audit_result === "人工核验异常") return "pending"; + if (item.audit_result === "人工核验通过") return "done"; + if (item.audit_result === "暂不确定") return "warn"; + if (item.audit_machine_verdict === "异常") return "pending"; + if (item.audit_machine_verdict) return "ai"; + return "warn"; +} + +function auditBadgeText(item) { + if (item.audit_result) return item.audit_result; + if (item.audit_machine_verdict === "通过") return "AI核验通过"; + if (item.audit_machine_verdict === "异常") return "AI核验异常"; + if (item.audit_machine_verdict === "不确定") return "AI无法确定"; + return "未核验"; +} + +function auditHistoryBadgeText(item) { + if (item.audit_result) return item.audit_result; + return item.audit_ai_feedback || item.audit_ai_raw_output || item.audit_machine_verdict ? "未人工核验" : "未填写人工结论"; +} + +function auditSourceLabel(value) { + const labels = { + manual_review_passed: "已处理-人工复核通过项", + db_auto_passed: "数据库-自动复核通过", + }; + return labels[value] || value || ""; +} + +function validationBadgeHtml(item) { + const validationWarnings = validatePatientLocal(item.patient || {}); + return validationWarnings.length + ? `当前字段未通过校验` + : ""; +} + +function auditTitleHtml(item, badgeText = auditBadgeText(item)) { + return ` + ${escapeHtml(item.patient["姓名"] || "未识别姓名")} · ${escapeHtml(item.patient["住院号"] || "无住院号")} + + ${escapeHtml(badgeText)} + ${validationBadgeHtml(item)} + + `; +} + +function auditQueueItemHtml(item, activeKey, extraRows = [], badgeText = auditBadgeText(item)) { + const metaRows = [ + item.patient["住院号"] || "无住院号", + ...extraRows, + ].filter(Boolean); + return ` + + `; +} + +function auditProgressText() { + const total = state.auditItems.length; + if (!total) return ""; + const aiDone = state.auditItems.filter((item) => item.audit_machine_verdict || item.audit_ai_feedback).length; + const manualDone = state.auditItems.filter((item) => item.audit_result).length; + return `抽查进度：AI ${aiDone}/${total}，人工 ${manualDone}/${total}`; +} + +function setSaveState(text) { + $("#saveState").textContent = text || ""; +} + +function setStatusFilter(value, reload = true) { + if (value === "ai_pending" || value === "still_issue") value = "confirming"; + $("#statusFilter").value = value; + document.querySelectorAll(".status-pill").forEach((button) => { + button.classList.toggle("active", button.dataset.status === value); + }); + const more = $("#moreStatusFilter"); + more.value = value === "pending" ? "" : value; + if (value === "done") { + $("#sortFilter").value = "updated_desc"; + } + if (reload) loadItems(true); +} + +function reviewConfirmCount(summary) { + return summary.state_counts?.["待确认"] ?? ( + (summary.state_counts?.["AI修改-待确认"] || 0) + (summary.state_counts?.["修订后仍需确认"] || 0) + ); +} + +function setAiBusy(label, busy) { + state.aiBusy = busy; + state.aiTaskLabel = busy ? label : ""; + document.body.classList.toggle("ai-busy", busy); + document.querySelectorAll("#kimiCurrentButton, #kimiFiveButton, #kimiRemainingButton, #auditKimiButton, #auditKimiAllButton").forEach((button) => { + button.disabled = busy; + }); +} + +function updateAiVisibility() { + document.body.classList.toggle("ai-disabled", !state.kimiEnabled); + document.querySelectorAll(".review-ai-panel, .audit-ai-panel").forEach((panel) => { + panel.classList.toggle("hidden", !state.kimiEnabled); + }); +} + +function ensureAiReady(label) { + if (state.aiBusy) { + setSaveState(`${state.aiTaskLabel || "AI处理"}中，请等待完成后再操作`); + window.alert(`${state.aiTaskLabel || "AI处理"}中，请等待完成后再切换页面或发起新的AI任务。`); + return false; + } + if (!state.kimiEnabled) { + setSaveState("目前无AI功能"); + return false; + } + setAiBusy(label, true); + return true; +} + +function finishAiTask() { + setAiBusy("", false); +} + +function extractJsonData(text) { + const raw = String(text ?? "").trim(); + if (!raw) return null; + const candidates = []; + const fencePattern = /```(?:json)?\s*([\s\S]*?)```/gi; + let fenced; + while ((fenced = fencePattern.exec(raw))) { + candidates.push(fenced[1].trim()); + } + candidates.push(raw); + for (let index = 0; index < raw.length; index += 1) { + if (raw[index] === "{") candidates.push(raw.slice(index)); + } + for (const candidate of candidates) { + try { + const data = JSON.parse(candidate); + if (data && typeof data === "object" && !Array.isArray(data)) return data; + } catch (error) { + const parsed = parseJsonPrefix(candidate); + if (parsed) return parsed; + } + } + return null; +} + +function parseJsonPrefix(text) { + let depth = 0; + let inString = false; + let escaped = false; + for (let index = 0; index < text.length; index += 1) { + const char = text[index]; + if (inString) { + if (escaped) { + escaped = false; + } else if (char === "\\") { + escaped = true; + } else if (char === '"') { + inString = false; + } + continue; + } + if (char === '"') { + inString = true; + } else if (char === "{") { + depth += 1; + } else if (char === "}") { + depth -= 1; + if (depth === 0) { + try { + const data = JSON.parse(text.slice(0, index + 1)); + return data && typeof data === "object" && !Array.isArray(data) ? data : null; + } catch (error) { + return null; + } + } + } + } + return null; +} + +function equivalentFieldValue(column, left, right) { + const leftText = normalizeFieldValue(column, left); + const rightText = normalizeFieldValue(column, right); + return leftText === rightText; +} + +function cleanAuditJsonData(data, patient = null) { + if (!data) return null; + const cleaned = { ...data }; + const patterns = [ + "最后书写时间与出院时间顺序异常", + "最后书写时间晚于出院时间", + "出院后仍有书写记录", + "已出院后仍有书写记录", + "实际值一致", + "时间值一致", + "格式差异", + "补零差异", + "小时格式", + "日期格式补零", + "仅格式", + "无需修正", + ]; + if (cleaned.建议修正 && typeof cleaned.建议修正 === "object" && !Array.isArray(cleaned.建议修正) && patient) { + const keptSuggestions = {}; + Object.entries(cleaned.建议修正).forEach(([column, value]) => { + if (columns.includes(column) && equivalentFieldValue(column, value, patient[column] || "")) return; + keptSuggestions[column] = value; + }); + cleaned.建议修正 = keptSuggestions; + } + if (Array.isArray(cleaned.问题)) { + const kept = cleaned.问题.filter((issue) => !patterns.some((pattern) => String(issue).includes(pattern))); + cleaned.问题 = kept; + } + if ((!cleaned.问题 || !cleaned.问题.length) && (!cleaned.建议修正 || !Object.keys(cleaned.建议修正).length)) { + cleaned.结论 = "通过"; + } + return cleaned; +} + +function normalizeModelResultText(text, patient = null) { + const data = cleanAuditJsonData(extractJsonData(text), patient); + if (data) return JSON.stringify(data, null, 2); + return ""; +} + +function renderModelResult(element, jsonText, rawText = jsonText, patient = null) { + const normalized = normalizeModelResultText(jsonText || rawText, patient); + element.dataset.json = normalized; + element.dataset.raw = String(rawText ?? ""); + const data = extractJsonData(normalized); + if (!String(jsonText || rawText || "").trim()) { + element.innerHTML = `
暂无AI输出
`; + return; + } + if (!data) { + element.innerHTML = `
无法解析
`; + return; + } + const rows = Object.entries(data).map(([key, value]) => { + const display = typeof value === "string" ? value : JSON.stringify(value, null, 2); + return `${escapeHtml(key)}
${escapeHtml(display)}
`; + }).join(""); + element.innerHTML = `${rows}
`; +} + +function setAuditModelResult(jsonText, rawText = jsonText, patient = state.auditCurrent?.patient || null) { + renderModelResult($("#auditModelResult"), jsonText, rawText, patient); + $("#auditRawOutput").textContent = String(rawText || jsonText || ""); + const hasStructuredResult = Boolean($("#auditModelResult").dataset.json || String(jsonText || rawText || "").trim()); + $("#auditPromptDetails").open = !hasStructuredResult; + $("#auditRawOutputDetails").open = !hasStructuredResult; +} + +function setAuditPrompt(text) { + const value = String(text || ""); + $("#auditPrompt").value = value; + $("#auditPromptText").textContent = value || "暂无询问提示词"; +} + +function getAuditModelResult() { + return $("#auditModelResult").dataset.json || ""; +} + +function getAuditRawOutput() { + return $("#auditModelResult").dataset.raw || ""; +} + +function setAuditHistoryModelResult(jsonText, rawText = jsonText, patient = state.auditHistoryCurrent?.patient || null) { + renderModelResult($("#auditHistoryAiResult"), jsonText, rawText, patient); + $("#auditHistoryRawOutput").textContent = String(rawText || jsonText || ""); +} + +function modelVerdict(rawText, patient = null) { + const data = cleanAuditJsonData(extractJsonData(rawText), patient); + const verdict = String(data?.结论 || data?.verdict || data?.result || "").trim(); + return ["通过", "异常", "不确定"].includes(verdict) ? verdict : verdict; +} + +function formatTime(value) { + if (!value) return ""; + return String(value).replace("T", " "); +} + +async function api(path, options = {}) { + let response; + try { + response = await fetch(path, { + headers: { "Content-Type": "application/json", ...(options.headers || {}) }, + ...options, + }); + } catch (error) { + throw new Error("网络连接中断或网页端服务暂不可用，请确认服务仍在运行后重试"); + } + if (response.status === 401) { + window.location.href = "/login"; + return null; + } + if (!response.ok) { + const text = await response.text(); + try { + const data = JSON.parse(text); + throw new Error(data.error || text || response.statusText); + } catch (error) { + if (error instanceof SyntaxError) throw new Error(text || response.statusText); + throw error; + } + } + return response.json(); +} + +async function loadSummary() { + const summary = await api("/api/summary"); + if (!summary) return; + renderOverview(summary); + $("#statsGrid").innerHTML = ` +
${summary.total}复核条目
+
${summary.state_counts["待处理"]}待处理
+
${reviewConfirmCount(summary)}待确认
+
${summary.state_counts["人工复核通过"]}人工通过条目
+ `; + if (summary.postgres) { + renderPostgresState(summary.postgres); + } + const current = $("#batchFilter").value; + $("#batchFilter").innerHTML = `` + summary.batches + .map((batch) => ``) + .join(""); + $("#batchFilter").value = current; +} + +function renderOverview(summary) { + $("#overviewGrid").innerHTML = ` +
${summary.total}复核条目
+
${summary.state_counts["待处理"]}待处理
+
${reviewConfirmCount(summary)}待确认
+
${summary.state_counts["人工复核通过"]}人工通过
+
${summary.postgres?.pending_sync_count ?? 0}PG待同步
+ `; + $("#overviewBatches").innerHTML = summary.batches.map((batch) => ` +
+

${escapeHtml(batch.batch_name)}

+
+ ${escapeHtml(batch.pending)}待处理 + ${escapeHtml(batch.still_issue)}待确认 + ${escapeHtml(batch.done)}人工通过 + ${escapeHtml(batch.total)}总条目 +
+
+ `).join(""); + renderAuditOverview(summary.audit_summary || {}); +} + +function renderAuditOverview(auditSummary) { + const cards = [ + ["抽查记录", auditSummary.total || 0], + ["未人工核验", auditSummary.unreviewed || 0], + ["人工通过", auditSummary.passed || 0], + ["人工异常", auditSummary.failed || 0], + ["暂不确定", auditSummary.unsure || 0], + ]; + $("#overviewAudit").innerHTML = cards.map(([label, value]) => ` +
+

${escapeHtml(label)}

+
+ ${escapeHtml(value)}${escapeHtml(label)} +
+
+ `).join(""); +} + +function renderProcessingOverview(data) { + state.processingOverview = data; + const grid = $("#processingFinderGrid"); + if (!grid) return; + const cards = [ + ["复核待处理", data.pending_review || 0], + ["待确认", data.still_issue || 0], + ["抽查未人工核验", data.audit_unreviewed || 0], + ["抽查异常", data.audit_failed || 0], + ["待同步", data.postgres_pending_sync || 0], + ]; + grid.innerHTML = cards.map(([label, value]) => ` +
+ ${escapeHtml(value)} + ${escapeHtml(label)} +
+ `).join(""); +} + +async function findProcessingItems(quiet = false) { + const stateBox = $("#processingFinderState"); + if (!quiet && stateBox) stateBox.textContent = "寻找中"; + try { + const data = await api("/api/processing-items"); + if (!data) return; + renderProcessingOverview(data); + if (stateBox) { + stateBox.textContent = `复核待处理${data.pending_review || 0}，抽查未人工核验${data.audit_unreviewed || 0}，待同步${data.postgres_pending_sync || 0}`; + } + } catch (error) { + if (stateBox) stateBox.textContent = `寻找失败：${error.message}`; + } +} + +function escapeHtml(value) { + return String(value ?? "") + .replaceAll("&", "&") + .replaceAll("<", "<") + .replaceAll(">", ">") + .replaceAll('"', """) + .replaceAll("'", "'"); +} + +function itemSubtitle(item) { + const rowText = displayRowText(item); + return [ + item.batch_name, + item.source_folder, + item.image_name ? `${item.image_name} ${rowText}` : "", + ].filter(Boolean).join(" / "); +} + +function displayRowNo(item) { + const row = Number(item.image_row_no || 0); + if (!row) return ""; + return Math.max(1, row - 2); +} + +function displayRowText(item) { + const row = Number(item.image_row_no || 0); + if (!row) return ""; + const patientRow = displayRowNo(item); + return patientRow === row ? `第${row}行` : `第${patientRow}条（图片第${row}行）`; +} + +async function loadItems(selectFirst = false, append = false) { + if (state.loadingItems) return; + state.loadingItems = true; + if (!append) { + state.page = 1; + } + const params = new URLSearchParams({ + page: state.page, + page_size: state.pageSize, + status: $("#statusFilter").value, + sort: $("#sortFilter").value, + batch: $("#batchFilter").value, + q: $("#searchInput").value, + }); + try { + const data = await api(`/api/items?${params.toString()}`); + if (!data) return; + state.totalItems = data.total; + state.page = data.page; + if (append) { + const seen = new Set(state.items.map((item) => item.key)); + state.items = state.items.concat((data.items || []).filter((item) => !seen.has(item.key))); + } else { + state.items = data.items || []; + } + renderQueue(); + if (selectFirst && state.items.length) { + selectItem(state.items[0].key); + } else if (state.current) { + highlightActive(state.current.key); + } + } catch (error) { + setSaveState(`列表加载失败：${error.message}`); + if (!append) { + $("#queueList").innerHTML = `
${escapeHtml(error.message)}
`; + } + } finally { + state.loadingItems = false; + } +} + +async function loadNextItems() { + if (state.loadingItems || state.items.length >= state.totalItems) return; + state.page += 1; + await loadItems(false, true); +} + +async function reloadLoadedItems() { + const pageCount = Math.max(1, Math.ceil(Math.max(state.items.length, state.pageSize) / state.pageSize)); + state.items = []; + state.totalItems = 0; + for (let page = 1; page <= pageCount; page += 1) { + state.page = page; + await loadItems(false, page > 1); + if (state.items.length >= state.totalItems) break; + } +} + +function renderQueue() { + const list = $("#queueList"); + if (!state.items.length) { + list.innerHTML = `
当前筛选下没有记录
`; + return; + } + const footer = state.items.length < state.totalItems ? "继续向下滚动加载更多" : "已加载全部"; + list.innerHTML = state.items.map((item) => ` + + `).join("") + `
${state.items.length} / ${state.totalItems} · ${footer}
`; + list.querySelectorAll(".queue-item").forEach((button) => { + button.addEventListener("click", () => { + $("#queuePanel").focus({ preventScroll: true }); + selectItem(button.dataset.key); + }); + }); + if (state.current) highlightActive(state.current.key); +} + +function highlightActive(key) { + document.querySelectorAll(".queue-item").forEach((button) => { + const active = button.dataset.key === key; + button.classList.toggle("active", active); + if (active) { + button.scrollIntoView({ block: "nearest" }); + } + }); +} + +async function selectItem(key) { + const item = await api(`/api/items/${key}`); + if (!item) return; + state.current = item; + highlightActive(key); + renderCurrent(); +} + +async function selectAdjacentItem(offset) { + if (!state.current) return; + let index = state.items.findIndex((item) => item.key === state.current.key); + if (index < 0) return; + let nextIndex = index + offset; + if (nextIndex >= state.items.length && state.items.length < state.totalItems) { + await loadNextItems(); + index = state.items.findIndex((item) => item.key === state.current.key); + nextIndex = index + offset; + } + if (nextIndex < 0 || nextIndex >= state.items.length) return; + await selectItem(state.items[nextIndex].key); +} + +function renderCurrent() { + const item = state.current; + if (!item) return; + $("#recordTitle").textContent = `${item.patient["姓名"] || "未识别姓名"} · ${item.patient["住院号"] || "无住院号"}`; + const handledAt = item.updated_at ? ` / 处理时间 ${formatTime(item.updated_at)}` : ""; + $("#recordMeta").textContent = `${item.major_department} / ${item.sub_department} / ${item.source_folder} / ${item.image_name} ${displayRowText(item)}${handledAt}`; + const cropUrl = `/api/crop?path=${encodeURIComponent(item.image_path)}&row=${encodeURIComponent(item.image_row_no)}&context=2&v=${Date.now()}`; + $("#cropImage").src = cropUrl; + resetImageView(); + $("#fullImageLink").href = `/api/image?path=${encodeURIComponent(item.image_path)}`; + columns.forEach((column) => { + const field = document.querySelector(`[name="${column}"]`); + if (field) field.value = item.patient[column] ?? ""; + }); + $("#manualNote").value = item.manual_note || ""; + renderReviewChangeLog(); + renderTips(); +} + +function renderTips() { + const item = state.current; + const currentTips = (item.validation_warnings || []).map((tip) => ({ text: `当前修订: ${tip}`, cls: "error" })); + const historyTips = (item.review_tips || []).map((tip) => ({ text: `原始OCR提示: ${tip}`, cls: "info" })); + if (!currentTips.length && !historyTips.length) { + $("#tipsBox").innerHTML = `
当前字段通过校验
`; + return; + } + const currentHtml = currentTips.length + ? currentTips.map((tip) => `
${escapeHtml(tip.text)}
`).join("") + : `
当前字段通过校验
`; + $("#tipsBox").innerHTML = currentHtml + historyTips.map((tip) => `
${escapeHtml(tip.text)}
`).join(""); +} + +function formPatient(writeBack = false) { + const patient = {}; + columns.forEach((column) => { + const field = document.querySelector(`[name="${column}"]`); + patient[column] = field ? normalizeFieldValue(column, field.value) : ""; + if (writeBack && field && datetimeColumns.has(column)) field.value = patient[column]; + }); + return patient; +} + +function normalizeDateTime(value) { + const text = String(value ?? "").replaceAll("：", ":").trim().replace(/\s+/g, " "); + if (!text) return ""; + const match = text.match(/^(\d{4})-(\d{1,2})-(\d{1,2})\s+(\d{1,2}):(\d{1,2}):(\d{1,2})$/); + if (!match) return text; + const [, year, month, day, hour, minute, second] = match; + const date = new Date(Number(year), Number(month) - 1, Number(day), Number(hour), Number(minute), Number(second)); + if ( + date.getFullYear() !== Number(year) || + date.getMonth() !== Number(month) - 1 || + date.getDate() !== Number(day) || + date.getHours() !== Number(hour) || + date.getMinutes() !== Number(minute) || + date.getSeconds() !== Number(second) + ) { + return text; + } + const pad = (value) => String(value).padStart(2, "0"); + return `${year}-${pad(month)}-${pad(day)} ${pad(hour)}:${pad(minute)}:${pad(second)}`; +} + +function normalizeFieldValue(column, value) { + const text = String(value ?? "").replaceAll("：", ":").trim(); + return datetimeColumns.has(column) ? normalizeDateTime(text) : text; +} + +function patientFromEditor(selector) { + const patient = {}; + document.querySelectorAll(`${selector} [data-patient-field]`).forEach((field) => { + patient[field.dataset.patientField] = normalizeFieldValue(field.dataset.patientField, field.value); + }); + return patient; +} + +function parseDateTime(value) { + const text = normalizeDateTime(value); + const match = text.match(/^(\d{4})-(\d{2})-(\d{2}) (\d{2}):(\d{2}):(\d{2})$/); + if (!match) return null; + const [, year, month, day, hour, minute, second] = match.map(Number); + const date = new Date(year, month - 1, day, hour, minute, second); + if ( + date.getFullYear() !== year || + date.getMonth() !== month - 1 || + date.getDate() !== day || + date.getHours() !== hour || + date.getMinutes() !== minute || + date.getSeconds() !== second + ) return null; + return date; +} + +function validatePatientLocal(patient) { + const warnings = []; + if (!String(patient["姓名"] || "").trim()) warnings.push("缺少姓名"); + if (!["男", "女"].includes(String(patient["性别"] || "").trim())) warnings.push("性别异常"); + const age = String(patient["年龄"] || "").trim(); + if (age && !/^\d{1,3}岁$/.test(age)) warnings.push("年龄格式异常"); + if (!String(patient["住院号"] || "").trim()) warnings.push("缺少住院号"); + const admission = String(patient["入院时间"] || "").trim(); + const discharge = String(patient["出院时间"] || "").trim(); + const lastWrite = String(patient["最后书写时间"] || "").trim(); + const admissionTime = parseDateTime(admission); + const dischargeTime = parseDateTime(discharge); + const lastWriteTime = parseDateTime(lastWrite); + if (!admission) warnings.push("缺少入院时间"); + else if (!admissionTime) warnings.push("入院时间格式异常"); + if (discharge && !dischargeTime) warnings.push("出院时间格式异常"); + if (lastWrite && !lastWriteTime) warnings.push("最后书写时间格式异常"); + if (admissionTime && dischargeTime && dischargeTime < admissionTime) warnings.push("出院时间早于入院时间"); + if (admissionTime && lastWriteTime && lastWriteTime < admissionTime) warnings.push("最后书写时间早于入院时间"); + const hospitalDays = String(patient["住院天数"] || "").trim(); + if (hospitalDays && !/^\d+$/.test(hospitalDays)) warnings.push("住院天数格式异常"); + const postoperativeDays = String(patient["手术后天数"] || "").trim(); + if (postoperativeDays && !/^后\d+天$/.test(postoperativeDays)) warnings.push("手术后天数格式异常"); + return warnings; +} + +function renderValidationBox(selector, warnings, prefix = "当前修订") { + const box = $(selector); + if (!box) return; + if (!warnings.length) { + box.innerHTML = `
当前字段通过校验
`; + return; + } + box.innerHTML = warnings.map((warning) => `
${escapeHtml(prefix)}: ${escapeHtml(warning)}
`).join(""); +} + +function updateAuditValidation(editorSelector, tipsSelector, targetItem = null) { + const patient = patientFromEditor(editorSelector); + const warnings = validatePatientLocal(patient); + if (targetItem) targetItem.validation_warnings = warnings; + renderValidationBox(tipsSelector, warnings); + return warnings; +} + +function patientChanges(before = {}, after = {}) { + return columns + .map((column) => ({ + column, + before: normalizeFieldValue(column, before[column] || ""), + after: normalizeFieldValue(column, after[column] || ""), + })) + .filter((item) => item.before !== item.after); +} + +function displayChangeLogEntries(entries = []) { + if (!Array.isArray(entries)) return []; + return entries + .map((entry) => ({ + column: entry.column || entry["字段"] || "", + before: normalizeFieldValue(entry.column || entry["字段"] || "", entry.before ?? entry["修改前"] ?? ""), + after: normalizeFieldValue(entry.column || entry["字段"] || "", entry.after ?? entry["修改后"] ?? ""), + source: entry.source || entry["修改者"] || "", + })) + .filter((entry) => columns.includes(entry.column)); +} + +function renderChangeLog(selector, item, title = "修改记录", sourceLabel = "", explicitChanges = null) { + const box = $(selector); + if (!box) return; + const changes = displayChangeLogEntries(explicitChanges || item.change_log || []) + .concat(explicitChanges || item.change_log ? [] : patientChanges(item.audit_original_patient || item.original_patient || {}, item.patient || {})); + if (!changes.length && !item.audit_change_summary) { + box.innerHTML = ""; + return; + } + const showSource = changes.some((change) => change.source); + const rows = changes.length + ? changes.map((change) => ` + + ${escapeHtml(change.column)} + ${escapeHtml(change.before || "空")} + ${escapeHtml(change.after || "空")} + ${showSource ? `${escapeHtml(change.source || sourceLabel || "人工")}` : ""} + + `).join("") + : `${escapeHtml(item.audit_change_summary || "")}`; + box.innerHTML = ` +
+ ${escapeHtml(title)} + + ${showSource ? "" : ""} + ${rows} +
字段修改前修改后修改者
+
+ `; +} + +function reviewPatientChangesWithSource(draftPatient) { + const currentPatient = state.current?.patient || {}; + const savedLog = displayChangeLogEntries(state.current?.change_log || []); + const pendingManualChanges = patientChanges(currentPatient, draftPatient).map((change) => ({ ...change, source: "人工" })); + if (savedLog.length || pendingManualChanges.length) return savedLog.concat(pendingManualChanges); + const original = state.current?.original_patient || {}; + const source = state.current?.ai_corrected ? "AI" : "人工"; + return patientChanges(original, draftPatient).map((change) => ({ ...change, source })); +} + +function renderReviewChangeLog() { + if (!state.current) { + $("#reviewChangeLog").innerHTML = ""; + return; + } + const draft = { + ...state.current, + patient: formPatient(), + }; + const changes = reviewPatientChangesWithSource(draft.patient); + renderChangeLog("#reviewChangeLog", draft, "修改记录", "", changes); +} + +function replaceFullwidthColon(field) { + if (!field.value.includes("：")) return; + const start = field.selectionStart; + const end = field.selectionEnd; + field.value = field.value.replaceAll("：", ":"); + if (typeof start === "number" && typeof end === "number") { + field.setSelectionRange(start, end); + } +} + +function formOptions() { + return {}; +} + +async function validateForm() { + if (!state.current) return; + const data = await api(`/api/items/${state.current.key}/validate`, { + method: "POST", + body: JSON.stringify({ patient: formPatient(), options: formOptions() }), + }); + if (!data) return; + if (data.patient) fillPatientFields(data.patient); + if (state.current) state.current.validation_warnings = data.warnings || []; + renderTips(); + renderReviewChangeLog(); + return data.warnings || []; +} + +function updateLocalReviewValidation() { + if (!state.current) return; + state.current.validation_warnings = validatePatientLocal(formPatient(false)); + renderTips(); + renderReviewChangeLog(); +} + +function setImageZoom(value, anchorEvent = null) { + const frame = $("#cropFrame"); + const image = $("#cropImage"); + const previousZoom = state.imageZoom; + const nextZoom = Math.min(3, Math.max(0.4, value)); + if (!image.naturalWidth) { + state.imageZoom = nextZoom; + $("#zoomLabel").textContent = `${Math.round(state.imageZoom * 100)}%`; + return; + } + const rect = frame.getBoundingClientRect(); + const anchorX = anchorEvent ? anchorEvent.clientX - rect.left : frame.clientWidth / 2; + const anchorY = anchorEvent ? anchorEvent.clientY - rect.top : frame.clientHeight / 2; + const scrollX = frame.scrollLeft + anchorX; + const scrollY = frame.scrollTop + anchorY; + const ratio = nextZoom / previousZoom; + state.imageZoom = nextZoom; + image.style.width = `${Math.round(image.naturalWidth * state.imageZoom)}px`; + $("#zoomLabel").textContent = `${Math.round(state.imageZoom * 100)}%`; + requestAnimationFrame(() => { + frame.scrollLeft = scrollX * ratio - anchorX; + frame.scrollTop = scrollY * ratio - anchorY; + }); +} + +function resetImageView() { + const frame = $("#cropFrame"); + state.imageZoom = 1; + $("#cropImage").style.width = ""; + $("#zoomLabel").textContent = "100%"; + requestAnimationFrame(() => { + frame.scrollLeft = 0; + frame.scrollTop = 0; + }); +} + +function startImagePan(event) { + if (event.button !== 0) return; + const frame = $("#cropFrame"); + state.imagePan = { + pointerId: event.pointerId, + startX: event.clientX, + startY: event.clientY, + scrollLeft: frame.scrollLeft, + scrollTop: frame.scrollTop, + }; + frame.classList.add("is-dragging"); + frame.setPointerCapture(event.pointerId); +} + +function moveImagePan(event) { + const pan = state.imagePan; + if (!pan || pan.pointerId !== event.pointerId) return; + event.preventDefault(); + const frame = $("#cropFrame"); + frame.scrollLeft = pan.scrollLeft - (event.clientX - pan.startX); + frame.scrollTop = pan.scrollTop - (event.clientY - pan.startY); +} + +function endImagePan(event) { + const pan = state.imagePan; + if (!pan || pan.pointerId !== event.pointerId) return; + const frame = $("#cropFrame"); + state.imagePan = null; + frame.classList.remove("is-dragging"); + if (frame.hasPointerCapture(event.pointerId)) { + frame.releasePointerCapture(event.pointerId); + } +} + +function setAuditImageZoom(value, anchorEvent = null) { + const frame = $("#auditFrame"); + const image = $("#auditImage"); + const previousZoom = state.auditImageZoom; + const nextZoom = Math.min(3, Math.max(0.4, value)); + if (!image.naturalWidth) { + state.auditImageZoom = nextZoom; + $("#auditZoomLabel").textContent = `${Math.round(state.auditImageZoom * 100)}%`; + return; + } + const rect = frame.getBoundingClientRect(); + const anchorX = anchorEvent ? anchorEvent.clientX - rect.left : frame.clientWidth / 2; + const anchorY = anchorEvent ? anchorEvent.clientY - rect.top : frame.clientHeight / 2; + const scrollX = frame.scrollLeft + anchorX; + const scrollY = frame.scrollTop + anchorY; + const ratio = nextZoom / previousZoom; + state.auditImageZoom = nextZoom; + image.style.width = `${Math.round(image.naturalWidth * state.auditImageZoom)}px`; + $("#auditZoomLabel").textContent = `${Math.round(state.auditImageZoom * 100)}%`; + requestAnimationFrame(() => { + frame.scrollLeft = scrollX * ratio - anchorX; + frame.scrollTop = scrollY * ratio - anchorY; + }); +} + +function resetAuditImageView() { + const frame = $("#auditFrame"); + state.auditImageZoom = 1; + $("#auditImage").style.width = ""; + $("#auditZoomLabel").textContent = "100%"; + requestAnimationFrame(() => { + frame.scrollLeft = Math.max(0, (frame.scrollWidth - frame.clientWidth) / 2); + frame.scrollTop = Math.max(0, (frame.scrollHeight - frame.clientHeight) / 2); + }); +} + +function startAuditImagePan(event) { + if (event.button !== 0) return; + const frame = $("#auditFrame"); + state.auditImagePan = { + pointerId: event.pointerId, + startX: event.clientX, + startY: event.clientY, + scrollLeft: frame.scrollLeft, + scrollTop: frame.scrollTop, + }; + frame.classList.add("is-dragging"); + frame.setPointerCapture(event.pointerId); +} + +function moveAuditImagePan(event) { + const pan = state.auditImagePan; + if (!pan || pan.pointerId !== event.pointerId) return; + event.preventDefault(); + const frame = $("#auditFrame"); + frame.scrollLeft = pan.scrollLeft - (event.clientX - pan.startX); + frame.scrollTop = pan.scrollTop - (event.clientY - pan.startY); +} + +function endAuditImagePan(event) { + const pan = state.auditImagePan; + if (!pan || pan.pointerId !== event.pointerId) return; + const frame = $("#auditFrame"); + state.auditImagePan = null; + frame.classList.remove("is-dragging"); + if (frame.hasPointerCapture(event.pointerId)) { + frame.releasePointerCapture(event.pointerId); + } +} + +function setAuditHistoryImageZoom(value, anchorEvent = null) { + const frame = $("#auditHistoryFrame"); + const image = $("#auditHistoryImage"); + const previousZoom = state.auditHistoryImageZoom; + const nextZoom = Math.min(3, Math.max(0.4, value)); + if (!image.naturalWidth) { + state.auditHistoryImageZoom = nextZoom; + $("#auditHistoryZoomLabel").textContent = `${Math.round(state.auditHistoryImageZoom * 100)}%`; + return; + } + const rect = frame.getBoundingClientRect(); + const anchorX = anchorEvent ? anchorEvent.clientX - rect.left : frame.clientWidth / 2; + const anchorY = anchorEvent ? anchorEvent.clientY - rect.top : frame.clientHeight / 2; + const scrollX = frame.scrollLeft + anchorX; + const scrollY = frame.scrollTop + anchorY; + const ratio = nextZoom / previousZoom; + state.auditHistoryImageZoom = nextZoom; + image.style.width = `${Math.round(image.naturalWidth * state.auditHistoryImageZoom)}px`; + $("#auditHistoryZoomLabel").textContent = `${Math.round(state.auditHistoryImageZoom * 100)}%`; + requestAnimationFrame(() => { + frame.scrollLeft = scrollX * ratio - anchorX; + frame.scrollTop = scrollY * ratio - anchorY; + }); +} + +function resetAuditHistoryImageView() { + const frame = $("#auditHistoryFrame"); + state.auditHistoryImageZoom = 1; + $("#auditHistoryImage").style.width = ""; + $("#auditHistoryZoomLabel").textContent = "100%"; + requestAnimationFrame(() => { + frame.scrollLeft = Math.max(0, (frame.scrollWidth - frame.clientWidth) / 2); + frame.scrollTop = Math.max(0, (frame.scrollHeight - frame.clientHeight) / 2); + }); +} + +function startAuditHistoryImagePan(event) { + if (event.button !== 0) return; + const frame = $("#auditHistoryFrame"); + state.auditHistoryImagePan = { + pointerId: event.pointerId, + startX: event.clientX, + startY: event.clientY, + scrollLeft: frame.scrollLeft, + scrollTop: frame.scrollTop, + }; + frame.classList.add("is-dragging"); + frame.setPointerCapture(event.pointerId); +} + +function moveAuditHistoryImagePan(event) { + const pan = state.auditHistoryImagePan; + if (!pan || pan.pointerId !== event.pointerId) return; + event.preventDefault(); + const frame = $("#auditHistoryFrame"); + frame.scrollLeft = pan.scrollLeft - (event.clientX - pan.startX); + frame.scrollTop = pan.scrollTop - (event.clientY - pan.startY); +} + +function endAuditHistoryImagePan(event) { + const pan = state.auditHistoryImagePan; + if (!pan || pan.pointerId !== event.pointerId) return; + const frame = $("#auditHistoryFrame"); + state.auditHistoryImagePan = null; + frame.classList.remove("is-dragging"); + if (frame.hasPointerCapture(event.pointerId)) { + frame.releasePointerCapture(event.pointerId); + } +} + +function switchView(view) { + if (state.permissions && state.permissions[permissionKeyForView(view)] === false) return; + if (state.aiBusy && view !== state.activeView) { + window.alert(`${state.aiTaskLabel || "AI处理"}中，请保持当前页面打开，完成后再切换到其他页面。`); + setSaveState(`${state.aiTaskLabel || "AI处理"}中，请保持当前页面打开`); + return; + } + document.querySelectorAll(".page-view").forEach((panel) => { + panel.classList.toggle("hidden", panel.id !== `${view}View`); + }); + document.querySelectorAll(".nav-button").forEach((button) => { + button.classList.toggle("active", button.dataset.view === view); + }); + state.activeView = view; + if (view === "review" && !state.items.length) loadItems(true); + if (view === "auditHistory") loadAuditHistory(); + if (view === "settings") loadSettings(); +} + +async function loadSession() { + const data = await api("/api/session"); + if (!data) return; + state.permissions = data.permissions || {}; + state.permissionLabels = data.permission_labels || {}; + state.kimiEnabled = data.kimi_enabled !== false; + updateAiVisibility(); + document.querySelectorAll(".nav-button").forEach((button) => { + const view = button.dataset.view; + button.classList.toggle("hidden", state.permissions[permissionKeyForView(view)] === false); + }); + const active = document.querySelector(".nav-button.active:not(.hidden)"); + if (!active) { + const first = document.querySelector(".nav-button:not(.hidden)"); + if (first) switchView(first.dataset.view); + } +} + +async function saveCorrection(event) { + event.preventDefault(); + if (!state.current) return; + const currentKey = state.current.key; + const wasAiPending = state.current.manual_state === "AI修改-待确认"; + const currentIndex = state.items.findIndex((item) => item.key === currentKey); + const nextKey = state.items[currentIndex + 1]?.key || state.items[currentIndex - 1]?.key || ""; + const precheckWarnings = await validateForm(); + if (precheckWarnings && precheckWarnings.length) { + const message = `当前修订仍有 ${precheckWarnings.length} 个提示：\n\n${precheckWarnings.join("\n")}\n\n仍然保存吗？`; + if (!window.confirm(message)) { + setSaveState("已取消保存"); + return; + } + } + setSaveState("保存中"); + const result = await api(`/api/items/${state.current.key}/correction`, { + method: "POST", + body: JSON.stringify({ + patient: formPatient(), + options: formOptions(), + manual_note: $("#manualNote").value.trim(), + }), + }); + if (!result) return; + if (result.postgres?.error) { + window.alert(`修订已保存到本地，但同步 PostgreSQL 失败，已纳入待同步：\n${result.postgres.error}`); + } else if (result.postgres?.enabled && result.postgres.updated === 0) { + window.alert("修订已保存到本地，但 PostgreSQL 暂未找到匹配记录，已纳入待同步。"); + } + await loadSummary(); + if (wasAiPending) { + setStatusFilter("confirming", false); + } + await reloadLoadedItems(); + if (!result.warnings.length && state.items.length) { + const fallbackIndex = currentIndex >= 0 ? Math.min(currentIndex, state.items.length - 1) : 0; + const targetKey = state.items.some((item) => item.key === nextKey) ? nextKey : state.items[fallbackIndex].key; + await selectItem(targetKey); + } else { + state.current = result.item; + renderCurrent(); + highlightActive(result.item.key); + } + setSaveState(result.warnings.length ? "已保存，仍需确认" : "已保存"); +} + +async function deleteCorrection() { + if (!state.current) return; + setSaveState("删除中"); + const result = await api(`/api/items/${state.current.key}/correction`, { method: "DELETE" }); + if (result?.postgres?.error) { + window.alert(`本地修订已删除，但同步 PostgreSQL 失败：\n${result.postgres.error}`); + } + const item = await api(`/api/items/${state.current.key}`); + state.current = item; + renderCurrent(); + await loadSummary(); + await reloadLoadedItems(); + setSaveState("已删除修订"); +} + +function resetToOriginal() { + if (!state.current) return; + columns.forEach((column) => { + const field = document.querySelector(`[name="${column}"]`); + if (field) field.value = state.current.original_patient[column] ?? ""; + }); + $("#manualNote").value = ""; + renderReviewChangeLog(); + validateForm(); +} + +function fillPatientFields(patient) { + columns.forEach((column) => { + const field = document.querySelector(`[name="${column}"]`); + if (field && Object.prototype.hasOwnProperty.call(patient, column)) { + field.value = patient[column] ?? ""; + } + }); +} + +function bindEvents() { + document.querySelectorAll(".nav-button").forEach((button) => { + button.addEventListener("click", () => switchView(button.dataset.view)); + }); + $("#batchFilter").addEventListener("change", () => loadItems(true)); + document.querySelectorAll(".status-pill").forEach((button) => { + button.addEventListener("click", () => setStatusFilter(button.dataset.status)); + }); + $("#moreStatusFilter").addEventListener("change", () => { + if ($("#moreStatusFilter").value) { + setStatusFilter($("#moreStatusFilter").value); + } + }); + $("#sortFilter").addEventListener("change", () => loadItems(true)); + $("#queuePanel").addEventListener("scroll", () => { + const panel = $("#queuePanel"); + if (panel.scrollTop + panel.clientHeight >= panel.scrollHeight - 180) { + loadNextItems(); + } + }); + $("#queuePanel").addEventListener("keydown", (event) => { + if (event.key === "ArrowDown") { + event.preventDefault(); + selectAdjacentItem(1); + } else if (event.key === "ArrowUp") { + event.preventDefault(); + selectAdjacentItem(-1); + } + }); + $("#pgTestButton").addEventListener("click", () => testPostgres(false)); + $("#pgSyncButton").addEventListener("click", syncPostgres); + $("#searchInput").addEventListener("input", () => { + clearTimeout(state.debounce); + state.debounce = setTimeout(() => loadItems(true), 220); + }); + $("#editForm").addEventListener("submit", saveCorrection); + $("#deleteButton").addEventListener("click", deleteCorrection); + $("#resetButton").addEventListener("click", resetToOriginal); + $("#zoomInButton").addEventListener("click", () => setImageZoom(state.imageZoom + 0.15)); + $("#zoomOutButton").addEventListener("click", () => setImageZoom(state.imageZoom - 0.15)); + $("#zoomResetButton").addEventListener("click", resetImageView); + $("#cropFrame").addEventListener("wheel", (event) => { + event.preventDefault(); + setImageZoom(state.imageZoom + (event.deltaY < 0 ? 0.12 : -0.12), event); + }, { passive: false }); + $("#cropFrame").addEventListener("pointerdown", startImagePan); + $("#cropFrame").addEventListener("pointermove", moveImagePan); + $("#cropFrame").addEventListener("pointerup", endImagePan); + $("#cropFrame").addEventListener("pointercancel", endImagePan); + $("#cropImage").addEventListener("load", resetImageView); + $("#cropImage").addEventListener("error", () => { + $("#tipsBox").innerHTML = `
截图加载失败，请检查原图路径或点击“原图”查看。
`; + }); + $("#auditZoomInButton").addEventListener("click", () => setAuditImageZoom(state.auditImageZoom + 0.15)); + $("#auditZoomOutButton").addEventListener("click", () => setAuditImageZoom(state.auditImageZoom - 0.15)); + $("#auditZoomResetButton").addEventListener("click", resetAuditImageView); + $("#auditFrame").addEventListener("wheel", (event) => { + event.preventDefault(); + setAuditImageZoom(state.auditImageZoom + (event.deltaY < 0 ? 0.12 : -0.12), event); + }, { passive: false }); + $("#auditFrame").addEventListener("pointerdown", startAuditImagePan); + $("#auditFrame").addEventListener("pointermove", moveAuditImagePan); + $("#auditFrame").addEventListener("pointerup", endAuditImagePan); + $("#auditFrame").addEventListener("pointercancel", endAuditImagePan); + $("#auditImage").addEventListener("load", resetAuditImageView); + $("#auditImage").addEventListener("error", () => { + $("#auditState").textContent = "截图加载失败，请检查原图路径"; + }); + $("#auditSampleButton").addEventListener("click", loadAuditSample); + $("#auditKimiButton").addEventListener("click", () => runKimiAudit()); + $("#auditKimiAllButton").addEventListener("click", runKimiAuditAll); + $("#auditHistoryRefreshButton").addEventListener("click", loadAuditHistory); + $("#auditHistorySourceFilter").addEventListener("change", loadAuditHistory); + $("#auditHistoryStatusFilter").addEventListener("change", loadAuditHistory); + $("#auditHistorySortFilter").addEventListener("change", loadAuditHistory); + $("#auditHistorySearchInput").addEventListener("input", () => { + clearTimeout(state.debounce); + state.debounce = setTimeout(loadAuditHistory, 220); + }); + $("#auditHistoryZoomInButton").addEventListener("click", () => setAuditHistoryImageZoom(state.auditHistoryImageZoom + 0.15)); + $("#auditHistoryZoomOutButton").addEventListener("click", () => setAuditHistoryImageZoom(state.auditHistoryImageZoom - 0.15)); + $("#auditHistoryZoomResetButton").addEventListener("click", resetAuditHistoryImageView); + $("#auditHistoryFrame").addEventListener("wheel", (event) => { + event.preventDefault(); + setAuditHistoryImageZoom(state.auditHistoryImageZoom + (event.deltaY < 0 ? 0.12 : -0.12), event); + }, { passive: false }); + $("#auditHistoryFrame").addEventListener("pointerdown", startAuditHistoryImagePan); + $("#auditHistoryFrame").addEventListener("pointermove", moveAuditHistoryImagePan); + $("#auditHistoryFrame").addEventListener("pointerup", endAuditHistoryImagePan); + $("#auditHistoryFrame").addEventListener("pointercancel", endAuditHistoryImagePan); + $("#auditHistoryImage").addEventListener("load", resetAuditHistoryImageView); + $("#auditHistoryImage").addEventListener("error", () => { + $("#auditHistoryState").textContent = "截图加载失败，请检查原图路径"; + }); + $("#auditPassButton").addEventListener("click", () => saveAuditResult("人工核验通过")); + $("#auditFailButton").addEventListener("click", () => saveAuditResult("人工核验异常")); + $("#auditUnsureButton").addEventListener("click", () => saveAuditResult("暂不确定")); + $("#auditHistoryPassButton").addEventListener("click", () => saveAuditHistoryResult("人工核验通过")); + $("#auditHistoryFailButton").addEventListener("click", () => saveAuditHistoryResult("人工核验异常")); + $("#auditHistoryUnsureButton").addEventListener("click", () => saveAuditHistoryResult("暂不确定")); + $("#userForm").addEventListener("submit", saveUser); + $("#kimiForm").addEventListener("submit", saveKimiSettings); + $("#kimiToggleButton").addEventListener("click", toggleKimiEnabled); + $("#processingFindButton").addEventListener("click", () => findProcessingItems(false)); + $("#commitManualPassedButton").addEventListener("click", commitManualPassed); + $("#resetAuditHistoryButton").addEventListener("click", resetAuditHistory); + $("#kimiCurrentButton").addEventListener("click", kimiCorrectCurrent); + $("#kimiFiveButton").addEventListener("click", kimiCorrectFive); + $("#kimiRemainingButton").addEventListener("click", kimiCorrectRemaining); + document.querySelectorAll(".edit-form input, .edit-form textarea").forEach((field) => { + field.addEventListener("input", () => { + const column = field.getAttribute("name"); + if (datetimeColumns.has(column)) { + replaceFullwidthColon(field); + } + updateLocalReviewValidation(); + }); + field.addEventListener("keydown", (event) => { + const column = field.getAttribute("name"); + if (event.key === "Enter" && datetimeColumns.has(column)) { + event.preventDefault(); + field.value = normalizeDateTime(field.value); + validateForm(); + } + }); + field.addEventListener("blur", () => { + const column = field.getAttribute("name"); + if (datetimeColumns.has(column)) { + field.value = normalizeDateTime(field.value); + validateForm(); + } else { + updateLocalReviewValidation(); + } + }); + }); +} + +async function loadAuditSample() { + $("#auditState").textContent = "抽取中"; + const count = Math.min(20, Math.max(1, Number($("#auditCount").value || 5))); + $("#auditCount").value = count; + const source = $("#auditSource").value; + try { + const data = await api("/api/audit/sample", { + method: "POST", + body: JSON.stringify({ count, source }), + }); + if (!data) return; + state.auditItems = data.items || []; + renderAuditList(); + if (state.auditItems.length) selectAuditItem(state.auditItems[0].key); + $("#auditState").textContent = `已抽取 ${state.auditItems.length} 条，${auditProgressText()}`; + } catch (error) { + $("#auditState").textContent = `抽取失败：${error.message}`; + } +} + +function renderAuditList() { + $("#auditList").innerHTML = state.auditItems + .map((item) => auditQueueItemHtml(item, state.auditCurrent?.key, item.audit_checked_at ? [`抽查时间 ${formatTime(item.audit_checked_at)}`] : [])) + .join(""); + $("#auditList").querySelectorAll(".queue-item").forEach((button) => { + button.addEventListener("click", () => selectAuditItem(button.dataset.key)); + }); +} + +function selectAuditItem(key) { + const item = state.auditItems.find((candidate) => candidate.key === key); + if (!item) return; + if (!item.audit_original_patient) { + item.audit_original_patient = { ...(item.patient || {}) }; + } + state.auditCurrent = item; + $("#auditTitle").innerHTML = auditTitleHtml(item); + $("#auditMeta").textContent = `${item.major_department} / ${item.sub_department} / ${item.source_folder} / ${item.image_name} ${displayRowText(item)}`; + $("#auditImage").src = `/api/crop?path=${encodeURIComponent(item.image_path)}&row=${encodeURIComponent(item.image_row_no)}&context=2&v=${Date.now()}`; + $("#auditFullImageLink").href = `/api/image?path=${encodeURIComponent(item.image_path)}`; + resetAuditImageView(); + setAuditPrompt(item.audit_prompt || ""); + setAuditModelResult(item.audit_ai_feedback || item.audit_result_text || "", item.audit_ai_raw_output || item.audit_ai_feedback || item.audit_result_text || "", item.patient || null); + $("#auditMachineVerdict").value = item.audit_machine_verdict || ""; + $("#auditManualFeedback").value = item.audit_manual_feedback || ""; + renderAuditPatientInfo(item); + renderAuditVerdicts(item); + updateAuditValidation("#auditPatientInfo", "#auditTipsBox", item); + renderChangeLog("#auditChangeLog", item); + renderAuditList(); +} + +function renderAuditPatientInfo(item) { + renderPatientEditor("#auditPatientInfo", item, "#auditTipsBox", "#auditChangeLog"); +} + +function renderPatientEditor(selector, item, tipsSelector = "", changeLogSelector = "") { + const wideColumns = new Set(["诊断", "手术后天数"]); + $(selector).innerHTML = columns.map((column) => { + const value = escapeHtml(item.patient[column] || ""); + const wide = wideColumns.has(column) ? "wide" : ""; + const control = column === "诊断" + ? `` + : ``; + return ``; + }).join(""); + document.querySelectorAll(`${selector} [data-patient-field]`).forEach((field) => { + field.addEventListener("input", () => { + const column = field.dataset.patientField; + if (datetimeColumns.has(column)) replaceFullwidthColon(field); + readPatientEditor(selector, item, false); + if (tipsSelector) updateAuditValidation(selector, tipsSelector, item); + if (changeLogSelector) renderChangeLog(changeLogSelector, item); + refreshAuditIdentity(selector, item); + }); + field.addEventListener("blur", () => { + const column = field.dataset.patientField; + field.value = normalizeFieldValue(column, field.value); + readPatientEditor(selector, item); + if (tipsSelector) updateAuditValidation(selector, tipsSelector, item); + if (changeLogSelector) renderChangeLog(changeLogSelector, item); + refreshAuditIdentity(selector, item); + }); + }); +} + +function refreshAuditIdentity(selector, item) { + if (selector === "#auditPatientInfo") { + $("#auditTitle").innerHTML = auditTitleHtml(item); + renderAuditList(); + } else if (selector === "#auditHistoryPatientInfo") { + $("#auditHistoryTitle").innerHTML = auditTitleHtml(item, auditHistoryBadgeText(item)); + renderAuditHistoryList(); + } +} + +function readPatientEditor(selector, targetItem, writeBack = true) { + const patient = { ...(targetItem.patient || {}) }; + document.querySelectorAll(`${selector} [data-patient-field]`).forEach((field) => { + const column = field.dataset.patientField; + patient[column] = normalizeFieldValue(column, field.value); + if (writeBack) field.value = patient[column]; + }); + targetItem.patient = patient; + return patient; +} + +function renderAuditVerdicts(item) { + $("#auditMachineVerdictText").textContent = item.audit_machine_verdict || "未判断"; + $("#auditManualVerdictText").textContent = item.audit_result || "未选择"; +} + +async function runKimiAudit(options = {}) { + if (!state.auditCurrent) return; + if (!ensureAiReady("AI抽查")) return; + readPatientEditor("#auditPatientInfo", state.auditCurrent); + setSaveState("AI抽查中，请保持当前页面打开"); + $("#auditState").textContent = auditProgressText(); + try { + await runKimiAuditForItem(state.auditCurrent); + setAuditPrompt(state.auditCurrent.audit_prompt); + setAuditModelResult(state.auditCurrent.audit_ai_feedback, state.auditCurrent.audit_ai_raw_output, state.auditCurrent.patient || null); + $("#auditMachineVerdict").value = state.auditCurrent.audit_machine_verdict; + renderAuditVerdicts(state.auditCurrent); + renderAuditList(); + await persistAuditCurrent(state.auditCurrent.audit_result || ""); + $("#auditState").textContent = auditProgressText(); + setSaveState("AI抽查完成，AI反馈已保存"); + } catch (error) { + setSaveState(`AI抽查失败：${error.message}`); + $("#auditState").textContent = auditProgressText(); + if (options.bubble) throw error; + } finally { + finishAiTask(); + } +} + +async function runKimiAuditForItem(item) { + const result = await api("/api/audit/kimi", { + method: "POST", + body: JSON.stringify({ key: item.key, item }), + }); + if (!result) return null; + item.audit_prompt = result.prompt || ""; + item.audit_ai_feedback = normalizeModelResultText(result.result || "", item.patient || null); + item.audit_ai_raw_output = result.raw_result || result.result || ""; + item.audit_machine_verdict = result.machine_verdict || modelVerdict(item.audit_ai_feedback || item.audit_ai_raw_output, item.patient || null) || ""; + return result; +} + +async function persistAuditItem(item, status, source = $("#auditSource").value) { + const originalPatient = item.audit_original_patient || item.original_patient || item.patient || {}; + const changes = patientChanges(originalPatient, item.patient || {}); + const updates = { + audit_result: status, + audit_ai_feedback: normalizeModelResultText(item.audit_ai_feedback || item.audit_ai_raw_output || "", item.patient || null), + audit_ai_raw_output: item.audit_ai_raw_output || item.audit_ai_feedback || "", + audit_manual_feedback: item.audit_manual_feedback || "", + audit_machine_verdict: item.audit_machine_verdict || "", + audit_source: source || item.audit_source || "", + audit_change_summary: changes.map((change) => `${change.column}: ${change.before || "空"} -> ${change.after || "空"}`).join("；"), + }; + const payload = { + key: item.key, + item, + original_patient: originalPatient, + ...updates, + }; + const result = await api("/api/audit/result", { + method: "POST", + body: JSON.stringify(payload), + }); + if (!result) return null; + Object.assign(item, updates, { + audit_checked_at: new Date().toISOString().slice(0, 19), + }); + return result; +} + +async function persistAuditCurrent(status) { + if (!state.auditCurrent) return; + readPatientEditor("#auditPatientInfo", state.auditCurrent); + state.auditCurrent.audit_result = status; + state.auditCurrent.audit_ai_feedback = getAuditModelResult(); + state.auditCurrent.audit_ai_raw_output = getAuditRawOutput(); + state.auditCurrent.audit_manual_feedback = $("#auditManualFeedback").value.trim(); + state.auditCurrent.audit_machine_verdict = $("#auditMachineVerdict").value; + const result = await persistAuditItem(state.auditCurrent, status, $("#auditSource").value); + if (!result) return; + renderAuditList(); + renderAuditVerdicts(state.auditCurrent); + return result; +} + +async function saveAuditResult(status) { + if (!state.auditCurrent) return; + state.auditCurrent.audit_result = status; + renderAuditVerdicts(state.auditCurrent); + const result = await persistAuditCurrent(status); + if (!result) return; + $("#auditState").textContent = result.postgres?.updated ? `当前项已保存：${status}，${auditProgressText()}` : `当前项未匹配数据库，${auditProgressText()}`; + renderChangeLog("#auditChangeLog", state.auditCurrent); +} + +async function runKimiAuditAll() { + if (!state.auditItems.length) return; + if (!window.confirm(`将对当前抽取的 ${state.auditItems.length} 条逐条调用 AI，确定继续？`)) return; + if (!ensureAiReady("AI抽查")) return; + let success = 0; + let failed = 0; + const currentKey = state.auditCurrent?.key || ""; + try { + for (const item of state.auditItems) { + try { + await runKimiAuditForItem(item); + await persistAuditItem(item, item.audit_result || "", item.audit_source || $("#auditSource").value); + success += 1; + if (item.key === currentKey) { + setAuditPrompt(item.audit_prompt || ""); + setAuditModelResult(item.audit_ai_feedback || "", item.audit_ai_raw_output || item.audit_ai_feedback || "", item.patient || null); + $("#auditMachineVerdict").value = item.audit_machine_verdict || ""; + renderAuditVerdicts(item); + } + renderAuditList(); + const progress = `AI抽查中 ${success + failed}/${state.auditItems.length}，${auditProgressText()}`; + $("#auditState").textContent = auditProgressText(); + setSaveState(progress); + } catch (error) { + failed += 1; + item.audit_ai_feedback = `调用失败：${error.message}`; + item.audit_ai_raw_output = item.audit_ai_feedback; + item.audit_machine_verdict = "不确定"; + if (state.auditCurrent?.key === item.key) { + setAuditModelResult("", item.audit_ai_raw_output); + $("#auditMachineVerdict").value = item.audit_machine_verdict; + renderAuditVerdicts(item); + } + renderAuditList(); + } + } + } finally { + finishAiTask(); + } + const doneText = `AI抽查所有项完成：成功${success}，失败${failed}，${auditProgressText()}`; + $("#auditState").textContent = auditProgressText(); + setSaveState(doneText); +} + +async function loadAuditHistory() { + $("#auditHistoryState").textContent = "加载中"; + try { + const params = new URLSearchParams({ + page: "1", + page_size: "80", + source: $("#auditHistorySourceFilter").value, + status: $("#auditHistoryStatusFilter").value, + sort: $("#auditHistorySortFilter").value, + q: $("#auditHistorySearchInput").value, + }); + const data = await api(`/api/audit/history?${params.toString()}`); + if (!data) return; + state.auditHistory = data.items || []; + $("#auditHistoryState").textContent = `${state.auditHistory.length} / ${data.total}`; + renderAuditHistoryList(); + if (state.auditHistory.length) { + selectAuditHistoryItem(state.auditHistory[0].key); + } else { + clearAuditHistoryDetail(); + } + } catch (error) { + $("#auditHistoryState").textContent = `加载失败：${error.message}`; + } +} + +function renderAuditHistoryList() { + $("#auditHistoryList").innerHTML = state.auditHistory.length + ? state.auditHistory.map((item) => auditQueueItemHtml(item, state.auditHistoryCurrent?.key, [ + `抽查源 ${auditSourceLabel(item.audit_source)}`, + `抽查时间 ${formatTime(item.audit_checked_at)}`, + ], auditHistoryBadgeText(item))).join("") + : `
暂无抽查记录
`; + $("#auditHistoryList").querySelectorAll(".queue-item").forEach((button) => { + button.addEventListener("click", () => selectAuditHistoryItem(button.dataset.key)); + }); +} + +function clearAuditHistoryDetail() { + state.auditHistoryCurrent = null; + $("#auditHistoryTitle").textContent = "未选择抽查记录"; + $("#auditHistoryMeta").textContent = ""; + $("#auditHistoryFullImageLink").removeAttribute("href"); + $("#auditHistoryImage").removeAttribute("src"); + setAuditHistoryModelResult("", ""); + $("#auditHistoryPatientInfo").innerHTML = ""; + $("#auditHistoryTipsBox").innerHTML = ""; + $("#auditHistoryChangeLog").innerHTML = ""; + $("#auditHistoryAiVerdict").textContent = ""; + $("#auditHistoryManualVerdict").textContent = ""; + $("#auditHistorySource").textContent = ""; + $("#auditHistoryUser").textContent = ""; + $("#auditHistoryCheckedAt").textContent = ""; + resetAuditHistoryImageView(); +} + +function selectAuditHistoryItem(key) { + const item = state.auditHistory.find((candidate) => candidate.key === key); + if (!item) return; + if (!item.audit_original_patient) { + item.audit_original_patient = { ...(item.patient || {}) }; + } + state.auditHistoryCurrent = item; + $("#auditHistoryTitle").innerHTML = auditTitleHtml(item, auditHistoryBadgeText(item)); + $("#auditHistoryMeta").textContent = `${item.major_department} / ${item.sub_department} / ${item.source_folder} / ${item.image_name} ${displayRowText(item)}`; + $("#auditHistoryImage").src = `/api/crop?path=${encodeURIComponent(item.image_path)}&row=${encodeURIComponent(item.image_row_no)}&context=2&v=${Date.now()}`; + $("#auditHistoryFullImageLink").href = `/api/image?path=${encodeURIComponent(item.image_path)}`; + resetAuditHistoryImageView(); + setAuditHistoryModelResult(item.audit_ai_feedback || "", item.audit_ai_raw_output || item.audit_ai_feedback || "", item.patient || null); + renderPatientEditor("#auditHistoryPatientInfo", item, "#auditHistoryTipsBox", "#auditHistoryChangeLog"); + updateAuditValidation("#auditHistoryPatientInfo", "#auditHistoryTipsBox", item); + renderChangeLog("#auditHistoryChangeLog", item); + $("#auditHistoryAiVerdict").textContent = item.audit_machine_verdict || "未判断"; + $("#auditHistoryManualVerdict").textContent = item.audit_result || "未人工核验"; + $("#auditHistorySource").textContent = auditSourceLabel(item.audit_source); + $("#auditHistoryUser").textContent = item.audit_checked_by || ""; + $("#auditHistoryCheckedAt").textContent = formatTime(item.audit_checked_at); + renderAuditHistoryList(); +} + +async function saveAuditHistoryResult(status) { + if (!state.auditHistoryCurrent) return; + readPatientEditor("#auditHistoryPatientInfo", state.auditHistoryCurrent); + state.auditHistoryCurrent.audit_result = status; + state.auditHistoryCurrent.audit_manual_feedback = ""; + const result = await persistAuditItem( + state.auditHistoryCurrent, + status, + state.auditHistoryCurrent.audit_source || $("#auditHistorySourceFilter").value, + ); + if (!result) return; + $("#auditHistoryManualVerdict").textContent = status; + renderChangeLog("#auditHistoryChangeLog", state.auditHistoryCurrent); + renderAuditHistoryList(); + $("#auditHistoryState").textContent = result.postgres?.updated ? `已保存：${status}` : "保存成功，本次未匹配数据库"; +} + +async function loadSettings() { + const data = await api("/api/settings"); + if (!data) return; + state.permissionLabels = data.permission_labels || state.permissionLabels; + state.kimiEnabled = data.kimi_enabled !== false; + updateAiVisibility(); + $("#kimiModel").value = data.kimi_model || "kimi-k2.6"; + $("#kimiApiKey").placeholder = data.kimi_api_key_set ? "已设置，留空表示不修改" : "未设置"; + $("#kimiToggleButton").textContent = state.kimiEnabled ? "关闭API" : "开启API"; + const defaultPermissions = { overview: true, review: true, audit: true, audit_history: true, settings: false }; + $("#settingsUserCount").textContent = `${(data.users || []).length} 个用户`; + $("#newUserPermissions").innerHTML = renderPermissionControls("new-perm", defaultPermissions); + $("#userList").innerHTML = (data.users || []).map((user, index) => ` +
+
+ ${escapeHtml(user.username)} + ${user.builtin ? "环境变量用户" : "配置用户"} +
+
${escapeHtml(user.created_at || "")}
+
${renderPermissionControls(`user-perm-${index}`, user.permissions || {}, user.builtin)}
+ ${user.builtin ? "" : ``} +
+ `).join(""); + document.querySelectorAll(".user-permission-save").forEach((button) => { + button.addEventListener("click", async () => { + const wrapper = button.closest(".settings-user"); + await saveUserPermissions(wrapper.dataset.username, wrapper.dataset.prefix); + }); + }); + if (state.processingOverview) { + renderProcessingOverview(state.processingOverview); + } else { + findProcessingItems(true); + } +} + +async function saveUser(event) { + event.preventDefault(); + const username = $("#newUsername").value.trim(); + const password = $("#newPassword").value; + if (!username || !password) return; + await api("/api/settings/users", { + method: "POST", + body: JSON.stringify({ username, password, permissions: readPermissionControls("new-perm") }), + }); + $("#newUsername").value = ""; + $("#newPassword").value = ""; + await loadSettings(); +} + +function renderPermissionControls(prefix, permissions = {}, disabled = false) { + return permissionEntries().map(([key, label]) => { + const checked = permissions[key] !== false; + return ` + + `; + }).join(""); +} + +function readPermissionControls(prefix) { + const permissions = {}; + document.querySelectorAll(`.${prefix}[data-permission]`).forEach((input) => { + permissions[input.dataset.permission] = input.checked; + }); + return permissions; +} + +async function saveUserPermissions(username, prefix) { + if (!username || !prefix) return; + await api(`/api/settings/users/${encodeURIComponent(username)}/permissions`, { + method: "POST", + body: JSON.stringify({ permissions: readPermissionControls(prefix) }), + }); + $("#settingsState").textContent = "用户权限已保存"; + await loadSettings(); +} + +async function saveKimiSettings(event) { + event.preventDefault(); + await api("/api/settings/kimi", { + method: "POST", + body: JSON.stringify({ api_key: $("#kimiApiKey").value.trim(), model: $("#kimiModel").value.trim() }), + }); + $("#kimiApiKey").value = ""; + $("#settingsState").textContent = "已保存"; + await loadSettings(); +} + +async function toggleKimiEnabled() { + const enabled = !state.kimiEnabled; + await api("/api/settings/kimi/enabled", { + method: "POST", + body: JSON.stringify({ enabled }), + }); + state.kimiEnabled = enabled; + updateAiVisibility(); + $("#kimiToggleButton").textContent = enabled ? "关闭API" : "开启API"; + $("#settingsState").textContent = enabled ? "AI功能已开启" : "AI功能已关闭"; +} + +async function commitManualPassed() { + if (!window.confirm("确定提交已人工复核通过的数据？提交后这些人工通过条目会归档隐藏，不再出现在复核工作台。")) return; + $("#commitManualPassedState").textContent = "提交中"; + try { + const result = await api("/api/settings/commit-manual-passed", { + method: "POST", + body: JSON.stringify({}), + }); + if (!result) return; + const message = [ + `可提交${result.eligible}`, + `合并结果更新${result.merged_updated}`, + `合并已是最新${result.merged_already_current ?? 0}`, + `数据库成功${result.postgres_success}`, + `已同步${result.postgres_already_success ?? 0}`, + `未匹配${result.postgres_not_found}`, + `失败${result.postgres_failed}`, + `跳过${result.skipped}（AI待确认${result.skipped_ai ?? 0}，仍有问题${result.skipped_issue ?? 0}）`, + `已归档隐藏${result.archived_corrections ?? 0}`, + ].join("，"); + $("#commitManualPassedState").textContent = message; + setSaveState("已提交人工通过数据"); + await loadSummary(); + window.alert(message); + } catch (error) { + $("#commitManualPassedState").textContent = `提交失败：${error.message}`; + } +} + +async function resetAuditHistory() { + if (!window.confirm("确定清空抽查一览结果？这只会删除抽查结论、AI反馈和人工核验结果，不会删除患者基础信息。")) return; + $("#resetAuditHistoryState").textContent = "重置中"; + try { + const result = await api("/api/settings/reset-audit-history", { + method: "POST", + body: JSON.stringify({}), + }); + if (!result) return; + const message = `已重置${result.updated ?? 0}条抽查记录`; + $("#resetAuditHistoryState").textContent = message; + state.auditHistory = []; + clearAuditHistoryDetail(); + renderAuditHistoryList(); + $("#auditHistoryState").textContent = "暂无抽查记录"; + } catch (error) { + $("#resetAuditHistoryState").textContent = `重置失败：${error.message}`; + } +} + +function renderPostgresState(pg) { + const pending = `待同步${pg.pending_sync_count ?? 0}`; + const text = pg.ok + ? `数据库已连接，${pending}` + : `数据库${pg.enabled ? "未连接" : "未启用"}，${pending}${pg.error ? `：${pg.error}` : ""}`; + $("#pgState").textContent = text; + $("#pgState").classList.toggle("warn", !pg.ok && pg.enabled); +} + +async function testPostgres(quiet = false) { + if (!quiet) setSaveState("测试PG中"); + try { + const status = await api("/api/postgres/status"); + if (!status) return; + renderPostgresState(status); + if (!quiet) setSaveState(status.ok ? `待同步${status.pending_sync_count ?? 0}` : "数据库未连接"); + } catch (error) { + $("#pgState").textContent = `数据库未连接：${error.message}`; + $("#pgState").classList.add("warn"); + if (!quiet) setSaveState("数据库连接失败"); + } +} + +async function syncPostgres() { + setSaveState("提交待同步中"); + try { + const result = await api("/api/postgres/sync", { method: "POST", body: JSON.stringify({}) }); + if (!result) return; + if (result.postgres) renderPostgresState(result.postgres); + await loadSummary(); + await loadItems(false); + const message = `成功${result.success}，未匹配${result.not_found}，失败${result.failed}`; + window.alert(message); + const failedKeys = (result.items || []).filter((item) => item.status !== "success").map((item) => item.key); + if (failedKeys.length && window.confirm(`有 ${failedKeys.length} 条未成功同步。确定删除这些待同步记录？取消则继续保留。`)) { + const cleanup = await api("/api/postgres/sync/drop_failed", { + method: "POST", + body: JSON.stringify({ keys: failedKeys }), + }); + if (cleanup?.postgres) renderPostgresState(cleanup.postgres); + await loadSummary(); + setSaveState(`已删除${cleanup?.removed ?? 0}条未成功待同步`); + } else { + setSaveState(`待同步${result.postgres?.pending_sync_count ?? 0}`); + } + } catch (error) { + setSaveState(`同步失败：${error.message}`); + } +} + +async function applyKimiCorrection(key) { + let result; + try { + result = await api(`/api/items/${key}/kimi-correction`, { + method: "POST", + body: JSON.stringify({}), + }); + } catch (error) { + throw new Error(`AI修改失败：${error.message}`); + } + if (!result) return null; + const index = state.items.findIndex((item) => item.key === key); + if (index >= 0) state.items[index] = result.item; + if (state.current?.key === key) { + state.current = result.item; + renderCurrent(); + } + renderQueue(); + return result.item; +} + +async function refreshKimiReviewList(focusKey = "") { + setStatusFilter("confirming", false); + await loadSummary(); + await reloadLoadedItems(); + if (focusKey && state.items.some((item) => item.key === focusKey)) { + await selectItem(focusKey); + } else if (state.items.length) { + await selectItem(state.items[0].key); + } +} + +async function loadKimiTargetsFromCurrent(limit = null) { + if (!state.current) return []; + const start = state.items.findIndex((item) => item.key === state.current.key); + if (start < 0) return []; + const targetEnd = limit ? start + limit : Number.POSITIVE_INFINITY; + while (state.items.length < state.totalItems && state.items.length < targetEnd) { + await loadNextItems(); + } + return state.items.slice(start, limit ? start + limit : undefined); +} + +async function kimiCorrectCurrent() { + if (!state.current) return; + if (!window.confirm("确定调用 AI 修改当前项？修改后会标记为 AI修改-待确认。")) return; + if (!ensureAiReady("AI修改")) return; + const focusKey = state.current.key; + setSaveState("AI修改当前项中，请保持当前页面打开"); + try { + await applyKimiCorrection(focusKey); + await refreshKimiReviewList(focusKey); + setSaveState("已生成 AI 修改，待人工确认"); + } catch (error) { + setSaveState(error.message); + } finally { + finishAiTask(); + } +} + +async function kimiCorrectFive() { + if (!state.current) return; + if (!window.confirm("将从当前项开始调用 AI 修改下 5 项，确定继续？")) return; + if (!ensureAiReady("AI修改")) return; + setSaveState("AI修改5项中，请保持当前页面打开"); + const focusKey = state.current.key; + let success = 0; + let failed = 0; + const failureReasons = new Map(); + let targets = []; + try { + targets = await loadKimiTargetsFromCurrent(5); + for (const item of targets) { + try { + await applyKimiCorrection(item.key); + success += 1; + setSaveState(`AI修改中 ${success + failed}/${targets.length}`); + } catch (error) { + failed += 1; + failureReasons.set(error.message, (failureReasons.get(error.message) || 0) + 1); + setSaveState(`AI修改中 ${success + failed}/${targets.length}，失败原因：${error.message}`); + } + } + await refreshKimiReviewList(focusKey); + } catch (error) { + failureReasons.set(`刷新列表失败：${error.message}`, 1); + } finally { + finishAiTask(); + } + const reasonText = [...failureReasons.entries()] + .map(([reason, count]) => `${reason}${count > 1 ? ` x${count}` : ""}`) + .join("；"); + setSaveState(`AI修改完成：成功${success}，失败${failed}${reasonText ? `，失败原因：${reasonText}` : ""}`); +} + +async function kimiCorrectRemaining() { + if (!state.current) return; + const start = state.items.findIndex((item) => item.key === state.current.key); + const remainingCount = start >= 0 ? Math.max(0, state.totalItems - start) : 0; + if (!remainingCount) return; + const message = `将从当前项开始调用 AI 修改下方全部 ${remainingCount} 项，会产生约 ${remainingCount} 次 AI 调用，确定继续？`; + if (!window.confirm(message)) return; + if (!ensureAiReady("AI修改")) return; + const focusKey = state.current.key; + setSaveState(`AI修改全部剩余项中，请保持当前页面打开`); + let success = 0; + let failed = 0; + const failureReasons = new Map(); + let targets = []; + try { + targets = await loadKimiTargetsFromCurrent(null); + for (const item of targets) { + try { + await applyKimiCorrection(item.key); + success += 1; + setSaveState(`AI修改中 ${success + failed}/${targets.length}`); + } catch (error) { + failed += 1; + failureReasons.set(error.message, (failureReasons.get(error.message) || 0) + 1); + setSaveState(`AI修改中 ${success + failed}/${targets.length}，失败原因：${error.message}`); + } + } + await refreshKimiReviewList(focusKey); + } catch (error) { + failureReasons.set(`刷新列表失败：${error.message}`, 1); + } finally { + finishAiTask(); + } + const reasonText = [...failureReasons.entries()] + .map(([reason, count]) => `${reason}${count > 1 ? ` x${count}` : ""}`) + .join("；"); + setSaveState(`AI修改全部剩余项完成：成功${success}，失败${failed}${reasonText ? `，失败原因：${reasonText}` : ""}`); +} + +async function boot() { + bindEvents(); + await loadSession(); + await loadSummary(); + await findProcessingItems(true); + if (state.permissions.review !== false) await loadItems(true); + state.pgTimer = setInterval(() => testPostgres(true).catch(() => {}), 20000); +} + +boot().catch((error) => { + setSaveState("加载失败"); + $("#queueList").innerHTML = `
${escapeHtml(error.message)}
`; +}); diff --git a/患者列表处理/人工复核网页端/templates/index.html b/患者列表处理/人工复核网页端/templates/index.html new file mode 100644 index 0000000..d07003a --- /dev/null +++ b/患者列表处理/人工复核网页端/templates/index.html @@ -0,0 +1,382 @@ + + + + + + 患者目录图片集群人工复核及抽查 + + + +
+
+
HIS LIST REVIEW
+

患者目录图片集群人工复核及抽查

+
+ +
+ + + + +
+ +
+
+
+ +
+
+
+

复核工作台

+
+
+
+

抽查一览工作台

+
+
+
+ +
+ + +
+
核对与同步都以住院号作为唯一定位依据，姓名和科室只作辅助参考
+
+
+
未选择记录
+
+
+ 原图 +
+
+ 裁剪截图 +
+
+ + + + 100% +
+
+
+ +
+
+
+ + + + +
+ +
+ + + + +
+ +
+ +
+ + + +
+
+
+
+ +
+
+ + + +
+
AI智能抽查
+
+ + +
+
+ +
+
+
+
+
+
+
未选择抽查记录
+
+
+ 原图 +
+
+
+ 抽查截图 +
+
+
+ + + + 100% +
+
+ AI结论与建议 +
+
+
+ 询问提示词 + 

+          

+          

+            AI完整输出
+            

+          

+          
+        

+        
+      

+    

+
+    

+      

+        
+        
+      

+      

+        
+        

+          

+            

+              
未选择抽查记录

+              

+            

+            原图
+          

+          

+            

+              抽查截图
+            

+          

+          

+            
+            
+            
+            100%
+          

+          

+            AI结论与建议
+            

+          

+          

+            AI完整输出
+            

+          

+        

+        
+      

+    

+
+    

+      

+        

+          

+            
用户与权限

+            0 个用户
+          

+          

+            

+              
+              
+            

+            

+              
页面权限

+              

+            

+            
+          

+          

+        

+
+        
+      

+    

+
+    
+  
+
diff --git a/患者列表处理/人工复核网页端/templates/login.html b/患者列表处理/人工复核网页端/templates/login.html
new file mode 100644
index 0000000..6ce5ef1
--- /dev/null
+++ b/患者列表处理/人工复核网页端/templates/login.html
@@ -0,0 +1,29 @@
+
+
+  
+    
+    
+    患者目录图片集群人工复核及抽查
+    
+  
+  
+    

+      
+      
患者目录图片集群人工复核及抽查

+      
+    

+  
+
diff --git a/患者列表处理/工作流_Gitea版.md b/患者列表处理/工作流_Gitea版.md
new file mode 100644
index 0000000..2cefa05
--- /dev/null
+++ b/患者列表处理/工作流_Gitea版.md
@@ -0,0 +1,202 @@
+# HIS 患者列表 OCR 归档工作流
+
+这是可提交到 Gitea 的无密钥版本。不要在本文件中写入腾讯云密钥、数据库密码、Gitea 密码或患者处理结果。
+
+## 目标
+
+把 HIS 患者列表截图整理为结构化患者记录，并同步到 PostgreSQL 的 `"Patient_Lists"` 单表。
+
+最终数据库只保留正式查看和追溯需要的信息：
+
+- 数据定位：`batch_name`、`source_folder`、`image_path`、`image_name`、`image_row_no`
+- 科室信息：`major_department`、`sub_department`
+- 患者信息：姓名、性别、年龄、住院号、诊断、入院时间、最后书写时间、住院天数、出院时间、手术后天数
+- 复核信息：`review_status`、`review_notes`、`manual_corrected`
+
+OCR 请求号、OCR 缓存路径、拼接图路径、拼接组号等中间态字段不进入数据库正式表。
+
+## 目录结构
+
+- `待处理-患者目录图片集群/`：尚未处理的 HIS 图片批次。
+- `已处理-患者目录图片集群/`：已经完成处理的原始图片批次。
+- `数据处理工作区/`：程序、科室规则、数据库结构、说明文档。
+- `数据处理结果区/已处理-患者目录图片集群/`：各批次本地处理结果。
+- `数据处理结果区/信息记录/`：全局信息记录、批次汇总、待处理盘点等。
+
+仓库只提交程序和文档。图片、OCR 缓存、处理结果、人工修正数据都应被 `.gitignore` 排除。
+
+## 关键文件
+
+1. `数据处理工作区/01_科室分类规则.json`
+2. `数据处理工作区/02_患者列表OCR归档.py`
+3. `数据处理工作区/03_人工复核修正.template.json`
+4. `数据处理工作区/04_合并批次结果.py`
+5. `数据处理工作区/05_同步PostgreSQL单表.py`
+6. `数据处理工作区/06_PostgreSQL建表结构.sql`
+7. `数据处理工作区/07_处理程序说明.md`
+8. `数据处理工作区/08_PostgreSQL调整Patient_Lists列顺序.sql`
+
+本地实际使用时，可以从模板复制出 `数据处理工作区/03_人工复核修正.json`。这个文件可能包含患者信息，不提交到 Git。
+
+患者列表字段顺序固定为：姓名、性别、年龄、住院号、诊断、入院时间、最后书写时间、住院天数、出院时间、手术后天数。
+
+## 处理前准备
+
+安装依赖：
+
+```bash
+python3 -m pip install pillow
+psql --version
+```
+
+设置腾讯云 OCR 环境变量：
+
+```bash
+export TENCENTCLOUD_SECRET_ID='填入腾讯云 SecretId'
+export TENCENTCLOUD_SECRET_KEY='填入腾讯云 SecretKey'
+```
+
+OCR 接口使用腾讯云 `RecognizeTableAccurateOCR` 表格识别 V3：
+
+- 表格识别 V3 文档：https://cloud.tencent.com/document/product/866/86721
+
+## 处理一个批次
+
+示例批次名用 `批次文件夹名` 表示：
+
+```bash
+python3 数据处理工作区/02_患者列表OCR归档.py \
+  --input "待处理-患者目录图片集群/批次文件夹名" \
+  --output "数据处理结果区/已处理-患者目录图片集群/批次文件夹名-列表归档结果" \
+  --ocr-engine table-v3 \
+  --batch-size 6 \
+  --image-padding-y 24 \
+  --workers 1 \
+  --folder-workers 2 \
+  --timeout 90 \
+  --max-retries 1
+```
+
+推荐正式处理优先使用 `--batch-size 6`。程序会根据行数校验和接口错误自动降到 4/3/2/单张；降档会增加调用次数，但能降低漏行风险。
+
+处理完成后检查：
+
+- `患者列表_结构化.json`
+- `患者列表_记录.csv`
+- `复核报告.json`
+- `重复住院号报告.json`
+- `信息记录/汇总.json`
+
+## 复核规则
+
+推荐使用 Docker 化网页端进行复核和抽查：
+
+```bash
+cd 人工复核网页端
+docker compose up -d --build
+```
+
+网页端包含概览、复核、抽查、设置四个区域；抽查功能使用 Kimi 多模态模型，配置项通过 `.env` 或网页端设置页提供，不在 Gitea 文档中保存密钥。
+
+`review_status` 常见值：
+
+- `自动复核通过`：OCR 结果经过程序规则校验后没有发现明显异常，`manual_corrected=false`。
+- `人工复核通过`：该行命中了 `03_人工复核修正.json` 中的人工修正项，修正后校验通过，`manual_corrected=true`。
+- `需人工复核`：程序无法可靠判断，需要人工查看图片和记录，原因写在 `review_notes`。
+
+`出院时间` 允许为空；若填写出院时间，则需要满足标准时间格式，且不应早于入院时间。PostgreSQL 约束会阻止“入院时间晚于出院时间”却仍被标记为通过的记录。
+
+人工修正流程：
+
+1. 打开该批次的 `复核报告.json`。
+2. 根据 `图片路径` 和 `图片内行号` 回看原图。
+3. 把确认后的患者字段写入 `数据处理工作区/03_人工复核修正.json`。
+4. 使用缓存重建批次：
+
+```bash
+python3 数据处理工作区/02_患者列表OCR归档.py \
+  --input "待处理-患者目录图片集群/批次文件夹名" \
+  --output "数据处理结果区/已处理-患者目录图片集群/批次文件夹名-列表归档结果" \
+  --ocr-engine table-v3 \
+  --batch-size 6 \
+  --image-padding-y 24 \
+  --workers 1 \
+  --folder-workers 2 \
+  --rebuild-from-cache
+```
+
+## 合并所有批次
+
+```bash
+python3 数据处理工作区/04_合并批次结果.py
+```
+
+输出包括：
+
+- `数据处理结果区/合并_患者列表_结构化.json`
+- `数据处理结果区/合并_患者列表_记录.csv`
+- `数据处理结果区/信息记录/全局汇总.json`
+- `数据处理结果区/信息记录/批次汇总.csv`
+- `数据处理结果区/信息记录/重复住院号报告.json`
+
+## 同步 PostgreSQL
+
+设置数据库环境变量：
+
+```bash
+export HIS_DB_HOST='数据库主机'
+export HIS_DB_PORT='5432'
+export HIS_DB_NAME='DB_NAME'
+export HIS_DB_USER='DB_USER'
+export HIS_DB_PASSWORD='填入数据库密码'
+```
+
+同步：
+
+```bash
+python3 数据处理工作区/05_同步PostgreSQL单表.py
+```
+
+核对：
+
+```sql
+SELECT count(*) FROM "Patient_Lists";
+SELECT review_status, manual_corrected, count(*) FROM "Patient_Lists" GROUP BY review_status, manual_corrected;
+SELECT inpatient_no FROM "Patient_Lists" GROUP BY inpatient_no HAVING count(*) > 1;
+```
+
+`"Patient_Lists"` 是带大小写的 PostgreSQL 表名，查询时必须加双引号。
+
+## 处理完成后
+
+确认批次结果和数据库无误后，将原始图片批次移动到：
+
+```bash
+mv "待处理-患者目录图片集群/批次文件夹名" "已处理-患者目录图片集群/"
+```
+
+## Gitea 提交
+
+仓库地址：
+
+```text
+https://gitea.huijutec.cn/admin/HIS_List.git
+```
+
+提交前检查不要包含数据和密钥：
+
+```bash
+git status --short --ignored
+git diff --cached --name-only
+grep -R "Secret\\|PASSWORD\\|密码" -n README.md 数据处理工作区 --exclude-dir='__pycache__' || true
+```
+
+提交：
+
+```bash
+git add .gitignore .env.example README.md 工作流_Gitea版.md 人工复核网页端 数据处理工作区/01_科室分类规则.json 数据处理工作区/02_患者列表OCR归档.py 数据处理工作区/03_人工复核修正.template.json 数据处理工作区/04_合并批次结果.py 数据处理工作区/05_同步PostgreSQL单表.py 数据处理工作区/06_PostgreSQL建表结构.sql 数据处理工作区/07_处理程序说明.md 数据处理工作区/08_PostgreSQL调整Patient_Lists列顺序.sql
+git commit -m "Update HIS patient list workflow"
+git push origin main
+```
+
+如果本地 `origin` 没有保存认证信息，需要使用一次性认证或交互式登录，但不要把密码写入文档或 Git 配置。
diff --git a/患者列表处理/数据处理工作区/01_科室分类规则.json b/患者列表处理/数据处理工作区/01_科室分类规则.json
new file mode 100644
index 0000000..da4c420
--- /dev/null
+++ b/患者列表处理/数据处理工作区/01_科室分类规则.json
@@ -0,0 +1,258 @@
+{
+  "说明": "每个具体子科室只归属一个大科室；aliases 用于把图片文件夹名归一到标准子科室名。",
+  "大科室列表": [
+    {
+      "大科室": "肝胆外科及肝移植相关",
+      "子科室": [
+        "肝胆外科1",
+        "肝胆外科2",
+        "肝胆外科3",
+        "肝胆外科4",
+        "肝胆外科5D",
+        "肝胆外科手术室",
+        "肝胆特种病区L",
+        "肝移植内",
+        "肝移植内N"
+      ]
+    },
+    {
+      "大科室": "普通外科及腹部外科",
+      "子科室": [
+        "普外科1",
+        "普外科2",
+        "普外科3",
+        "普外科4D",
+        "腹部外科L",
+        "普通腺体外科",
+        "普通胃肠外科"
+      ]
+    },
+    {
+      "大科室": "急诊医学科",
+      "子科室": [
+        "急诊中心",
+        "急诊中心L",
+        "急诊重症医学科L"
+      ]
+    },
+    {
+      "大科室": "重症医学科",
+      "子科室": [
+        "※重症医学1病区",
+        "重症医学2N病区",
+        "重症医学3病区",
+        "重症医学4D病区",
+        "外科ICU",
+        "肝胆ICU",
+        "肿瘤外科重症",
+        "感染重症"
+      ]
+    },
+    {
+      "大科室": "骨科",
+      "子科室": [
+        "骨科",
+        "骨科1",
+        "骨科2"
+      ]
+    },
+    {
+      "大科室": "泌尿外科",
+      "子科室": [
+        "泌尿外1",
+        "泌尿外2"
+      ]
+    },
+    {
+      "大科室": "呼吸内科",
+      "子科室": [
+        "呼吸内科3病区"
+      ]
+    },
+    {
+      "大科室": "耳鼻喉头颈外科",
+      "子科室": [
+        "耳鼻喉D",
+        "耳鼻喉头颈外科",
+        "耳鼻喉头颈外科L",
+        "耳鼻喉2病区"
+      ]
+    },
+    {
+      "大科室": "日间诊疗中心",
+      "子科室": [
+        "日间手术中心",
+        "日间中心病房L"
+      ]
+    },
+    {
+      "大科室": "乳腺外科",
+      "子科室": [
+        "乳腺外科1N",
+        "乳腺外科2N",
+        "乳腺外科2D"
+      ]
+    },
+    {
+      "大科室": "胸外科",
+      "子科室": [
+        "胸外1",
+        "胸外2",
+        "胸外3",
+        "胸外4D"
+      ]
+    },
+    {
+      "大科室": "肝病内科",
+      "子科室": [
+        "肝病内科"
+      ]
+    },
+    {
+      "大科室": "感染科",
+      "子科室": [
+        "感染1",
+        "感染2",
+        "感染3"
+      ]
+    },
+    {
+      "大科室": "肿瘤放疗科",
+      "子科室": [
+        "肿瘤放疗1",
+        "肿瘤放疗2",
+        "肿瘤放疗3",
+        "肿瘤放疗病区L",
+        "肿瘤放疗中心L【无人】",
+        "肿瘤放疗日间【无人】"
+      ]
+    },
+    {
+      "大科室": "特需/涉外病房",
+      "子科室": [
+        "健苑五涉外病房"
+      ]
+    },
+    {
+      "大科室": "老年外科",
+      "子科室": [
+        "老年外科"
+      ]
+    }
+  ],
+  "aliases": {
+    "呼吸内科3": "呼吸内科3病区",
+    "外科ICU": "外科ICU",
+    "急诊中心病房": "急诊中心",
+    "急诊中心": "急诊中心",
+    "急诊中心病房L": "急诊中心L",
+    "急诊中心L": "急诊中心L",
+    "急诊重症医学L": "急诊重症医学科L",
+    "急诊重症医学": "急诊重症医学科L",
+    "急诊重症L": "急诊重症医学科L",
+    "急诊重症医学科L": "急诊重症医学科L",
+    "感染科1": "感染1",
+    "感染科2": "感染2",
+    "感染科3": "感染3",
+    "感染1": "感染1",
+    "感染2": "感染2",
+    "感染3": "感染3",
+    "感染科重症": "感染重症",
+    "感染重症": "感染重症",
+    "肝病内科": "肝病内科",
+    "肝病内科N": "肝病内科",
+    "肝胆外科1": "肝胆外科1",
+    "肝胆外科2": "肝胆外科2",
+    "肝胆外科3": "肝胆外科3",
+    "肝胆外科4": "肝胆外科4",
+    "肝胆1": "肝胆外科1",
+    "肝胆2": "肝胆外科2",
+    "肝胆3": "肝胆外科3",
+    "肝胆4": "肝胆外科4",
+    "肝胆5": "肝胆外科5D",
+    "肝胆_5": "肝胆外科5D",
+    "肝胆5D": "肝胆外科5D",
+    "肝胆外科5病房D": "肝胆外科5D",
+    "肝胆特种L": "肝胆特种病区L",
+    "肝胆特种病区L": "肝胆特种病区L",
+    "肝胆移植内": "肝移植内",
+    "肝移植内": "肝移植内",
+    "肝移植内科": "肝移植内",
+    "肝移植内N": "肝移植内N",
+    "肝移植内科N": "肝移植内N",
+    "肿瘤放疗1": "肿瘤放疗1",
+    "肿瘤放疗2": "肿瘤放疗2",
+    "肿瘤放疗3": "肿瘤放疗3",
+    "肿放1": "肿瘤放疗1",
+    "肿放2": "肿瘤放疗2",
+    "肿放3": "肿瘤放疗3",
+    "肿瘤放疗L": "肿瘤放疗病区L",
+    "肿放L": "肿瘤放疗病区L",
+    "重症医学1": "※重症医学1病区",
+    "重症医学科1": "※重症医学1病区",
+    "重症1": "※重症医学1病区",
+    "重症医学2": "重症医学2N病区",
+    "重症医学科2": "重症医学2N病区",
+    "重症2": "重症医学2N病区",
+    "重症2N": "重症医学2N病区",
+    "重症医学3": "重症医学3病区",
+    "重症医学科3": "重症医学3病区",
+    "重症3": "重症医学3病区",
+    "重症医学4": "重症医学4D病区",
+    "重症医学科4": "重症医学4D病区",
+    "重症4": "重症医学4D病区",
+    "重症4D": "重症医学4D病区",
+    "胸外1": "胸外1",
+    "胸外2": "胸外2",
+    "胸外3": "胸外3",
+    "胸外4D": "胸外4D",
+    "普外1": "普外科1",
+    "普外2": "普外科2",
+    "普外3": "普外科3",
+    "普外4D": "普外科4D",
+    "普通外科1": "普外科1",
+    "普通外科2": "普外科2",
+    "普通外科3": "普外科3",
+    "普通外科4": "普外科4D",
+    "普通外科4D": "普外科4D",
+    "普通腺体外科": "普通腺体外科",
+    "普通胃肠": "普通胃肠外科",
+    "普通胃肠外科": "普通胃肠外科",
+    "腹部外科L": "腹部外科L",
+    "肝胆ICU": "肝胆ICU",
+    "日间手术": "日间手术中心",
+    "日间手术中心": "日间手术中心",
+    "日间手术L": "日间中心病房L",
+    "日间中心L": "日间中心病房L",
+    "日间中心病房L": "日间中心病房L",
+    "耳鼻喉": "耳鼻喉头颈外科",
+    "耳鼻喉D": "耳鼻喉D",
+    "耳鼻喉L": "耳鼻喉头颈外科L",
+    "耳鼻喉头颈外科": "耳鼻喉头颈外科",
+    "耳鼻喉头颈外科L": "耳鼻喉头颈外科L",
+    "耳鼻咽喉": "耳鼻喉头颈外科",
+    "耳鼻咽喉头颈外科": "耳鼻喉头颈外科",
+    "耳鼻喉2": "耳鼻喉2病区",
+    "耳鼻喉2D": "耳鼻喉2病区",
+    "耳鼻喉头颈2": "耳鼻喉2病区",
+    "乳腺外科": "乳腺外科1N",
+    "乳腺外科1": "乳腺外科1N",
+    "乳腺外科1N": "乳腺外科1N",
+    "乳腺外科2": "乳腺外科2N",
+    "乳腺外科2N": "乳腺外科2N",
+    "乳腺外科2D": "乳腺外科2D",
+    "乳腺1N": "乳腺外科1N",
+    "乳腺2N": "乳腺外科2N",
+    "乳腺2D": "乳腺外科2D",
+    "骨科": "骨科",
+    "骨科1": "骨科1",
+    "骨科2": "骨科2",
+    "骨科_1": "骨科1",
+    "骨科_2": "骨科2",
+    "泌尿1": "泌尿外1",
+    "泌尿2": "泌尿外2",
+    "健苑五涉外": "健苑五涉外病房",
+    "健苑五涉外病房": "健苑五涉外病房",
+    "老年外科": "老年外科"
+  }
+}
diff --git a/患者列表处理/数据处理工作区/02_患者列表OCR归档.py b/患者列表处理/数据处理工作区/02_患者列表OCR归档.py
new file mode 100644
index 0000000..c91fe50
--- /dev/null
+++ b/患者列表处理/数据处理工作区/02_患者列表OCR归档.py
@@ -0,0 +1,1346 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+"""Archive HIS patient-list screenshots into structured JSON.
+
+Tencent OCR credentials are read from environment variables:
+TENCENTCLOUD_SECRET_ID and TENCENTCLOUD_SECRET_KEY.
+"""
+
+from __future__ import annotations
+
+import argparse
+import base64
+import concurrent.futures
+import csv
+import datetime as dt
+import hashlib
+import hmac
+import json
+import os
+import re
+import signal
+import subprocess
+import threading
+import time
+import unicodedata
+import urllib.error
+import urllib.request
+from contextlib import contextmanager
+from dataclasses import dataclass
+from pathlib import Path
+from typing import Any
+
+from PIL import Image
+
+
+COLUMNS = [
+    "姓名",
+    "性别",
+    "年龄",
+    "住院号",
+    "诊断",
+    "入院时间",
+    "最后书写时间",
+    "住院天数",
+    "出院时间",
+    "手术后天数",
+]
+
+IMAGE_EXTENSIONS = {".png", ".jpg", ".jpeg", ".bmp"}
+OCR_ENGINE_LABELS = {
+    "table-v3": "腾讯云 RecognizeTableAccurateOCR 表格识别 V3",
+}
+GENERAL_COLUMN_ANCHORS = [0, 96, 160, 230, 450, 820, 1075, 1305, 1395, 1590]
+GENERAL_REFERENCE_WIDTH = 1842
+GENERAL_ROW_Y_THRESHOLD = 18
+
+
+class NonRetryableOcrError(RuntimeError):
+    pass
+
+
+@dataclass(frozen=True)
+class Department:
+    major: str
+    sub: str
+    raw_name: str
+
+
+def natural_key(path: Path) -> tuple[Any, ...]:
+    parts = re.split(r"(\d+)", path.stem)
+    key: list[Any] = []
+    for part in parts:
+        if part.isdigit():
+            key.append(int(part))
+        else:
+            key.append(part)
+    return tuple(key)
+
+
+def normalize_text(value: Any) -> str:
+    if value is None:
+        return ""
+    text = unicodedata.normalize("NFKC", str(value))
+    text = text.replace("\u3000", " ")
+    return re.sub(r"\s+", " ", text).strip()
+
+
+def normalize_folder_name(folder_name: str) -> str:
+    name = re.sub(r"^\d{4}[_-]\d{1,2}[_-]\d{1,2}~\d{4}[_-]\d{1,2}[_-]\d{1,2}_", "", folder_name)
+    name = re.sub(r"^\d{4}[_-]\d{1,2}[_-]\d{1,2}_", "", name)
+    name = re.sub(r"^\d{4}年\d{1,2}月\d{1,2}日[_-]?", "", name)
+    name = re.sub(r"病房|病区", "", name)
+    return normalize_text(name)
+
+
+def load_departments(path: Path) -> tuple[dict[str, str], dict[str, str]]:
+    data = json.loads(path.read_text(encoding="utf-8"))
+    sub_to_major: dict[str, str] = {}
+    for group in data["大科室列表"]:
+        major = group["大科室"]
+        for sub in group["子科室"]:
+            if sub in sub_to_major:
+                raise ValueError(f"重复子科室: {sub}")
+            sub_to_major[sub] = major
+
+    aliases: dict[str, str] = {}
+    for alias, sub in data.get("aliases", {}).items():
+        if sub not in sub_to_major:
+            raise ValueError(f"别名 {alias} 指向不存在的子科室 {sub}")
+        aliases[normalize_text(alias)] = sub
+    for sub in sub_to_major:
+        aliases[normalize_text(sub)] = sub
+    return sub_to_major, aliases
+
+
+def classify_department(folder_name: str, sub_to_major: dict[str, str], aliases: dict[str, str]) -> Department:
+    raw = normalize_folder_name(folder_name)
+    if raw in aliases:
+        sub = aliases[raw]
+        return Department(sub_to_major[sub], sub, raw)
+
+    compact = raw.replace("科", "").replace("外", "外")
+    for alias, sub in sorted(aliases.items(), key=lambda item: len(item[0]), reverse=True):
+        alias_compact = alias.replace("科", "")
+        if alias and (alias in raw or alias_compact in compact):
+            return Department(sub_to_major[sub], sub, raw)
+    return Department("未分类", "未分类", raw)
+
+
+def batched(items: list[Path], size: int) -> list[list[Path]]:
+    return [items[i : i + size] for i in range(0, len(items), size)]
+
+
+def merge_images(image_paths: list[Path], output_path: Path, padding_y: int = 0) -> dict[str, Any]:
+    padding_y = max(0, int(padding_y))
+    opened = [Image.open(path).convert("RGB") for path in image_paths]
+    try:
+        width = max(image.width for image in opened)
+        height = sum(image.height + padding_y * 2 for image in opened)
+        merged = Image.new("RGB", (width, height), "white")
+        y = 0
+        source_images = []
+        for path, image in zip(image_paths, opened):
+            image_y = y + padding_y
+            merged.paste(image, (0, image_y))
+            source_images.append(
+                {
+                    "path": str(path),
+                    "name": path.name,
+                    "width": image.width,
+                    "height": image.height,
+                    "y_offset": image_y,
+                    "block_y_offset": y,
+                    "block_height": image.height + padding_y * 2,
+                    "padding_top": padding_y,
+                    "padding_bottom": padding_y,
+                }
+            )
+            y += image.height + padding_y * 2
+        output_path.parent.mkdir(parents=True, exist_ok=True)
+        merged.save(output_path)
+        return {
+            "path": str(output_path),
+            "width": width,
+            "height": height,
+            "padding_y": padding_y,
+            "source_images": source_images,
+        }
+    finally:
+        for image in opened:
+            image.close()
+
+
+def infer_rows_for_image(image_path: Path, override_rows: int = 0) -> int:
+    if override_rows > 0:
+        return override_rows
+    with Image.open(image_path) as image:
+        # HIS list screenshots in this batch use about 39 px per row:
+        # 700 px -> 18 rows, 784 px -> 20 rows.
+        return max(1, round(image.height / 39.0))
+
+
+def locate_source_row(table_row_index: int, row_counts: list[int]) -> tuple[int, int]:
+    offset = 0
+    for image_index, row_count in enumerate(row_counts):
+        if table_row_index < offset + row_count:
+            return image_index, table_row_index - offset
+        offset += row_count
+    return len(row_counts) - 1, max(0, table_row_index - sum(row_counts[:-1]))
+
+
+def get_credentials() -> tuple[str, str]:
+    secret_id = os.getenv("TENCENTCLOUD_SECRET_ID") or os.getenv("TENCENT_SECRET_ID")
+    secret_key = os.getenv("TENCENTCLOUD_SECRET_KEY") or os.getenv("TENCENT_SECRET_KEY")
+    if not secret_id or not secret_key:
+        raise RuntimeError("请先设置 TENCENTCLOUD_SECRET_ID 和 TENCENTCLOUD_SECRET_KEY 环境变量")
+    return secret_id, secret_key
+
+
+def tc3_request(
+    action: str,
+    payload: dict[str, Any],
+    secret_id: str,
+    secret_key: str,
+    region: str,
+    timeout: int,
+) -> dict[str, Any]:
+    service = "ocr"
+    host = "ocr.tencentcloudapi.com"
+    endpoint = f"https://{host}"
+    version = "2018-11-19"
+    body = json.dumps(payload, ensure_ascii=False, separators=(",", ":"))
+    algorithm = "TC3-HMAC-SHA256"
+    timestamp = int(dt.datetime.now(dt.timezone.utc).timestamp())
+    date = dt.datetime.fromtimestamp(timestamp, dt.timezone.utc).strftime("%Y-%m-%d")
+    content_type = "application/json; charset=utf-8"
+
+    canonical_headers = f"content-type:{content_type}\nhost:{host}\n"
+    signed_headers = "content-type;host"
+    hashed_payload = hashlib.sha256(body.encode("utf-8")).hexdigest()
+    canonical_request = "\n".join(["POST", "/", "", canonical_headers, signed_headers, hashed_payload])
+    credential_scope = f"{date}/{service}/tc3_request"
+    string_to_sign = "\n".join(
+        [
+            algorithm,
+            str(timestamp),
+            credential_scope,
+            hashlib.sha256(canonical_request.encode("utf-8")).hexdigest(),
+        ]
+    )
+
+    def sign(key: bytes, message: str) -> bytes:
+        return hmac.new(key, message.encode("utf-8"), hashlib.sha256).digest()
+
+    secret_date = sign(("TC3" + secret_key).encode("utf-8"), date)
+    secret_service = sign(secret_date, service)
+    secret_signing = sign(secret_service, "tc3_request")
+    signature = hmac.new(secret_signing, string_to_sign.encode("utf-8"), hashlib.sha256).hexdigest()
+    authorization = (
+        f"{algorithm} Credential={secret_id}/{credential_scope}, "
+        f"SignedHeaders={signed_headers}, Signature={signature}"
+    )
+
+    headers = {
+        "Authorization": authorization,
+        "Content-Type": content_type,
+        "Host": host,
+        "X-TC-Action": action,
+        "X-TC-Timestamp": str(timestamp),
+        "X-TC-Version": version,
+        "X-TC-Region": region,
+    }
+    command = [
+        "curl",
+        "-sS",
+        "--connect-timeout",
+        str(min(10, max(1, timeout))),
+        "--max-time",
+        str(max(1, timeout)),
+        "-X",
+        "POST",
+        endpoint,
+    ]
+    for key, value in headers.items():
+        command.extend(["-H", f"{key}: {value}"])
+    command.extend(["--data-binary", "@-"])
+    completed = subprocess.run(
+        command,
+        input=body.encode("utf-8"),
+        capture_output=True,
+        timeout=max(1, timeout) + 5,
+        check=False,
+    )
+    if completed.returncode != 0:
+        error_text = completed.stderr.decode("utf-8", errors="replace").strip()
+        raise urllib.error.URLError(error_text or f"curl return code {completed.returncode}")
+    return json.loads(completed.stdout.decode("utf-8"))
+
+
+@contextmanager
+def wall_clock_timeout(seconds: int):
+    if seconds <= 0 or threading.current_thread() is not threading.main_thread():
+        yield
+        return
+
+    def handle_timeout(_signum: int, _frame: Any) -> None:
+        raise TimeoutError(f"OCR请求超过 {seconds} 秒")
+
+    previous_handler = signal.getsignal(signal.SIGALRM)
+    previous_timer = signal.setitimer(signal.ITIMER_REAL, 0)
+    signal.signal(signal.SIGALRM, handle_timeout)
+    signal.setitimer(signal.ITIMER_REAL, seconds)
+    try:
+        yield
+    finally:
+        signal.setitimer(signal.ITIMER_REAL, 0)
+        signal.signal(signal.SIGALRM, previous_handler)
+        if previous_timer[0] > 0:
+            signal.setitimer(signal.ITIMER_REAL, previous_timer[0], previous_timer[1])
+
+
+def call_tencent_ocr(
+    action: str,
+    image_path: Path,
+    cache_path: Path,
+    secret_id: str,
+    secret_key: str,
+    region: str,
+    timeout: int,
+    force: bool,
+    max_retries: int,
+) -> dict[str, Any]:
+    if cache_path.exists() and not force:
+        return json.loads(cache_path.read_text(encoding="utf-8"))
+    if not secret_id or not secret_key:
+        raise RuntimeError(f"OCR缓存不存在，且当前为仅重建模式: {cache_path}")
+
+    image_base64 = base64.b64encode(image_path.read_bytes()).decode("ascii")
+    payload = {"ImageBase64": image_base64, "UseNewModel": True}
+    if action == "GeneralAccurateOCR":
+        payload = {"ImageBase64": image_base64}
+    last_error: str | None = None
+    for attempt in range(max_retries + 1):
+        try:
+            with wall_clock_timeout(timeout):
+                data = tc3_request(action, payload, secret_id, secret_key, region, timeout)
+            response = data.get("Response", {})
+            if "Error" in response:
+                error_text = json.dumps(response["Error"], ensure_ascii=False)
+                if response["Error"].get("Code") == "FailedOperation.OcrFailed":
+                    raise NonRetryableOcrError(error_text)
+                raise RuntimeError(error_text)
+            response.pop("Data", None)
+            cache_path.parent.mkdir(parents=True, exist_ok=True)
+            cache_path.write_text(json.dumps(response, ensure_ascii=False, indent=2), encoding="utf-8")
+            return response
+        except NonRetryableOcrError:
+            raise
+        except (urllib.error.URLError, TimeoutError, OSError, RuntimeError) as exc:
+            last_error = str(exc)
+            if attempt >= max_retries:
+                break
+            time.sleep(2 ** attempt)
+    raise RuntimeError(f"OCR 调用失败: {image_path} {last_error}")
+
+
+def call_table_v3(
+    image_path: Path,
+    cache_path: Path,
+    secret_id: str,
+    secret_key: str,
+    region: str,
+    timeout: int,
+    force: bool,
+    max_retries: int,
+) -> dict[str, Any]:
+    return call_tencent_ocr(
+        "RecognizeTableAccurateOCR",
+        image_path,
+        cache_path,
+        secret_id,
+        secret_key,
+        region,
+        timeout,
+        force,
+        max_retries,
+    )
+
+
+def call_general_accurate(
+    image_path: Path,
+    cache_path: Path,
+    secret_id: str,
+    secret_key: str,
+    region: str,
+    timeout: int,
+    force: bool,
+    max_retries: int,
+) -> dict[str, Any]:
+    return call_tencent_ocr(
+        "GeneralAccurateOCR",
+        image_path,
+        cache_path,
+        secret_id,
+        secret_key,
+        region,
+        timeout,
+        force,
+        max_retries,
+    )
+
+
+def cells_to_rows(table_response: dict[str, Any]) -> list[list[str]]:
+    tables = table_response.get("TableDetections") or []
+    cells: list[dict[str, Any]] = []
+    for table in tables:
+        cells.extend(table.get("Cells") or [])
+    if not cells:
+        return []
+
+    max_row = max(int(cell.get("RowTl", 0)) for cell in cells)
+    max_col = max(int(cell.get("ColTl", 0)) for cell in cells)
+    rows = [["" for _ in range(max(max_col + 1, len(COLUMNS)))] for _ in range(max_row + 1)]
+    for cell in cells:
+        row = int(cell.get("RowTl", 0))
+        col = int(cell.get("ColTl", 0))
+        text = normalize_text(cell.get("Text", ""))
+        if col >= len(rows[row]):
+            rows[row].extend([""] * (col - len(rows[row]) + 1))
+        if rows[row][col]:
+            rows[row][col] = normalize_text(rows[row][col] + " " + text)
+        else:
+            rows[row][col] = text
+    return [row[: len(COLUMNS)] + [""] * max(0, len(COLUMNS) - len(row)) for row in rows]
+
+
+def general_column_index(x: float, image_width: int) -> int:
+    scale = image_width / GENERAL_REFERENCE_WIDTH if image_width else 1.0
+    anchors = [value * scale for value in GENERAL_COLUMN_ANCHORS]
+    boundaries = [(anchors[index] + anchors[index + 1]) / 2 for index in range(len(anchors) - 1)]
+    for index, boundary in enumerate(boundaries):
+        if x < boundary:
+            return index
+    return len(anchors) - 1
+
+
+def general_detections_to_rows(general_response: dict[str, Any], image_width: int) -> list[list[str]]:
+    detections: list[dict[str, Any]] = []
+    for item in general_response.get("TextDetections") or []:
+        text = normalize_text(item.get("DetectedText", ""))
+        if not text:
+            continue
+        polygon = item.get("ItemPolygon") or {}
+        x = float(polygon.get("X", 0) or 0)
+        y = float(polygon.get("Y", 0) or 0)
+        height = float(polygon.get("Height", 0) or 0)
+        detections.append({"x": x, "y_center": y + height / 2, "text": text})
+
+    detections.sort(key=lambda item: (item["y_center"], item["x"]))
+    grouped_rows: list[dict[str, Any]] = []
+    for detection in detections:
+        if not grouped_rows or abs(detection["y_center"] - grouped_rows[-1]["y_center"]) > GENERAL_ROW_Y_THRESHOLD:
+            grouped_rows.append({"y_center": detection["y_center"], "items": []})
+        else:
+            count = len(grouped_rows[-1]["items"])
+            grouped_rows[-1]["y_center"] = (grouped_rows[-1]["y_center"] * count + detection["y_center"]) / (count + 1)
+        grouped_rows[-1]["items"].append(detection)
+
+    rows: list[list[str]] = []
+    for grouped_row in grouped_rows:
+        columns: list[list[str]] = [[] for _ in COLUMNS]
+        for detection in sorted(grouped_row["items"], key=lambda item: item["x"]):
+            columns[general_column_index(detection["x"], image_width)].append(detection["text"])
+        rows.append([normalize_text(" ".join(values)) for values in columns])
+    return rows
+
+
+def ocr_response_to_rows(response: dict[str, Any], engine: str, image_width: int) -> list[list[str]]:
+    if engine == "general-accurate":
+        return general_detections_to_rows(response, image_width)
+    return cells_to_rows(response)
+
+
+def normalize_date(text: str) -> str:
+    text = normalize_text(text).replace("/", "-").replace(".", "-")
+    text = re.sub(r"(\d{4})-(\d{1,2})-(\d{1,2})\s+(\d{1,2}):(\d{1,2}):(\d{1,2})", date_repl, text)
+    return text
+
+
+def date_repl(match: re.Match[str]) -> str:
+    year, month, day, hour, minute, second = match.groups()
+    return f"{int(year):04d}-{int(month):02d}-{int(day):02d} {int(hour):02d}:{int(minute):02d}:{int(second):02d}"
+
+
+def clean_patient_row(row: list[str]) -> dict[str, Any]:
+    values = {column: normalize_text(row[index]) for index, column in enumerate(COLUMNS)}
+    if values["姓名"] and not values["性别"]:
+        name_gender = re.fullmatch(r"(.+?)(男|女)", values["姓名"])
+        if name_gender:
+            values["姓名"] = normalize_text(name_gender.group(1))
+            values["性别"] = name_gender.group(2)
+    if values["性别"] not in {"男", "女"}:
+        sex_age = re.fullmatch(r"(男|女)\s*(\d{1,3}岁)", values["性别"])
+        if sex_age:
+            values["性别"] = sex_age.group(1)
+            if not values["年龄"]:
+                values["年龄"] = sex_age.group(2)
+    values["住院号"] = re.sub(r"\s+", "", values["住院号"]).upper()
+    if values["住院号"].startswith("ZV"):
+        values["住院号"] = "ZY" + values["住院号"][2:]
+    if values["住院号"].startswith("ZYS"):
+        values["住院号"] = "ZY5" + values["住院号"][3:]
+    if len(values["住院号"]) > 2:
+        prefix, number = values["住院号"][:2], values["住院号"][2:]
+        values["住院号"] = prefix + number.translate(str.maketrans({"O": "0", "I": "1", "L": "1", "S": "5"}))
+    def looks_datetime(value: Any) -> bool:
+        return bool(
+            re.fullmatch(
+                r"\d{4}[-/.]\d{1,2}[-/.]\d{1,2}\s+\d{1,2}:\d{1,2}:\d{1,2}",
+                str(value),
+            )
+        )
+
+    def looks_days(value: Any) -> bool:
+        return bool(re.fullmatch(r"\d+", str(value)))
+
+    def looks_postop(value: Any) -> bool:
+        return bool(re.fullmatch(r"后\d+天", str(value)))
+
+    values["入院时间"] = normalize_date(values["入院时间"])
+    values["最后书写时间"] = normalize_date(values["最后书写时间"])
+    values["出院时间"] = normalize_date(values["出院时间"])
+    shifted_by_long_diagnosis = re.search(
+        r"(\d{4}[-/.]\d{1,2}[-/.]\d{1,2}\s+\d{1,2}:\d{1,2}:\d{1,2})$",
+        values["诊断"],
+    )
+    if (
+        shifted_by_long_diagnosis
+        and looks_datetime(values["入院时间"])
+        and looks_days(values["最后书写时间"])
+        and (not values["住院天数"] or looks_datetime(values["住院天数"]) or looks_postop(values["住院天数"]))
+    ):
+        shifted_admission = normalize_date(shifted_by_long_diagnosis.group(1))
+        old_last_write = values["入院时间"]
+        old_hospital_days = values["最后书写时间"]
+        old_discharge_or_postop = values["住院天数"]
+        old_postop = values["出院时间"]
+        values["诊断"] = values["诊断"][: shifted_by_long_diagnosis.start()].strip()
+        values["入院时间"] = shifted_admission
+        values["最后书写时间"] = normalize_date(old_last_write)
+        values["住院天数"] = old_hospital_days
+        values["出院时间"] = normalize_date(old_discharge_or_postop) if looks_datetime(old_discharge_or_postop) else ""
+        if not values["手术后天数"]:
+            if looks_postop(old_discharge_or_postop):
+                values["手术后天数"] = old_discharge_or_postop
+            elif looks_postop(old_postop):
+                values["手术后天数"] = old_postop
+    if shifted_by_long_diagnosis and not values["入院时间"]:
+        values["诊断"] = values["诊断"][: shifted_by_long_diagnosis.start()].strip()
+        values["入院时间"] = normalize_date(shifted_by_long_diagnosis.group(1))
+
+    days_with_discharge = re.fullmatch(
+        r"(\d+)\s+(\d{4}[-/.]\d{1,2}[-/.]\d{1,2}\s+\d{1,2}:\d{1,2}:\d{1,2})",
+        str(values["住院天数"]),
+    )
+    if days_with_discharge:
+        values["住院天数"] = days_with_discharge.group(1)
+        if not values["出院时间"]:
+            values["出院时间"] = normalize_date(days_with_discharge.group(2))
+
+    if looks_postop(values["出院时间"]) and not values["手术后天数"]:
+        values["手术后天数"] = values["出院时间"]
+        values["出院时间"] = ""
+    if looks_postop(values["住院天数"]) and not values["手术后天数"]:
+        values["手术后天数"] = values["住院天数"]
+        values["住院天数"] = ""
+
+    if looks_days(values["最后书写时间"]) and looks_datetime(values["住院天数"]):
+        old_days = values["最后书写时间"]
+        old_discharge = values["住院天数"]
+        old_postop = values["出院时间"]
+        values["最后书写时间"] = ""
+        values["住院天数"] = old_days
+        values["出院时间"] = normalize_date(old_discharge)
+        if looks_postop(old_postop) and not values["手术后天数"]:
+            values["手术后天数"] = old_postop
+    elif looks_datetime(values["住院天数"]) and (not values["出院时间"] or looks_postop(values["出院时间"])):
+        old_discharge = values["住院天数"]
+        old_postop = values["出院时间"]
+        values["住院天数"] = ""
+        values["出院时间"] = normalize_date(old_discharge)
+        if looks_postop(old_postop) and not values["手术后天数"]:
+            values["手术后天数"] = old_postop
+
+    last_write_with_postop = re.fullmatch(r"(.+?)\s+(后\d+天)", values["最后书写时间"])
+    if last_write_with_postop and not values["手术后天数"]:
+        values["最后书写时间"] = last_write_with_postop.group(1)
+        values["手术后天数"] = last_write_with_postop.group(2)
+    if re.fullmatch(r"后\d+天", values["最后书写时间"]) and not values["手术后天数"]:
+        values["手术后天数"] = values["最后书写时间"]
+        values["最后书写时间"] = ""
+    if values["住院天数"].isdigit():
+        values["住院天数"] = int(values["住院天数"])
+    return values
+
+
+def validate_patient_row(values: dict[str, Any]) -> list[str]:
+    warnings: list[str] = []
+    if not values.get("姓名"):
+        warnings.append("缺少姓名")
+    if values.get("性别") not in {"男", "女"}:
+        warnings.append("性别异常")
+    if values.get("年龄") and not re.fullmatch(r"\d{1,3}岁", str(values["年龄"])):
+        warnings.append("年龄格式异常")
+    if not normalize_text(values.get("住院号", "")):
+        warnings.append("缺少住院号")
+    if not values.get("入院时间"):
+        warnings.append("缺少入院时间")
+    elif not re.fullmatch(r"\d{4}-\d{2}-\d{2} \d{2}:\d{2}:\d{2}", str(values["入院时间"])):
+        warnings.append("入院时间格式异常")
+    if values.get("出院时间") and not re.fullmatch(
+        r"\d{4}-\d{2}-\d{2} \d{2}:\d{2}:\d{2}", str(values["出院时间"])
+    ):
+        warnings.append("出院时间格式异常")
+    if (
+        re.fullmatch(r"\d{4}-\d{2}-\d{2} \d{2}:\d{2}:\d{2}", str(values.get("入院时间", "")))
+        and re.fullmatch(r"\d{4}-\d{2}-\d{2} \d{2}:\d{2}:\d{2}", str(values.get("出院时间", "")))
+        and str(values["入院时间"]) > str(values["出院时间"])
+    ):
+        warnings.append("出院时间早于入院时间")
+    if values.get("最后书写时间") and not re.fullmatch(
+        r"\d{4}-\d{2}-\d{2} \d{2}:\d{2}:\d{2}", str(values["最后书写时间"])
+    ):
+        warnings.append("最后书写时间格式异常")
+    if values.get("住院天数") != "" and not isinstance(values.get("住院天数"), int):
+        warnings.append("住院天数格式异常")
+    return warnings
+
+
+def is_blank_or_footer(row: dict[str, Any]) -> bool:
+    if row.get("住院号"):
+        return False
+    filled = [value for value in row.values() if value not in ("", None)]
+    return len(filled) == 0
+
+
+def write_json(path: Path, data: Any) -> None:
+    path.parent.mkdir(parents=True, exist_ok=True)
+    path.write_text(json.dumps(data, ensure_ascii=False, indent=2), encoding="utf-8")
+
+
+def write_jsonl(path: Path, records: list[dict[str, Any]]) -> None:
+    path.parent.mkdir(parents=True, exist_ok=True)
+    with path.open("w", encoding="utf-8") as file:
+        for record in records:
+            file.write(json.dumps(record, ensure_ascii=False) + "\n")
+
+
+def write_csv(path: Path, records: list[dict[str, Any]]) -> None:
+    fieldnames = [
+        "大科室",
+        "子科室",
+        "来源文件夹",
+        "图片名",
+        "图片内行号",
+        *COLUMNS,
+        "复核状态",
+        "复核提示",
+    ]
+    path.parent.mkdir(parents=True, exist_ok=True)
+    with path.open("w", encoding="utf-8-sig", newline="") as file:
+        writer = csv.DictWriter(file, fieldnames=fieldnames)
+        writer.writeheader()
+        for record in records:
+            patient = record["患者信息"]
+            writer.writerow(
+                {
+                    "大科室": record["大科室"],
+                    "子科室": record["子科室"],
+                    "来源文件夹": record["来源文件夹"],
+                    "图片名": record["图片信息"]["图片名"],
+                    "图片内行号": record["图片信息"]["图片内行号"],
+                    **patient,
+                    "复核状态": record["复核"]["状态"],
+                    "复核提示": "；".join(record["复核"]["提示"]),
+                }
+            )
+
+
+def load_corrections(path: Path) -> dict[tuple[str, int], dict[str, Any]]:
+    if not path.exists():
+        return {}
+    data = json.loads(path.read_text(encoding="utf-8"))
+    corrections: dict[tuple[str, int], dict[str, Any]] = {}
+    for item in data:
+        image_path = item["图片路径"]
+        row_no = int(item["图片内行号"])
+        corrections[(image_path, row_no)] = item
+        for prefix in ["已处理-患者目录图片集群/", "待处理-患者目录图片集群/"]:
+            corrections[(prefix + image_path, row_no)] = item
+    return corrections
+
+
+def apply_review_options(warnings: list[str], correction: dict[str, Any]) -> list[str]:
+    return warnings
+
+
+def apply_corrections(records: list[dict[str, Any]], corrections: dict[tuple[str, int], dict[str, Any]]) -> None:
+    for record in records:
+        key = (record["图片信息"]["图片路径"], int(record["图片信息"]["图片内行号"]))
+        if key not in corrections:
+            record["复核"]["提示"] = validate_patient_row(record["患者信息"])
+            record["复核"]["状态"] = "需人工复核" if record["复核"]["提示"] else "自动复核通过"
+            continue
+        correction = corrections[key]
+        record["患者信息"].update(correction.get("患者信息", {}))
+        record["复核"]["人工修正"] = True
+        record["复核"]["复核选项"] = correction.get("复核选项", {})
+        if correction.get("复核备注"):
+            record["复核"]["人工备注"] = correction.get("复核备注", "")
+        record["复核"]["提示"] = apply_review_options(validate_patient_row(record["患者信息"]), correction)
+        record["复核"]["状态"] = "需人工复核" if record["复核"]["提示"] else "人工复核通过"
+
+
+def record_quality_rank(record: dict[str, Any]) -> tuple[int, int, int, int]:
+    patient = record["患者信息"]
+    review = record["复核"]
+    review_ok = 0 if review.get("状态") == "需人工复核" else 1
+    manual_corrected = 1 if review.get("人工修正") else 0
+    date_count = int(bool(patient.get("入院时间"))) + int(bool(patient.get("出院时间")))
+    filled_count = sum(1 for column in COLUMNS if patient.get(column) not in ("", None))
+    return (review_ok, manual_corrected, date_count, filled_count)
+
+
+def summarize_record_for_duplicate(record: dict[str, Any]) -> dict[str, Any]:
+    patient = record["患者信息"]
+    image = record["图片信息"]
+    return {
+        "大科室": record["大科室"],
+        "子科室": record["子科室"],
+        "来源文件夹": record["来源文件夹"],
+        "图片路径": image["图片路径"],
+        "图片名": image["图片名"],
+        "图片内行号": image["图片内行号"],
+        "姓名": patient.get("姓名", ""),
+        "住院号": patient.get("住院号", ""),
+        "入院时间": patient.get("入院时间", ""),
+        "出院时间": patient.get("出院时间", ""),
+        "复核状态": record["复核"].get("状态", ""),
+    }
+
+
+def deduplicate_records_by_inpatient_no(
+    records: list[dict[str, Any]],
+) -> tuple[list[dict[str, Any]], list[dict[str, Any]]]:
+    kept_by_no: dict[str, dict[str, Any]] = {}
+    kept_order: list[str] = []
+    dropped: list[dict[str, Any]] = []
+    for record in records:
+        inpatient_no = normalize_text(record["患者信息"].get("住院号", ""))
+        if not inpatient_no:
+            dropped.append(
+                {
+                    "住院号": "",
+                    "保留记录": {},
+                    "剔除记录": summarize_record_for_duplicate(record),
+                    "规则": "住院号为空，未纳入归档结果和数据库",
+                }
+            )
+            continue
+        if inpatient_no not in kept_by_no:
+            kept_order.append(inpatient_no)
+        else:
+            dropped.append(
+                {
+                    "住院号": inpatient_no,
+                    "保留记录": summarize_record_for_duplicate(record),
+                    "剔除记录": summarize_record_for_duplicate(kept_by_no[inpatient_no]),
+                    "规则": "住院号重复，后出现记录覆盖先出现记录",
+                }
+            )
+        kept_by_no[inpatient_no] = record
+    return [kept_by_no[inpatient_no] for inpatient_no in kept_order], dropped
+
+
+def process_folder(
+    folder: Path,
+    output_root: Path,
+    department: Department,
+    args: argparse.Namespace,
+    secret_id: str,
+    secret_key: str,
+) -> tuple[list[dict[str, Any]], dict[str, Any]]:
+    images = sorted(
+        [path for path in folder.iterdir() if path.is_file() and path.suffix.lower() in IMAGE_EXTENSIONS],
+        key=natural_key,
+    )
+    folder_key = folder.name
+    composites_dir = output_root / "merged_images" / folder_key
+    raw_dir = output_root / "raw_ocr" / folder_key
+    records: list[dict[str, Any]] = []
+    folder_warnings: list[str] = []
+    group_infos: list[dict[str, Any]] = []
+
+    def cache_label(label: str) -> str:
+        if args.ocr_engine != "table-v3":
+            label = f"{args.ocr_engine.replace('-', '_')}_{label}"
+        if args.image_padding_y > 0:
+            return f"{label}_pady{args.image_padding_y}"
+        return label
+
+    def records_from_rows(
+        rows: list[list[str]],
+        group_paths: list[Path],
+        group_index_value: int,
+        composite_path_value: Path,
+        cache_path_value: Path,
+        request_id: str | None,
+    ) -> list[dict[str, Any]]:
+        built_records: list[dict[str, Any]] = []
+        row_counts = [infer_rows_for_image(path, args.rows_per_image) for path in group_paths]
+        for table_row_index, row in enumerate(rows):
+            patient = clean_patient_row(row)
+            if is_blank_or_footer(patient):
+                continue
+            source_index, image_row = locate_source_row(table_row_index, row_counts)
+            source_path = group_paths[source_index]
+            warnings = validate_patient_row(patient)
+            record = {
+                "大科室": department.major,
+                "子科室": department.sub,
+                "来源文件夹": folder.name,
+                "标准化文件夹科室名": department.raw_name,
+                "患者信息": patient,
+                "图片信息": {
+                    "图片路径": str(source_path),
+                    "图片名": source_path.name,
+                    "图片序号": natural_key(source_path),
+                    "图片内行号": image_row + 1,
+                    "拼接组序号": group_index_value,
+                    "拼接图片路径": str(composite_path_value),
+                    "OCR缓存路径": str(cache_path_value),
+                    "OCR请求ID": request_id,
+                },
+                "复核": {
+                    "状态": "需人工复核" if warnings else "自动复核通过",
+                    "提示": warnings,
+                },
+            }
+            built_records.append(record)
+        return built_records
+
+    def expected_row_count(group_paths: list[Path]) -> int:
+        return sum(infer_rows_for_image(path, args.rows_per_image) for path in group_paths)
+
+    def attempt_ocr_group(
+        group_paths: list[Path],
+        group_index_value: int,
+        label: str,
+        display_label: str,
+        prebuilt_merge_info: dict[str, Any] | None = None,
+    ) -> tuple[dict[str, Any], list[dict[str, Any]]]:
+        composite_path = composites_dir / f"{label}.png"
+        merge_info = prebuilt_merge_info or merge_images(group_paths, composite_path, args.image_padding_y)
+        cache_path = raw_dir / f"{label}.json"
+        print(f"    {display_label}: {len(group_paths)} images -> {composite_path}", flush=True)
+        if args.ocr_engine == "general-accurate":
+            response = call_general_accurate(
+                composite_path,
+                cache_path,
+                secret_id,
+                secret_key,
+                args.region,
+                args.timeout,
+                args.force,
+                args.max_retries,
+            )
+        else:
+            response = call_table_v3(
+                composite_path,
+                cache_path,
+                secret_id,
+                secret_key,
+                args.region,
+                args.timeout,
+                args.force,
+                args.max_retries,
+            )
+        rows = ocr_response_to_rows(response, args.ocr_engine, int(merge_info.get("width", 0)))
+        expected_rows = expected_row_count(group_paths)
+        if len(group_paths) > 1 and len(rows) < expected_rows:
+            raise RuntimeError(f"识别行数偏少: {len(rows)} / {expected_rows}")
+        request_id = response.get("RequestId")
+        print(f"      rows: {len(rows)} request: {request_id}", flush=True)
+        info = {
+            "label": label,
+            "merged_image": merge_info,
+            "ocr_cache": str(cache_path),
+            "ocr_request_id": request_id,
+            "row_count": len(rows),
+            "expected_row_count": expected_rows,
+            "image_count": len(group_paths),
+        }
+        return info, records_from_rows(rows, group_paths, group_index_value, composite_path, cache_path, request_id)
+
+    def run_chunked_fallback(
+        group: list[Path],
+        group_index_value: int,
+        main_label: str,
+        fallback_size: int,
+    ) -> tuple[list[dict[str, Any]], list[dict[str, Any]]]:
+        candidate_infos: list[dict[str, Any]] = []
+        candidate_records: list[dict[str, Any]] = []
+        chunks = list(enumerate(batched(group, fallback_size)))
+
+        def run_chunk(item: tuple[int, list[Path]]) -> tuple[dict[str, Any], list[dict[str, Any]]]:
+            chunk_index, chunk = item
+            chunk_label = f"{main_label}_fallback{fallback_size}_{chunk_index:02d}"
+            chunk_info, chunk_records = attempt_ocr_group(
+                chunk,
+                group_index_value,
+                chunk_label,
+                chunk_label,
+            )
+            chunk_info["part_index"] = chunk_index
+            return chunk_info, chunk_records
+
+        if args.workers > 1 and len(chunks) > 1:
+            with concurrent.futures.ThreadPoolExecutor(max_workers=min(args.workers, len(chunks))) as executor:
+                results = list(executor.map(run_chunk, chunks))
+        else:
+            results = [run_chunk(item) for item in chunks]
+
+        for chunk_info, chunk_records in results:
+            candidate_infos.append(chunk_info)
+            candidate_records.extend(chunk_records)
+        return candidate_infos, candidate_records
+
+    def single_cache_paths(main_label: str, group: list[Path]) -> list[Path]:
+        return [raw_dir / f"{main_label}_part_{part_index:02d}.json" for part_index in range(len(group))]
+
+    def chunk_cache_paths(main_label: str, group: list[Path], fallback_size: int) -> list[Path]:
+        chunks = batched(group, fallback_size)
+        return [raw_dir / f"{main_label}_fallback{fallback_size}_{chunk_index:02d}.json" for chunk_index in range(len(chunks))]
+
+    def run_single_fallback(
+        group: list[Path],
+        group_index_value: int,
+        main_label: str,
+        main_merge_info: dict[str, Any],
+        main_cache_path: Path,
+        initial_error: str,
+        fallback_errors: list[str],
+        adaptive_from_previous: bool = False,
+    ) -> None:
+        print(f"    尝试单张OCR回退", flush=True)
+        single_infos: list[dict[str, Any]] = []
+        for part_index, single_path in enumerate(group):
+            single_label = f"{main_label}_part_{part_index:02d}"
+            single_composite_path = composites_dir / f"{single_label}.png"
+            single_cache_path = raw_dir / f"{single_label}.json"
+            single_merge_info = merge_images([single_path], single_composite_path, args.image_padding_y)
+            try:
+                single_info, single_records = attempt_ocr_group(
+                    [single_path],
+                    group_index_value,
+                    single_label,
+                    single_label,
+                    single_merge_info,
+                )
+                single_info["part_index"] = part_index
+                records.extend(single_records)
+                single_infos.append(single_info)
+            except RuntimeError as single_exc:
+                single_message = f"{single_label} OCR失败: {single_exc}"
+                print(f"    {single_message}", flush=True)
+                folder_warnings.append(single_message)
+                single_infos.append(
+                    {
+                        "label": single_label,
+                        "part_index": part_index,
+                        "merged_image": single_merge_info,
+                        "ocr_cache": str(single_cache_path),
+                        "ocr_request_id": None,
+                        "row_count": 0,
+                        "expected_row_count": expected_row_count([single_path]),
+                        "image_count": 1,
+                        "error": str(single_exc),
+                    }
+                )
+        group_infos.append(
+            {
+                "group_index": group_index_value,
+                "label": main_label,
+                "merged_image": main_merge_info,
+                "ocr_cache": str(main_cache_path),
+                "ocr_request_id": None,
+                "row_count": sum(item["row_count"] for item in single_infos),
+                "expected_row_count": expected_row_count(group),
+                "fallback": True,
+                "fallback_strategy": "single_image",
+                "fallback_parts": single_infos,
+                "initial_error": initial_error,
+                "fallback_errors": fallback_errors,
+                "adaptive_from_previous": adaptive_from_previous,
+            }
+        )
+
+    groups = batched(images, args.batch_size)
+    if args.limit_groups_per_folder:
+        groups = groups[: args.limit_groups_per_folder]
+
+    preferred_fallback_size: int | None = None
+    preferred_fallback_failures = 0
+    for group_index, group in enumerate(groups):
+        main_label = cache_label(f"group_{group_index:04d}")
+        main_composite_path = composites_dir / f"{main_label}.png"
+        main_cache_path = raw_dir / f"{main_label}.json"
+        main_merge_info = merge_images(group, main_composite_path, args.image_padding_y)
+        part_cache_paths = single_cache_paths(main_label, group)
+
+        if not main_cache_path.exists() and all(path.exists() for path in part_cache_paths):
+            message = f"{main_label} 使用已有单图缓存重建"
+            print(f"    {message}", flush=True)
+            run_single_fallback(
+                group,
+                group_index,
+                main_label,
+                main_merge_info,
+                main_cache_path,
+                message,
+                [],
+                adaptive_from_previous=True,
+            )
+            time.sleep(args.sleep)
+            continue
+
+        if preferred_fallback_size and len(group) > 1:
+            use_preferred_fallback = True
+            if args.rebuild_from_cache:
+                if preferred_fallback_size == 1:
+                    preferred_paths = single_cache_paths(main_label, group)
+                else:
+                    preferred_paths = chunk_cache_paths(main_label, group, int(preferred_fallback_size))
+                use_preferred_fallback = all(path.exists() for path in preferred_paths)
+                if not use_preferred_fallback:
+                    print(f"    {main_label} 当前回退缓存不完整，先尝试主缓存", flush=True)
+            if not use_preferred_fallback:
+                preferred_fallback_size = None
+            elif preferred_fallback_size == 1:
+                message = f"{main_label} 依据前序组结果，直接使用单张OCR"
+                print(f"    {message}", flush=True)
+                run_single_fallback(
+                    group,
+                    group_index,
+                    main_label,
+                    main_merge_info,
+                    main_cache_path,
+                    message,
+                    [],
+                    adaptive_from_previous=True,
+                )
+                time.sleep(args.sleep)
+                continue
+            elif 1 < preferred_fallback_size < len(group):
+                message = f"{main_label} 依据前序组结果，直接使用 {preferred_fallback_size} 张拼接"
+                print(f"    {message}", flush=True)
+                try:
+                    candidate_infos, candidate_records = run_chunked_fallback(
+                        group,
+                        group_index,
+                        main_label,
+                        int(preferred_fallback_size),
+                    )
+                    records.extend(candidate_records)
+                    group_infos.append(
+                        {
+                            "group_index": group_index,
+                            "label": main_label,
+                            "merged_image": main_merge_info,
+                            "ocr_cache": str(main_cache_path),
+                            "ocr_request_id": None,
+                            "row_count": sum(item["row_count"] for item in candidate_infos),
+                            "expected_row_count": expected_row_count(group),
+                            "fallback": True,
+                            "fallback_strategy": f"{preferred_fallback_size}_images",
+                            "fallback_parts": candidate_infos,
+                            "initial_error": message,
+                            "fallback_errors": [],
+                            "adaptive_from_previous": True,
+                        }
+                    )
+                    time.sleep(args.sleep)
+                    preferred_fallback_failures = 0
+                    continue
+                except RuntimeError as adaptive_exc:
+                    failed_size = int(preferred_fallback_size)
+                    preferred_fallback_failures += 1
+                    adaptive_message = f"{main_label} 前序 {failed_size} 张策略未通过，本组改用单张OCR: {adaptive_exc}"
+                    print(f"    {adaptive_message}", flush=True)
+                    folder_warnings.append(adaptive_message)
+                    if preferred_fallback_failures >= args.fallback_demote_threshold:
+                        preferred_fallback_size = 1
+                        demote_message = (
+                            f"{main_label} 连续 {preferred_fallback_failures} 个拼接组未通过，"
+                            "后续直接使用单张OCR"
+                        )
+                        print(f"    {demote_message}", flush=True)
+                        folder_warnings.append(demote_message)
+                    else:
+                        preferred_fallback_size = failed_size
+                    run_single_fallback(
+                        group,
+                        group_index,
+                        main_label,
+                        main_merge_info,
+                        main_cache_path,
+                        adaptive_message,
+                        [],
+                        adaptive_from_previous=True,
+                    )
+                    time.sleep(args.sleep)
+                    continue
+        try:
+            main_info, main_records = attempt_ocr_group(
+                group,
+                group_index,
+                main_label,
+                main_label,
+                main_merge_info,
+            )
+            main_info.update(
+                {
+                    "group_index": group_index,
+                    "fallback": False,
+                    "fallback_strategy": None,
+                }
+            )
+            group_infos.append(main_info)
+            records.extend(main_records)
+        except RuntimeError as exc:
+            if len(group) == 1:
+                message = f"{main_label} OCR失败: {exc}"
+                print(f"    {message}", flush=True)
+                folder_warnings.append(message)
+                group_infos.append(
+                    {
+                        "group_index": group_index,
+                        "label": main_label,
+                        "merged_image": main_merge_info,
+                        "ocr_cache": str(main_cache_path),
+                        "ocr_request_id": None,
+                        "row_count": 0,
+                        "expected_row_count": expected_row_count(group),
+                        "fallback": False,
+                        "fallback_strategy": None,
+                        "error": str(exc),
+                    }
+                )
+                time.sleep(args.sleep)
+                continue
+
+            message = f"{main_label} {len(group)}张拼接OCR未通过，开始降档: {exc}"
+            print(f"    {message}", flush=True)
+            folder_warnings.append(message)
+            fallback_errors: list[str] = []
+            fallback_success = False
+
+            for fallback_size in (4, 3, 2):
+                if not 1 < fallback_size < len(group):
+                    continue
+                print(f"    尝试 {fallback_size} 张拼接回退", flush=True)
+                try:
+                    candidate_infos, candidate_records = run_chunked_fallback(group, group_index, main_label, fallback_size)
+                except RuntimeError as fallback_exc:
+                    fallback_message = f"{main_label} {fallback_size}张拼接仍未通过: {fallback_exc}"
+                    print(f"    {fallback_message}", flush=True)
+                    folder_warnings.append(fallback_message)
+                    fallback_errors.append(fallback_message)
+                    continue
+
+                records.extend(candidate_records)
+                preferred_fallback_size = fallback_size
+                preferred_fallback_failures = 0
+                group_infos.append(
+                    {
+                        "group_index": group_index,
+                        "label": main_label,
+                        "merged_image": main_merge_info,
+                        "ocr_cache": str(main_cache_path),
+                        "ocr_request_id": None,
+                        "row_count": sum(item["row_count"] for item in candidate_infos),
+                        "expected_row_count": expected_row_count(group),
+                        "fallback": True,
+                        "fallback_strategy": f"{fallback_size}_images",
+                        "fallback_parts": candidate_infos,
+                        "initial_error": str(exc),
+                        "fallback_errors": fallback_errors,
+                    }
+                )
+                fallback_success = True
+                break
+
+            if fallback_success:
+                time.sleep(args.sleep)
+                continue
+
+            preferred_fallback_size = 1
+            run_single_fallback(group, group_index, main_label, main_merge_info, main_cache_path, str(exc), fallback_errors)
+        time.sleep(args.sleep)
+
+    summary = {
+        "来源文件夹": folder.name,
+        "大科室": department.major,
+        "子科室": department.sub,
+        "图片数": len(images),
+        "记录数": len(records),
+        "拼接组数": len(group_infos),
+        "拼接组": group_infos,
+        "提示": folder_warnings,
+    }
+    return records, summary
+
+
+def parse_args() -> argparse.Namespace:
+    parser = argparse.ArgumentParser(description=__doc__)
+    parser.add_argument("--input", default="列表", help="患者列表图片根目录")
+    parser.add_argument("--output", default="数据处理工作区/列表归档结果", help="输出目录")
+    parser.add_argument("--departments", default="数据处理工作区/01_科室分类规则.json", help="科室分类 JSON")
+    parser.add_argument("--batch-size", type=int, default=3, help="每次拼接图片数量，建议 3 或 4")
+    parser.add_argument("--image-padding-y", type=int, default=24, help="每张图拼接前添加的上下白色边界像素数")
+    parser.add_argument("--rows-per-image", type=int, default=0, help="每张 HIS 列表截图的行数；0 表示按图片高度自动推断")
+    parser.add_argument("--corrections", default="数据处理工作区/03_人工复核修正.json", help="人工复核修正 JSON")
+    parser.add_argument(
+        "--ocr-engine",
+        choices=sorted(OCR_ENGINE_LABELS),
+        default="table-v3",
+        help="OCR 引擎：table-v3 表格识别 V3",
+    )
+    parser.add_argument("--region", default="ap-shanghai", help="腾讯云 OCR 地域")
+    parser.add_argument("--timeout", type=int, default=60, help="单次 OCR 超时秒数")
+    parser.add_argument("--sleep", type=float, default=0.2, help="OCR 调用间隔秒数")
+    parser.add_argument("--max-retries", type=int, default=0, help="OCR 调用失败重试次数")
+    parser.add_argument(
+        "--fallback-demote-threshold",
+        type=int,
+        default=3,
+        help="同一科室连续多少个拼接组失败后，才把后续组整体降为单张OCR",
+    )
+    parser.add_argument("--workers", type=int, default=1, help="同一拼接组内并发OCR分片数；建议 1-2")
+    parser.add_argument("--folder-workers", type=int, default=1, help="并发处理科室目录数；OCR接口稳定时可设为 2-4")
+    parser.add_argument("--force", action="store_true", help="忽略 OCR 缓存重新识别")
+    parser.add_argument("--limit-folders", type=int, default=0, help="调试用：只处理前 N 个科室目录")
+    parser.add_argument("--limit-groups-per-folder", type=int, default=0, help="调试用：每个科室只处理前 N 个拼接组")
+    parser.add_argument("--rebuild-from-cache", action="store_true", help="只用已有 OCR 缓存重建结果，不发起 OCR 请求")
+    return parser.parse_args()
+
+
+def main() -> None:
+    args = parse_args()
+    if args.batch_size < 1:
+        raise ValueError("--batch-size 必须大于 0")
+    if args.workers < 1:
+        raise ValueError("--workers 必须大于 0")
+    if args.folder_workers < 1:
+        raise ValueError("--folder-workers 必须大于 0")
+    if args.image_padding_y < 0:
+        raise ValueError("--image-padding-y 不能小于 0")
+
+    input_root = Path(args.input)
+    output_root = Path(args.output)
+    sub_to_major, aliases = load_departments(Path(args.departments))
+    if args.rebuild_from_cache:
+        secret_id, secret_key = "", ""
+    else:
+        secret_id, secret_key = get_credentials()
+    corrections = load_corrections(Path(args.corrections))
+
+    folders = sorted([path for path in input_root.iterdir() if path.is_dir()], key=lambda path: path.name)
+    if args.limit_folders:
+        folders = folders[: args.limit_folders]
+
+    all_records: list[dict[str, Any]] = []
+    folder_summaries: list[dict[str, Any]] = []
+    classifications: list[dict[str, Any]] = []
+
+    folder_jobs: list[tuple[Path, Department]] = []
+    for folder in folders:
+        department = classify_department(folder.name, sub_to_major, aliases)
+        classifications.append(
+            {
+                "来源文件夹": folder.name,
+                "标准化文件夹科室名": department.raw_name,
+                "大科室": department.major,
+                "子科室": department.sub,
+            }
+        )
+        folder_jobs.append((folder, department))
+
+    def run_folder_job(item: tuple[Path, Department]) -> tuple[list[dict[str, Any]], dict[str, Any]]:
+        folder, department = item
+        print(f"[{folder.name}] -> {department.major} / {department.sub}", flush=True)
+        return process_folder(folder, output_root, department, args, secret_id, secret_key)
+
+    if args.folder_workers > 1 and len(folder_jobs) > 1:
+        with concurrent.futures.ThreadPoolExecutor(max_workers=min(args.folder_workers, len(folder_jobs))) as executor:
+            folder_results = list(executor.map(run_folder_job, folder_jobs))
+    else:
+        folder_results = [run_folder_job(item) for item in folder_jobs]
+
+    for records, summary in folder_results:
+        all_records.extend(records)
+        folder_summaries.append(summary)
+
+    apply_corrections(all_records, corrections)
+    record_count_before_dedup = len(all_records)
+    all_records, duplicate_records = deduplicate_records_by_inpatient_no(all_records)
+    issue_records = [record for record in all_records if record["复核"]["状态"] == "需人工复核"]
+    corrected_records = [record for record in all_records if record["复核"].get("人工修正")]
+    archive = {
+        "生成时间": dt.datetime.now().isoformat(timespec="seconds"),
+        "输入目录": str(input_root),
+        "OCR引擎": OCR_ENGINE_LABELS[args.ocr_engine],
+        "拼接设置": {
+            "batch_size": args.batch_size,
+            "rows_per_image": args.rows_per_image,
+            "image_padding_y": args.image_padding_y,
+        },
+        "科室归类": classifications,
+        "汇总": {
+            "科室目录数": len(folders),
+            "图片数": sum(item["图片数"] for item in folder_summaries),
+            "去重前患者记录数": record_count_before_dedup,
+            "患者记录数": len(all_records),
+            "需人工复核记录数": len(issue_records),
+            "人工修正记录数": len(corrected_records),
+            "重复住院号剔除记录数": len(duplicate_records),
+        },
+        "科室汇总": folder_summaries,
+        "重复住院号剔除记录": duplicate_records,
+        "患者记录": all_records,
+    }
+    review_report = {
+        "生成时间": archive["生成时间"],
+        "汇总": archive["汇总"],
+        "需人工复核记录": issue_records,
+        "人工修正记录": corrected_records,
+        "重复住院号剔除记录": duplicate_records,
+        "科室级提示": [
+            {"来源文件夹": item["来源文件夹"], "提示": item["提示"]}
+            for item in folder_summaries
+            if item["提示"]
+        ],
+    }
+
+    write_json(output_root / "列表_科室归类.json", classifications)
+    write_json(output_root / "患者列表_结构化.json", archive)
+    write_jsonl(output_root / "患者列表_记录.jsonl", all_records)
+    write_csv(output_root / "患者列表_记录.csv", all_records)
+    write_json(output_root / "复核报告.json", review_report)
+    write_json(output_root / "重复住院号报告.json", duplicate_records)
+    print(json.dumps(archive["汇总"], ensure_ascii=False, indent=2), flush=True)
+
+
+if __name__ == "__main__":
+    main()
diff --git a/患者列表处理/数据处理工作区/03_人工复核修正.template.json b/患者列表处理/数据处理工作区/03_人工复核修正.template.json
new file mode 100644
index 0000000..fe51488
--- /dev/null
+++ b/患者列表处理/数据处理工作区/03_人工复核修正.template.json
@@ -0,0 +1 @@
+[]
diff --git a/患者列表处理/数据处理工作区/04_合并批次结果.py b/患者列表处理/数据处理工作区/04_合并批次结果.py
new file mode 100644
index 0000000..8baab0e
--- /dev/null
+++ b/患者列表处理/数据处理工作区/04_合并批次结果.py
@@ -0,0 +1,247 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+"""Consolidate patient archive result folders."""
+
+from __future__ import annotations
+
+import csv
+import json
+from collections import Counter
+from pathlib import Path
+from typing import Any
+
+
+ROOT = Path("数据处理结果区")
+RESULT_SUFFIX = "-列表归档结果"
+INFO_DIR_NAME = "信息记录"
+
+
+def write_json(path: Path, data: Any) -> None:
+    path.parent.mkdir(parents=True, exist_ok=True)
+    path.write_text(json.dumps(data, ensure_ascii=False, indent=2), encoding="utf-8")
+
+
+def write_jsonl(path: Path, records: list[dict[str, Any]]) -> None:
+    path.parent.mkdir(parents=True, exist_ok=True)
+    with path.open("w", encoding="utf-8") as file:
+        for record in records:
+            file.write(json.dumps(record, ensure_ascii=False) + "\n")
+
+
+def flatten_record(batch_name: str, record: dict[str, Any]) -> dict[str, Any]:
+    patient = record["患者信息"]
+    image = record["图片信息"]
+    review = record["复核"]
+    return {
+        "处理批次": batch_name,
+        "大科室": record["大科室"],
+        "子科室": record["子科室"],
+        "来源文件夹": record["来源文件夹"],
+        "图片路径": image["图片路径"],
+        "图片名": image["图片名"],
+        "图片内行号": image["图片内行号"],
+        "拼接组序号": image["拼接组序号"],
+        "OCR请求ID": image.get("OCR请求ID", ""),
+        "姓名": patient["姓名"],
+        "性别": patient["性别"],
+        "年龄": patient["年龄"],
+        "住院号": patient["住院号"],
+        "诊断": patient["诊断"],
+        "入院时间": patient["入院时间"],
+        "最后书写时间": patient["最后书写时间"],
+        "住院天数": patient["住院天数"],
+        "出院时间": patient["出院时间"],
+        "手术后天数": patient["手术后天数"],
+        "复核状态": review["状态"],
+        "复核提示": review_note_text(review),
+        "人工修正": bool(review.get("人工修正")),
+    }
+
+
+def review_note_text(review: dict[str, Any]) -> str:
+    notes = [str(item) for item in review.get("提示", []) if str(item) and str(item) != "缺少出院时间"]
+    if review.get("人工备注"):
+        notes.append(f"人工备注: {review['人工备注']}")
+    return "；".join(notes)
+
+
+def record_quality_rank(record: dict[str, Any]) -> tuple[int, int, int, int]:
+    patient = record["患者信息"]
+    review = record["复核"]
+    review_ok = 0 if review.get("状态") == "需人工复核" else 1
+    manual_corrected = 1 if review.get("人工修正") else 0
+    date_count = int(bool(patient.get("入院时间"))) + int(bool(patient.get("出院时间")))
+    filled_count = sum(1 for value in patient.values() if value not in ("", None))
+    return (review_ok, manual_corrected, date_count, filled_count)
+
+
+def summarize_record_for_duplicate(record: dict[str, Any]) -> dict[str, Any]:
+    patient = record["患者信息"]
+    image = record["图片信息"]
+    review = record["复核"]
+    return {
+        "处理批次": record["处理批次"],
+        "大科室": record["大科室"],
+        "子科室": record["子科室"],
+        "来源文件夹": record["来源文件夹"],
+        "图片路径": image["图片路径"],
+        "图片名": image["图片名"],
+        "图片内行号": image["图片内行号"],
+        "姓名": patient.get("姓名", ""),
+        "住院号": patient.get("住院号", ""),
+        "入院时间": patient.get("入院时间", ""),
+        "出院时间": patient.get("出院时间", ""),
+        "复核状态": review.get("状态", ""),
+    }
+
+
+def deduplicate_records_by_inpatient_no(
+    records: list[dict[str, Any]],
+) -> tuple[list[dict[str, Any]], list[dict[str, Any]], list[dict[str, Any]]]:
+    kept_by_no: dict[str, dict[str, Any]] = {}
+    kept_order: list[str] = []
+    duplicates: list[dict[str, Any]] = []
+    missing_inpatient_no_records: list[dict[str, Any]] = []
+    for record in records:
+        inpatient_no = str(record["患者信息"].get("住院号", "")).strip()
+        if not inpatient_no:
+            missing_inpatient_no_records.append(
+                {
+                    "剔除记录": summarize_record_for_duplicate(record),
+                    "规则": "住院号为空，未纳入合并总表和数据库",
+                }
+            )
+            continue
+        if inpatient_no not in kept_by_no:
+            kept_order.append(inpatient_no)
+        else:
+            previous = kept_by_no[inpatient_no]
+            duplicates.append(
+                {
+                    "住院号": inpatient_no,
+                    "保留记录": summarize_record_for_duplicate(record),
+                    "剔除记录": summarize_record_for_duplicate(previous),
+                    "规则": "住院号重复，后出现记录覆盖先出现记录",
+                }
+            )
+        kept_by_no[inpatient_no] = record
+    return [kept_by_no[inpatient_no] for inpatient_no in kept_order], duplicates, missing_inpatient_no_records
+
+
+def slim_department_summary(batch_name: str, item: dict[str, Any]) -> dict[str, Any]:
+    raw_tips = item.get("提示", [])
+    if isinstance(raw_tips, list):
+        tips = "；".join(str(tip) for tip in raw_tips)
+    else:
+        tips = str(raw_tips)
+    return {
+        "处理批次": batch_name,
+        "来源文件夹": item["来源文件夹"],
+        "大科室": item["大科室"],
+        "子科室": item["子科室"],
+        "图片数": item["图片数"],
+        "记录数": item["记录数"],
+        "拼接组数": item["拼接组数"],
+        "提示": tips,
+    }
+
+
+def write_csv(path: Path, rows: list[dict[str, Any]]) -> None:
+    if not rows:
+        return
+    path.parent.mkdir(parents=True, exist_ok=True)
+    with path.open("w", encoding="utf-8-sig", newline="") as file:
+        writer = csv.DictWriter(file, fieldnames=list(rows[0].keys()))
+        writer.writeheader()
+        writer.writerows(rows)
+
+
+def main() -> None:
+    ROOT.mkdir(exist_ok=True)
+    raw_records: list[dict[str, Any]] = []
+    batch_summaries: list[dict[str, Any]] = []
+
+    for result_dir in sorted(ROOT.rglob(f"*{RESULT_SUFFIX}")):
+        archive_path = result_dir / "患者列表_结构化.json"
+        review_path = result_dir / "复核报告.json"
+        if not archive_path.exists():
+            continue
+        batch_name = result_dir.name[: -len(RESULT_SUFFIX)]
+        archive = json.loads(archive_path.read_text(encoding="utf-8"))
+        review = json.loads(review_path.read_text(encoding="utf-8")) if review_path.exists() else {}
+        department_summaries = [
+            slim_department_summary(batch_name, item) for item in archive.get("科室汇总", [])
+        ]
+        summary = {
+            "处理批次": batch_name,
+            "结果目录": str(result_dir),
+            **archive.get("汇总", {}),
+            "科室归类": archive.get("科室归类", []),
+            "科室汇总": department_summaries,
+            "需人工复核记录数": len(review.get("需人工复核记录", [])),
+            "人工修正记录数": len(review.get("人工修正记录", [])),
+        }
+        batch_summaries.append(summary)
+
+        summary_dir = result_dir / INFO_DIR_NAME
+        write_json(summary_dir / "汇总.json", {**summary, "科室汇总": department_summaries})
+        write_csv(summary_dir / "科室汇总.csv", department_summaries)
+
+        for record in archive.get("患者记录", []):
+            enriched = {"处理批次": batch_name, **record}
+            raw_records.append(enriched)
+
+    all_records, duplicate_records, missing_inpatient_no_records = deduplicate_records_by_inpatient_no(raw_records)
+    flat_records = [flatten_record(record["处理批次"], record) for record in all_records]
+    kept_by_batch = Counter(record["处理批次"] for record in all_records)
+    dropped_by_batch = Counter(item["剔除记录"]["处理批次"] for item in duplicate_records)
+    missing_by_batch = Counter(item["剔除记录"]["处理批次"] for item in missing_inpatient_no_records)
+    batch_level_duplicate_count = sum(int(item.get("重复住院号剔除记录数", 0)) for item in batch_summaries)
+    batch_level_raw_count = sum(
+        int(item.get("去重前患者记录数", int(item.get("患者记录数", 0)) + int(item.get("重复住院号剔除记录数", 0))))
+        for item in batch_summaries
+    )
+
+    for summary in batch_summaries:
+        batch_name = summary["处理批次"]
+        original_count = int(summary.get("患者记录数", 0))
+        summary["原始患者记录数"] = original_count
+        summary["合并阶段重复住院号剔除记录数"] = int(dropped_by_batch.get(batch_name, 0))
+        summary["合并阶段缺少住院号剔除记录数"] = int(missing_by_batch.get(batch_name, 0))
+        summary["患者记录数"] = int(kept_by_batch.get(batch_name, 0))
+        write_json(Path(summary["结果目录"]) / INFO_DIR_NAME / "汇总.json", summary)
+
+    global_summary = {
+        "批次数": len(batch_summaries),
+        "总图片数": sum(int(item.get("图片数", 0)) for item in batch_summaries),
+        "批次内去重前患者记录数": batch_level_raw_count,
+        "批次内重复住院号剔除记录数": batch_level_duplicate_count,
+        "合并前患者记录数": len(raw_records),
+        "总患者记录数": len(all_records),
+        "合并阶段重复住院号剔除记录数": len(duplicate_records),
+        "合并阶段缺少住院号剔除记录数": len(missing_inpatient_no_records),
+        "重复住院号剔除记录数": batch_level_duplicate_count + len(duplicate_records),
+        "需人工复核记录数": sum(int(item.get("需人工复核记录数", 0)) for item in batch_summaries),
+        "人工修正记录数": sum(int(item.get("人工修正记录数", 0)) for item in batch_summaries),
+        "批次汇总": batch_summaries,
+    }
+    write_json(ROOT / INFO_DIR_NAME / "全局汇总.json", global_summary)
+    write_csv(ROOT / INFO_DIR_NAME / "批次汇总.csv", batch_summaries)
+    write_json(
+        ROOT / "合并_患者列表_结构化.json",
+        {
+            "汇总": global_summary,
+            "重复住院号剔除记录": duplicate_records,
+            "缺少住院号剔除记录": missing_inpatient_no_records,
+            "患者记录": all_records,
+        },
+    )
+    write_jsonl(ROOT / "合并_患者列表_记录.jsonl", all_records)
+    write_csv(ROOT / "合并_患者列表_记录.csv", flat_records)
+    write_json(ROOT / INFO_DIR_NAME / "重复住院号报告.json", duplicate_records)
+    write_json(ROOT / INFO_DIR_NAME / "缺少住院号报告.json", missing_inpatient_no_records)
+    print(json.dumps(global_summary, ensure_ascii=False, indent=2))
+
+
+if __name__ == "__main__":
+    main()
diff --git a/患者列表处理/数据处理工作区/05_同步PostgreSQL单表.py b/患者列表处理/数据处理工作区/05_同步PostgreSQL单表.py
new file mode 100644
index 0000000..277da04
--- /dev/null
+++ b/患者列表处理/数据处理工作区/05_同步PostgreSQL单表.py
@@ -0,0 +1,233 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+"""Sync consolidated HIS patient records into PostgreSQL single table.
+
+Database credentials are read from environment variables by default:
+HIS_DB_HOST, HIS_DB_PORT, HIS_DB_NAME, HIS_DB_USER, HIS_DB_PASSWORD.
+"""
+
+from __future__ import annotations
+
+import argparse
+import csv
+import json
+import os
+import subprocess
+import tempfile
+from pathlib import Path
+from typing import Any
+
+
+DEFAULT_RESULT_JSON = Path("数据处理结果区/合并_患者列表_结构化.json")
+DEFAULT_SCHEMA_SQL = Path("数据处理工作区/06_PostgreSQL建表结构.sql")
+
+TABLE_NAME = '"Patient_Lists"'
+
+PATIENT_FIELDS = [
+    "batch_name",
+    "major_department",
+    "sub_department",
+    "source_folder",
+    "image_path",
+    "image_name",
+    "image_row_no",
+    "patient_name",
+    "gender",
+    "age",
+    "inpatient_no",
+    "diagnosis",
+    "admission_time",
+    "last_write_time",
+    "hospital_days",
+    "discharge_time",
+    "postoperative_days",
+    "review_status",
+    "review_notes",
+    "manual_corrected",
+]
+
+
+def scalar(value: Any) -> str:
+    if value is None:
+        return ""
+    if isinstance(value, bool):
+        return "true" if value else "false"
+    return str(value)
+
+
+def clean_integer(value: Any) -> tuple[str, str]:
+    text = scalar(value).strip()
+    if not text:
+        return "", ""
+    if text.isdigit():
+        return text, ""
+    return "", f"住院天数非数字，原识别值: {text}"
+
+
+def review_note_items(review: dict[str, Any]) -> list[str]:
+    tips = review.get("提示", [])
+    if isinstance(tips, list):
+        note_items = [str(tip) for tip in tips if str(tip) and str(tip) != "缺少出院时间"]
+    else:
+        note_items = [str(tips)] if str(tips) else []
+    if review.get("人工备注"):
+        note_items.append(f"人工备注: {review['人工备注']}")
+    return note_items
+
+
+def psql_quote(path: Path) -> str:
+    return "'" + str(path.resolve()).replace("'", "''") + "'"
+
+
+def flatten_patient_record(record: dict[str, Any]) -> dict[str, Any]:
+    patient = record["患者信息"]
+    image = record["图片信息"]
+    review = record["复核"]
+    note_items = review_note_items(review)
+    hospital_days, hospital_days_note = clean_integer(patient["住院天数"])
+    if hospital_days_note:
+        note_items.append(hospital_days_note)
+    review_notes = "；".join(note_items)
+    return {
+        "batch_name": record["处理批次"],
+        "major_department": record["大科室"],
+        "sub_department": record["子科室"],
+        "source_folder": record["来源文件夹"],
+        "image_path": image["图片路径"],
+        "image_name": image["图片名"],
+        "image_row_no": image["图片内行号"],
+        "patient_name": patient["姓名"],
+        "gender": patient["性别"],
+        "age": patient["年龄"],
+        "inpatient_no": patient["住院号"],
+        "diagnosis": patient["诊断"],
+        "admission_time": patient["入院时间"],
+        "last_write_time": patient["最后书写时间"],
+        "hospital_days": hospital_days,
+        "discharge_time": patient["出院时间"],
+        "postoperative_days": patient["手术后天数"],
+        "review_status": review["状态"],
+        "review_notes": review_notes,
+        "manual_corrected": bool(review.get("人工修正")),
+    }
+
+
+def prepare_unique_rows(rows: list[dict[str, Any]]) -> tuple[list[dict[str, Any]], list[str], list[str]]:
+    kept_by_no: dict[str, dict[str, Any]] = {}
+    kept_order: list[str] = []
+    replacements: list[str] = []
+    invalid: list[str] = []
+    for row in rows:
+        inpatient_no = str(row.get("inpatient_no", "")).strip()
+        location = f"{row['batch_name']} / {row['image_path']} / 第{row['image_row_no']}行"
+        if not inpatient_no:
+            invalid.append(f"{location}: 空")
+            continue
+        if inpatient_no not in kept_by_no:
+            kept_order.append(inpatient_no)
+        else:
+            previous = kept_by_no[inpatient_no]
+            replacements.append(
+                f"{inpatient_no}: {previous['batch_name']} / {previous['image_path']} / 第{previous['image_row_no']}行 -> {location}"
+            )
+        kept_by_no[inpatient_no] = row
+    return [kept_by_no[inpatient_no] for inpatient_no in kept_order], replacements, invalid
+
+
+def load_rows(result_json: Path) -> tuple[list[dict[str, Any]], list[str], list[str]]:
+    merged = json.loads(result_json.read_text(encoding="utf-8"))
+    rows = [flatten_patient_record(record) for record in merged.get("患者记录", [])]
+    return prepare_unique_rows(rows)
+
+
+def write_temp_csv(rows: list[dict[str, Any]]) -> Path:
+    temp_file = tempfile.NamedTemporaryFile(
+        mode="w",
+        encoding="utf-8",
+        newline="",
+        suffix="_his_patient_lists.csv",
+        delete=False,
+    )
+    with temp_file:
+        writer = csv.DictWriter(temp_file, fieldnames=PATIENT_FIELDS)
+        writer.writeheader()
+        for row in rows:
+            writer.writerow({field: scalar(row.get(field, "")) for field in PATIENT_FIELDS})
+    return Path(temp_file.name)
+
+
+def sync_to_postgres(args: argparse.Namespace, csv_path: Path, row_count: int) -> None:
+    sql = f"""\\set ON_ERROR_STOP on
+\\i {psql_quote(Path(args.schema))}
+TRUNCATE TABLE {TABLE_NAME} RESTART IDENTITY;
+\\copy {TABLE_NAME}({','.join(PATIENT_FIELDS)}) FROM {psql_quote(csv_path)} WITH (FORMAT csv, HEADER true, NULL '')
+"""
+    env = os.environ.copy()
+    password = args.password or env.get("HIS_DB_PASSWORD")
+    if password:
+        env["PGPASSWORD"] = password
+    command = [
+        "psql",
+        "-h",
+        args.host,
+        "-p",
+        str(args.port),
+        "-U",
+        args.user,
+        "-d",
+        args.dbname,
+        "-v",
+        "ON_ERROR_STOP=1",
+    ]
+    completed = subprocess.run(command, input=sql, text=True, env=env, capture_output=True)
+    print(completed.stdout, end="")
+    if completed.returncode != 0:
+        print(completed.stderr, end="")
+        completed.check_returncode()
+    print(json.dumps({"synced_table": "Patient_Lists", "records": row_count}, ensure_ascii=False))
+
+
+def parse_args() -> argparse.Namespace:
+    parser = argparse.ArgumentParser(description=__doc__)
+    parser.add_argument("--input", default=str(DEFAULT_RESULT_JSON), help="合并后的患者记录 JSON")
+    parser.add_argument("--schema", default=str(DEFAULT_SCHEMA_SQL), help="单表建表 SQL")
+    parser.add_argument("--host", default=os.getenv("HIS_DB_HOST", "DB_HOST"))
+    parser.add_argument("--port", default=os.getenv("HIS_DB_PORT", "5432"))
+    parser.add_argument("--dbname", default=os.getenv("HIS_DB_NAME", "DB_NAME"))
+    parser.add_argument("--user", default=os.getenv("HIS_DB_USER", "DB_USER"))
+    parser.add_argument("--password", default="", help="数据库密码；建议改用 HIS_DB_PASSWORD 环境变量")
+    return parser.parse_args()
+
+
+def main() -> None:
+    args = parse_args()
+    rows, replacements, invalid = load_rows(Path(args.input))
+    csv_path = write_temp_csv(rows)
+    try:
+        sync_to_postgres(args, csv_path, len(rows))
+        if invalid:
+            print(
+                json.dumps(
+                    {
+                        "skipped_empty_inpatient_no": len(invalid),
+                        "empty_examples": invalid[:20],
+                    },
+                    ensure_ascii=False,
+                )
+            )
+        if replacements:
+            print(
+                json.dumps(
+                    {
+                        "deduplicated_by_inpatient_no": len(replacements),
+                        "replacement_examples": replacements[:20],
+                    },
+                    ensure_ascii=False,
+                )
+            )
+    finally:
+        csv_path.unlink(missing_ok=True)
+
+
+if __name__ == "__main__":
+    main()
diff --git a/患者列表处理/数据处理工作区/06_PostgreSQL建表结构.sql b/患者列表处理/数据处理工作区/06_PostgreSQL建表结构.sql
new file mode 100644
index 0000000..967b8a9
--- /dev/null
+++ b/患者列表处理/数据处理工作区/06_PostgreSQL建表结构.sql
@@ -0,0 +1,84 @@
+DROP TABLE IF EXISTS patient_records CASCADE;
+
+CREATE TABLE IF NOT EXISTS "Patient_Lists" (
+  record_id bigserial PRIMARY KEY,
+  batch_name text NOT NULL,
+  major_department text NOT NULL,
+  sub_department text NOT NULL,
+  source_folder text NOT NULL,
+  image_path text NOT NULL,
+  image_name text NOT NULL,
+  image_row_no integer NOT NULL,
+  patient_name text NOT NULL,
+  gender text,
+  age text,
+  inpatient_no text NOT NULL,
+  diagnosis text,
+  admission_time text,
+  last_write_time text,
+  hospital_days integer,
+  discharge_time text,
+  postoperative_days text,
+  review_status text NOT NULL,
+  review_notes text,
+  manual_corrected boolean NOT NULL DEFAULT false,
+  imported_at timestamptz NOT NULL DEFAULT now(),
+  audit_result text,
+  audit_ai_feedback text,
+  audit_manual_feedback text,
+  audit_machine_verdict text,
+  audit_source text,
+  audit_checked_by text,
+  audit_checked_at timestamptz,
+  CONSTRAINT uq_patient_lists_inpatient_no UNIQUE (inpatient_no),
+  CONSTRAINT ck_patient_lists_inpatient_no_present CHECK (btrim(inpatient_no) <> ''),
+  CONSTRAINT ck_patient_lists_admission_before_discharge CHECK (
+    COALESCE(discharge_time, '') = ''
+    OR COALESCE(admission_time, '') = ''
+    OR admission_time !~ '^\d{4}-\d{2}-\d{2} \d{2}:\d{2}:\d{2}$'
+    OR discharge_time !~ '^\d{4}-\d{2}-\d{2} \d{2}:\d{2}:\d{2}$'
+    OR admission_time <= discharge_time
+    OR review_status = '需人工复核'
+  )
+);
+
+CREATE INDEX IF NOT EXISTS idx_patient_lists_batch_name ON "Patient_Lists"(batch_name);
+CREATE INDEX IF NOT EXISTS idx_patient_lists_department ON "Patient_Lists"(major_department, sub_department);
+CREATE INDEX IF NOT EXISTS idx_patient_lists_source_folder ON "Patient_Lists"(source_folder);
+CREATE INDEX IF NOT EXISTS idx_patient_lists_patient_name ON "Patient_Lists"(patient_name);
+CREATE INDEX IF NOT EXISTS idx_patient_lists_review_status ON "Patient_Lists"(review_status);
+CREATE INDEX IF NOT EXISTS idx_patient_lists_audit_result ON "Patient_Lists"(audit_result);
+
+COMMENT ON TABLE "Patient_Lists" IS 'HIS患者列表图片OCR归档后的正式患者记录单表';
+COMMENT ON COLUMN "Patient_Lists".record_id IS '患者记录主键，数据库自动生成';
+COMMENT ON COLUMN "Patient_Lists".batch_name IS '处理批次名称，通常为原始图片集群文件夹名';
+COMMENT ON COLUMN "Patient_Lists".major_department IS '归类后的大科室';
+COMMENT ON COLUMN "Patient_Lists".sub_department IS '归类后的子科室';
+COMMENT ON COLUMN "Patient_Lists".source_folder IS '原始第一层科室文件夹名';
+COMMENT ON COLUMN "Patient_Lists".image_path IS '原始患者列表图片路径，用于定位数据来源';
+COMMENT ON COLUMN "Patient_Lists".image_name IS '原始患者列表图片文件名';
+COMMENT ON COLUMN "Patient_Lists".image_row_no IS '患者记录在原始图片内的行号';
+COMMENT ON COLUMN "Patient_Lists".patient_name IS '患者姓名';
+COMMENT ON COLUMN "Patient_Lists".gender IS '患者性别';
+COMMENT ON COLUMN "Patient_Lists".age IS '患者年龄，保留原始岁数字符串';
+COMMENT ON COLUMN "Patient_Lists".inpatient_no IS '住院号，不能为空且全库唯一；不强制校验格式';
+COMMENT ON COLUMN "Patient_Lists".diagnosis IS '诊断，可能为空';
+COMMENT ON COLUMN "Patient_Lists".admission_time IS '入院时间，按OCR/清洗后的文本保存';
+COMMENT ON COLUMN "Patient_Lists".last_write_time IS '最后书写时间，按OCR/清洗后的文本保存';
+COMMENT ON COLUMN "Patient_Lists".hospital_days IS '住院天数';
+COMMENT ON COLUMN "Patient_Lists".discharge_time IS '出院时间，按OCR/清洗后的文本保存';
+COMMENT ON COLUMN "Patient_Lists".postoperative_days IS '手术后天数，可能为空';
+COMMENT ON COLUMN "Patient_Lists".review_status IS '自动复核或人工复核状态';
+COMMENT ON COLUMN "Patient_Lists".review_notes IS '复核提示信息';
+COMMENT ON COLUMN "Patient_Lists".manual_corrected IS '是否经过人工修正';
+COMMENT ON COLUMN "Patient_Lists".imported_at IS '导入数据库的时间';
+COMMENT ON COLUMN "Patient_Lists".audit_result IS '抽查人工核验结果';
+COMMENT ON COLUMN "Patient_Lists".audit_ai_feedback IS '抽查项AI反馈原文';
+COMMENT ON COLUMN "Patient_Lists".audit_manual_feedback IS '抽查项人工反馈';
+COMMENT ON COLUMN "Patient_Lists".audit_machine_verdict IS '抽查机器核验判断';
+COMMENT ON COLUMN "Patient_Lists".audit_source IS '抽查来源';
+COMMENT ON COLUMN "Patient_Lists".audit_checked_by IS '抽查操作用户';
+COMMENT ON COLUMN "Patient_Lists".audit_checked_at IS '抽查结果保存时间';
+COMMENT ON CONSTRAINT uq_patient_lists_inpatient_no ON "Patient_Lists" IS '保证同一个住院号只归档一条患者记录';
+COMMENT ON CONSTRAINT ck_patient_lists_inpatient_no_present ON "Patient_Lists" IS '住院号不能为空；格式不做强制校验';
+COMMENT ON CONSTRAINT ck_patient_lists_admission_before_discharge ON "Patient_Lists" IS '出院时间为空时放行；入院时间晚于出院时间时必须标记为需人工复核';
diff --git a/患者列表处理/数据处理工作区/07_处理程序说明.md b/患者列表处理/数据处理工作区/07_处理程序说明.md
new file mode 100644
index 0000000..026aaa9
--- /dev/null
+++ b/患者列表处理/数据处理工作区/07_处理程序说明.md
@@ -0,0 +1,72 @@
+# 患者列表归档处理程序说明
+
+## 工作区文件
+
+1. `01_科室分类规则.json`：大科室、子科室、文件夹别名归类规则。
+2. `02_患者列表OCR归档.py`：单个图片集群的 OCR、字段清洗、去重、复核报告生成程序。
+3. `03_人工复核修正.json`：少量无法靠规则自动判断的人工修正项，可能包含患者信息，不提交到 Git。
+4. `04_合并批次结果.py`：合并所有 `*-列表归档结果`，生成全局 JSON/CSV 和 `信息记录`。
+5. `05_同步PostgreSQL单表.py`：把合并结果同步到 PostgreSQL 的 `"Patient_Lists"` 单表。
+6. `06_PostgreSQL建表结构.sql`：数据库单表结构、索引、表注释、字段注释。
+7. `07_处理程序说明.md`：当前说明。
+8. `08_PostgreSQL调整Patient_Lists列顺序.sql`：保留数据的前提下，把现有 `"Patient_Lists"` 物理列顺序调整为当前字段顺序。
+9. `09_PostgreSQL住院号非空唯一约束.sql`：删除住院号为空的记录，重复住院号保留最新一条，并让 PostgreSQL 强制住院号不能为空且唯一。
+
+## 智能体处理顺序
+
+1. 检查待处理批次目录和图片数量。
+2. 必要时更新 `01_科室分类规则.json` 的别名。
+3. 运行 `02_患者列表OCR归档.py` 处理一个批次。
+4. 查看 `复核报告.json` 和 `重复住院号报告.json`。
+5. 若有需人工复核项，优先补自动清洗规则；仍不能自动判断时，写入 `03_人工复核修正.json`。
+6. 用 `--rebuild-from-cache` 重建该批次结果，确认复核项和重复住院号符合预期。
+7. 运行 `04_合并批次结果.py` 更新全局 `信息记录`。
+8. 运行 `05_同步PostgreSQL单表.py` 同步到 PostgreSQL。
+9. 导入 PostgreSQL 后查询记录数、需复核记录数、重复住院号组数。
+
+## 人工复核网页端
+
+- 代码目录：`人工复核网页端/`。
+- 启动方式：`cd 人工复核网页端 && docker compose up -d --build`。
+- 登录配置从项目根目录 `.env` 读取：`REVIEW_APP_USERNAME`、`REVIEW_APP_PASSWORD`、`REVIEW_APP_SECRET_KEY`、`REVIEW_APP_PORT`。
+- 网页端读取各批次 `复核报告.json`，按原始图片路径与图片内行号裁剪截图；住院号不能为空且唯一，格式是否规范交由人工复核时判断。
+- 保存修订后写入 `数据处理工作区/03_人工复核修正.json`，并尽量同步 PostgreSQL；若 PostgreSQL 暂不可用，会保留待同步状态，之后可在网页端点击“提交待同步”。
+- 出院时间默认允许为空；确有出院时间时，网页端和处理程序会校验时间格式以及是否早于入院时间。
+- 页面包含概览、复核、抽查、抽查查看、设置区域；抽查功能可随机抽取患者行，调用 Kimi 多模态模型二次核验，并把 AI 反馈、机器判断、人工抽查结论写入 PostgreSQL。
+
+## 识别策略
+
+- OCR 接口：腾讯云 `RecognizeTableAccurateOCR` 表格识别 V3。
+- 新批次拼接图默认给每张原图上下各加 `24px` 白色边界，减少贴边表格行被裁掉或和相邻图片粘连的概率。
+- 正式处理优先用 6 张拼接；接口超时、失败或行数偏少时，程序会自动降到 4/3/2/单张。
+- 行数校验：按图片高度推断应有行数，返回行数偏少即判定该拼接档位不可靠。
+- 漏行不一定只发生在最上方或最下方两行；长拼接图中间也可能漏行，白边只能缓解贴边问题。
+- 同一科室内会记忆稳定回退档位；跨科室不共享该记忆，避免重建历史缓存时误用。
+- 只用缓存重建时，若前序回退策略需要的缓存不完整，会先尝试该组主缓存，避免因缓存形态不同造成少行。
+- 字段顺序固定为：姓名、性别、年龄、住院号、诊断、入院时间、最后书写时间、住院天数、出院时间、手术后天数。
+- 字段清洗包含日期规范化、住院号常见误识别修正、长诊断吞入入院时间、空时间字段导致列左移、`后XX天` 错位等。
+
+## 住院号核验规则
+
+- `住院号` 是唯一强制校验条件：不能为空且全库唯一。
+- 单批次、全局合并和 PostgreSQL 同步都会剔除住院号为空的记录；重复住院号按后出现记录覆盖先出现记录；格式异常但非空的住院号保留。
+- 网页端保存时优先按图片路径和图片内行号定位数据库记录；如果当前住院号已存在于另一条记录，会用当前修订覆盖既有住院号记录，并删除当前重复图片记录。
+- 数据库约束允许 `出院时间` 为空；若入院时间和出院时间均为标准格式且入院时间晚于出院时间，该记录必须标记为 `需人工复核`。
+
+## PostgreSQL 单表字段
+
+`"Patient_Lists"` 只保留正式查看和追溯需要的字段：
+
+- 数据定位：`batch_name`、`source_folder`、`image_path`、`image_name`、`image_row_no`
+- 科室信息：`major_department`、`sub_department`
+- 患者信息：姓名、性别、年龄、住院号、诊断、入院时间、最后书写时间、住院天数、出院时间、手术后天数
+- 复核信息：`review_status`、`review_notes`、`manual_corrected`
+- 抽查信息：`audit_result`、`audit_ai_feedback`、`audit_manual_feedback`、`audit_machine_verdict`、`audit_source`、`audit_checked_by`、`audit_checked_at`
+
+复核状态含义：
+
+- `自动复核通过`：OCR 结果经过程序规则校验后没有发现明显异常，`manual_corrected=false`。
+- `人工复核通过`：该行命中了 `03_人工复核修正.json` 中的人工修正项，修正后校验通过，`manual_corrected=true`。
+- `AI修改-待确认`：网页端 Kimi 辅助生成的草稿状态，不会被待同步流程写入 PostgreSQL，需人工保存确认。
+
+OCR 请求号、OCR 缓存路径、拼接图路径、拼接组号等中间态字段不进入数据库正式表。
diff --git a/患者列表处理/数据处理工作区/08_PostgreSQL调整Patient_Lists列顺序.sql b/患者列表处理/数据处理工作区/08_PostgreSQL调整Patient_Lists列顺序.sql
new file mode 100644
index 0000000..84d6cde
--- /dev/null
+++ b/患者列表处理/数据处理工作区/08_PostgreSQL调整Patient_Lists列顺序.sql
@@ -0,0 +1,168 @@
+-- 保留现有数据，重建 "Patient_Lists" 的物理列顺序。
+-- 只调整列顺序，不交换 last_write_time/discharge_time 的值。
+-- 执行前建议先确认已备份数据库。
+
+BEGIN;
+
+LOCK TABLE "Patient_Lists" IN ACCESS EXCLUSIVE MODE;
+
+DROP TABLE IF EXISTS "Patient_Lists__reordered";
+
+CREATE TABLE "Patient_Lists__reordered" (
+  record_id bigint NOT NULL,
+  batch_name text NOT NULL,
+  major_department text NOT NULL,
+  sub_department text NOT NULL,
+  source_folder text NOT NULL,
+  image_path text NOT NULL,
+  image_name text NOT NULL,
+  image_row_no integer NOT NULL,
+  patient_name text NOT NULL,
+  gender text,
+  age text,
+  inpatient_no text NOT NULL,
+  diagnosis text,
+  admission_time text,
+  last_write_time text,
+  hospital_days integer,
+  discharge_time text,
+  postoperative_days text,
+  review_status text NOT NULL,
+  review_notes text,
+  manual_corrected boolean NOT NULL DEFAULT false,
+  imported_at timestamptz NOT NULL DEFAULT now(),
+  audit_result text,
+  audit_ai_feedback text,
+  audit_manual_feedback text,
+  audit_machine_verdict text,
+  audit_source text,
+  audit_checked_by text,
+  audit_checked_at timestamptz,
+  CONSTRAINT ck_patient_lists_inpatient_no_present CHECK (btrim(inpatient_no) <> ''),
+  CONSTRAINT ck_patient_lists_admission_before_discharge CHECK (
+    COALESCE(discharge_time, '') = ''
+    OR COALESCE(admission_time, '') = ''
+    OR admission_time !~ '^\d{4}-\d{2}-\d{2} \d{2}:\d{2}:\d{2}$'
+    OR discharge_time !~ '^\d{4}-\d{2}-\d{2} \d{2}:\d{2}:\d{2}$'
+    OR admission_time <= discharge_time
+    OR review_status = '需人工复核'
+  )
+);
+
+INSERT INTO "Patient_Lists__reordered" (
+  record_id,
+  batch_name,
+  major_department,
+  sub_department,
+  source_folder,
+  image_path,
+  image_name,
+  image_row_no,
+  patient_name,
+  gender,
+  age,
+  inpatient_no,
+  diagnosis,
+  admission_time,
+  last_write_time,
+  hospital_days,
+  discharge_time,
+  postoperative_days,
+  review_status,
+  review_notes,
+  manual_corrected,
+  imported_at,
+  audit_result,
+  audit_ai_feedback,
+  audit_manual_feedback,
+  audit_machine_verdict,
+  audit_source,
+  audit_checked_by,
+  audit_checked_at
+)
+SELECT
+  record_id,
+  batch_name,
+  major_department,
+  sub_department,
+  source_folder,
+  image_path,
+  image_name,
+  image_row_no,
+  patient_name,
+  gender,
+  age,
+  inpatient_no,
+  diagnosis,
+  admission_time,
+  last_write_time,
+  hospital_days,
+  discharge_time,
+  postoperative_days,
+  review_status,
+  review_notes,
+  manual_corrected,
+  imported_at,
+  audit_result,
+  audit_ai_feedback,
+  audit_manual_feedback,
+  audit_machine_verdict,
+  audit_source,
+  audit_checked_by,
+  audit_checked_at
+FROM "Patient_Lists";
+
+DROP TABLE "Patient_Lists";
+ALTER TABLE "Patient_Lists__reordered" RENAME TO "Patient_Lists";
+
+DROP SEQUENCE IF EXISTS "Patient_Lists_record_id_seq";
+CREATE SEQUENCE "Patient_Lists_record_id_seq";
+ALTER SEQUENCE "Patient_Lists_record_id_seq" OWNED BY "Patient_Lists".record_id;
+ALTER TABLE "Patient_Lists" ALTER COLUMN record_id SET DEFAULT nextval('"Patient_Lists_record_id_seq"'::regclass);
+SELECT setval('"Patient_Lists_record_id_seq"', COALESCE((SELECT max(record_id) FROM "Patient_Lists"), 1), true);
+
+ALTER TABLE "Patient_Lists" ADD CONSTRAINT "Patient_Lists_pkey" PRIMARY KEY (record_id);
+ALTER TABLE "Patient_Lists" ADD CONSTRAINT uq_patient_lists_inpatient_no UNIQUE (inpatient_no);
+
+CREATE INDEX idx_patient_lists_batch_name ON "Patient_Lists"(batch_name);
+CREATE INDEX idx_patient_lists_department ON "Patient_Lists"(major_department, sub_department);
+CREATE INDEX idx_patient_lists_source_folder ON "Patient_Lists"(source_folder);
+CREATE INDEX idx_patient_lists_patient_name ON "Patient_Lists"(patient_name);
+CREATE INDEX idx_patient_lists_review_status ON "Patient_Lists"(review_status);
+CREATE INDEX idx_patient_lists_audit_result ON "Patient_Lists"(audit_result);
+
+COMMENT ON TABLE "Patient_Lists" IS 'HIS患者列表图片OCR归档后的正式患者记录单表';
+COMMENT ON COLUMN "Patient_Lists".record_id IS '患者记录主键，数据库自动生成';
+COMMENT ON COLUMN "Patient_Lists".batch_name IS '处理批次名称，通常为原始图片集群文件夹名';
+COMMENT ON COLUMN "Patient_Lists".major_department IS '归类后的大科室';
+COMMENT ON COLUMN "Patient_Lists".sub_department IS '归类后的子科室';
+COMMENT ON COLUMN "Patient_Lists".source_folder IS '原始第一层科室文件夹名';
+COMMENT ON COLUMN "Patient_Lists".image_path IS '原始患者列表图片路径，用于定位数据来源';
+COMMENT ON COLUMN "Patient_Lists".image_name IS '原始患者列表图片文件名';
+COMMENT ON COLUMN "Patient_Lists".image_row_no IS '患者记录在原始图片内的行号';
+COMMENT ON COLUMN "Patient_Lists".patient_name IS '患者姓名';
+COMMENT ON COLUMN "Patient_Lists".gender IS '患者性别';
+COMMENT ON COLUMN "Patient_Lists".age IS '患者年龄，保留原始岁数字符串';
+COMMENT ON COLUMN "Patient_Lists".inpatient_no IS '住院号，不能为空且全库唯一；不强制校验格式';
+COMMENT ON COLUMN "Patient_Lists".diagnosis IS '诊断，可能为空';
+COMMENT ON COLUMN "Patient_Lists".admission_time IS '入院时间，按OCR/清洗后的文本保存';
+COMMENT ON COLUMN "Patient_Lists".last_write_time IS '最后书写时间，按OCR/清洗后的文本保存';
+COMMENT ON COLUMN "Patient_Lists".hospital_days IS '住院天数';
+COMMENT ON COLUMN "Patient_Lists".discharge_time IS '出院时间，按OCR/清洗后的文本保存';
+COMMENT ON COLUMN "Patient_Lists".postoperative_days IS '手术后天数，可能为空，格式通常为“后X天”';
+COMMENT ON COLUMN "Patient_Lists".review_status IS '自动复核或人工复核状态';
+COMMENT ON COLUMN "Patient_Lists".review_notes IS '复核提示信息';
+COMMENT ON COLUMN "Patient_Lists".manual_corrected IS '是否经过人工修正';
+COMMENT ON COLUMN "Patient_Lists".imported_at IS '导入数据库或人工同步更新的时间';
+COMMENT ON COLUMN "Patient_Lists".audit_result IS '抽查人工核验结果';
+COMMENT ON COLUMN "Patient_Lists".audit_ai_feedback IS '抽查项AI反馈原文';
+COMMENT ON COLUMN "Patient_Lists".audit_manual_feedback IS '抽查项人工反馈';
+COMMENT ON COLUMN "Patient_Lists".audit_machine_verdict IS '抽查机器核验判断';
+COMMENT ON COLUMN "Patient_Lists".audit_source IS '抽查来源';
+COMMENT ON COLUMN "Patient_Lists".audit_checked_by IS '抽查操作用户';
+COMMENT ON COLUMN "Patient_Lists".audit_checked_at IS '抽查结果保存时间';
+COMMENT ON CONSTRAINT uq_patient_lists_inpatient_no ON "Patient_Lists" IS '保证同一个住院号只归档一条患者记录';
+COMMENT ON CONSTRAINT ck_patient_lists_inpatient_no_present ON "Patient_Lists" IS '住院号不能为空；格式不做强制校验';
+COMMENT ON CONSTRAINT ck_patient_lists_admission_before_discharge ON "Patient_Lists" IS '出院时间为空时放行；入院时间晚于出院时间时必须标记为需人工复核';
+
+COMMIT;
diff --git a/患者列表处理/数据处理工作区/09_PostgreSQL住院号非空唯一约束.sql b/患者列表处理/数据处理工作区/09_PostgreSQL住院号非空唯一约束.sql
new file mode 100644
index 0000000..681c07f
--- /dev/null
+++ b/患者列表处理/数据处理工作区/09_PostgreSQL住院号非空唯一约束.sql
@@ -0,0 +1,102 @@
+-- 住院号为唯一定位条件：数据库强制“不能为空且唯一”。
+-- 不校验 ZY 位数格式；空住院号记录会先删除，重复住院号保留 id 最大/最新的一条。
+
+DELETE FROM "Patient_Lists"
+WHERE inpatient_no IS NULL OR btrim(inpatient_no) = '';
+
+DELETE FROM "Patient_Lists" p
+USING (
+  SELECT record_id
+  FROM (
+    SELECT record_id, row_number() OVER (PARTITION BY inpatient_no ORDER BY record_id DESC) AS rn
+    FROM "Patient_Lists"
+  ) ranked
+  WHERE rn > 1
+) duplicate_rows
+WHERE p.record_id = duplicate_rows.record_id;
+
+ALTER TABLE "Patient_Lists"
+  DROP CONSTRAINT IF EXISTS ck_patient_lists_inpatient_no_format;
+
+ALTER TABLE "Patient_Lists"
+  DROP CONSTRAINT IF EXISTS uq_patient_lists_inpatient_no;
+
+DROP INDEX IF EXISTS uq_patient_lists_inpatient_no;
+
+ALTER TABLE "Patient_Lists"
+  DROP CONSTRAINT IF EXISTS ck_patient_lists_inpatient_no_present;
+
+ALTER TABLE "Patient_Lists"
+  DROP CONSTRAINT IF EXISTS ck_patient_lists_inpatient_no_required;
+
+ALTER TABLE "Patient_Lists"
+  ALTER COLUMN inpatient_no SET NOT NULL;
+
+ALTER TABLE "Patient_Lists"
+  ADD CONSTRAINT ck_patient_lists_inpatient_no_present
+  CHECK (btrim(inpatient_no) <> '');
+
+ALTER TABLE "Patient_Lists"
+  ADD CONSTRAINT uq_patient_lists_inpatient_no UNIQUE (inpatient_no);
+
+COMMENT ON COLUMN "Patient_Lists".inpatient_no
+  IS '住院号，不能为空且全库唯一；不强制校验格式';
+
+COMMENT ON CONSTRAINT ck_patient_lists_inpatient_no_present ON "Patient_Lists"
+  IS '住院号不能为空；格式不做强制校验';
+
+COMMENT ON CONSTRAINT uq_patient_lists_inpatient_no ON "Patient_Lists"
+  IS '保证同一个住院号只归档一条患者记录';
+
+DELETE FROM "Patient_FrontPages"
+WHERE inpatient_no IS NULL OR btrim(inpatient_no) = '';
+
+DELETE FROM "Patient_FrontPages" p
+USING (
+  SELECT id
+  FROM (
+    SELECT id, row_number() OVER (PARTITION BY inpatient_no ORDER BY id DESC) AS rn
+    FROM "Patient_FrontPages"
+  ) ranked
+  WHERE rn > 1
+) duplicate_rows
+WHERE p.id = duplicate_rows.id;
+
+ALTER TABLE "Patient_FrontPages"
+  DROP CONSTRAINT IF EXISTS "ck_Patient_FrontPages_inpatient_no_format";
+
+ALTER TABLE "Patient_FrontPages"
+  DROP CONSTRAINT IF EXISTS ck_patient_frontpages_inpatient_no_format;
+
+ALTER TABLE "Patient_FrontPages"
+  DROP CONSTRAINT IF EXISTS uq_patient_frontpages_inpatient_no;
+
+DROP INDEX IF EXISTS uq_patient_frontpages_inpatient_no;
+
+ALTER TABLE "Patient_FrontPages"
+  DROP CONSTRAINT IF EXISTS ck_patient_frontpages_inpatient_no_present;
+
+ALTER TABLE "Patient_FrontPages"
+  DROP CONSTRAINT IF EXISTS "ck_Patient_FrontPages_inpatient_no_required";
+
+ALTER TABLE "Patient_FrontPages"
+  DROP CONSTRAINT IF EXISTS ck_patient_frontpages_inpatient_no_required;
+
+ALTER TABLE "Patient_FrontPages"
+  ALTER COLUMN inpatient_no SET NOT NULL;
+
+ALTER TABLE "Patient_FrontPages"
+  ADD CONSTRAINT ck_patient_frontpages_inpatient_no_present
+  CHECK (btrim(inpatient_no) <> '');
+
+ALTER TABLE "Patient_FrontPages"
+  ADD CONSTRAINT uq_patient_frontpages_inpatient_no UNIQUE (inpatient_no);
+
+COMMENT ON COLUMN "Patient_FrontPages".inpatient_no
+  IS '住院号，不能为空且全库唯一；不强制校验格式';
+
+COMMENT ON CONSTRAINT ck_patient_frontpages_inpatient_no_present ON "Patient_FrontPages"
+  IS '住院号不能为空；格式不做强制校验';
+
+COMMENT ON CONSTRAINT uq_patient_frontpages_inpatient_no ON "Patient_FrontPages"
+  IS '保证同一个住院号只保留一条首页记录';
diff --git a/患者首页处理/.gitignore b/患者首页处理/.gitignore
new file mode 100644
index 0000000..4c55e29
--- /dev/null
+++ b/患者首页处理/.gitignore
@@ -0,0 +1,22 @@
+# 原始 PDF 与处理结果不进入 Gitea
+待处理-患者首页PDF/
+已处理-患者首页PDF/
+数据处理结果区/
+
+# 旧参考程序不作为本次工作流代码提交
+参考-过去很烂的程序/
+参考-病案首页图片/
+
+# 本地工作流不进入 Gitea
+工作流_本地使用版.md
+工作流_本地使用版*.zip
+
+# 本地配置、缓存和临时文件
+.env
+.env.*
+!.env.example
+数据可视化网页端/review_settings.local.json
+__pycache__/
+*.pyc
+*.tmp
+*.log
diff --git a/患者首页处理/README.md b/患者首页处理/README.md
new file mode 100644
index 0000000..ca289ed
--- /dev/null
+++ b/患者首页处理/README.md
@@ -0,0 +1,259 @@
+# 患者首页 PDF 本地处理工作流
+
+这个目录用于批量处理“患者首页 PDF”，把每份 PDF 抽取成结构化记录，并输出到 `数据处理结果区/`。
+
+## 目录约定
+
+- `待处理-患者首页PDF/`：放入等待处理的 PDF，文件名建议保留首页病案号或住院号。
+- `数据处理工作区/`：存放处理脚本和后续工作流工具。
+- `数据处理结果区/`：脚本输出 CSV、JSONL、单份 JSON、抽取文本、PDF 图片对照页和复核清单。
+- `数据可视化网页端/`：Docker 化复核网页端，包含概览、复核、抽查、抽查一览、设置；设置内统一管理用户与权限、数据库状态和目录信息。
+- `已处理-患者首页PDF/`：人工确认后，可把已处理 PDF 移到这里归档。
+
+## 工作区文件
+
+- `数据处理工作区/01_配置规则/01_科室分类规则.json`：子科室到大科室的分类规则。
+- `数据处理工作区/02_解析入库/02_患者首页PDF解析与入库.py`：解析 PDF，生成 CSV/JSON，并可写入 PostgreSQL。
+- `数据处理工作区/03_人工复核/03_人工复核导出与回写.py`：从 PostgreSQL 导出人工复核表，并把人工修正结果回写。
+- `数据处理工作区/04_质量体检/04_字段核验与数据库体检.py`：检查字段注释、关键字段空值和疑似错位记录。
+- `数据处理工作区/05_备用读取/05_备用PDF转Markdown_Mineru.py`：备用 PDF 转 Markdown 工具，用于扫描件或 `pdftotext` 读取异常时兜底。
+- `数据处理工作区/06_图片对照核验/06_PDF转图片与对照核验.py`：把 PDF 转为图片，生成图片-字段对照 HTML 和缺项核验索引。
+- `数据处理工作区/07_Kimi视觉兜底/07_Kimi图片识别辅助.py`：可选的 Kimi 视觉识别兜底，仅在文本/Markdown 解析仍不稳定时人工触发。
+
+## 环境文件
+
+本地 `.env` 保存 PostgreSQL 连接信息和网页端目录映射，已被 `.gitignore` 忽略，不会提交到 Gitea。网页端模板见：
+
+```text
+数据可视化网页端/.env.example
+```
+
+`.env` 也可配置 `MOONSHOT_API_KEY` 和 `KIMI_MODEL`。涉及患者信息的图片不会自动上传，Kimi 视觉步骤只在人工运行 `07_Kimi图片识别辅助.py` 时调用。
+
+## 快速运行
+
+在当前项目根目录执行：
+
+```bash
+python3 数据处理工作区/02_解析入库/02_患者首页PDF解析与入库.py
+```
+
+默认 `--text-source auto`：先用 `pdftotext`，如果抽取文本过短或缺少首页关键标记，再调用 Mineru 转 Markdown。
+
+强制使用 Mineru：
+
+```bash
+python3 数据处理工作区/02_解析入库/02_患者首页PDF解析与入库.py \
+  --text-source mineru \
+  --mineru-url "http://MINERU_HOST:4000/extract"
+```
+
+单独运行 PDF 转 Markdown：
+
+```bash
+python3 数据处理工作区/05_备用读取/05_备用PDF转Markdown_Mineru.py \
+  -s 待处理-患者首页PDF \
+  -t 数据处理结果区/06_Mineru_MD \
+  -u "http://MINERU_HOST:4000/extract"
+```
+
+正式流程建议同时写入 PostgreSQL：
+
+```bash
+export PGPASSWORD='请填写数据库密码'
+python3 数据处理工作区/02_解析入库/02_患者首页PDF解析与入库.py \
+  --write-postgres \
+  --pg-host DB_HOST \
+  --pg-port 5432 \
+  --pg-database DB_NAME \
+  --pg-user DB_USER \
+  --pg-table Patient_FrontPages
+unset PGPASSWORD
+```
+
+也可以指定输入和输出目录：
+
+```bash
+python3 数据处理工作区/02_解析入库/02_患者首页PDF解析与入库.py \
+  --input-dir 待处理-患者首页PDF \
+  --output-dir 数据处理结果区
+```
+
+生成 PDF 图片对照页：
+
+```bash
+python3 数据处理工作区/06_图片对照核验/06_PDF转图片与对照核验.py --force
+```
+
+该步骤会读取 `02_单份JSON/` 的结构化结果，把每份 PDF 转为 PNG，并生成：
+
+- `数据处理结果区/06_PDF图片对照/图片/`：每份 PDF 一个图片目录。
+- `数据处理结果区/06_PDF图片对照/患者首页_PDF图片对照.html`：浏览器打开后可左侧看原图、右侧看结构化字段和缺项提示。
+- `数据处理结果区/06_PDF图片对照/患者首页_PDF图片对照索引.csv`：按 PDF 汇总页数、复核状态、核心缺项、建议核对缺项。
+
+启动可视化网页端：
+
+```bash
+docker compose -f 数据可视化网页端/docker-compose.yml up -d --build
+```
+
+打开：
+
+```text
+http://localhost:8501
+```
+
+首次进入会显示登录界面；默认管理员为 `admin`，默认密码为 `change-me`。设置页创建的本地用户会同步作为登录账号，并按权限控制各页面。
+
+网页端会显示数据库连接情况、记录数、需复核数、PDF 数量；人工修改右侧字段并点击“复核并保存”后，会同步更新 PostgreSQL，并把 `manual_corrected` 标记为 `true`。
+
+网页端说明：
+
+- PDF 以 `inline` 方式在页面内打开，不再作为附件触发保存；浏览器内核无法彻底禁止用户保存已展示内容。
+- “复核并保存”后，复核状态由后端自动标记为 `reviewed`，复核状态和质控状态只读显示在 PDF 下方。
+- 诊断、手术、费用等 JSON 字段在右侧以表格方式编辑，保存时再写回 JSONB。
+- 手术操作按编码、日期、级别、名称、术者、I助、II助、切口愈合等级、麻醉方式、麻醉医师拆分展示。
+
+可选 Kimi 视觉兜底：
+
+```bash
+python3 数据处理工作区/07_Kimi视觉兜底/07_Kimi图片识别辅助.py \
+  --case ZY010001672803_ori \
+  --page page-001.png
+```
+
+输出位于 `数据处理结果区/07_Kimi视觉识别/`，默认不直接覆盖 PostgreSQL，需要人工确认后再回写。
+
+## 人工复核流程
+
+从 PostgreSQL 导出待复核表：
+
+```bash
+export PGPASSWORD='请填写数据库密码'
+python3 数据处理工作区/03_人工复核/03_人工复核导出与回写.py export \
+  --pg-host DB_HOST \
+  --pg-port 5432 \
+  --pg-database DB_NAME \
+  --pg-user DB_USER \
+  --pg-table Patient_FrontPages
+unset PGPASSWORD
+```
+
+人工在 `数据处理结果区/04_复核与人工校正/患者首页_人工复核表.csv` 中填写：
+
+- `manual_review_notes`：人工复核说明。
+- `corrected_medical_record_no`：修正后的病案号。
+- `corrected_patient_name`：修正后的姓名。
+- `corrected_major_department`：修正后的大科室。
+- `corrected_primary_diagnosis`、`corrected_primary_diagnosis_code`：修正后的主要诊断。
+- `corrected_total_cost`、`corrected_self_pay_amount`：修正后的费用。
+- `mark_review_status`：建议填 `reviewed`；确实仍有问题可填 `needs_review`。
+
+回写 PostgreSQL：
+
+```bash
+export PGPASSWORD='请填写数据库密码'
+python3 数据处理工作区/03_人工复核/03_人工复核导出与回写.py import \
+  --pg-host DB_HOST \
+  --pg-port 5432 \
+  --pg-database DB_NAME \
+  --pg-user DB_USER \
+  --pg-table Patient_FrontPages
+unset PGPASSWORD
+```
+
+回写后建议做数据库体检：
+
+```bash
+export PGPASSWORD='请填写数据库密码'
+python3 数据处理工作区/04_质量体检/04_字段核验与数据库体检.py \
+  --pg-host DB_HOST \
+  --pg-port 5432 \
+  --pg-database DB_NAME \
+  --pg-user DB_USER \
+  --pg-table Patient_FrontPages
+unset PGPASSWORD
+```
+
+体检会生成：
+
+- `数据处理结果区/05_质量体检/患者首页_数据库体检报告.txt`
+- `数据处理结果区/05_质量体检/患者首页_字段空值统计.csv`
+- `数据处理结果区/05_质量体检/患者首页_疑似异常记录.csv`
+
+## 输出文件
+
+运行后会生成：
+
+- `数据处理结果区/01_结构化结果/患者首页_结构化结果.csv`：适合 Excel/WPS 查看的一行一病例总表。
+- `数据处理结果区/01_结构化结果/患者首页_结构化结果.jsonl`：每行一个完整 JSON，适合后续程序继续处理。
+- `数据处理结果区/02_单份JSON/*.json`：每份 PDF 一个完整 JSON，保留嵌套诊断、手术、费用明细和原始文本。
+- `数据处理结果区/03_提取文本/*.txt`：`pdftotext` 抽出的原始文本，用于人工核查。
+- `数据处理结果区/04_复核与人工校正/患者首页_复核清单.csv`：脚本认为需要人工复核的病例。
+- `数据处理结果区/04_复核与人工校正/患者首页_处理失败.csv`：只有处理失败时才会生成。
+- `数据处理结果区/06_PDF图片对照/患者首页_PDF图片对照.html`：PDF 图片与结构化字段的对照复核页。
+- `数据处理结果区/06_PDF图片对照/患者首页_PDF图片对照索引.csv`：图片对照与字段缺项索引。
+
+## PostgreSQL 表结构
+
+`--write-postgres` 会写入 `public."Patient_FrontPages"` 宽表，不再使用 `payload` 总 JSON 字段，也不写入 `parsed_at/created_at/updated_at`。
+
+入库时会同步写入表注释和字段注释，便于在数据库客户端中查看每列含义。
+
+网页端会按需把人工复核和抽查记录写入主表 JSONB 列：`review_logs` 记录人工保存时间、修改字段和备注；`audit_logs` 记录抽查来源、结论和备注。历史辅助表 `"Patient_FrontPages_review_logs"`、`"Patient_FrontPages_audit_logs"` 会自动迁移并删除，数据库最终只保留 `"Patient_FrontPages"` 一张业务表。
+
+患者号/住院号 `inpatient_no` 是首页与患者列表联动唯一键，也是本程序唯一强制校验条件：不能为空。程序不校验编号格式，同一非空 `inpatient_no` 只保留一条首页记录；迁移时会删除历史重复住院号并保留最新记录，后续重复入库或网页写入会按 `inpatient_no` 覆盖旧首页。若 PDF 文件名和首页病案号可推导，程序仍会自动生成类似 `ZY020001447443` 的默认住院号，后续格式问题交由患者目录核验网页端处理。
+
+病案号和首页病案号都按 10 位文本保存，保留前导 0。若 PDF 首页病案号缺前导 0，但文件名可还原，例如 `01447443` -> `0001447443`，脚本会自动修正，并写入：
+
+- `inpatient_no`
+- `front_page_medical_record_no`
+- `major_department`
+- `text_extraction_method`
+- `mineru_markdown_dir`
+- `review_status`
+- `review_notes`
+- `auto_corrections`
+- `manual_corrected`
+
+`Patient_Lists` 会同步增加 `has_front_page`、`front_page_id`、`front_page_source_file`：凡是患者号非空的首页记录，都会通过 PostgreSQL 触发器自动和 `Patient_Lists.inpatient_no` 关联；如果患者列表没有对应患者号，会由首页表自动补建一条列表记录并标记 `has_front_page = true`，同时写入主要诊断、入院时间和出院时间。空患者号记录会在迁移和入库时被删除，PGSQL 会阻止再次写入空患者号。当首页记录删除，或 `inpatient_no` 改为其他患者号时，触发器会把旧患者号对应的 `has_front_page` 置回 `false` 并清空首页引用。若同一患者号的姓名不一致，以首页表中的非空姓名覆盖 `Patient_Lists.patient_name`，保证患者号、姓名对齐；已有列表记录的诊断和入出院时间不被首页联动覆盖。
+
+## 环境要求
+
+脚本使用 Python 标准库，不需要安装 pandas/pdfplumber。
+
+系统需要有 `pdftotext` 和 `pdftoppm` 命令。Linux 上通常都来自 `poppler-utils`：
+
+```bash
+sudo apt-get install poppler-utils
+```
+
+`05_备用PDF转Markdown_Mineru.py` 需要 `requests`；只在单独使用 Mineru 客户端时安装即可：
+
+```bash
+python3 -m pip install requests
+```
+
+Docker 网页端依赖已写入 `数据可视化网页端/Dockerfile`，不再单独维护 `requirements.txt`。
+
+## 当前解析范围
+
+脚本会抽取：
+
+- 基本信息：住院号、病案号、首页病案号、姓名、性别、出生日期、年龄、身份证号、付费方式等。
+- 入出院信息：入院/出院时间、科别、病房、实际住院天数、门急诊诊断。
+- 科室分类：根据 `数据处理工作区/01_配置规则/01_科室分类规则.json` 生成 CSV 的 `大科室` 和 PostgreSQL 的 `major_department`。
+- 诊断信息：`discharge_diagnoses` 出院诊断 JSONB 数组，包含主要诊断、其他诊断、疾病编码、入院病情。
+- 手术操作：编码、日期、名称、原始行内容。
+- 首页质控信息：病理、药物过敏、尸检、血型/Rh、医师、护士、编码员、病案质量、质控日期、离院方式、再住院计划等。
+- 费用信息：总费用、自付金额、各费用明细；若 PDF 首页未填写费用，不强制判为失败。
+
+## 建议审核
+
+每次运行后先看：
+
+```bash
+sed -n '1,20p' 数据处理结果区/04_复核与人工校正/患者首页_复核清单.csv
+```
+
+随后打开 `数据处理结果区/06_PDF图片对照/患者首页_PDF图片对照.html`，用 PDF 图片逐项对照病案号、姓名、入出院信息、诊断、手术、质控和费用。遇到新医院、新模板、扫描件或字段错位时，把 PDF、对应 TXT、图片对照索引中的缺项提示留在目录中，再让 Codex 继续补规则。
diff --git a/患者首页处理/工作流_Gitea版.md b/患者首页处理/工作流_Gitea版.md
new file mode 100644
index 0000000..03ba9c9
--- /dev/null
+++ b/患者首页处理/工作流_Gitea版.md
@@ -0,0 +1,306 @@
+# 患者首页 PDF 处理工作流_Gitea版
+
+## Gitea 仓库
+
+仓库地址：
+
+```text
+https://gitea.huijutec.cn/admin/HIS_Front_Page.git
+```
+
+仓库只保存程序、文档和配置说明。不要提交原始 PDF、处理结果、明文密码、`.env` 文件，也不再提交加密本地工作流压缩包。
+
+## 仓库内容
+
+```text
+README.md
+工作流_Gitea版.md
+数据可视化网页端/.env.example
+数据可视化网页端/Dockerfile
+数据可视化网页端/docker-compose.yml
+数据可视化网页端/README.md
+数据可视化网页端/app/
+数据处理工作区/01_配置规则/01_科室分类规则.json
+数据处理工作区/02_解析入库/02_患者首页PDF解析与入库.py
+数据处理工作区/03_人工复核/03_人工复核导出与回写.py
+数据处理工作区/04_质量体检/04_字段核验与数据库体检.py
+数据处理工作区/05_备用读取/05_备用PDF转Markdown_Mineru.py
+数据处理工作区/06_图片对照核验/06_PDF转图片与对照核验.py
+数据处理工作区/07_Kimi视觉兜底/07_Kimi图片识别辅助.py
+```
+
+## PostgreSQL 存储约定
+
+目标数据库：
+
+```text
+主机/IP: DB_HOST
+端口: 5432
+数据库名: DB_NAME
+用户名: DB_USER
+表名: Patient_FrontPages
+```
+
+脚本会在首次写入时自动创建或迁移表 `"Patient_FrontPages"`。表是宽表结构，`payload` 中的顶层信息会展开为独立列；不再保存 `payload`、`parsed_at`、`created_at`、`updated_at`。
+
+脚本会为表和全部字段写入 PostgreSQL 注释，数据库客户端中可直接查看每列含义。
+
+网页端会按需在主表中创建两个 JSONB 日志列：
+
+- `review_logs`：记录人工保存时间、修改字段、修改前后值和人工备注。
+- `audit_logs`：记录抽查来源、抽查状态、抽查备注和更新时间。
+
+历史辅助表 `"Patient_FrontPages_review_logs"`、`"Patient_FrontPages_audit_logs"` 会自动迁移进主表并删除，最终只保留 `"Patient_FrontPages"` 一张业务表。
+
+核心列包括：
+
+- `inpatient_no`：患者号/住院号，作为首页与患者列表联动唯一键；这是本程序唯一强制校验条件，不能为空。程序不校验格式，同一非空患者号只允许一条首页，迁移时保留最新记录，后续同号入库或网页写入会覆盖旧首页。
+- `source_file`：PDF 文件名；重复运行以 `inpatient_no` 为准更新。
+- `medical_record_no`：10 位病案号，文本保存，保留前导 0。
+- `front_page_medical_record_no`：10 位首页病案号，文本保存，保留前导 0。
+- `patient_name`、`gender`、`birth_date`、`id_card_no` 等基本信息列。
+- `admission_time`、`discharge_time`、`primary_diagnosis`、`total_cost` 等业务字段列。
+- `major_department`：根据 `数据处理工作区/01_配置规则/01_科室分类规则.json` 从出院科别优先、入院科别兜底映射出的大科室。
+- `discharge_diagnoses`、`operations`、`fee_details`：复杂明细仍以 JSONB 保存；出院诊断已包含主要诊断和其他诊断，不再单独保存 `other_diagnoses`。
+- `quality_status`、`quality_notes`：程序质控结果。
+- `text_extraction_method`：本次使用的文本抽取方式，如 `pdftotext` 或 `mineru_markdown`。
+- `mineru_markdown_dir`：Mineru Markdown 输出目录。
+- `review_status`：`auto_pass`、`auto_corrected` 或 `needs_review`。
+- `review_notes`：需要复核的原因。
+- `manual_corrected`：人工是否已修正，默认 `false`。
+- `auto_corrections`：程序自动修正记录。
+- `review_logs`、`audit_logs`：网页端人工复核与抽查归类记录。
+- `raw_text`：PDF 抽取出的原始文本。
+
+`Patient_Lists` 会自动增加并维护 `has_front_page`、`front_page_id`、`front_page_source_file`。首页 `inpatient_no` 非空时，PostgreSQL 触发器按 `Patient_FrontPages.inpatient_no = Patient_Lists.inpatient_no` 联动；列表没有对应患者号时会自动补建列表记录，并写入主要诊断、入院时间和出院时间；患者号重复时以唯一键覆盖更新，保证一个患者号只有一个病案首页。空患者号记录会在迁移和入库时被删除，PGSQL 会阻止再次写入空患者号。删除首页记录，或把首页 `inpatient_no` 改成其他患者号时，旧患者号的 `has_front_page` 会自动回落为 `false` 并清空首页引用。若同一患者号姓名不一致，以首页表中的非空姓名覆盖 `Patient_Lists.patient_name`；已有列表记录的诊断和入出院时间不被首页联动覆盖。患者号格式核验交由患者目录核验网页端处理。
+
+病案号规则：
+
+- 住院号 = `ZY` + 住院次数 2 位 + 首页病案号 10 位，例如 `ZY020001447443`。
+- `0001906820` 是合法 10 位病案号，前导 0 不能删除。
+- 若 PDF 首页病案号少前导 0，但文件名可还原，例如 `ZY020001447443_ori.pdf` 中可得到 `0001447443`，脚本会把 PDF 值 `01447443` 自动修正为 `0001447443`，并记录到 `auto_corrections`。
+
+## 从 Gitea 拉取后运行
+
+1. 克隆仓库：
+
+   ```bash
+   git clone https://gitea.huijutec.cn/admin/HIS_Front_Page.git
+   cd HIS_Front_Page
+   ```
+
+2. 准备目录：
+
+   ```bash
+   mkdir -p 待处理-患者首页PDF 数据处理结果区 已处理-患者首页PDF
+   ```
+
+3. 放入 PDF：
+
+   ```text
+   待处理-患者首页PDF/
+   ```
+
+4. 本地生成 CSV/JSON：
+
+   ```bash
+   python3 数据处理工作区/02_解析入库/02_患者首页PDF解析与入库.py
+   ```
+
+   默认使用 `--text-source auto`：先用 `pdftotext`，若读取结果异常再调用 Mineru。需要强制 Mineru 时：
+
+   ```bash
+   python3 数据处理工作区/02_解析入库/02_患者首页PDF解析与入库.py \
+     --text-source mineru \
+     --mineru-url "http://MINERU_HOST:4000/extract"
+   ```
+
+   也可以先单独转 Markdown：
+
+   ```bash
+   python3 数据处理工作区/05_备用读取/05_备用PDF转Markdown_Mineru.py \
+     -s 待处理-患者首页PDF \
+     -t 数据处理结果区/06_Mineru_MD \
+     -u "http://MINERU_HOST:4000/extract"
+   ```
+
+5. 正式处理并写入 PostgreSQL：
+
+   ```bash
+   export PGPASSWORD='请填写数据库密码'
+   python3 数据处理工作区/02_解析入库/02_患者首页PDF解析与入库.py \
+     --write-postgres \
+     --pg-host DB_HOST \
+     --pg-port 5432 \
+     --pg-database DB_NAME \
+     --pg-user DB_USER \
+     --pg-table Patient_FrontPages
+   unset PGPASSWORD
+   ```
+
+6. 生成 PDF 图片对照核验页：
+
+   ```bash
+   python3 数据处理工作区/06_图片对照核验/06_PDF转图片与对照核验.py --force
+   ```
+
+   输出位于：
+
+   ```text
+   数据处理结果区/06_PDF图片对照/图片/
+   数据处理结果区/06_PDF图片对照/患者首页_PDF图片对照.html
+   数据处理结果区/06_PDF图片对照/患者首页_PDF图片对照索引.csv
+   ```
+
+   HTML 对照页用于人工左侧看首页图片、右侧看结构化字段；CSV 索引用于快速筛出核心缺项、建议核对缺项和已有复核状态。
+
+7. 启动可视化网页端：
+
+   ```bash
+   docker compose -f 数据可视化网页端/docker-compose.yml up -d --build
+   ```
+
+   浏览器打开：
+
+   ```text
+   http://localhost:8501
+   ```
+
+   首次进入会显示登录界面；默认管理员为 `admin`，默认密码为 `change-me`。设置页创建的本地用户会同步作为登录账号，并按权限控制各页面。
+
+   页面顶部显示数据库连接状态、总记录数、需复核数、PDF 数量；顶部页签包括概览、复核、抽查、抽查一览、设置，设置内合并用户与权限、数据库状态和目录信息。复核页默认进入“全部复核数据”，包含需复核、自动修正、已人工复核/已人工修改；左侧选择病例，中间查看 PDF，右侧以表格方式编辑摘录字段。点击“复核并保存”后会同步写回 PostgreSQL，把 `manual_corrected` 标记为 `true`，自动把 `review_status` 标记为 `reviewed`；`Patient_Lists.has_front_page/front_page_id/front_page_source_file` 由数据库触发器按住院号自动维护。
+
+   复核提示会定位到右侧模块和字段：需要核对的位置标红，无关模块默认收起。人工备注下方会显示修改记录表，主表 `review_logs` 记录修改时间、修改字段、修改前后值和人工备注。
+
+   抽查页可从已人工复核或自动通过记录中随机抽样；右侧和复核页一样显示并可编辑全部结构化信息，下方有人工备注和修改记录。随机抽取不会写入“待抽查”记录，只有点击“归类通过并保存”“归类异常并保存”“归类不确定并保存”后，才把通过、异常、不确定记录写入主表 `audit_logs`；抽查一览展示修改内容、人工备注和更新时间。
+
+   PDF 以 `inline` 方式展示，不再作为附件触发保存。浏览器内核不能彻底禁止用户保存已展示内容。
+
+8. 可选 Kimi 视觉兜底：
+
+   ```bash
+   python3 数据处理工作区/07_Kimi视觉兜底/07_Kimi图片识别辅助.py \
+     --case ZY010001672803_ori \
+     --page page-001.png
+   ```
+
+   Kimi 调用需要本地 `.env` 中配置 `MOONSHOT_API_KEY`。该步骤只在人工触发时上传指定图片，输出到 `数据处理结果区/07_Kimi视觉识别/`，默认不直接覆盖 PostgreSQL。
+
+## 审核流程
+
+每批数据处理后执行：
+
+```bash
+sed -n '1,50p' 数据处理结果区/04_复核与人工校正/患者首页_复核清单.csv
+```
+
+如果复核清单为空，仍建议抽查：
+
+- `数据处理结果区/01_结构化结果/患者首页_结构化结果.csv`
+- `数据处理结果区/02_单份JSON/`
+- `数据处理结果区/03_提取文本/`
+- `数据处理结果区/06_PDF图片对照/患者首页_PDF图片对照.html`
+- `数据处理结果区/06_PDF图片对照/患者首页_PDF图片对照索引.csv`
+- `http://localhost:8501` 可视化网页端
+
+如果出现新模板、扫描件、字段错位、少项漏项或手术操作跨行异常，把对应 PDF 文件名、抽取文本、图片对照索引中的缺项提示交给 Codex 继续补规则。
+
+## PostgreSQL 人工复核流程
+
+1. 导出待复核表：
+
+   ```bash
+   export PGPASSWORD='请填写数据库密码'
+   python3 数据处理工作区/03_人工复核/03_人工复核导出与回写.py export \
+     --pg-host DB_HOST \
+     --pg-port 5432 \
+     --pg-database DB_NAME \
+     --pg-user DB_USER \
+     --pg-table Patient_FrontPages
+   unset PGPASSWORD
+   ```
+
+2. 打开 `数据处理结果区/04_复核与人工校正/患者首页_人工复核表.csv`，人工填写这些列：
+
+   ```text
+   manual_review_notes
+   corrected_medical_record_no
+   corrected_patient_name
+   corrected_major_department
+   corrected_primary_diagnosis
+   corrected_primary_diagnosis_code
+   corrected_total_cost
+   corrected_self_pay_amount
+   mark_review_status
+   ```
+
+3. 回写 PostgreSQL：
+
+   ```bash
+   export PGPASSWORD='请填写数据库密码'
+   python3 数据处理工作区/03_人工复核/03_人工复核导出与回写.py import \
+     --pg-host DB_HOST \
+     --pg-port 5432 \
+     --pg-database DB_NAME \
+     --pg-user DB_USER \
+     --pg-table Patient_FrontPages
+   unset PGPASSWORD
+   ```
+
+回写后，程序会更新可修正字段，并把 `manual_corrected` 标记为 `true`；`manual_review_notes` 会追加进入 `review_notes`。
+
+4. 做数据库体检：
+
+   ```bash
+   export PGPASSWORD='请填写数据库密码'
+   python3 数据处理工作区/04_质量体检/04_字段核验与数据库体检.py \
+     --pg-host DB_HOST \
+     --pg-port 5432 \
+     --pg-database DB_NAME \
+     --pg-user DB_USER \
+     --pg-table Patient_FrontPages
+   unset PGPASSWORD
+   ```
+
+体检输出：
+
+```text
+数据处理结果区/05_质量体检/患者首页_数据库体检报告.txt
+数据处理结果区/05_质量体检/患者首页_字段空值统计.csv
+数据处理结果区/05_质量体检/患者首页_疑似异常记录.csv
+```
+
+## 提交规则
+
+提交到 Gitea 前务必检查：
+
+```bash
+git status --short
+```
+
+不应出现这些内容：
+
+- `待处理-患者首页PDF/`
+- `数据处理结果区/`
+- `已处理-患者首页PDF/`
+- `.env`
+- 含真实密码的环境文件
+- 明文 `工作流_本地使用版.md`
+
+可以提交：
+
+- 脚本
+- `01_~0X_` 前缀的工作区文件
+- README
+- Gitea 工作流文档
+- 不含密码的配置模板
+
+依赖说明：
+
+- 主解析脚本使用 Python 标准库和系统命令 `pdftotext/pdftoppm`。
+- Mineru 备用读取单独需要 `python3 -m pip install requests`。
+- Docker 网页端依赖写在 `数据可视化网页端/Dockerfile` 中，不再维护独立 `requirements.txt`。
+
+## 本地工作流
+
+明文 `工作流_本地使用版.md` 只用于本机执行，不提交到 Gitea；加密压缩包已从仓库删除，后续不再维护该 zip。
diff --git a/患者首页处理/数据可视化网页端/.dockerignore b/患者首页处理/数据可视化网页端/.dockerignore
new file mode 100644
index 0000000..b405bf5
--- /dev/null
+++ b/患者首页处理/数据可视化网页端/.dockerignore
@@ -0,0 +1,4 @@
+__pycache__/
+*.pyc
+.env
+.env.*
diff --git a/患者首页处理/数据可视化网页端/.env.example b/患者首页处理/数据可视化网页端/.env.example
new file mode 100644
index 0000000..de4e941
--- /dev/null
+++ b/患者首页处理/数据可视化网页端/.env.example
@@ -0,0 +1,17 @@
+PGHOST=DB_HOST
+PGPORT=5432
+PGDATABASE=DB_NAME
+PGUSER=DB_USER
+PGPASSWORD=请填写数据库密码
+PGTABLE=Patient_FrontPages
+
+APP_HOST=0.0.0.0
+APP_PORT=8501
+REVIEW_ADMIN_USER=admin
+REVIEW_ADMIN_PASSWORD=change-me
+REVIEW_STATUS_CHECK_TIME=03:00
+
+PDF_DIR=/data/pdfs
+
+MOONSHOT_API_KEY=请填写 Kimi API Key
+KIMI_MODEL=kimi-k2.6
diff --git a/患者首页处理/数据可视化网页端/Dockerfile b/患者首页处理/数据可视化网页端/Dockerfile
new file mode 100644
index 0000000..5a26844
--- /dev/null
+++ b/患者首页处理/数据可视化网页端/Dockerfile
@@ -0,0 +1,19 @@
+FROM python:3.12-slim
+
+ENV PYTHONDONTWRITEBYTECODE=1
+ENV PYTHONUNBUFFERED=1
+
+WORKDIR /app
+
+RUN pip install --no-cache-dir \
+    fastapi==0.115.6 \
+    "uvicorn[standard]==0.34.0" \
+    psycopg2-binary==2.9.10 \
+    python-dotenv==1.0.1 \
+    PyMuPDF==1.24.14
+
+COPY app /app/app
+
+EXPOSE 8501
+
+CMD ["sh", "-c", "uvicorn app.main:app --host ${APP_HOST:-0.0.0.0} --port ${APP_PORT:-8501}"]
diff --git a/患者首页处理/数据可视化网页端/README.md b/患者首页处理/数据可视化网页端/README.md
new file mode 100644
index 0000000..0ff9f0a
--- /dev/null
+++ b/患者首页处理/数据可视化网页端/README.md
@@ -0,0 +1,58 @@
+# 患者首页可视化对照网页端
+
+这个网页端用于人工复核患者首页：概览看批次状态，复核页只处理需复核记录，中间查看 PDF，右侧编辑结构化摘录信息；保存后直接同步到 PostgreSQL 表 `"Patient_FrontPages"`。
+
+## 启动
+
+在项目根目录确认 `.env` 已存在，然后运行：
+
+```bash
+docker compose -f 数据可视化网页端/docker-compose.yml up -d --build
+```
+
+浏览器打开：
+
+```text
+http://localhost:8501
+```
+
+首次进入会显示登录界面。默认管理员为 `admin`，默认密码为 `change-me`；可在 `.env` 中通过 `REVIEW_ADMIN_USER`、`REVIEW_ADMIN_PASSWORD` 修改。设置页创建的本地用户会同步作为登录账号，并按权限控制概览、复核、抽查、抽查一览、设置页面。
+
+## 目录挂载
+
+- `已处理-患者首页PDF/2026_5_25_处理/` 挂载为容器内 `/data/pdfs`，只读。
+- 网页端只预览 PDF，不再挂载或依赖 PDF 转图片目录。
+
+## 保存规则
+
+右侧字段保存后会：
+
+- 更新被编辑的 PostgreSQL 字段。
+- 设置 `manual_corrected = true`。
+- 自动将 `review_status` 标记为 `reviewed`。
+- 把人工备注追加到 `review_notes` JSONB 数组中。
+- 在主表 `review_logs` JSONB 列中记录修改时间、修改字段、修改前后值和人工备注。
+- PostgreSQL 触发器会按非空 `inpatient_no` 自动同步 `Patient_Lists.has_front_page/front_page_id/front_page_source_file`；本网页端只校验患者号不能为空，不校验编号格式，同一患者号只保留一个首页，迁移时保留最新记录，后续同号写入会覆盖旧首页，删除首页记录时旧关联会自动解除；同一患者号姓名不一致时，以首页表中的非空姓名覆盖 `Patient_Lists.patient_name`。只有自动补建新列表记录时，才把主要诊断、入院时间、出院时间写入 `Patient_Lists`。
+
+`.env` 含数据库密码，只保留本地，不提交到 Gitea；可参考 `.env.example` 创建。
+
+## 页面规则
+
+- 顶部页签包括：概览、复核、抽查、抽查一览、设置；设置内合并用户与权限、数据库状态和目录信息。
+- 复核页以住院号作为第一识别项；住院号由住院次数 2 位和首页病案号 10 位自动生成，例如 `ZY020001447443`。
+- 复核页默认进入“全部复核数据”，包含需复核、自动修正、已人工复核/已人工修改，不处理无需复核的自动通过记录。
+- 复核提示会自动定位到右侧模块；命中的模块和字段标红，无关模块默认收起。
+- 人工备注下方展示修改记录表，时间由 PostgreSQL 后台记录。
+- PDF 以 `inline` 方式展示，不再作为附件触发保存；浏览器不能彻底禁止保存已展示内容。
+- 顶部数据库/PDF 状态读取缓存；设置页可手动“立即检查”，也可设置每日凌晨等固定时间自动检查。
+- 复核状态、质控状态显示在 PDF 下方，只读。
+- 诊断、手术、费用等 JSONB 字段以表格方式展示和编辑。
+- 手术操作按编码、日期、级别、名称、术者、I助、II助、切口愈合等级、麻醉方式、麻醉医师拆分。
+
+## 抽查规则
+
+- 抽查页可从 `reviewed` 或 `auto_pass` 中随机抽样。
+- 抽查页右侧和复核页一样展示并可编辑全部结构化信息，下方有人工备注和修改记录。
+- 随机抽取不会写入“待抽查”记录；只有点击“归类通过并保存”“归类异常并保存”“归类不确定并保存”后，才写入主表 `audit_logs` JSONB 列。
+- 抽查一览只展示已归类的通过、异常、不确定记录，并显示修改内容、人工备注和更新时间。
+- 涉及患者资料的视觉识别仍需人工运行 `07_Kimi图片识别辅助.py`，不会由网页端自动上传。
diff --git a/患者首页处理/数据可视化网页端/app/__init__.py b/患者首页处理/数据可视化网页端/app/__init__.py
new file mode 100644
index 0000000..82aeed3
--- /dev/null
+++ b/患者首页处理/数据可视化网页端/app/__init__.py
@@ -0,0 +1 @@
+"""Patient front page visual review web app."""
diff --git a/患者首页处理/数据可视化网页端/app/main.py b/患者首页处理/数据可视化网页端/app/main.py
new file mode 100644
index 0000000..b909ea3
--- /dev/null
+++ b/患者首页处理/数据可视化网页端/app/main.py
@@ -0,0 +1,2779 @@
+from __future__ import annotations
+
+import json
+import os
+import base64
+import hashlib
+import secrets
+import re
+import threading
+import time
+import urllib.error
+import urllib.request
+from datetime import date, datetime, timedelta
+from decimal import Decimal
+from pathlib import Path
+from queue import Empty, Queue
+from typing import Any
+from urllib.parse import urlparse
+from zoneinfo import ZoneInfo
+
+import psycopg2
+import psycopg2.extras
+from dotenv import load_dotenv
+from fastapi import FastAPI, HTTPException, Request, Response
+from fastapi.responses import FileResponse, JSONResponse
+from fastapi.staticfiles import StaticFiles
+from pydantic import BaseModel
+from psycopg2 import sql
+
+
+PROJECT_ROOT = Path(__file__).resolve().parents[2]
+load_dotenv(PROJECT_ROOT / ".env")
+
+APP_DIR = Path(__file__).resolve().parent
+STATIC_DIR = APP_DIR / "static"
+
+
+def env(name: str, default: str = "") -> str:
+    return os.getenv(name, default).strip()
+
+
+DB_CONFIG = {
+    "host": env("PGHOST", "DB_HOST"),
+    "port": int(env("PGPORT", "5432")),
+    "dbname": env("PGDATABASE", "DB_NAME"),
+    "user": env("PGUSER", "DB_USER"),
+    "password": env("PGPASSWORD"),
+}
+PGTABLE = env("PGTABLE", "Patient_FrontPages")
+PDF_DIR = Path(env("PDF_DIR", str(PROJECT_ROOT / "待处理-患者首页PDF"))).resolve()
+
+SETTINGS_PATH = Path(env("REVIEW_SETTINGS_PATH", str(PROJECT_ROOT / "数据可视化网页端/review_settings.local.json"))).resolve()
+APP_TIMEZONE = ZoneInfo(env("APP_TIMEZONE", "Asia/Shanghai") or "Asia/Shanghai")
+
+
+FIELD_GROUPS: list[dict[str, Any]] = [
+    {
+        "name": "基本信息",
+        "fields": [
+            ("inpatient_no", "住院号", "text", None),
+            ("medical_record_no", "病案号", "text", None),
+            ("front_page_medical_record_no", "首页病案号", "text", None),
+            ("patient_name", "姓名", "text", None),
+            ("gender", "性别", "text", None),
+            ("birth_date", "出生日期", "date", None),
+            ("age", "年龄", "text", None),
+            ("nationality", "国籍", "text", None),
+            ("id_card_no", "身份证号", "text", None),
+            ("payment_method", "医疗付费方式", "text", None),
+            ("health_card_no", "健康卡号", "text", None),
+            ("admission_count", "住院次数", "integer", None),
+            ("occupation", "职业", "text", None),
+            ("marital_status_code", "婚姻代码", "text", None),
+            ("admission_path_code", "入院途径代码", "text", None),
+            ("admission_time", "入院时间", "datetime", None),
+            ("admission_dept", "入院科别", "text", None),
+            ("admission_ward", "入院病房", "text", None),
+            ("transfer_dept", "转科科别", "text", None),
+            ("transfer_time", "转科时间", "text", None),
+            ("discharge_time", "出院时间", "datetime", None),
+            ("discharge_dept", "出院科别", "text", None),
+            ("discharge_ward", "出院病房", "text", None),
+            ("hospital_days", "实际住院天数", "integer", None),
+            ("major_department", "大科室", "text", None),
+        ],
+    },
+    {
+        "name": "地址联系人",
+        "fields": [
+            ("current_address", "现住址", "text", None),
+            ("current_address_phone", "现住址电话", "text", None),
+            ("current_address_postcode", "现住址邮编", "text", None),
+            ("household_address", "户口地址", "text", None),
+            ("household_postcode", "户口地址邮编", "text", None),
+            ("employer_address", "工作单位及地址", "text", None),
+            ("employer_phone", "单位电话", "text", None),
+            ("employer_postcode", "单位邮编", "text", None),
+            ("contact_name", "联系人姓名", "text", None),
+            ("contact_relationship", "联系人关系", "text", None),
+            ("contact_address", "联系人地址", "text", None),
+            ("contact_phone", "联系人电话", "text", None),
+        ],
+    },
+    {
+        "name": "诊断表格",
+        "fields": [
+            ("outpatient_diagnosis", "门急诊诊断", "text", None),
+            ("outpatient_diagnosis_code", "门急诊诊断编码", "text", None),
+            ("primary_diagnosis", "主要诊断", "text", None),
+            ("primary_diagnosis_code", "主要诊断编码", "text", None),
+            ("primary_admission_condition", "主要诊断入院病情", "text", None),
+            ("discharge_diagnoses", "出院诊断", "json", None),
+            ("injury_poisoning_external_cause", "损伤中毒外部原因", "text", None),
+            ("injury_poisoning_code", "损伤中毒疾病编码", "text", None),
+            ("pathology_diagnosis", "病理诊断", "text", None),
+            ("pathology_diagnosis_code", "病理诊断编码", "text", None),
+            ("pathology_no", "病理号", "text", None),
+        ],
+    },
+    {
+        "name": "手术表格",
+        "fields": [
+            ("operations", "手术操作 JSON", "json", None),
+        ],
+    },
+    {
+        "name": "离院费用",
+        "fields": [
+            ("discharge_disposition_code", "离院方式代码", "text", None),
+            ("receiving_org_name", "拟接收医疗机构名称", "text", None),
+            ("readmission_plan_code", "出院31天内再住院计划代码", "text", None),
+            ("readmission_plan_purpose", "再住院计划目的", "text", None),
+            ("coma_before_days", "入院前昏迷天数", "integer", None),
+            ("coma_before_hours", "入院前昏迷小时", "integer", None),
+            ("coma_before_minutes", "入院前昏迷分钟", "integer", None),
+            ("coma_after_days", "入院后昏迷天数", "integer", None),
+            ("coma_after_hours", "入院后昏迷小时", "integer", None),
+            ("coma_after_minutes", "入院后昏迷分钟", "integer", None),
+            ("total_cost", "总费用", "numeric", None),
+            ("self_pay_amount", "自付金额", "numeric", None),
+            ("fee_details", "费用明细 JSON", "json", None),
+        ],
+    },
+]
+
+FIELD_META: dict[str, dict[str, Any]] = {}
+for group in FIELD_GROUPS:
+    for name, label, field_type, options in group["fields"]:
+        FIELD_META[name] = {"name": name, "label": label, "type": field_type, "options": options}
+
+EDITABLE_FIELDS = set(FIELD_META)
+JSON_FIELDS = {name for name, meta in FIELD_META.items() if meta["type"] == "json"}
+JSON_DB_FIELDS = JSON_FIELDS | {"review_notes", "quality_notes", "auto_corrections"}
+INTEGER_FIELDS = {name for name, meta in FIELD_META.items() if meta["type"] == "integer"}
+NUMERIC_FIELDS = {name for name, meta in FIELD_META.items() if meta["type"] == "numeric"}
+
+
+class UpdatePayload(BaseModel):
+    fields: dict[str, Any]
+    manual_note: str = ""
+    note_prefix: str = "人工复核"
+
+
+class AuditPayload(BaseModel):
+    audit_status: str = "pending"
+    audit_notes: str = ""
+    ai_result: Any = None
+
+
+class AuditClassifyPayload(BaseModel):
+    record_id: int
+    audit_source: str = "reviewed"
+    audit_status: str
+    audit_notes: str = ""
+    fields: dict[str, Any] = {}
+
+
+class UserPayload(BaseModel):
+    username: str
+    password: str = ""
+    permissions: dict[str, bool] = {}
+
+
+class UserUpdatePayload(BaseModel):
+    username: str = ""
+    password: str = ""
+    permissions: dict[str, bool] = {}
+
+
+class PasswordPayload(BaseModel):
+    password: str = ""
+
+
+class PermissionPayload(BaseModel):
+    permissions: dict[str, bool] = {}
+
+
+class LoginPayload(BaseModel):
+    username: str = ""
+    password: str = ""
+
+
+class SystemSettingsPayload(BaseModel):
+    status_check_time: str = ""
+
+
+class KimiSettingsPayload(BaseModel):
+    enabled: bool = True
+    model: str = ""
+    api_base: str = ""
+    concurrency: int = 3
+
+
+class AiReviewPayload(BaseModel):
+    scope: str = "current"
+    record_id: int | None = None
+
+
+app = FastAPI(title="Patient Front Page Visual Review")
+app.mount("/static", StaticFiles(directory=STATIC_DIR), name="static")
+
+STATUS_CHECK_LOCK = threading.Lock()
+WORKFLOW_LOCK = threading.Lock()
+WORKFLOW_READY = False
+AI_JOB_LOCK = threading.Lock()
+AI_REVIEW_JOB: dict[str, Any] = {
+    "running": False,
+    "scope": "",
+    "total": 0,
+    "processed": 0,
+    "ok": 0,
+    "pending": 0,
+    "failed": 0,
+    "concurrency": 0,
+    "message": "",
+    "errors": [],
+    "started_at": "",
+    "finished_at": "",
+    "last_record_id": None,
+}
+DEFAULT_STATUS_CHECK_TIME = env("REVIEW_STATUS_CHECK_TIME", "03:00") or "03:00"
+DEFAULT_KIMI_API_BASE = env("MOONSHOT_API_BASE", env("KIMI_API_BASE", "https://api.moonshot.cn/v1")) or "https://api.moonshot.cn/v1"
+DEFAULT_KIMI_MODEL = env("KIMI_MODEL", "kimi-k2.6") or "kimi-k2.6"
+AI_NO_ISSUE_STATUS = "AI复核-无问题"
+AI_PENDING_STATUS = "AI复核-待确认"
+AI_CONFIRMED_PROBLEM_KEYWORDS = (
+    "缺少",
+    "缺失",
+    "为空白",
+    "为空",
+    "空白",
+    "无编码",
+    "未填写",
+    "不清晰",
+    "不一致",
+    "错位",
+    "混乱",
+    "需人工",
+    "需要人工",
+    "待确认",
+)
+AI_CONFIRMED_PROBLEM_QUALIFIERS = ("确实", "证实", "属实", "仍", "依然", "存在", "需要", "需人工", "待确认")
+AI_STOP_ERROR_MARKERS = (
+    "exceeded_current_quota_error",
+    "insufficient_quota",
+    "consumption budget",
+    "billing details",
+    "quota",
+    "余额",
+    "额度",
+)
+AI_JOB_ERROR_LIMIT = 50
+PDF_MODULE_DEFINITIONS: list[dict[str, Any]] = [
+    {
+        "name": "基本信息",
+        "keywords": ["住院病案首页", "姓名", "性别", "出生日期", "入院时间", "出院时间", "住院号"],
+        "note_keywords": ["姓名", "性别", "出生", "入院", "出院", "住院号", "病案号", "年龄"],
+        "tail": 330,
+    },
+    {
+        "name": "地址联系人",
+        "keywords": ["现住址", "户口地址", "工作单位及地址", "联系人姓名", "联系人地址"],
+        "note_keywords": ["地址", "电话", "联系人", "户口", "单位", "邮编"],
+        "tail": 260,
+    },
+    {
+        "name": "诊断表格",
+        "keywords": ["门（急）诊诊断", "门急诊诊断", "出院诊断", "疾病编码", "入院病情", "其他诊断"],
+        "note_keywords": ["诊断", "疾病编码", "编码格式", "入院病情", "病理"],
+        "tail": 520,
+    },
+    {
+        "name": "手术表格",
+        "keywords": ["手术及操作编码", "手术及操作日期", "手术及操作名称", "手术级别", "麻醉方式", "术者"],
+        "note_keywords": ["手术", "操作", "麻醉", "切口", "术者"],
+        "tail": 520,
+    },
+    {
+        "name": "离院费用",
+        "keywords": ["离院方式", "出院31天内再住院计划", "住院费用", "总费用", "自付金额", "综合医疗服务类"],
+        "note_keywords": ["离院", "费用", "总费用", "自付", "金额", "再住院", "昏迷"],
+        "tail": 520,
+    },
+]
+AI_REVIEWABLE_STATUSES = ("needs_review", AI_PENDING_STATUS)
+SUBMITTED_STATUS = "已提交"
+
+
+@app.on_event("startup")
+def start_status_scheduler() -> None:
+    threading.Thread(target=status_scheduler_loop, name="status-check-scheduler", daemon=True).start()
+
+
+def table_identifier() -> sql.Composable:
+    if "." in PGTABLE:
+        schema, table = PGTABLE.split(".", 1)
+        return sql.Identifier(schema, table)
+    return sql.Identifier(PGTABLE)
+
+
+def patient_lists_identifier() -> sql.Composable:
+    return sql.Identifier("Patient_Lists")
+
+
+def patient_list_trigger_function_identifier(base_table: str) -> sql.Composable:
+    function_name = f"{base_table}_sync_patient_lists_trigger_fn"
+    if "." in PGTABLE:
+        schema = PGTABLE.split(".", 1)[0]
+        return sql.Identifier(schema, function_name)
+    return sql.Identifier(function_name)
+
+
+def patient_dedupe_trigger_function_identifier(base_table: str) -> sql.Composable:
+    function_name = f"{base_table}_dedupe_inpatient_no_trigger_fn"
+    if "." in PGTABLE:
+        schema = PGTABLE.split(".", 1)[0]
+        return sql.Identifier(schema, function_name)
+    return sql.Identifier(function_name)
+
+
+def related_table_identifier(suffix: str) -> sql.Composable:
+    if "." in PGTABLE:
+        schema, table = PGTABLE.split(".", 1)
+        return sql.Identifier(schema, f"{table}{suffix}")
+    return sql.Identifier(f"{PGTABLE}{suffix}")
+
+
+def connect():
+    return psycopg2.connect(**DB_CONFIG, cursor_factory=psycopg2.extras.RealDictCursor)
+
+
+def json_ready(value: Any) -> Any:
+    if isinstance(value, (date, datetime)):
+        return value.isoformat(sep=" ") if isinstance(value, datetime) else value.isoformat()
+    if isinstance(value, Decimal):
+        return str(value)
+    return value
+
+
+def row_to_json(row: dict[str, Any]) -> dict[str, Any]:
+    return {key: json_ready(value) for key, value in row.items()}
+
+
+def json_ready_deep(value: Any) -> Any:
+    if isinstance(value, dict):
+        return {key: json_ready_deep(item) for key, item in value.items()}
+    if isinstance(value, list):
+        return [json_ready_deep(item) for item in value]
+    return json_ready(value)
+
+
+def comparable(value: Any) -> str:
+    return json.dumps(json_ready_deep(value), ensure_ascii=False, sort_keys=True, default=str)
+
+
+PERMISSION_LABELS = {
+    "overview": "概览",
+    "review": "复核",
+    "audit": "抽查",
+    "audit_history": "抽查一览",
+    "settings": "设置",
+}
+DEFAULT_PERMISSIONS = {key: True for key in PERMISSION_LABELS}
+SESSION_COOKIE = "frontpage_review_session"
+SESSIONS: dict[str, dict[str, Any]] = {}
+
+
+def password_hash(password: str) -> dict[str, str]:
+    salt = secrets.token_hex(12)
+    digest = hashlib.sha256((salt + password).encode("utf-8")).hexdigest()
+    return {"salt": salt, "password_hash": digest}
+
+
+def verify_password(password: str, salt: str, digest: str) -> bool:
+    expected = hashlib.sha256((salt + password).encode("utf-8")).hexdigest()
+    return secrets.compare_digest(expected, digest or "")
+
+
+def admin_username() -> str:
+    return env("REVIEW_ADMIN_USER", "admin") or "admin"
+
+
+def admin_password() -> str:
+    return env("REVIEW_ADMIN_PASSWORD", "change-me") or "change-me"
+
+
+def public_user(username: str, permissions: dict[str, bool], source: str) -> dict[str, Any]:
+    return {
+        "username": username,
+        "permissions": {**DEFAULT_PERMISSIONS, **(permissions or {})},
+        "source": source,
+    }
+
+
+def load_local_settings() -> dict[str, Any]:
+    if not SETTINGS_PATH.exists():
+        return {
+            "users": [],
+            "permission_labels": PERMISSION_LABELS,
+            "system": default_system_settings(),
+            "kimi": default_kimi_settings(),
+            "status_snapshot": default_status_snapshot(),
+        }
+    try:
+        data = json.loads(SETTINGS_PATH.read_text(encoding="utf-8"))
+    except json.JSONDecodeError:
+        data = {
+            "users": [],
+            "permission_labels": PERMISSION_LABELS,
+            "system": default_system_settings(),
+            "kimi": default_kimi_settings(),
+            "status_snapshot": default_status_snapshot(),
+        }
+    data.setdefault("users", [])
+    data.setdefault("permission_labels", PERMISSION_LABELS)
+    data["system"] = normalize_system_settings(data.get("system") or {})
+    data["kimi"] = normalize_kimi_settings(data.get("kimi") or {})
+    data.setdefault("status_snapshot", default_status_snapshot())
+    return data
+
+
+def save_local_settings(data: dict[str, Any]) -> None:
+    SETTINGS_PATH.parent.mkdir(parents=True, exist_ok=True)
+    temp = SETTINGS_PATH.with_suffix(SETTINGS_PATH.suffix + ".tmp")
+    temp.write_text(json.dumps(data, ensure_ascii=False, indent=2), encoding="utf-8")
+    temp.replace(SETTINGS_PATH)
+
+
+def public_settings() -> dict[str, Any]:
+    data = load_local_settings()
+    users = []
+    env_admin_seen = False
+    for user in data.get("users", []):
+        users.append(
+            {
+                "username": user.get("username", ""),
+                "permissions": {**DEFAULT_PERMISSIONS, **(user.get("permissions") or {})},
+                "source": "local",
+                "created_at": user.get("created_at", ""),
+                "updated_at": user.get("updated_at", ""),
+                "has_password": bool(user.get("password_hash")),
+            }
+        )
+        if user.get("username") == admin_username():
+            env_admin_seen = True
+    if not env_admin_seen:
+        users.insert(
+            0,
+            {
+                "username": admin_username(),
+                "permissions": dict(DEFAULT_PERMISSIONS),
+                "source": "env",
+                "created_at": "",
+                "updated_at": "",
+                "has_password": True,
+            },
+        )
+    return {
+        "users": users,
+        "permission_labels": PERMISSION_LABELS,
+        "system": normalize_system_settings(data.get("system") or {}),
+        "kimi": public_kimi_settings(data.get("kimi") or {}),
+        "status_snapshot": data.get("status_snapshot") or default_status_snapshot(),
+    }
+
+
+def normalize_status_check_time(value: str) -> str:
+    value = (value or DEFAULT_STATUS_CHECK_TIME).strip()
+    match = re.fullmatch(r"([01]?\d|2[0-3]):([0-5]\d)", value)
+    if not match:
+        raise HTTPException(status_code=400, detail="状态检查时间必须是 HH:MM，例如 03:00")
+    return f"{int(match.group(1)):02d}:{match.group(2)}"
+
+
+def kimi_api_key() -> str:
+    return env("MOONSHOT_API_KEY") or env("KIMI_API_KEY")
+
+
+def normalize_kimi_concurrency(value: Any) -> int:
+    try:
+        concurrency = int(value)
+    except (TypeError, ValueError):
+        concurrency = int(env("KIMI_CONCURRENCY", "3") or 3)
+    return max(1, min(concurrency, 6))
+
+
+def default_kimi_settings() -> dict[str, Any]:
+    return {
+        "enabled": bool(kimi_api_key()),
+        "api_base": DEFAULT_KIMI_API_BASE,
+        "model": DEFAULT_KIMI_MODEL,
+        "concurrency": normalize_kimi_concurrency(env("KIMI_CONCURRENCY", "3")),
+    }
+
+
+def normalize_kimi_settings(kimi: dict[str, Any]) -> dict[str, Any]:
+    defaults = default_kimi_settings()
+    api_base = str(kimi.get("api_base") or defaults["api_base"]).strip().rstrip("/")
+    model = str(kimi.get("model") or defaults["model"]).strip()
+    return {
+        "enabled": bool(kimi.get("enabled", defaults["enabled"])),
+        "api_base": api_base or DEFAULT_KIMI_API_BASE,
+        "model": model or DEFAULT_KIMI_MODEL,
+        "concurrency": normalize_kimi_concurrency(kimi.get("concurrency", defaults["concurrency"])),
+    }
+
+
+def public_kimi_settings(kimi: dict[str, Any] | None = None) -> dict[str, Any]:
+    settings = normalize_kimi_settings(kimi or {})
+    settings["api_key_configured"] = bool(kimi_api_key())
+    settings["available"] = settings["enabled"] and settings["api_key_configured"]
+    return settings
+
+
+def status_now() -> datetime:
+    return datetime.now(APP_TIMEZONE)
+
+
+def default_system_settings() -> dict[str, Any]:
+    return {
+        "status_check_time": normalize_status_check_time(DEFAULT_STATUS_CHECK_TIME),
+        "last_status_check_date": "",
+        "last_status_checked_at": "",
+    }
+
+
+def normalize_system_settings(system: dict[str, Any]) -> dict[str, Any]:
+    defaults = default_system_settings()
+    merged = {**defaults, **(system or {})}
+    merged["status_check_time"] = normalize_status_check_time(str(merged.get("status_check_time") or defaults["status_check_time"]))
+    return merged
+
+
+def next_status_check_at(system: dict[str, Any] | None = None, now: datetime | None = None) -> str:
+    now = now or status_now()
+    system = normalize_system_settings(system or {})
+    hour, minute = [int(part) for part in system["status_check_time"].split(":")]
+    next_run = now.replace(hour=hour, minute=minute, second=0, microsecond=0)
+    if next_run <= now:
+        next_run += timedelta(days=1)
+    return next_run.isoformat(timespec="seconds")
+
+
+def default_status_snapshot() -> dict[str, Any]:
+    system = default_system_settings()
+    return {
+        "database": "unchecked",
+        "host": DB_CONFIG["host"],
+        "port": DB_CONFIG["port"],
+        "database_name": DB_CONFIG["dbname"],
+        "table": PGTABLE,
+        "pdf_dir": str(PDF_DIR),
+        "pdf_count": None,
+        "total": None,
+        "review_needed": None,
+        "needs_review": None,
+        "auto_passed": None,
+        "ai_passed": None,
+        "ai_pending": None,
+        "reviewed": None,
+        "submitted": None,
+        "manual_corrected": None,
+        "audit_total": None,
+        "message": "尚未执行状态检查",
+        "checked_at": "",
+        "check_source": "",
+        "next_check_at": next_status_check_at(system),
+    }
+
+
+def compute_status_snapshot(source: str = "manual") -> dict[str, Any]:
+    result: dict[str, Any] = {
+        "database": "offline",
+        "host": DB_CONFIG["host"],
+        "port": DB_CONFIG["port"],
+        "database_name": DB_CONFIG["dbname"],
+        "table": PGTABLE,
+        "pdf_dir": str(PDF_DIR),
+        "pdf_count": len(list(PDF_DIR.glob("*.pdf"))) if PDF_DIR.exists() else 0,
+        "checked_at": status_now().isoformat(timespec="seconds"),
+        "check_source": source,
+    }
+    try:
+        query = sql.SQL(
+            """
+            SELECT
+              count(*) AS total,
+              count(*) FILTER (WHERE review_status IN ('needs_review', 'AI复核-待确认')) AS review_needed,
+              count(*) FILTER (WHERE review_status = 'needs_review') AS needs_review,
+              count(*) FILTER (WHERE review_status = 'auto_pass') AS auto_passed,
+              count(*) FILTER (WHERE review_status = 'AI复核-无问题') AS ai_passed,
+              count(*) FILTER (WHERE review_status = 'AI复核-待确认') AS ai_pending,
+              count(*) FILTER (WHERE review_status = 'reviewed') AS reviewed,
+              count(*) FILTER (WHERE review_status = '已提交') AS submitted,
+              count(*) FILTER (WHERE manual_corrected IS TRUE) AS manual_corrected
+            FROM {table}
+            """
+        ).format(table=table_identifier())
+        with connect() as conn, conn.cursor() as cur:
+            cur.execute(query)
+            row = cur.fetchone()
+        result.update(row_to_json(dict(row)))
+        result["audit_total"] = None
+        result["database"] = "online"
+        result["message"] = "连接正常"
+    except Exception as exc:  # noqa: BLE001
+        result["message"] = str(exc)
+    return result
+
+
+def refresh_status_snapshot(source: str = "manual") -> dict[str, Any]:
+    with STATUS_CHECK_LOCK:
+        snapshot = compute_status_snapshot(source=source)
+        data = load_local_settings()
+        system = normalize_system_settings(data.get("system") or {})
+        now = status_now()
+        if source == "scheduled":
+            system["last_status_check_date"] = now.date().isoformat()
+        system["last_status_checked_at"] = snapshot.get("checked_at", now.isoformat(timespec="seconds"))
+        snapshot["next_check_at"] = next_status_check_at(system, now)
+        data["system"] = system
+        data["status_snapshot"] = snapshot
+        save_local_settings(data)
+        return snapshot
+
+
+def status_scheduler_loop() -> None:
+    while True:
+        try:
+            data = load_local_settings()
+            system = normalize_system_settings(data.get("system") or {})
+            now = status_now()
+            hour, minute = [int(part) for part in system["status_check_time"].split(":")]
+            due_time = now.replace(hour=hour, minute=minute, second=0, microsecond=0)
+            already_ran = system.get("last_status_check_date") == now.date().isoformat()
+            if now >= due_time and not already_ran:
+                refresh_status_snapshot(source="scheduled")
+        except Exception:
+            pass
+        time.sleep(60)
+
+
+def clean_permissions(permissions: dict[str, bool]) -> dict[str, bool]:
+    return {key: bool(permissions.get(key, DEFAULT_PERMISSIONS[key])) for key in PERMISSION_LABELS}
+
+
+def local_user_index(data: dict[str, Any], username: str) -> int | None:
+    for index, user in enumerate(data.get("users", [])):
+        if user.get("username") == username:
+            return index
+    return None
+
+
+def validate_local_username(username: str, data: dict[str, Any], current_username: str = "") -> str:
+    username = username.strip()
+    if not username:
+        raise HTTPException(status_code=400, detail="用户名不能为空")
+    if username == admin_username() and username != current_username:
+        raise HTTPException(status_code=400, detail="不能覆盖环境变量管理员")
+    for user in data.get("users", []):
+        if user.get("username") == username and user.get("username") != current_username:
+            raise HTTPException(status_code=400, detail="用户已存在")
+    return username
+
+
+def authenticate_user(username: str, password: str) -> dict[str, Any] | None:
+    username = username.strip()
+    if username == admin_username():
+        if secrets.compare_digest(password, admin_password()):
+            return public_user(username, DEFAULT_PERMISSIONS, "env")
+        return None
+    data = load_local_settings()
+    for user in data.get("users", []):
+        if user.get("username") != username:
+            continue
+        if not user.get("password_hash") or not user.get("salt"):
+            return None
+        if verify_password(password, user.get("salt", ""), user.get("password_hash", "")):
+            return public_user(username, clean_permissions(user.get("permissions") or {}), "local")
+        return None
+    return None
+
+
+def session_from_request(request: Request) -> dict[str, Any] | None:
+    token = request.cookies.get(SESSION_COOKIE, "")
+    if not token:
+        return None
+    return SESSIONS.get(token)
+
+
+def page_permission_for_path(path: str, method: str) -> str | tuple[str, ...] | None:
+    if path in {"/api/status", "/api/schema"}:
+        return None
+    if path.startswith("/api/settings"):
+        return "settings"
+    if path.startswith("/api/overview"):
+        return "overview"
+    if path.startswith("/api/audit/logs") and method == "GET":
+        return "audit_history"
+    if path.startswith("/api/audit"):
+        return "audit"
+    if path.startswith("/api/ai"):
+        return "review"
+    if path.startswith("/api/pdf/"):
+        return ("review", "audit")
+    if path == "/api/records":
+        return "review"
+    if path.startswith("/api/records/"):
+        return "review" if method != "GET" else ("review", "audit")
+    return None
+
+
+def has_page_permission(user: dict[str, Any], requirement: str | tuple[str, ...] | None) -> bool:
+    if requirement is None:
+        return True
+    permissions = user.get("permissions") or {}
+    if isinstance(requirement, tuple):
+        return any(permissions.get(item, False) for item in requirement)
+    return bool(permissions.get(requirement, False))
+
+
+@app.middleware("http")
+async def auth_middleware(request: Request, call_next):
+    path = request.url.path
+    if path.startswith("/api/") and not path.startswith("/api/auth/"):
+        user = session_from_request(request)
+        if not user:
+            return JSONResponse({"detail": "请先登录"}, status_code=401)
+        requirement = page_permission_for_path(path, request.method)
+        if not has_page_permission(user, requirement):
+            return JSONResponse({"detail": "当前用户没有访问权限"}, status_code=403)
+        request.state.user = user
+    return await call_next(request)
+
+
+def digits(value: Any, width: int) -> str:
+    text = re.sub(r"\D", "", str(value or ""))
+    return text[-width:].zfill(width) if text else ""
+
+
+def source_file_inpatient_no(source_file: str) -> str:
+    match = re.match(r"^(ZY\d{12})", Path(source_file or "").stem, flags=re.IGNORECASE)
+    return match.group(1).upper() if match else ""
+
+
+def source_file_admission_count(source_file: str) -> str:
+    match = re.match(r"^ZY(\d{2})\d{10}", Path(source_file or "").stem, flags=re.IGNORECASE)
+    return match.group(1) if match else ""
+
+
+def source_file_medical_record_no(source_file: str) -> str:
+    match = re.match(r"^ZY\d{2}(\d{10})", Path(source_file or "").stem, flags=re.IGNORECASE)
+    return match.group(1) if match else ""
+
+
+def build_inpatient_no_from_record(record: dict[str, Any]) -> str:
+    source_file = str(record.get("source_file") or "")
+    admission = digits(record.get("admission_count"), 2) or source_file_admission_count(source_file)
+    page_no = (
+        digits(record.get("front_page_medical_record_no"), 10)
+        or digits(record.get("medical_record_no"), 10)
+        or source_file_medical_record_no(source_file)
+    )
+    if admission and page_no:
+        return f"ZY{admission}{page_no}"
+    return source_file_inpatient_no(source_file)
+
+
+def ensure_workflow_tables(force: bool = False) -> None:
+    global WORKFLOW_READY
+    if WORKFLOW_READY and not force:
+        return
+    with WORKFLOW_LOCK:
+        if WORKFLOW_READY and not force:
+            return
+        _ensure_workflow_tables_uncached()
+        WORKFLOW_READY = True
+
+
+def _ensure_workflow_tables_uncached() -> None:
+    table = table_identifier()
+    old_review_logs = related_table_identifier("_review_logs")
+    old_audit_logs = related_table_identifier("_audit_logs")
+    schema = PGTABLE.split(".", 1)[0] if "." in PGTABLE else "public"
+    base_table = PGTABLE.split(".", 1)[-1]
+    old_review_regclass = f'{schema}."{base_table}_review_logs"'
+    old_audit_regclass = f'{schema}."{base_table}_audit_logs"'
+    with connect() as conn, conn.cursor() as cur:
+        cur.execute("SELECT pg_advisory_xact_lock(hashtext(%s))", (f"{PGTABLE}:workflow_storage",))
+        cur.execute(sql.SQL("ALTER TABLE {table} ADD COLUMN IF NOT EXISTS inpatient_no TEXT").format(table=table))
+        cur.execute(sql.SQL("ALTER TABLE {table} ADD COLUMN IF NOT EXISTS major_department TEXT").format(table=table))
+        cur.execute(
+            sql.SQL("ALTER TABLE {table} ADD COLUMN IF NOT EXISTS review_logs JSONB NOT NULL DEFAULT '[]'::jsonb").format(
+                table=table
+            )
+        )
+        cur.execute(
+            sql.SQL("ALTER TABLE {table} ADD COLUMN IF NOT EXISTS audit_logs JSONB NOT NULL DEFAULT '[]'::jsonb").format(
+                table=table
+            )
+        )
+        cur.execute(
+            sql.SQL("COMMENT ON COLUMN {table}.review_logs IS '人工复核修改记录，JSONB数组，已合并到患者首页主表'").format(
+                table=table
+            )
+        )
+        cur.execute(
+            sql.SQL("COMMENT ON COLUMN {table}.audit_logs IS '抽查归类记录，JSONB数组，已合并到患者首页主表'").format(
+                table=table
+            )
+        )
+        cur.execute(
+            sql.SQL("COMMENT ON COLUMN {table}.inpatient_no IS '患者号/住院号，作为首页与患者列表联动唯一键；不能为空，格式由患者目录核验端处理。'").format(
+                table=table
+            )
+        )
+        cur.execute(
+            sql.SQL("COMMENT ON COLUMN {table}.major_department IS '大科室分类，来源于01_科室分类规则.json。'").format(
+                table=table
+            )
+        )
+        cur.execute(
+            sql.SQL(
+                """
+                UPDATE {table}
+                SET front_page_medical_record_no = RIGHT(LPAD(regexp_replace(front_page_medical_record_no, '\\D', '', 'g'), 10, '0'), 10)
+                WHERE front_page_medical_record_no IS NOT NULL
+                  AND front_page_medical_record_no <> ''
+                  AND front_page_medical_record_no !~ '^\\d{{10}}$'
+                  AND regexp_replace(front_page_medical_record_no, '\\D', '', 'g') <> ''
+                """
+            ).format(table=table)
+        )
+        cur.execute(
+            sql.SQL(
+                """
+                UPDATE {table}
+                SET inpatient_no =
+                    'ZY'
+                    || COALESCE(
+                        LPAD(admission_count::text, 2, '0'),
+                        substring(source_file from '^ZY([0-9]{{2}})[0-9]{{10}}')
+                    )
+                    || RIGHT(
+                        LPAD(
+                            COALESCE(
+                                NULLIF(regexp_replace(COALESCE(front_page_medical_record_no, ''), '\\D', '', 'g'), ''),
+                                NULLIF(regexp_replace(COALESCE(medical_record_no, ''), '\\D', '', 'g'), ''),
+                                substring(source_file from '^ZY[0-9]{{2}}([0-9]{{10}})')
+                            ),
+                            10,
+                            '0'
+                        ),
+                        10
+                    )
+                WHERE (inpatient_no IS NULL OR BTRIM(inpatient_no) = '')
+                  AND COALESCE(
+                        LPAD(admission_count::text, 2, '0'),
+                        substring(source_file from '^ZY([0-9]{{2}})[0-9]{{10}}')
+                      ) IS NOT NULL
+                  AND COALESCE(
+                        NULLIF(regexp_replace(COALESCE(front_page_medical_record_no, ''), '\\D', '', 'g'), ''),
+                        NULLIF(regexp_replace(COALESCE(medical_record_no, ''), '\\D', '', 'g'), ''),
+                        substring(source_file from '^ZY[0-9]{{2}}([0-9]{{10}})')
+                      ) IS NOT NULL
+                """
+            ).format(table=table)
+        )
+        cur.execute(
+            sql.SQL("ALTER TABLE {table} DROP CONSTRAINT IF EXISTS {constraint}").format(
+                table=table,
+                constraint=sql.Identifier(f"{base_table}_source_file_key"),
+            )
+        )
+        for constraint_name in {
+            f"ck_{base_table}_inpatient_no_format",
+            f"ck_{base_table.lower()}_inpatient_no_format",
+            f"ck_{base_table}_inpatient_no_required",
+        }:
+            cur.execute(
+                sql.SQL("ALTER TABLE {table} DROP CONSTRAINT IF EXISTS {constraint}").format(
+                    table=table,
+                    constraint=sql.Identifier(constraint_name),
+                )
+            )
+        cur.execute(
+            sql.SQL("DELETE FROM {table} WHERE NULLIF(BTRIM(inpatient_no), '') IS NULL").format(table=table)
+        )
+        cur.execute(
+            sql.SQL("UPDATE {table} SET inpatient_no = BTRIM(inpatient_no) WHERE inpatient_no <> BTRIM(inpatient_no)").format(
+                table=table
+            )
+        )
+        cur.execute(
+            sql.SQL(
+                """
+                WITH ranked AS (
+                  SELECT
+                    id,
+                    ROW_NUMBER() OVER (PARTITION BY BTRIM(inpatient_no) ORDER BY id DESC) AS duplicate_rank
+                  FROM {table}
+                  WHERE NULLIF(BTRIM(inpatient_no), '') IS NOT NULL
+                )
+                DELETE FROM {table} p
+                USING ranked
+                WHERE p.id = ranked.id
+                  AND ranked.duplicate_rank > 1
+                """
+            ).format(table=table)
+        )
+        cur.execute(sql.SQL("ALTER TABLE {table} ALTER COLUMN inpatient_no SET NOT NULL").format(table=table))
+        cur.execute("SELECT 1 FROM pg_constraint WHERE conname = %s", (f"ck_{base_table}_inpatient_no_required",))
+        if not cur.fetchone():
+            cur.execute(
+                sql.SQL("ALTER TABLE {table} ADD CONSTRAINT {constraint} CHECK (NULLIF(BTRIM(inpatient_no), '') IS NOT NULL)").format(
+                    table=table,
+                    constraint=sql.Identifier(f"ck_{base_table}_inpatient_no_required"),
+                )
+            )
+        cur.execute(
+            sql.SQL("CREATE UNIQUE INDEX IF NOT EXISTS {index_name} ON {table}(inpatient_no)").format(
+                index_name=sql.Identifier(f"{base_table}_inpatient_no_uidx"),
+                table=table,
+            )
+        )
+        cur.execute("SELECT to_regclass(%s) AS table_oid", (old_review_regclass,))
+        if cur.fetchone()["table_oid"]:
+            cur.execute(
+                sql.SQL(
+                    """
+                    WITH grouped AS (
+                      SELECT
+                        record_id,
+                        jsonb_agg(
+                          jsonb_build_object(
+                            'id', id::text,
+                            'record_id', record_id,
+                            'source_file', source_file,
+                            'changed_at', changed_at,
+                            'changed_by', changed_by,
+                            'manual_note', manual_note,
+                            'changed_fields', changed_fields
+                          )
+                          ORDER BY changed_at DESC, id DESC
+                        ) AS logs
+                      FROM {old_review_logs}
+                      GROUP BY record_id
+                    )
+                    UPDATE {table} p
+                    SET review_logs = COALESCE(p.review_logs, '[]'::jsonb) || grouped.logs
+                    FROM grouped
+                    WHERE p.id = grouped.record_id
+                    """
+                ).format(table=table, old_review_logs=old_review_logs)
+            )
+            cur.execute(sql.SQL("DROP TABLE IF EXISTS {old_review_logs}").format(old_review_logs=old_review_logs))
+        cur.execute("SELECT to_regclass(%s) AS table_oid", (old_audit_regclass,))
+        if cur.fetchone()["table_oid"]:
+            cur.execute(
+                sql.SQL(
+                    """
+                    WITH grouped AS (
+                      SELECT
+                        record_id,
+                        jsonb_agg(
+                          jsonb_build_object(
+                            'id', id::text,
+                            'record_id', record_id,
+                            'source_file', source_file,
+                            'audit_source', audit_source,
+                            'audit_status', audit_status,
+                            'audit_notes', audit_notes,
+                            'ai_result', ai_result,
+                            'snapshot', snapshot,
+                            'created_at', created_at,
+                            'updated_at', updated_at
+                          )
+                          ORDER BY updated_at DESC, id DESC
+                        ) AS logs
+                      FROM {old_audit_logs}
+                      GROUP BY record_id
+                    )
+                    UPDATE {table} p
+                    SET audit_logs = COALESCE(p.audit_logs, '[]'::jsonb) || grouped.logs
+                    FROM grouped
+                    WHERE p.id = grouped.record_id
+                    """
+                ).format(table=table, old_audit_logs=old_audit_logs)
+            )
+            cur.execute(sql.SQL("DROP TABLE IF EXISTS {old_audit_logs}").format(old_audit_logs=old_audit_logs))
+        ensure_patient_frontpage_dedupe_trigger(cur)
+        sync_patient_lists(cur)
+        ensure_patient_lists_trigger(cur)
+        conn.commit()
+
+
+def sync_patient_lists(cur) -> None:
+    table = table_identifier()
+    list_table = patient_lists_identifier()
+    cur.execute(
+        sql.SQL(
+            """
+            CREATE TABLE IF NOT EXISTS {list_table} (
+              record_id BIGSERIAL PRIMARY KEY,
+              batch_name TEXT NOT NULL DEFAULT 'Patient_FrontPages',
+              major_department TEXT NOT NULL DEFAULT '',
+              sub_department TEXT NOT NULL DEFAULT '',
+              source_folder TEXT NOT NULL DEFAULT 'Patient_FrontPages',
+              image_path TEXT NOT NULL DEFAULT '',
+              image_name TEXT NOT NULL DEFAULT '',
+              image_row_no INTEGER NOT NULL DEFAULT 0,
+              patient_name TEXT NOT NULL DEFAULT '',
+              gender TEXT,
+              age TEXT,
+              inpatient_no TEXT NOT NULL,
+              diagnosis TEXT,
+              admission_time TEXT,
+              last_write_time TEXT,
+              hospital_days INTEGER,
+              discharge_time TEXT,
+              postoperative_days TEXT,
+              review_status TEXT NOT NULL DEFAULT '首页自动关联',
+              review_notes TEXT,
+              manual_corrected BOOLEAN NOT NULL DEFAULT false,
+              imported_at TIMESTAMPTZ NOT NULL DEFAULT now()
+            )
+            """
+        ).format(list_table=list_table)
+    )
+    for name, column_type in [
+        ("has_front_page", "BOOLEAN NOT NULL DEFAULT false"),
+        ("front_page_id", "BIGINT"),
+        ("front_page_source_file", "TEXT"),
+    ]:
+        cur.execute(
+            sql.SQL("ALTER TABLE {list_table} ADD COLUMN IF NOT EXISTS {column} " + column_type).format(
+                list_table=list_table,
+                column=sql.Identifier(name),
+            )
+        )
+    for constraint_name in {
+        "ck_patient_lists_inpatient_no_format",
+        "ck_Patient_Lists_inpatient_no_format",
+        "ck_patient_lists_inpatient_no_required",
+    }:
+        cur.execute(
+            sql.SQL("ALTER TABLE {list_table} DROP CONSTRAINT IF EXISTS {constraint}").format(
+                list_table=list_table,
+                constraint=sql.Identifier(constraint_name),
+            )
+        )
+    cur.execute(
+        sql.SQL("DELETE FROM {list_table} WHERE NULLIF(BTRIM(inpatient_no), '') IS NULL").format(
+            list_table=list_table
+        )
+    )
+    cur.execute(
+        sql.SQL("UPDATE {list_table} SET inpatient_no = BTRIM(inpatient_no) WHERE inpatient_no <> BTRIM(inpatient_no)").format(
+            list_table=list_table
+        )
+    )
+    cur.execute(
+        sql.SQL(
+            """
+            WITH ranked AS (
+              SELECT
+                record_id,
+                ROW_NUMBER() OVER (PARTITION BY BTRIM(inpatient_no) ORDER BY record_id DESC) AS duplicate_rank
+              FROM {list_table}
+              WHERE NULLIF(BTRIM(inpatient_no), '') IS NOT NULL
+            )
+            DELETE FROM {list_table} pl
+            USING ranked
+            WHERE pl.record_id = ranked.record_id
+              AND ranked.duplicate_rank > 1
+            """
+        ).format(list_table=list_table)
+    )
+    cur.execute(sql.SQL("ALTER TABLE {list_table} ALTER COLUMN inpatient_no SET NOT NULL").format(list_table=list_table))
+    cur.execute("SELECT 1 FROM pg_constraint WHERE conname = %s", ("ck_patient_lists_inpatient_no_required",))
+    if not cur.fetchone():
+        cur.execute(
+            sql.SQL(
+                "ALTER TABLE {list_table} ADD CONSTRAINT {constraint} CHECK (NULLIF(BTRIM(inpatient_no), '') IS NOT NULL)"
+            ).format(
+                list_table=list_table,
+                constraint=sql.Identifier("ck_patient_lists_inpatient_no_required"),
+            )
+        )
+    cur.execute(sql.SQL("COMMENT ON COLUMN {list_table}.has_front_page IS '是否有患者首页：由Patient_FrontPages按住院号自动联动。'").format(list_table=list_table))
+    cur.execute(sql.SQL("COMMENT ON COLUMN {list_table}.front_page_id IS '关联的Patient_FrontPages.id。'").format(list_table=list_table))
+    cur.execute(sql.SQL("COMMENT ON COLUMN {list_table}.front_page_source_file IS '关联患者首页PDF文件名。'").format(list_table=list_table))
+    cur.execute(
+        sql.SQL("CREATE UNIQUE INDEX IF NOT EXISTS {index_name} ON {list_table}(inpatient_no)").format(
+            index_name=sql.Identifier("uq_patient_lists_inpatient_no"),
+            list_table=list_table,
+        )
+    )
+    cur.execute(
+        sql.SQL(
+            """
+            WITH front_pages AS (
+              SELECT DISTINCT ON (BTRIM(inpatient_no))
+                id,
+                BTRIM(inpatient_no) AS inpatient_no,
+                source_file,
+                COALESCE(patient_name, '') AS patient_name,
+                gender,
+                age,
+                COALESCE(major_department, '') AS major_department,
+                COALESCE(discharge_dept, admission_dept, '') AS sub_department,
+                primary_diagnosis,
+                admission_time,
+                discharge_time,
+                hospital_days,
+                manual_corrected
+              FROM {table}
+              WHERE NULLIF(BTRIM(inpatient_no), '') IS NOT NULL
+              ORDER BY BTRIM(inpatient_no), id DESC
+            )
+            INSERT INTO {list_table} (
+              batch_name, major_department, sub_department, source_folder, image_path, image_name,
+              image_row_no, patient_name, gender, age, inpatient_no, diagnosis, admission_time,
+              hospital_days, discharge_time, review_status, review_notes, manual_corrected,
+              has_front_page, front_page_id, front_page_source_file, imported_at
+            )
+            SELECT
+              'Patient_FrontPages',
+              major_department,
+              sub_department,
+              'Patient_FrontPages',
+              source_file,
+              source_file,
+              0,
+              patient_name,
+              gender,
+              age,
+              inpatient_no,
+              primary_diagnosis,
+              to_char(admission_time, 'YYYY-MM-DD HH24:MI:SS'),
+              hospital_days,
+              to_char(discharge_time, 'YYYY-MM-DD HH24:MI:SS'),
+              '首页自动关联',
+              '由Patient_FrontPages按住院号自动关联',
+              manual_corrected,
+              true,
+              id,
+              source_file,
+              now()
+            FROM front_pages
+            ON CONFLICT (inpatient_no) DO UPDATE SET
+              has_front_page = true,
+              front_page_id = EXCLUDED.front_page_id,
+              front_page_source_file = EXCLUDED.front_page_source_file,
+              patient_name = COALESCE(NULLIF(EXCLUDED.patient_name, ''), {list_table}.patient_name),
+              gender = EXCLUDED.gender,
+              age = EXCLUDED.age,
+              major_department = EXCLUDED.major_department,
+              sub_department = EXCLUDED.sub_department,
+              manual_corrected = EXCLUDED.manual_corrected,
+              imported_at = now()
+            """
+        ).format(table=table, list_table=list_table)
+    )
+    cur.execute(
+        sql.SQL(
+            """
+            UPDATE {list_table} AS pl
+            SET has_front_page = false,
+                front_page_id = NULL,
+                front_page_source_file = NULL,
+                imported_at = now()
+            WHERE has_front_page IS TRUE
+              AND NOT EXISTS (
+                SELECT 1 FROM {table} fp
+                WHERE BTRIM(fp.inpatient_no) = pl.inpatient_no
+              )
+            """
+        ).format(table=table, list_table=list_table)
+    )
+
+
+def ensure_patient_frontpage_dedupe_trigger(cur) -> None:
+    table = table_identifier()
+    schema = PGTABLE.split(".", 1)[0] if "." in PGTABLE else "public"
+    base_table = PGTABLE.split(".", 1)[-1]
+    trigger_name = sql.Identifier(f"trg_{base_table}_dedupe_inpatient_no")
+    trigger_function = patient_dedupe_trigger_function_identifier(base_table)
+    cur.execute(
+        """
+        SELECT column_name
+        FROM information_schema.columns
+        WHERE table_schema = %s
+          AND table_name = %s
+          AND column_name NOT IN ('id', 'inpatient_no', 'review_logs', 'audit_logs')
+        ORDER BY ordinal_position
+        """,
+        (schema, base_table),
+    )
+    update_columns = [row["column_name"] for row in cur.fetchall()]
+    update_assignments = sql.SQL(",\n                  ").join(
+        sql.SQL("{column} = NEW.{column}").format(column=sql.Identifier(column_name))
+        for column_name in update_columns
+    )
+    cur.execute(
+        sql.SQL(
+            """
+            CREATE OR REPLACE FUNCTION {trigger_function}()
+            RETURNS trigger
+            LANGUAGE plpgsql
+            AS $trigger$
+            DECLARE
+              existing_id BIGINT;
+            BEGIN
+              NEW.inpatient_no := BTRIM(NEW.inpatient_no);
+              IF NULLIF(NEW.inpatient_no, '') IS NULL THEN
+                RETURN NEW;
+              END IF;
+
+              IF TG_OP = 'INSERT' THEN
+                SELECT id
+                INTO existing_id
+                FROM {table}
+                WHERE BTRIM(inpatient_no) = NEW.inpatient_no
+                ORDER BY id DESC
+                LIMIT 1;
+
+                IF existing_id IS NOT NULL THEN
+                  DELETE FROM {table}
+                  WHERE BTRIM(inpatient_no) = NEW.inpatient_no
+                    AND id <> existing_id;
+
+                  UPDATE {table}
+                  SET {update_assignments}
+                  WHERE id = existing_id;
+
+                  RETURN NULL;
+                END IF;
+              END IF;
+
+              DELETE FROM {table}
+              WHERE BTRIM(inpatient_no) = NEW.inpatient_no
+                AND id <> NEW.id;
+
+              RETURN NEW;
+            END;
+            $trigger$;
+            """
+        ).format(trigger_function=trigger_function, table=table, update_assignments=update_assignments)
+    )
+    cur.execute(
+        sql.SQL("DROP TRIGGER IF EXISTS {trigger_name} ON {table}").format(
+            trigger_name=trigger_name,
+            table=table,
+        )
+    )
+    cur.execute(
+        sql.SQL(
+            """
+            CREATE TRIGGER {trigger_name}
+            BEFORE INSERT OR UPDATE OF inpatient_no ON {table}
+            FOR EACH ROW EXECUTE FUNCTION {trigger_function}()
+            """
+        ).format(trigger_name=trigger_name, table=table, trigger_function=trigger_function)
+    )
+
+
+def ensure_patient_lists_trigger(cur) -> None:
+    table = table_identifier()
+    list_table = patient_lists_identifier()
+    base_table = PGTABLE.split(".", 1)[-1]
+    trigger_name = sql.Identifier(f"trg_{base_table}_sync_patient_lists")
+    trigger_function = patient_list_trigger_function_identifier(base_table)
+    cur.execute(
+        sql.SQL(
+            """
+            CREATE OR REPLACE FUNCTION {trigger_function}()
+            RETURNS trigger
+            LANGUAGE plpgsql
+            AS $trigger$
+            BEGIN
+              IF TG_OP = 'DELETE' THEN
+                IF NULLIF(BTRIM(OLD.inpatient_no), '') IS NOT NULL THEN
+                  UPDATE {list_table} AS pl
+                  SET has_front_page = false,
+                      front_page_id = NULL,
+                      front_page_source_file = NULL,
+                      imported_at = now()
+                  WHERE pl.inpatient_no = BTRIM(OLD.inpatient_no)
+                    AND NOT EXISTS (
+                      SELECT 1 FROM {table} fp
+                      WHERE BTRIM(fp.inpatient_no) = BTRIM(OLD.inpatient_no)
+                    );
+                END IF;
+                RETURN OLD;
+              END IF;
+
+              IF TG_OP = 'UPDATE'
+                 AND NULLIF(BTRIM(OLD.inpatient_no), '') IS NOT NULL
+                 AND BTRIM(OLD.inpatient_no) IS DISTINCT FROM BTRIM(NEW.inpatient_no) THEN
+                UPDATE {list_table} AS pl
+                SET has_front_page = false,
+                    front_page_id = NULL,
+                    front_page_source_file = NULL,
+                    imported_at = now()
+                WHERE pl.inpatient_no = BTRIM(OLD.inpatient_no)
+                  AND NOT EXISTS (
+                    SELECT 1 FROM {table} fp
+                    WHERE BTRIM(fp.inpatient_no) = BTRIM(OLD.inpatient_no)
+                  );
+              END IF;
+
+              IF NULLIF(BTRIM(NEW.inpatient_no), '') IS NOT NULL THEN
+                INSERT INTO {list_table} (
+                  batch_name, major_department, sub_department, source_folder, image_path, image_name,
+                  image_row_no, patient_name, gender, age, inpatient_no, diagnosis, admission_time,
+                  hospital_days, discharge_time, review_status, review_notes, manual_corrected,
+                  has_front_page, front_page_id, front_page_source_file, imported_at
+                )
+                VALUES (
+                  'Patient_FrontPages',
+                  COALESCE(NEW.major_department, ''),
+                  COALESCE(NEW.discharge_dept, NEW.admission_dept, ''),
+                  'Patient_FrontPages',
+                  COALESCE(NEW.source_file, ''),
+                  COALESCE(NEW.source_file, ''),
+                  0,
+                  COALESCE(NEW.patient_name, ''),
+                  NEW.gender,
+                  NEW.age,
+                  BTRIM(NEW.inpatient_no),
+                  NEW.primary_diagnosis,
+                  to_char(NEW.admission_time, 'YYYY-MM-DD HH24:MI:SS'),
+                  NEW.hospital_days,
+                  to_char(NEW.discharge_time, 'YYYY-MM-DD HH24:MI:SS'),
+                  '首页自动关联',
+                  '由Patient_FrontPages触发器按住院号自动关联',
+                  COALESCE(NEW.manual_corrected, false),
+                  true,
+                  NEW.id,
+                  NEW.source_file,
+                  now()
+                )
+                ON CONFLICT (inpatient_no) DO UPDATE SET
+                  has_front_page = true,
+                  front_page_id = EXCLUDED.front_page_id,
+                  front_page_source_file = EXCLUDED.front_page_source_file,
+                  patient_name = COALESCE(NULLIF(EXCLUDED.patient_name, ''), {list_table}.patient_name),
+                  gender = EXCLUDED.gender,
+                  age = EXCLUDED.age,
+                  major_department = EXCLUDED.major_department,
+                  sub_department = EXCLUDED.sub_department,
+                  manual_corrected = EXCLUDED.manual_corrected,
+                  imported_at = now();
+              END IF;
+
+              RETURN NEW;
+            END;
+            $trigger$;
+            """
+        ).format(trigger_function=trigger_function, table=table, list_table=list_table)
+    )
+    cur.execute(
+        sql.SQL("DROP TRIGGER IF EXISTS {trigger_name} ON {table}").format(
+            trigger_name=trigger_name,
+            table=table,
+        )
+    )
+    cur.execute(
+        sql.SQL(
+            """
+            CREATE TRIGGER {trigger_name}
+            AFTER INSERT OR UPDATE OR DELETE ON {table}
+            FOR EACH ROW EXECUTE FUNCTION {trigger_function}()
+            """
+        ).format(trigger_name=trigger_name, table=table, trigger_function=trigger_function)
+    )
+
+
+def fetch_review_logs(record_id: int, limit: int = 30) -> list[dict[str, Any]]:
+    query = sql.SQL("SELECT review_logs FROM {table} WHERE id = %s").format(table=table_identifier())
+    with connect() as conn, conn.cursor() as cur:
+        cur.execute(query, (record_id,))
+        row = cur.fetchone()
+    if not row:
+        return []
+    logs = row.get("review_logs") or []
+    if not isinstance(logs, list):
+        return []
+    normalized = [row_to_json(dict(item)) for item in logs if isinstance(item, dict)]
+    normalized.sort(key=lambda item: (str(item.get("changed_at") or ""), str(item.get("id") or "")), reverse=True)
+    return normalized[:limit]
+
+
+def fetch_audit_logs(limit: int = 100) -> list[dict[str, Any]]:
+    ensure_workflow_tables()
+    query = sql.SQL(
+        """
+        SELECT
+          p.id AS record_id,
+          p.source_file,
+          p.inpatient_no,
+          p.medical_record_no,
+          p.patient_name,
+          p.primary_diagnosis,
+          p.review_status,
+          log_item.value AS log,
+          log_item.ordinality AS log_order
+        FROM {table} p
+        CROSS JOIN LATERAL jsonb_array_elements(COALESCE(p.audit_logs, '[]'::jsonb)) WITH ORDINALITY AS log_item(value, ordinality)
+        WHERE COALESCE(log_item.value->>'audit_status', '') <> 'pending'
+        ORDER BY COALESCE(log_item.value->>'updated_at', log_item.value->>'created_at', '') DESC, log_item.ordinality DESC
+        LIMIT %s
+        """
+    ).format(table=table_identifier())
+    with connect() as conn, conn.cursor() as cur:
+        cur.execute(query, (limit,))
+        rows = cur.fetchall()
+    logs: list[dict[str, Any]] = []
+    for row in rows:
+        base = row_to_json({key: value for key, value in dict(row).items() if key not in {"log", "log_order"}})
+        item = row.get("log") or {}
+        if isinstance(item, dict):
+            log = {**base, **row_to_json(dict(item))}
+            log.setdefault("record_id", base.get("record_id"))
+            log.setdefault("source_file", base.get("source_file"))
+            logs.append(log)
+    return logs
+
+
+def insert_review_log(
+    cur,
+    record_id: int,
+    source_file: str,
+    changed_fields: list[dict[str, Any]],
+    manual_note: str,
+    changed_by: str = "web",
+    ai_result: Any = None,
+) -> None:
+    changed_at = datetime.now().isoformat(timespec="seconds")
+    log = {
+        "id": datetime.now().strftime("%Y%m%d%H%M%S%f"),
+        "record_id": record_id,
+        "source_file": source_file,
+        "changed_at": changed_at,
+        "changed_by": changed_by,
+        "manual_note": manual_note,
+        "changed_fields": json_ready_deep(changed_fields),
+    }
+    if ai_result is not None:
+        log["ai_result"] = json_ready_deep(ai_result)
+    cur.execute(
+        sql.SQL(
+            """
+            UPDATE {table}
+            SET review_logs = COALESCE(review_logs, '[]'::jsonb) || %s::jsonb
+            WHERE id = %s
+            """
+        ).format(table=table_identifier()),
+        (
+            json.dumps([log], ensure_ascii=False),
+            record_id,
+        ),
+    )
+
+
+def build_record_updates(
+    record_id: int,
+    fields: dict[str, Any],
+    manual_note: str = "",
+    note_prefix: str = "人工复核",
+    force_reviewed: bool = True,
+) -> tuple[dict[str, Any], list[dict[str, Any]], dict[str, Any]]:
+    before = fetch_record(record_id)
+    updates: dict[str, Any] = {}
+    for field, value in fields.items():
+        if field not in EDITABLE_FIELDS:
+            continue
+        updates[field] = parse_field_value(field, value)
+
+    for number_field in ["medical_record_no", "front_page_medical_record_no"]:
+        if updates.get(number_field):
+            normalized_number = digits(updates[number_field], 10)
+            if normalized_number:
+                updates[number_field] = normalized_number
+
+    if "inpatient_no" in updates and updates["inpatient_no"] is not None:
+        updates["inpatient_no"] = str(updates["inpatient_no"]).strip()
+    else:
+        preview = {**before, **updates}
+        derived_inpatient_no = build_inpatient_no_from_record(preview)
+        if derived_inpatient_no:
+            updates["inpatient_no"] = derived_inpatient_no
+    if "inpatient_no" in updates and not str(updates["inpatient_no"] or "").strip():
+        raise HTTPException(status_code=400, detail="患者号不能为空")
+
+    manual_note = manual_note.strip()
+    if manual_note:
+        current = before.get("review_notes") or []
+        current_notes = current if isinstance(current, list) else [current]
+        current_notes.append(f"{note_prefix}({datetime.now().strftime('%Y-%m-%d %H:%M:%S')}): {manual_note}")
+        updates["review_notes"] = current_notes
+
+    changed_fields: list[dict[str, Any]] = []
+    for field, value in updates.items():
+        if field not in FIELD_META:
+            continue
+        old_value = before.get(field)
+        if comparable(old_value) == comparable(value):
+            continue
+        changed_fields.append(
+            {
+                "field": field,
+                "label": FIELD_META[field]["label"],
+                "old": json_ready_deep(old_value),
+                "new": json_ready_deep(value),
+            }
+        )
+
+    should_mark_manual = force_reviewed or bool(changed_fields)
+    if force_reviewed and before.get("review_status") != "reviewed":
+        changed_fields.append(
+            {
+                "field": "review_status",
+                "label": "复核状态",
+                "old": json_ready_deep(before.get("review_status")),
+                "new": "reviewed",
+            }
+        )
+    if should_mark_manual and before.get("manual_corrected") is not True:
+        changed_fields.append(
+            {
+                "field": "manual_corrected",
+                "label": "人工修正",
+                "old": json_ready_deep(before.get("manual_corrected")),
+                "new": True,
+            }
+        )
+
+    return updates, changed_fields, before
+
+
+def apply_record_updates(
+    cur,
+    record_id: int,
+    updates: dict[str, Any],
+    changed_fields: list[dict[str, Any]],
+    before: dict[str, Any],
+    manual_note: str = "",
+    force_reviewed: bool = True,
+) -> None:
+    assignments = []
+    values: list[Any] = []
+    for field, value in updates.items():
+        assignments.append(sql.SQL("{} = %s").format(sql.Identifier(field)))
+        values.append(psycopg2.extras.Json(value, dumps=lambda obj: json.dumps(obj, ensure_ascii=False)) if field in JSON_DB_FIELDS else value)
+    if force_reviewed:
+        assignments.append(sql.SQL("review_status = 'reviewed'"))
+        assignments.append(sql.SQL("manual_corrected = TRUE"))
+    elif changed_fields:
+        assignments.append(sql.SQL("manual_corrected = TRUE"))
+        if before.get("review_status") == "auto_pass":
+            assignments.append(sql.SQL("review_status = 'reviewed'"))
+
+    if assignments:
+        query = sql.SQL("UPDATE {table} SET {assignments} WHERE id = %s").format(
+            table=table_identifier(),
+            assignments=sql.SQL(", ").join(assignments),
+        )
+        cur.execute(query, [*values, record_id])
+        if cur.rowcount != 1:
+            raise HTTPException(status_code=404, detail="记录不存在")
+    if changed_fields or manual_note.strip():
+        insert_review_log(cur, record_id, before.get("source_file", ""), changed_fields, manual_note.strip())
+
+
+def safe_child(root: Path, child_name: str) -> Path:
+    if Path(child_name).name != child_name:
+        raise HTTPException(status_code=400, detail="非法文件名")
+    path = (root / child_name).resolve()
+    if root not in path.parents and path != root:
+        raise HTTPException(status_code=400, detail="非法路径")
+    return path
+
+
+def get_pdf_path(source_file: str) -> Path | None:
+    path = safe_child(PDF_DIR, source_file)
+    return path if path.exists() and path.is_file() else None
+
+
+def parse_json_content(content: str) -> Any:
+    text = content.strip()
+    if text.startswith("```"):
+        text = text.strip("`")
+        if text.startswith("json"):
+            text = text[4:].strip()
+    try:
+        return json.loads(text)
+    except json.JSONDecodeError as exc:
+        start = text.find("{")
+        end = text.rfind("}")
+        if start >= 0 and end > start:
+            try:
+                return json.loads(text[start : end + 1])
+            except json.JSONDecodeError:
+                pass
+        return {
+            "decision": "confirm",
+            "confidence": 0,
+            "summary": f"Kimi 返回 JSON 无法解析：{exc.msg}",
+            "raw_text": content[:4000],
+        }
+
+
+def render_pdf_page_data_url(pdf_path: Path, dpi: int = 96, page_index: int = 0, clip: Any | None = None) -> str:
+    try:
+        import fitz  # type: ignore[import-not-found]
+    except ImportError as exc:
+        raise HTTPException(status_code=500, detail="Web 容器缺少 PyMuPDF，无法渲染 PDF 供 AI 核验") from exc
+
+    with fitz.open(str(pdf_path)) as document:
+        if document.page_count < 1:
+            raise HTTPException(status_code=400, detail="PDF 没有可渲染页面")
+        page = document.load_page(max(0, min(page_index, document.page_count - 1)))
+        matrix = fitz.Matrix(dpi / 72, dpi / 72)
+        pixmap = page.get_pixmap(matrix=matrix, alpha=False, clip=clip)
+        encoded = base64.b64encode(pixmap.tobytes("png")).decode("utf-8")
+    return f"data:image/png;base64,{encoded}"
+
+
+def normalize_pdf_text(text: str) -> str:
+    return re.sub(r"\s+", "", text or "")
+
+
+def record_review_text(record: dict[str, Any]) -> str:
+    parts: list[str] = []
+    for key in ("review_notes", "quality_notes", "auto_corrections"):
+        value = record.get(key)
+        if isinstance(value, list):
+            parts.extend(json.dumps(item, ensure_ascii=False) if not isinstance(item, str) else item for item in value)
+        elif value:
+            parts.append(str(value))
+    return "；".join(parts)
+
+
+def relevant_pdf_modules(record: dict[str, Any]) -> list[dict[str, Any]]:
+    text = record_review_text(record)
+    normalized = normalize_pdf_text(text)
+    selected: list[dict[str, Any]] = []
+    for module in PDF_MODULE_DEFINITIONS:
+        if any(normalize_pdf_text(keyword) in normalized for keyword in module["note_keywords"]):
+            selected.append(module)
+    if not selected:
+        selected = [module for module in PDF_MODULE_DEFINITIONS if module["name"] in {"诊断表格", "手术表格"}]
+    return selected[:4]
+
+
+def clip_text_from_blocks(blocks: list[dict[str, Any]], page_index: int, clip: Any) -> str:
+    lines: list[str] = []
+    for block in blocks:
+        if block["page_index"] != page_index:
+            continue
+        rect = block["rect"]
+        if rect.intersects(clip):
+            text = " ".join(str(block["text"]).split())
+            if text:
+                lines.append(text)
+    joined = "\n".join(lines)
+    return joined[:2200]
+
+
+def add_keyword_from_value(keywords: list[str], value: Any) -> None:
+    if value in {None, ""}:
+        return
+    text = str(value).strip()
+    if len(text) >= 2:
+        keywords.append(text)
+
+
+def record_module_keywords(record: dict[str, Any], module_name: str) -> list[str]:
+    keywords: list[str] = []
+    if module_name == "诊断表格":
+        for field in ("outpatient_diagnosis", "outpatient_diagnosis_code", "primary_diagnosis", "primary_diagnosis_code", "pathology_diagnosis", "pathology_diagnosis_code"):
+            add_keyword_from_value(keywords, record.get(field))
+        diagnoses = record.get("discharge_diagnoses")
+        if isinstance(diagnoses, list):
+            for row in diagnoses[:8]:
+                if isinstance(row, dict):
+                    for key in ("出院诊断", "疾病编码", "诊断名称", "诊断编码"):
+                        add_keyword_from_value(keywords, row.get(key))
+    elif module_name == "手术表格":
+        operations = record.get("operations")
+        if isinstance(operations, list):
+            for row in operations[:6]:
+                if isinstance(row, dict):
+                    for key in ("手术操作名称", "手术操作编码", "手术操作日期"):
+                        add_keyword_from_value(keywords, row.get(key))
+    elif module_name == "地址联系人":
+        for field in ("current_address", "household_address", "employer_address", "contact_name", "contact_address", "contact_phone"):
+            add_keyword_from_value(keywords, record.get(field))
+    elif module_name == "基本信息":
+        for field in ("patient_name", "medical_record_no", "inpatient_no", "id_card_no", "admission_time", "discharge_time"):
+            add_keyword_from_value(keywords, record.get(field))
+    elif module_name == "离院费用":
+        for field in ("total_cost", "self_pay_amount", "discharge_disposition_code", "receiving_org_name"):
+            add_keyword_from_value(keywords, record.get(field))
+    return keywords
+
+
+def extract_pdf_module_context(pdf_path: Path, record: dict[str, Any]) -> dict[str, Any]:
+    try:
+        import fitz  # type: ignore[import-not-found]
+    except ImportError as exc:
+        raise HTTPException(status_code=500, detail="Web 容器缺少 PyMuPDF，无法提取 PDF 文本") from exc
+
+    modules = relevant_pdf_modules(record)
+    contexts: list[dict[str, Any]] = []
+    all_blocks: list[dict[str, Any]] = []
+    with fitz.open(str(pdf_path)) as document:
+        for page_index in range(document.page_count):
+            page = document.load_page(page_index)
+            for block in page.get_text("blocks"):
+                if len(block) < 5:
+                    continue
+                text = str(block[4] or "").strip()
+                if not text:
+                    continue
+                all_blocks.append(
+                    {
+                        "page_index": page_index,
+                        "rect": fitz.Rect(block[:4]),
+                        "text": text,
+                        "normalized": normalize_pdf_text(text),
+                    }
+                )
+
+        for module in modules:
+            module_keywords = [*module["keywords"], *record_module_keywords(record, module["name"])]
+            keywords = [normalize_pdf_text(keyword) for keyword in module_keywords]
+            hits = [block for block in all_blocks if any(keyword and keyword in block["normalized"] for keyword in keywords)]
+            if not hits:
+                continue
+            page_counts: dict[int, int] = {}
+            for hit in hits:
+                page_counts[hit["page_index"]] = page_counts.get(hit["page_index"], 0) + 1
+            page_index = max(page_counts, key=page_counts.get)
+            page = document.load_page(page_index)
+            page_hits = [hit for hit in hits if hit["page_index"] == page_index]
+            min_y = min(hit["rect"].y0 for hit in page_hits)
+            max_y = max(hit["rect"].y1 for hit in page_hits)
+            tail = float(module.get("tail") or 420)
+            clip = fitz.Rect(
+                max(page.rect.x0, page.rect.x0 + 18),
+                max(page.rect.y0, min_y - 60),
+                min(page.rect.x1, page.rect.x1 - 18),
+                min(page.rect.y1, max(max_y + tail, min_y + 220)),
+            )
+            contexts.append(
+                {
+                    "name": module["name"],
+                    "page": page_index + 1,
+                    "bbox": [round(clip.x0, 1), round(clip.y0, 1), round(clip.x1, 1), round(clip.y1, 1)],
+                    "text": clip_text_from_blocks(all_blocks, page_index, clip),
+                    "image_url": render_pdf_page_data_url(pdf_path, dpi=120, page_index=page_index, clip=clip),
+                }
+            )
+
+    full_text = "\n".join(" ".join(str(block["text"]).split()) for block in all_blocks)
+    return {
+        "modules": contexts,
+        "full_text_excerpt": full_text[:3500],
+    }
+
+
+def ai_record_snapshot(record: dict[str, Any]) -> dict[str, Any]:
+    keys = [
+        "source_file",
+        "inpatient_no",
+        "medical_record_no",
+        "front_page_medical_record_no",
+        "patient_name",
+        "gender",
+        "birth_date",
+        "age",
+        "admission_time",
+        "admission_dept",
+        "discharge_time",
+        "discharge_dept",
+        "hospital_days",
+        "outpatient_diagnosis",
+        "outpatient_diagnosis_code",
+        "primary_diagnosis",
+        "primary_diagnosis_code",
+        "primary_admission_condition",
+        "discharge_diagnoses",
+        "operations",
+        "pathology_diagnosis",
+        "pathology_diagnosis_code",
+        "total_cost",
+        "self_pay_amount",
+        "quality_status",
+        "quality_notes",
+        "review_notes",
+    ]
+    return {key: record.get(key) for key in keys}
+
+
+def build_ai_prompt(record: dict[str, Any], pdf_context: dict[str, Any]) -> str:
+    snapshot = json.dumps(ai_record_snapshot(record), ensure_ascii=False, indent=2)
+    context_for_prompt = {
+        "modules": [
+            {
+                "name": item["name"],
+                "page": item["page"],
+                "bbox": item["bbox"],
+                "pdf_text": item["text"],
+            }
+            for item in pdf_context.get("modules", [])
+        ],
+        "full_text_excerpt": pdf_context.get("full_text_excerpt", ""),
+    }
+    context_json = json.dumps(context_for_prompt, ensure_ascii=False, indent=2)
+    return f"""
+请对这份住院病案首页做视觉核验，只返回 JSON，不要输出 Markdown。
+
+核验目标：
+1. 优先核对“PDF定位文本”和对应的局部截图，局部截图是按 PDF 中文本特征截取出来的区域。
+2. 将 PDF 文本、局部截图中的可见内容，与下面的结构化字段逐项对比。
+3. 只有当复核提示被 PDF 明确证明为误报，且结构化字段与 PDF 可见内容一致、完整、无实质质量问题时，decision 才能返回 "ok"。
+4. 如果复核提示是编码缺失/格式异常/错位/不一致，而 PDF 对应区域也显示为空白、缺失、异常或不清晰，这说明问题被证实，decision 必须返回 "confirm"。
+5. 如果图片不清晰、字段不一致、编码缺失/错位、手术/诊断仍有疑点，decision 返回 "confirm"。
+6. 如果能从 PDF 读出明确修正值，请放入 suggested_updates，但 decision 仍返回 "confirm"，等待人工确认。
+7. 不要编造 PDF 中看不见的内容。
+8. 如果手术表格中能看到“手术及操作编码”列，但对应单元格为空，必须写“编码栏可见但为空白”，不要写“编码区域未在首页显示”。
+9. 手术操作名称可能因为换行被结构化解析截断；如果 PDF定位文本或局部截图中显示完整多行名称，请把完整名称放入 suggested_updates。
+
+必须返回这个 JSON 结构：
+{{
+  "decision": "ok 或 confirm",
+  "confidence": 0.0,
+  "issue_resolution": "false_positive/confirmed_problem/uncertain/update_suggested",
+  "confirmed_issue": false,
+  "summary": "一句话结论，60字以内",
+  "method": "Kimi视觉核验：PDF文本定位+局部截图，对照复核定位和结构化字段",
+  "evidence": [
+    {{"target": "核验点", "pdf_value": "PDF图片值", "structured_value": "结构化值", "result": "match/mismatch/uncertain", "note": "30字以内"}}
+  ],
+  "suggested_updates": [
+    {{"field": "字段名", "current": "当前值", "pdf_value": "PDF图片建议值", "reason": ""}}
+  ]
+}}
+
+evidence 最多返回 3 条，只保留最关键依据；没有明确修正值时 suggested_updates 返回 []。
+
+结构化字段：
+{snapshot}
+
+PDF定位文本：
+{context_json}
+""".strip()
+
+
+def call_kimi_ai_review(record: dict[str, Any]) -> dict[str, Any]:
+    settings = public_kimi_settings(load_local_settings().get("kimi") or {})
+    if not settings["available"]:
+        raise HTTPException(status_code=400, detail="Kimi AI 核验未启用或未配置 API Key")
+    pdf_path = get_pdf_path(record.get("source_file") or "")
+    if not pdf_path:
+        raise HTTPException(status_code=404, detail="PDF 文件不存在，无法 AI 核验")
+
+    pdf_context = extract_pdf_module_context(pdf_path, record)
+    content_parts: list[dict[str, Any]] = [{"type": "text", "text": build_ai_prompt(record, pdf_context)}]
+    for item in pdf_context.get("modules", [])[:4]:
+        content_parts.append({"type": "text", "text": f"局部截图：{item['name']}，第 {item['page']} 页，bbox={item['bbox']}"})
+        content_parts.append({"type": "image_url", "image_url": {"url": item["image_url"]}})
+    if not pdf_context.get("modules"):
+        content_parts.append({"type": "text", "text": "未能定位到局部模块，以下为首页整页截图。"})
+        content_parts.append({"type": "image_url", "image_url": {"url": render_pdf_page_data_url(pdf_path)}})
+
+    payload = {
+        "model": settings["model"],
+        "temperature": 0.6,
+        "max_tokens": 1600,
+        "thinking": {"type": "disabled"},
+        "response_format": {"type": "json_object"},
+        "messages": [
+            {"role": "system", "content": "你是严谨的病案首页视觉核验助手，只输出 JSON。"},
+            {
+                "role": "user",
+                "content": content_parts,
+            },
+        ],
+    }
+    request = urllib.request.Request(
+        f"{settings['api_base'].rstrip('/')}/chat/completions",
+        data=json.dumps(payload, ensure_ascii=False).encode("utf-8"),
+        headers={
+            "Content-Type": "application/json",
+            "Authorization": f"Bearer {kimi_api_key()}",
+        },
+        method="POST",
+    )
+    try:
+        with urllib.request.urlopen(request, timeout=90) as response:
+            data = json.loads(response.read().decode("utf-8"))
+    except urllib.error.HTTPError as exc:
+        detail = exc.read().decode("utf-8", errors="replace")
+        raise HTTPException(status_code=502, detail=f"Kimi API 返回错误 {exc.code}: {detail}") from exc
+    except urllib.error.URLError as exc:
+        raise HTTPException(status_code=502, detail=f"Kimi API 调用失败：{exc}") from exc
+
+    message = data["choices"][0].get("message") or {}
+    content = message.get("content") or message.get("reasoning_content") or ""
+    parsed = parse_json_content(content)
+    if not isinstance(parsed, dict):
+        parsed = {"decision": "confirm", "summary": "Kimi 返回 JSON 不是对象", "raw_response": content}
+    return {
+        "model": settings["model"],
+        "checked_at": datetime.now().isoformat(timespec="seconds"),
+        "pdf_context": {
+            "modules": [
+                {key: item[key] for key in ("name", "page", "bbox", "text")}
+                for item in pdf_context.get("modules", [])
+            ],
+            "full_text_excerpt": pdf_context.get("full_text_excerpt", ""),
+        },
+        "raw_response": content,
+        "parsed": parsed,
+    }
+
+
+def ai_bool(value: Any) -> bool:
+    if isinstance(value, bool):
+        return value
+    if isinstance(value, (int, float)):
+        return bool(value)
+    return str(value or "").strip().lower() in {"true", "yes", "1", "是", "有", "确认"}
+
+
+def ai_join_text(value: Any) -> str:
+    if isinstance(value, str):
+        return value
+    if isinstance(value, list):
+        return " ".join(ai_join_text(item) for item in value)
+    if isinstance(value, dict):
+        return " ".join(ai_join_text(item) for item in value.values())
+    return str(value or "")
+
+
+def ai_has_confirmed_problem(parsed: dict[str, Any]) -> bool:
+    resolution = str(parsed.get("issue_resolution") or "").strip().lower()
+    if ai_bool(parsed.get("confirmed_issue")):
+        return True
+    if resolution in {"confirmed_problem", "uncertain", "update_suggested", "problem", "待确认", "已证实"}:
+        return True
+
+    suggested_updates = parsed.get("suggested_updates")
+    if isinstance(suggested_updates, list) and suggested_updates:
+        return True
+
+    evidence = parsed.get("evidence")
+    if isinstance(evidence, list):
+        for item in evidence:
+            if not isinstance(item, dict):
+                continue
+            result = str(item.get("result") or "").strip().lower()
+            if result in {"mismatch", "uncertain", "missing", "problem"}:
+                return True
+
+    if resolution in {"false_positive", "ok", "no_issue", "误报", "无问题"}:
+        return False
+
+    text = ai_join_text(
+        {
+            "summary": parsed.get("summary"),
+            "evidence": parsed.get("evidence"),
+        }
+    )
+    if any(keyword in text for keyword in ("需人工", "需要人工", "待确认")):
+        return True
+    has_problem_word = any(keyword in text for keyword in AI_CONFIRMED_PROBLEM_KEYWORDS)
+    has_qualifier = any(keyword in text for keyword in AI_CONFIRMED_PROBLEM_QUALIFIERS)
+    return bool(has_problem_word and has_qualifier)
+
+
+def ai_status_from_result(result: dict[str, Any]) -> str:
+    parsed = result.get("parsed") if isinstance(result.get("parsed"), dict) else {}
+    decision = str(parsed.get("decision") or "").lower()
+    confidence = parsed.get("confidence")
+    try:
+        confidence_value = float(confidence)
+    except (TypeError, ValueError):
+        confidence_value = 0.0
+    if ai_has_confirmed_problem(parsed):
+        return AI_PENDING_STATUS
+    if decision in {"ok", "pass", "no_issue", "无问题", "通过"} and confidence_value >= 0.65:
+        return AI_NO_ISSUE_STATUS
+    return AI_PENDING_STATUS
+
+
+def ai_review_note(status: str, result: dict[str, Any]) -> str:
+    parsed = result.get("parsed") if isinstance(result.get("parsed"), dict) else {}
+    summary = str(parsed.get("summary") or "").strip()
+    confidence = parsed.get("confidence")
+    confidence_text = f"，置信度 {confidence}" if confidence not in {None, ""} else ""
+    verdict = "未发现问题" if status == AI_NO_ISSUE_STATUS else "需要人工确认"
+    return f"Kimi视觉核验({datetime.now().strftime('%Y-%m-%d %H:%M:%S')}): {verdict}{confidence_text}。{summary}".strip()
+
+
+def apply_ai_review(record_id: int) -> dict[str, Any]:
+    before = fetch_record(record_id)
+    result = call_kimi_ai_review(before)
+    new_status = ai_status_from_result(result)
+    note = ai_review_note(new_status, result)
+    previous_notes = before.get("review_notes") if isinstance(before.get("review_notes"), list) else []
+    new_notes = [note] if new_status == AI_NO_ISSUE_STATUS else [*previous_notes, note]
+    changed_fields = [
+        {
+            "field": "ai_method",
+            "label": "AI核验方式",
+            "old": "",
+            "new": "Kimi视觉核验：PDF文本定位+局部截图，对照复核定位和结构化字段",
+        },
+        {
+            "field": "review_status",
+            "label": "复核状态",
+            "old": json_ready_deep(before.get("review_status")),
+            "new": new_status,
+        },
+        {
+            "field": "review_notes",
+            "label": "复核备注",
+            "old": json_ready_deep(before.get("review_notes")),
+            "new": json_ready_deep(new_notes),
+        },
+    ]
+    with connect() as conn, conn.cursor() as cur:
+        cur.execute(
+            sql.SQL("UPDATE {table} SET review_status = %s, review_notes = %s::jsonb WHERE id = %s").format(table=table_identifier()),
+            (new_status, json.dumps(new_notes, ensure_ascii=False), record_id),
+        )
+        insert_review_log(cur, record_id, before.get("source_file", ""), changed_fields, note, changed_by="Kimi AI", ai_result=result)
+        conn.commit()
+    return {"record_id": record_id, "status": new_status, "result": result}
+
+
+def apply_ai_review_with_retry(record_id: int, attempts: int = 2) -> dict[str, Any]:
+    last_exc: Exception | None = None
+    for attempt in range(attempts):
+        try:
+            return apply_ai_review(record_id)
+        except Exception as exc:  # noqa: BLE001
+            last_exc = exc
+            if attempt + 1 < attempts:
+                time.sleep(4 + attempt * 4)
+    if last_exc:
+        raise last_exc
+    raise RuntimeError("AI核验失败")
+
+
+def ai_target_ids(scope: str, record_id: int | None) -> list[int]:
+    if scope not in {"current", "five", "all"}:
+        raise HTTPException(status_code=400, detail="AI核验范围只能是 current/five/all")
+    if scope in {"current", "five"} and not record_id:
+        raise HTTPException(status_code=400, detail="缺少当前记录 ID")
+    if scope == "current":
+        return [int(record_id)]
+
+    where = sql.SQL("review_status = 'needs_review'")
+    params: list[Any] = []
+    if scope == "five":
+        where = sql.SQL("review_status = 'needs_review' AND id > %s")
+        params.append(record_id)
+        limit_sql = sql.SQL("LIMIT 5")
+    else:
+        if record_id:
+            where = sql.SQL("review_status = 'needs_review' AND id > %s")
+            params.append(record_id)
+        limit_sql = sql.SQL("")
+    query = sql.SQL("SELECT id FROM {table} WHERE {where} ORDER BY id {limit}").format(
+        table=table_identifier(),
+        where=where,
+        limit=limit_sql,
+    )
+    with connect() as conn, conn.cursor() as cur:
+        cur.execute(query, params)
+        rows = cur.fetchall()
+    return [int(row["id"]) for row in rows]
+
+
+def update_ai_job(**updates: Any) -> dict[str, Any]:
+    with AI_JOB_LOCK:
+        AI_REVIEW_JOB.update(updates)
+        return dict(AI_REVIEW_JOB)
+
+
+def is_ai_stop_error(message: str) -> bool:
+    lower = message.lower()
+    return any(marker in lower for marker in AI_STOP_ERROR_MARKERS)
+
+
+def append_ai_job_error(record_id: int, message: str) -> None:
+    errors = AI_REVIEW_JOB.setdefault("errors", [])
+    errors.append({"record_id": record_id, "message": message})
+    if len(errors) > AI_JOB_ERROR_LIMIT:
+        del errors[: len(errors) - AI_JOB_ERROR_LIMIT]
+
+
+def run_ai_review_job(scope: str, ids: list[int]) -> None:
+    settings = public_kimi_settings(load_local_settings().get("kimi") or {})
+    concurrency = min(max(1, int(settings.get("concurrency") or 3)), max(1, len(ids)))
+    stop_event = threading.Event()
+    update_ai_job(
+        running=True,
+        scope=scope,
+        total=len(ids),
+        processed=0,
+        ok=0,
+        pending=0,
+        failed=0,
+        concurrency=concurrency,
+        message="AI核验中",
+        errors=[],
+        started_at=datetime.now().isoformat(timespec="seconds"),
+        finished_at="",
+        last_record_id=None,
+    )
+    record_queue: Queue[int] = Queue()
+    for record_id in ids:
+        record_queue.put(record_id)
+
+    def worker() -> None:
+        try:
+            while not stop_event.is_set():
+                try:
+                    record_id = record_queue.get_nowait()
+                except Empty:
+                    return
+                try:
+                    item = apply_ai_review_with_retry(record_id)
+                    status = item.get("status")
+                    with AI_JOB_LOCK:
+                        AI_REVIEW_JOB["processed"] += 1
+                        AI_REVIEW_JOB["last_record_id"] = record_id
+                        if status == AI_NO_ISSUE_STATUS:
+                            AI_REVIEW_JOB["ok"] += 1
+                        elif status == AI_PENDING_STATUS:
+                            AI_REVIEW_JOB["pending"] += 1
+                except Exception as exc:  # noqa: BLE001
+                    message = str(getattr(exc, "detail", exc))
+                    with AI_JOB_LOCK:
+                        AI_REVIEW_JOB["processed"] += 1
+                        AI_REVIEW_JOB["failed"] += 1
+                        AI_REVIEW_JOB["last_record_id"] = record_id
+                        AI_REVIEW_JOB["message"] = message
+                        append_ai_job_error(record_id, message)
+                        if is_ai_stop_error(message):
+                            AI_REVIEW_JOB["message"] = f"AI核验已暂停：{message}"
+                            stop_event.set()
+                finally:
+                    record_queue.task_done()
+        except Exception as exc:  # noqa: BLE001
+            with AI_JOB_LOCK:
+                AI_REVIEW_JOB["message"] = str(exc)
+
+    workers = [
+        threading.Thread(target=worker, name=f"kimi-ai-worker-{index + 1}", daemon=True)
+        for index in range(concurrency)
+    ]
+    for thread in workers:
+        thread.start()
+    for thread in workers:
+        thread.join()
+    refresh_status_snapshot(source="ai")
+    with AI_JOB_LOCK:
+        failed = int(AI_REVIEW_JOB.get("failed") or 0)
+        stopped = stop_event.is_set()
+        message = str(AI_REVIEW_JOB.get("message") or "")
+    update_ai_job(
+        running=False,
+        message=message if stopped else ("AI核验完成" if failed == 0 else f"AI核验完成，失败 {failed} 条"),
+        finished_at=datetime.now().isoformat(timespec="seconds"),
+    )
+
+
+def mark_ai_no_issue_reviewed() -> int:
+    changed_at = datetime.now().isoformat(timespec="seconds")
+    note = f"批量确认AI复核-无问题({datetime.now().strftime('%Y-%m-%d %H:%M:%S')}): 确认为已人工复核"
+    query = sql.SQL(
+        """
+        UPDATE {table}
+        SET review_status = 'reviewed',
+            manual_corrected = TRUE,
+            review_logs = COALESCE(review_logs, '[]'::jsonb) || jsonb_build_array(
+                jsonb_build_object(
+                    'id', %s || id::text,
+                    'record_id', id,
+                    'source_file', source_file,
+                    'changed_at', %s,
+                    'changed_by', 'web',
+                    'manual_note', %s,
+                    'changed_fields', jsonb_build_array(
+                        jsonb_build_object('field', 'review_status', 'label', '复核状态', 'old', review_status, 'new', 'reviewed'),
+                        jsonb_build_object('field', 'manual_corrected', 'label', '人工修正', 'old', COALESCE(manual_corrected, false), 'new', true)
+                    )
+                )
+            )
+        WHERE review_status = 'AI复核-无问题'
+        """
+    ).format(table=table_identifier())
+    with connect() as conn, conn.cursor() as cur:
+        cur.execute(query, (datetime.now().strftime("%Y%m%d%H%M%S%f"), changed_at, note))
+        count = cur.rowcount
+        conn.commit()
+    refresh_status_snapshot(source="bulk")
+    return int(count)
+
+
+def submit_reviewed_records() -> int:
+    changed_at = datetime.now().isoformat(timespec="seconds")
+    note = f"一键提交已人工复核项目({datetime.now().strftime('%Y-%m-%d %H:%M:%S')}): 已从患者首页复核工作台隐藏"
+    query = sql.SQL(
+        """
+        UPDATE {table}
+        SET review_status = '已提交',
+            review_logs = COALESCE(review_logs, '[]'::jsonb) || jsonb_build_array(
+                jsonb_build_object(
+                    'id', %s || id::text,
+                    'record_id', id,
+                    'source_file', source_file,
+                    'changed_at', %s,
+                    'changed_by', 'web',
+                    'manual_note', %s,
+                    'changed_fields', jsonb_build_array(
+                        jsonb_build_object('field', 'review_status', 'label', '复核状态', 'old', review_status, 'new', '已提交')
+                    )
+                )
+            )
+        WHERE review_status = 'reviewed'
+        """
+    ).format(table=table_identifier())
+    with connect() as conn, conn.cursor() as cur:
+        cur.execute(query, (datetime.now().strftime("%Y%m%d%H%M%S%f"), changed_at, note))
+        count = cur.rowcount
+        conn.commit()
+    refresh_status_snapshot(source="submit")
+    return int(count)
+
+
+
+def parse_field_value(field: str, value: Any) -> Any:
+    meta = FIELD_META[field]
+    field_type = meta["type"]
+    if value == "":
+        return [] if field in JSON_FIELDS else None
+    if field in JSON_FIELDS:
+        if isinstance(value, str):
+            try:
+                return json.loads(value)
+            except json.JSONDecodeError as exc:
+                raise HTTPException(status_code=400, detail=f"{meta['label']} 不是合法 JSON：{exc}") from exc
+        return value
+    if field in INTEGER_FIELDS:
+        return None if value is None else int(value)
+    if field in NUMERIC_FIELDS:
+        return None if value is None else Decimal(str(value))
+    return value
+
+
+def fetch_record(record_id: int) -> dict[str, Any]:
+    query = sql.SQL("SELECT * FROM {table} WHERE id = %s").format(table=table_identifier())
+    with connect() as conn, conn.cursor() as cur:
+        cur.execute(query, (record_id,))
+        row = cur.fetchone()
+    if not row:
+        raise HTTPException(status_code=404, detail="记录不存在")
+    record = row_to_json(dict(row))
+    record["pdf_url"] = f"/api/pdf/{record['source_file']}" if get_pdf_path(record["source_file"]) else ""
+
+    record["review_logs"] = fetch_review_logs(record_id)
+    record["last_activity_at"] = record["review_logs"][0].get("changed_at") if record["review_logs"] else None
+    return record
+
+
+@app.get("/")
+def index():
+    return FileResponse(STATIC_DIR / "index.html")
+
+
+@app.get("/favicon.ico")
+def favicon():
+    return Response(status_code=204)
+
+
+@app.post("/api/auth/login")
+def login(payload: LoginPayload, response: Response):
+    user = authenticate_user(payload.username, payload.password)
+    if not user:
+        raise HTTPException(status_code=401, detail="用户名或密码错误")
+    token = secrets.token_urlsafe(32)
+    SESSIONS[token] = {
+        **user,
+        "login_at": datetime.now().isoformat(timespec="seconds"),
+    }
+    response.set_cookie(
+        SESSION_COOKIE,
+        token,
+        httponly=True,
+        samesite="lax",
+        max_age=12 * 60 * 60,
+    )
+    return {"authenticated": True, "user": SESSIONS[token]}
+
+
+@app.post("/api/auth/logout")
+def logout(request: Request, response: Response):
+    token = request.cookies.get(SESSION_COOKIE, "")
+    if token:
+        SESSIONS.pop(token, None)
+    response.delete_cookie(SESSION_COOKIE)
+    return {"ok": True}
+
+
+@app.get("/api/auth/me")
+def auth_me(request: Request):
+    user = session_from_request(request)
+    if not user:
+        return {"authenticated": False, "user": None}
+    return {"authenticated": True, "user": user}
+
+
+@app.get("/api/status")
+def status():
+    data = load_local_settings()
+    snapshot = data.get("status_snapshot") or default_status_snapshot()
+    system = normalize_system_settings(data.get("system") or {})
+    snapshot["next_check_at"] = next_status_check_at(system)
+    try:
+        query = sql.SQL(
+            """
+            SELECT
+              count(*) AS total,
+              count(*) FILTER (WHERE review_status IN ('needs_review', 'AI复核-待确认')) AS review_needed,
+              count(*) FILTER (WHERE review_status = 'needs_review') AS needs_review,
+              count(*) FILTER (WHERE review_status = 'AI复核-无问题') AS ai_passed,
+              count(*) FILTER (WHERE review_status = 'AI复核-待确认') AS ai_pending,
+              count(*) FILTER (WHERE review_status = 'reviewed') AS reviewed,
+              count(*) FILTER (WHERE review_status = '已提交') AS submitted,
+              count(*) FILTER (WHERE manual_corrected IS TRUE) AS manual_corrected
+            FROM {table}
+            """
+        ).format(table=table_identifier())
+        with connect() as conn, conn.cursor() as cur:
+            cur.execute(query)
+            snapshot.update(row_to_json(dict(cur.fetchone())))
+        snapshot["database"] = "online"
+        snapshot["message"] = snapshot.get("message") or "连接正常"
+        snapshot["checked_at"] = status_now().isoformat(timespec="seconds")
+    except Exception as exc:  # noqa: BLE001
+        snapshot["database"] = "offline"
+        snapshot["message"] = str(exc)
+    return snapshot
+
+
+@app.get("/api/schema")
+def schema():
+    return {"groups": FIELD_GROUPS}
+
+
+@app.get("/api/overview")
+def overview():
+    query = sql.SQL(
+        """
+        SELECT
+          count(*) AS total,
+          count(*) FILTER (WHERE review_status IN ('needs_review', 'AI复核-待确认')) AS review_queue,
+          count(*) FILTER (WHERE review_status = 'needs_review') AS needs_review,
+          count(*) FILTER (WHERE review_status = 'AI复核-无问题') AS ai_passed,
+          count(*) FILTER (WHERE review_status = 'AI复核-待确认') AS ai_pending,
+          count(*) FILTER (WHERE review_status = 'auto_corrected') AS auto_corrected,
+          count(*) FILTER (WHERE review_status = 'reviewed') AS reviewed,
+          count(*) FILTER (WHERE review_status = '已提交') AS submitted,
+          count(*) FILTER (WHERE review_status = 'auto_pass') AS auto_passed,
+          count(*) FILTER (WHERE manual_corrected IS TRUE) AS manual_corrected
+        FROM {table}
+        """
+    ).format(table=table_identifier())
+    with connect() as conn, conn.cursor() as cur:
+        cur.execute(query)
+        summary = row_to_json(dict(cur.fetchone()))
+    summary.update({"audit_total": 0, "audit_pending": 0, "audit_passed": 0, "audit_failed": 0, "audit_unsure": 0})
+    return {
+        "summary": summary,
+        "recent_logs": [],
+    }
+
+
+@app.get("/api/records")
+def list_records(q: str = "", status_filter: str = "review_all", limit: int = 300, offset: int = 0):
+    limit = max(50, min(int(limit), 500))
+    offset = max(0, int(offset))
+    clauses = []
+    params: list[Any] = []
+    if q:
+        like = f"%{q}%"
+        clauses.append("(source_file ILIKE %s OR inpatient_no ILIKE %s OR medical_record_no ILIKE %s OR patient_name ILIKE %s OR primary_diagnosis ILIKE %s)")
+        params.extend([like, like, like, like, like])
+    if status_filter == "review_all":
+        clauses.append("review_status IN ('needs_review', 'reviewed', 'AI复核-无问题', 'AI复核-待确认')")
+    elif status_filter != "all":
+        if status_filter == "reviewed":
+            clauses.append("review_status = 'reviewed'")
+        else:
+            clauses.append("review_status = %s")
+            params.append(status_filter)
+    where_sql = sql.SQL("")
+    if clauses:
+        where_sql = sql.SQL("WHERE ") + sql.SQL(" AND ").join(sql.SQL(clause) for clause in clauses)
+
+    query = sql.SQL(
+        """
+        SELECT
+          id, source_file, inpatient_no, medical_record_no, patient_name, review_status, manual_corrected,
+          major_department, discharge_dept, primary_diagnosis, primary_diagnosis_code,
+          contact_phone,
+          NULLIF(review_logs->-1->>'changed_at', '')::timestamp AS last_activity_at
+        FROM {table}
+        {where}
+        ORDER BY
+          last_activity_at DESC NULLS LAST,
+          CASE review_status
+            WHEN 'needs_review' THEN 1
+            WHEN 'AI复核-待确认' THEN 2
+            WHEN 'AI复核-无问题' THEN 3
+            WHEN 'auto_corrected' THEN 4
+            WHEN 'reviewed' THEN 5
+            WHEN '已提交' THEN 6
+            ELSE 7
+          END,
+          id
+        LIMIT %s OFFSET %s
+        """
+    ).format(table=table_identifier(), where=where_sql)
+    with connect() as conn, conn.cursor() as cur:
+        cur.execute(query, [*params, limit + 1, offset])
+        rows = [row_to_json(dict(row)) for row in cur.fetchall()]
+    has_more = len(rows) > limit
+    rows = rows[:limit]
+    for row in rows:
+        row["has_pdf"] = get_pdf_path(row["source_file"]) is not None
+
+    return {"records": rows, "limit": limit, "offset": offset, "has_more": has_more}
+
+
+@app.get("/api/records/{record_id}")
+def get_record(record_id: int):
+    return {"record": fetch_record(record_id)}
+
+
+@app.post("/api/records/{record_id}")
+def update_record(record_id: int, payload: UpdatePayload):
+    updates, changed_fields, before = build_record_updates(
+        record_id,
+        payload.fields,
+        payload.manual_note,
+        payload.note_prefix or "人工复核",
+        force_reviewed=True,
+    )
+    if not updates and not changed_fields:
+        raise HTTPException(status_code=400, detail="没有可保存字段")
+    with connect() as conn, conn.cursor() as cur:
+        apply_record_updates(cur, record_id, updates, changed_fields, before, payload.manual_note, force_reviewed=True)
+        conn.commit()
+    return {"ok": True, "record": fetch_record(record_id)}
+
+
+@app.post("/api/audit/sample")
+def create_audit_sample(source: str = "reviewed", count: int = 5):
+    ensure_workflow_tables()
+    if source not in {"reviewed", "auto_pass"}:
+        raise HTTPException(status_code=400, detail="抽查来源只能是 reviewed 或 auto_pass")
+    count = max(1, min(int(count), 50))
+    status_clause = "review_status = 'reviewed'" if source == "reviewed" else "review_status = 'auto_pass'"
+    query = sql.SQL(
+        """
+        SELECT *
+        FROM {table}
+        WHERE {status_clause}
+        ORDER BY random()
+        LIMIT %s
+        """
+    ).format(table=table_identifier(), status_clause=sql.SQL(status_clause))
+    with connect() as conn, conn.cursor() as cur:
+        cur.execute(query, (count,))
+        rows = [dict(row) for row in cur.fetchall()]
+    return {
+        "audit_source": source,
+        "records": [fetch_record(int(row["id"])) for row in rows],
+    }
+
+
+@app.get("/api/audit/logs")
+def list_audit_logs(limit: int = 100):
+    return {"logs": fetch_audit_logs(max(1, min(int(limit), 500)))}
+
+
+@app.post("/api/audit/classify")
+def classify_audit(payload: AuditClassifyPayload):
+    ensure_workflow_tables()
+    if payload.audit_status not in {"passed", "failed", "unsure"}:
+        raise HTTPException(status_code=400, detail="抽查归类只能是 passed/failed/unsure")
+    if payload.audit_source not in {"reviewed", "auto_pass"}:
+        raise HTTPException(status_code=400, detail="抽查来源只能是 reviewed 或 auto_pass")
+
+    updates, changed_fields, before = build_record_updates(
+        payload.record_id,
+        payload.fields,
+        "",
+        "抽查",
+        force_reviewed=False,
+    )
+    snapshot = {
+        key: json_ready_deep(before.get(key))
+        for key in [
+            "source_file",
+            "medical_record_no",
+            "patient_name",
+            "primary_diagnosis",
+            "primary_diagnosis_code",
+            "discharge_diagnoses",
+            "operations",
+            "review_status",
+        ]
+    }
+    now = datetime.now().isoformat(timespec="seconds")
+    log = {
+        "id": datetime.now().strftime("%Y%m%d%H%M%S%f"),
+        "record_id": payload.record_id,
+        "source_file": before.get("source_file"),
+        "audit_source": payload.audit_source,
+        "audit_status": payload.audit_status,
+        "audit_notes": payload.audit_notes.strip(),
+        "ai_result": None,
+        "snapshot": {**snapshot, "changed_fields": changed_fields},
+        "created_at": now,
+        "updated_at": now,
+    }
+    with connect() as conn, conn.cursor() as cur:
+        apply_record_updates(
+            cur,
+            payload.record_id,
+            updates,
+            changed_fields,
+            before,
+            payload.audit_notes if changed_fields else "",
+            force_reviewed=False,
+        )
+        cur.execute(
+            sql.SQL(
+                """
+                UPDATE {table}
+                SET audit_logs = COALESCE(audit_logs, '[]'::jsonb) || %s::jsonb
+                WHERE id = %s
+                """
+            ).format(table=table_identifier()),
+            (
+                json.dumps([json_ready_deep(log)], ensure_ascii=False),
+                payload.record_id,
+            ),
+        )
+        conn.commit()
+    return {"ok": True, "log": row_to_json(log), "record": fetch_record(payload.record_id)}
+
+
+@app.post("/api/audit/logs/{audit_id}")
+def update_audit_log(audit_id: int, payload: AuditPayload):
+    raise HTTPException(status_code=410, detail="抽查日志已并入主表，请使用归类保存接口")
+
+
+@app.get("/api/settings")
+def get_settings():
+    return public_settings()
+
+
+@app.post("/api/settings/status/check")
+def check_status_now():
+    return {"status_snapshot": refresh_status_snapshot(source="manual")}
+
+
+@app.post("/api/settings/system")
+def update_system_settings(payload: SystemSettingsPayload):
+    data = load_local_settings()
+    system = normalize_system_settings(data.get("system") or {})
+    system["status_check_time"] = normalize_status_check_time(payload.status_check_time)
+    data["system"] = system
+    snapshot = data.get("status_snapshot") or default_status_snapshot()
+    snapshot["next_check_at"] = next_status_check_at(system)
+    data["status_snapshot"] = snapshot
+    save_local_settings(data)
+    return public_settings()
+
+
+@app.post("/api/settings/kimi")
+def update_kimi_settings(payload: KimiSettingsPayload):
+    data = load_local_settings()
+    data["kimi"] = normalize_kimi_settings(
+        {
+            "enabled": payload.enabled,
+            "model": payload.model,
+            "api_base": payload.api_base,
+            "concurrency": payload.concurrency,
+        }
+    )
+    save_local_settings(data)
+    return public_settings()
+
+
+@app.get("/api/ai/config")
+def ai_config():
+    data = load_local_settings()
+    return {"kimi": public_kimi_settings(data.get("kimi") or {})}
+
+
+@app.get("/api/ai/review/status")
+def ai_review_status():
+    with AI_JOB_LOCK:
+        return dict(AI_REVIEW_JOB)
+
+
+@app.post("/api/ai/review/approve-no-issue")
+def approve_ai_no_issue():
+    count = mark_ai_no_issue_reviewed()
+    return {"ok": True, "updated": count, "status_snapshot": load_local_settings().get("status_snapshot")}
+
+
+@app.post("/api/ai/review")
+def start_ai_review(payload: AiReviewPayload):
+    settings = public_kimi_settings(load_local_settings().get("kimi") or {})
+    if not settings["available"]:
+        raise HTTPException(status_code=400, detail="Kimi AI 核验未启用或未配置 API Key")
+    with AI_JOB_LOCK:
+        if AI_REVIEW_JOB.get("running"):
+            raise HTTPException(status_code=409, detail="已有 AI 核验任务正在运行")
+    ids = ai_target_ids(payload.scope, payload.record_id)
+    if not ids:
+        raise HTTPException(status_code=400, detail="当前范围没有可 AI 核验的需复核记录")
+    thread = threading.Thread(target=run_ai_review_job, args=(payload.scope, ids), name="kimi-ai-review", daemon=True)
+    thread.start()
+    time.sleep(0.1)
+    with AI_JOB_LOCK:
+        return dict(AI_REVIEW_JOB)
+
+
+@app.post("/api/settings/submit-reviewed")
+def submit_reviewed():
+    count = submit_reviewed_records()
+    return {"ok": True, "updated": count, "status_snapshot": load_local_settings().get("status_snapshot")}
+
+
+@app.post("/api/settings/users")
+def create_user(payload: UserPayload):
+    data = load_local_settings()
+    username = validate_local_username(payload.username, data)
+    if not payload.password:
+        raise HTTPException(status_code=400, detail="新用户必须设置密码")
+    user = {
+        "username": username,
+        "permissions": clean_permissions(payload.permissions),
+        "created_at": datetime.now().isoformat(timespec="seconds"),
+    }
+    user.update(password_hash(payload.password))
+    data.setdefault("users", []).append(user)
+    save_local_settings(data)
+    return public_settings()
+
+
+@app.post("/api/settings/users/{username}")
+def update_user(username: str, payload: UserUpdatePayload):
+    if username == admin_username():
+        raise HTTPException(status_code=400, detail="环境变量管理员不能在网页端编辑")
+    data = load_local_settings()
+    index = local_user_index(data, username)
+    if index is None:
+        raise HTTPException(status_code=404, detail="只能编辑本地配置用户")
+    user = data["users"][index]
+    new_username = validate_local_username(payload.username or username, data, current_username=username)
+    user["username"] = new_username
+    user["permissions"] = clean_permissions(payload.permissions)
+    if payload.password:
+        user.update(password_hash(payload.password))
+    user["updated_at"] = datetime.now().isoformat(timespec="seconds")
+    save_local_settings(data)
+    return public_settings()
+
+
+@app.post("/api/settings/users/{username}/password")
+def update_user_password(username: str, payload: PasswordPayload):
+    if username == admin_username():
+        raise HTTPException(status_code=400, detail="环境变量管理员密码请在 .env 中修改")
+    if not payload.password:
+        raise HTTPException(status_code=400, detail="密码不能为空")
+    data = load_local_settings()
+    index = local_user_index(data, username)
+    if index is None:
+        raise HTTPException(status_code=404, detail="只能修改本地配置用户")
+    data["users"][index].update(password_hash(payload.password))
+    data["users"][index]["updated_at"] = datetime.now().isoformat(timespec="seconds")
+    save_local_settings(data)
+    return public_settings()
+
+
+@app.post("/api/settings/users/{username}/permissions")
+def update_user_permissions(username: str, payload: PermissionPayload):
+    if username == admin_username():
+        raise HTTPException(status_code=400, detail="环境变量管理员不能在网页端编辑")
+    data = load_local_settings()
+    for user in data.get("users", []):
+        if user.get("username") == username:
+            user["permissions"] = clean_permissions(payload.permissions)
+            user["updated_at"] = datetime.now().isoformat(timespec="seconds")
+            save_local_settings(data)
+            return public_settings()
+    raise HTTPException(status_code=404, detail="只能修改本地配置用户")
+
+
+@app.delete("/api/settings/users/{username}")
+def delete_user(username: str):
+    if username == admin_username():
+        raise HTTPException(status_code=400, detail="环境变量管理员不能删除")
+    data = load_local_settings()
+    users = data.get("users", [])
+    next_users = [user for user in users if user.get("username") != username]
+    if len(next_users) == len(users):
+        raise HTTPException(status_code=404, detail="只能删除本地配置用户")
+    data["users"] = next_users
+    save_local_settings(data)
+    return public_settings()
+
+
+@app.get("/api/pdf/{source_file:path}")
+def pdf_file(source_file: str, request: Request):
+    referer = request.headers.get("referer", "")
+    host = request.headers.get("host", "")
+    if not referer:
+        raise HTTPException(status_code=403, detail="PDF 只能在工作台内预览")
+    if host and urlparse(referer).netloc != host:
+        raise HTTPException(status_code=403, detail="PDF 只能在同源工作台内预览")
+    path = get_pdf_path(source_file)
+    if not path:
+        raise HTTPException(status_code=404, detail="PDF 文件不存在")
+    return FileResponse(
+        path,
+        media_type="application/pdf",
+        headers={
+            "Content-Disposition": "inline",
+            "Cache-Control": "no-store",
+            "X-Content-Type-Options": "nosniff",
+        },
+    )
diff --git a/患者首页处理/数据可视化网页端/app/static/app.js b/患者首页处理/数据可视化网页端/app/static/app.js
new file mode 100644
index 0000000..7d1b8c4
--- /dev/null
+++ b/患者首页处理/数据可视化网页端/app/static/app.js
@@ -0,0 +1,1393 @@
+const state = {
+  records: [],
+  recordLimit: 300,
+  recordOffset: 0,
+  recordHasMore: false,
+  recordsLoading: false,
+  selectedId: null,
+  selectedRecord: null,
+  schema: [],
+  view: localStorage.getItem("frontPageReviewView") || "pdf",
+  activePage: "overview",
+  overview: null,
+  auditSamples: [],
+  auditCurrentId: null,
+  auditView: localStorage.getItem("frontPageAuditView") || "pdf",
+  auditLogs: [],
+  settings: null,
+  ai: null,
+  aiPolling: null,
+  currentUser: null,
+  authenticated: false,
+  pdfZoom: Number(localStorage.getItem("frontPagePdfZoom")) || 110,
+};
+
+const $ = (id) => document.getElementById(id);
+const PDF_ZOOM_STEP = 15;
+const PDF_MIN_ZOOM = 55;
+const PDF_MAX_ZOOM = 220;
+const OPERATION_COLUMNS = ["手术操作编码", "手术操作日期", "手术级别", "手术操作名称", "术者", "I助", "II助", "切口愈合等级", "麻醉方式", "麻醉医师", "原始内容"];
+const DIAGNOSIS_COLUMNS = ["诊断类别", "出院诊断", "疾病编码", "入院病情"];
+
+const GROUP_KEYWORDS = {
+  基本信息: ["住院号", "病案号", "首页病案号", "姓名", "性别", "出生", "年龄", "身份证", "婚姻", "健康卡", "入院", "出院", "科别", "科室", "病房", "住院天数"],
+  地址联系人: ["地址", "电话", "联系人", "单位", "邮编", "户口"],
+  诊断表格: ["诊断", "疾病编码", "编码格式", "病理", "入院病情"],
+  手术表格: ["手术", "操作", "麻醉", "切口"],
+  离院费用: ["费用", "金额", "自付", "总费用", "离院", "再住院", "昏迷"],
+};
+
+const FIELD_KEYWORDS = {
+  inpatient_no: ["住院号", "ZY", "12位"],
+  medical_record_no: ["病案号", "10位", "前导0"],
+  front_page_medical_record_no: ["首页病案号"],
+  patient_name: ["姓名"],
+  id_card_no: ["身份证"],
+  contact_address: ["联系人地址"],
+  contact_phone: ["联系人电话"],
+  admission_time: ["入院时间"],
+  admission_dept: ["入院科别"],
+  discharge_time: ["出院时间"],
+  discharge_dept: ["出院科别", "科室"],
+  hospital_days: ["住院天数"],
+  major_department: ["大科室"],
+  primary_diagnosis: ["主要诊断"],
+  primary_diagnosis_code: ["主要诊断编码", "疾病编码"],
+  discharge_diagnoses: ["出院诊断", "其他诊断", "诊断编码", "入院病情"],
+  operations: ["手术", "操作", "麻醉", "切口"],
+  total_cost: ["总费用"],
+  self_pay_amount: ["自付"],
+  fee_details: ["费用明细"],
+};
+
+function statusTag(status) {
+  const value = status || "pending";
+  const meta = {
+    auto_pass: ["自动通过", "auto_pass"],
+    auto_corrected: ["自动修正", "auto_corrected"],
+    needs_review: ["需复核", "needs_review"],
+    reviewed: ["已复核", "reviewed"],
+    "AI复核-无问题": ["AI复核-无问题", "ai_passed"],
+    "AI复核-待确认": ["AI复核-待确认", "ai_pending"],
+    "已提交": ["已提交", "submitted"],
+    pending: ["待处理", "pending"],
+  };
+  const [label, className] = meta[value] || [value, "pending"];
+  return `${label}`;
+}
+
+function auditTag(status) {
+  const labels = { pending: "待抽查", passed: "抽查通过", failed: "抽查异常", unsure: "不确定" };
+  return `${labels[status] || status || "待抽查"}`;
+}
+
+async function api(path, options = {}) {
+  let response;
+  try {
+    response = await fetch(path, {
+      headers: { "Content-Type": "application/json" },
+      ...options,
+    });
+  } catch (_) {
+    throw new Error("无法连接网页服务，请确认 8501 服务正在运行后重试");
+  }
+  if (!response.ok) {
+    let message = response.statusText;
+    try {
+      const data = await response.json();
+      message = data.detail || message;
+    } catch (_) {
+      // ignore non-json errors
+    }
+    if (response.status === 401) {
+      showLogin(message || "请先登录");
+    }
+    throw new Error(message);
+  }
+  return response.json();
+}
+
+function userPermissions() {
+  return state.currentUser?.permissions || {};
+}
+
+function canOpenPage(page) {
+  const key = page === "auditHistory" ? "audit_history" : page;
+  return userPermissions()[key] !== false;
+}
+
+function firstAllowedPage() {
+  return ["overview", "review", "audit", "auditHistory", "settings"].find(canOpenPage) || "overview";
+}
+
+function applyPermissions() {
+  const permissions = userPermissions();
+  document.querySelectorAll(".nav-button").forEach((button) => {
+    const key = button.dataset.page === "auditHistory" ? "audit_history" : button.dataset.page;
+    button.classList.toggle("is-hidden", permissions[key] === false);
+  });
+}
+
+function showLogin(message = "") {
+  state.authenticated = false;
+  $("appShell")?.classList.add("is-hidden");
+  $("loginView")?.classList.remove("is-hidden");
+  if ($("loginMessage")) $("loginMessage").textContent = message;
+  setTimeout(() => $("loginPassword")?.focus(), 0);
+}
+
+function showApp() {
+  state.authenticated = true;
+  $("loginView")?.classList.add("is-hidden");
+  $("appShell")?.classList.remove("is-hidden");
+  $("currentUserLabel").textContent = state.currentUser?.username ? `当前用户：${state.currentUser.username}` : "已登录";
+  applyPermissions();
+}
+
+function showMessage(message, kind = "") {
+  const box = $("saveMessage");
+  if (!box) return;
+  box.textContent = message;
+  box.className = `save-message ${kind ? `is-${kind}` : ""}`;
+}
+
+function showAuditMessage(message, kind = "") {
+  const box = $("auditMessage");
+  box.textContent = message;
+  box.className = `inline-message ${kind ? `is-${kind}` : ""}`;
+}
+
+function pdfViewHash() {
+  return `#toolbar=0&navpanes=0&scrollbar=1&zoom=${state.pdfZoom}`;
+}
+
+function setPdfFrame(frameId, url) {
+  const frame = $(frameId);
+  if (!frame) return;
+  frame.src = "about:blank";
+  if (!url) return;
+  const separator = url.includes("?") ? "&" : "?";
+  requestAnimationFrame(() => {
+    frame.src = `${url}${separator}zoom_refresh=${Date.now()}${pdfViewHash()}`;
+  });
+}
+
+function refreshPdfFrames() {
+  if (state.selectedRecord?.pdf_url) {
+    setPdfFrame("pdfFrame", state.selectedRecord.pdf_url);
+  }
+  const auditRecord = state.auditSamples.find((sample) => sample.record.id === state.auditCurrentId)?.record;
+  if (auditRecord?.pdf_url) {
+    setPdfFrame("auditPdfFrame", auditRecord.pdf_url);
+  }
+}
+
+function updatePdfZoomLabels() {
+  ["pdfZoomLabel", "auditPdfZoomLabel"].forEach((id) => {
+    if ($(id)) $(id).textContent = `${state.pdfZoom}%`;
+  });
+}
+
+function setPdfZoom(value) {
+  state.pdfZoom = Math.max(PDF_MIN_ZOOM, Math.min(PDF_MAX_ZOOM, Number(value) || 110));
+  localStorage.setItem("frontPagePdfZoom", String(state.pdfZoom));
+  updatePdfZoomLabels();
+  refreshPdfFrames();
+}
+
+function changePdfZoom(delta) {
+  setPdfZoom(state.pdfZoom + delta);
+}
+
+function switchPage(page) {
+  if (!state.authenticated || !canOpenPage(page)) return;
+  state.activePage = page;
+  document.querySelectorAll(".page-view").forEach((el) => el.classList.add("is-hidden"));
+  $(`${page}Page`)?.classList.remove("is-hidden");
+  document.querySelectorAll(".nav-button").forEach((button) => {
+    button.classList.toggle("is-active", button.dataset.page === page);
+  });
+  if (page === "overview") loadOverview();
+  if (page === "review") loadRecords();
+  if (page === "auditHistory") loadAuditLogs();
+  if (page === "settings") loadSettings();
+}
+
+async function loadStatus() {
+  const data = await api("/api/status");
+  const stateLabel = data.database === "online" ? "在线" : data.database === "unchecked" ? "未检查" : "离线";
+  $("dbState").textContent = stateLabel;
+  $("dbState").className = data.database === "online" ? "status-online" : data.database === "unchecked" ? "status-muted" : "status-offline";
+  $("dbTotal").textContent = data.total ?? "--";
+  $("dbNeedsReview").textContent = data.needs_review ?? "--";
+  $("dbAiPassed").textContent = data.ai_passed ?? "--";
+  $("dbAiPending").textContent = data.ai_pending ?? "--";
+  $("dbReviewed").textContent = data.reviewed ?? "--";
+  $("statusGrid").title = `${data.host}:${data.port}/${data.database_name} · ${data.table} · ${data.message || ""} · ${data.checked_at ? `上次检查 ${formatTime(data.checked_at)}` : "尚未检查"}`;
+  state.status = data;
+  return data;
+}
+
+async function loadOverview() {
+  const data = await api("/api/overview");
+  state.overview = data;
+  renderOverview(data);
+}
+
+function renderOverview(data) {
+  const summary = data.summary || {};
+  const metrics = [
+    ["总记录", summary.total],
+    ["复核队列", summary.review_queue],
+    ["需复核", summary.needs_review],
+    ["AI待确认", summary.ai_pending],
+    ["AI无问题", summary.ai_passed],
+    ["已复核", summary.reviewed],
+    ["已提交", summary.submitted],
+    ["自动通过", summary.auto_passed],
+    ["抽查记录", summary.audit_total],
+    ["抽查异常", summary.audit_failed],
+  ];
+  $("overviewMetrics").innerHTML = metrics.map(([label, value]) => `
+    

+      ${escapeHtml(label)}
+      ${value ?? 0}
+    

+  `).join("");
+  $("overviewQueues").innerHTML = `
+    
复核工作台默认全部数据：需复核 ${summary.needs_review || 0} / AI待确认 ${summary.ai_pending || 0} / 已复核 ${summary.reviewed || 0}

+    
抽查候选池已复核 ${summary.reviewed || 0} 条 / 自动通过 ${summary.auto_passed || 0} 条

+    
日志存储人工复核和抽查记录均写入主表 JSONB 列

+  `;
+  $("overviewRecentLogs").innerHTML = renderLogTable(data.recent_logs || []);
+}
+
+async function loadSchema() {
+  const data = await api("/api/schema");
+  state.schema = data.groups;
+}
+
+async function loadRecords({ append = false } = {}) {
+  if (state.recordsLoading) return;
+  state.recordsLoading = true;
+  const query = encodeURIComponent($("searchInput").value.trim());
+  const status = encodeURIComponent($("statusFilter").value || "review_all");
+  const offset = append ? state.recordOffset : 0;
+  try {
+    const data = await api(`/api/records?q=${query}&status_filter=${status}&limit=${state.recordLimit || 300}&offset=${offset}`);
+    state.records = append ? [...state.records, ...data.records] : data.records;
+    state.recordLimit = data.limit || 300;
+    state.recordOffset = (data.offset || 0) + data.records.length;
+    state.recordHasMore = Boolean(data.has_more);
+    if (!append && !state.records.some((record) => record.id === state.selectedId)) {
+      state.selectedId = null;
+    }
+    renderRecords();
+    renderFilterActions();
+    if (!state.selectedId && state.records.length) {
+      await selectRecord(state.records[0].id);
+    }
+  } catch (error) {
+    showMessage(error.message || "加载列表失败", "error");
+  } finally {
+    state.recordsLoading = false;
+    renderRecords();
+    renderFilterActions();
+  }
+}
+
+function renderRecords() {
+  const list = $("recordList");
+  const previousScrollTop = list.scrollTop;
+  if (!state.records.length) {
+    list.innerHTML = `
当前筛选下没有需要处理的记录
`;
+    return;
+  }
+  const limitNotice = state.recordHasMore
+    ? `
已显示 ${state.records.length} 条，继续下拉加载更多
`
+    : "";
+  const loadingNotice = state.recordsLoading ? `
正在加载...
` : "";
+  const endNotice = !state.recordHasMore && state.records.length ? `
已加载全部 ${state.records.length} 条
` : "";
+  list.innerHTML = limitNotice + state.records.map((record) => {
+    const active = record.id === state.selectedId ? " is-active" : "";
+    const manual = record.manual_corrected ? " · 人工已改" : "";
+    const activity = record.last_activity_at ? `最近 ${escapeHtml(formatTime(record.last_activity_at))}` : "";
+    return `
+      
+    `;
+  }).join("") + loadingNotice + endNotice;
+  list.querySelectorAll(".record-item").forEach((button) => {
+    button.addEventListener("click", () => selectRecord(Number(button.dataset.id)));
+  });
+  list.scrollTop = previousScrollTop;
+}
+
+function renderFilterActions() {
+  const button = $("approveAiNoIssueBtn");
+  if (!button) return;
+  button.classList.toggle("is-hidden", $("statusFilter").value !== "AI复核-无问题");
+}
+
+function scrollSelectedRecordIntoView() {
+  const selected = $("recordList")?.querySelector(`.record-item[data-id="${state.selectedId}"]`);
+  selected?.scrollIntoView({ block: "nearest" });
+}
+
+function maybeLoadMoreRecords() {
+  const list = $("recordList");
+  if (!list || !state.recordHasMore || state.recordsLoading) return;
+  const distanceToBottom = list.scrollHeight - list.scrollTop - list.clientHeight;
+  if (distanceToBottom < 240) loadRecords({ append: true });
+}
+
+async function selectRecord(id, options = {}) {
+  state.selectedId = id;
+  renderRecords();
+  if (options.keepInView) scrollSelectedRecordIntoView();
+  showMessage("");
+  try {
+    const data = await api(`/api/records/${id}`);
+    if (state.selectedId !== id) return;
+    state.selectedRecord = data.record;
+    renderSelectedRecord();
+  } catch (error) {
+    showMessage(error.message || "记录加载失败，请刷新后重试", "error");
+  }
+}
+
+function recordListIndex() {
+  return state.records.findIndex((record) => record.id === state.selectedId);
+}
+
+function recordMatchesActiveFilter(record) {
+  const status = $("statusFilter").value || "review_all";
+  if (status === "review_all") {
+    return ["needs_review", "reviewed", "AI复核-无问题", "AI复核-待确认"].includes(record.review_status) || record.manual_corrected === true;
+  }
+  if (status === "needs_review") {
+    return record.review_status === "needs_review";
+  }
+  if (status === "reviewed") {
+    return record.review_status === "reviewed";
+  }
+  return status === "all" || record.review_status === status;
+}
+
+async function selectAdjacentRecord(delta) {
+  if (!state.records.length || state.recordsLoading) return;
+  let index = recordListIndex();
+  if (index < 0) index = delta > 0 ? -1 : state.records.length;
+  let nextIndex = index + delta;
+  if (nextIndex >= state.records.length && state.recordHasMore) {
+    await loadRecords({ append: true });
+    nextIndex = index + delta;
+  }
+  if (nextIndex < 0 || nextIndex >= state.records.length) return;
+  await selectRecord(state.records[nextIndex].id, { keepInView: true });
+}
+
+function handleRecordListKeydown(event) {
+  if (event.key !== "ArrowDown" && event.key !== "ArrowUp") return;
+  event.preventDefault();
+  selectAdjacentRecord(event.key === "ArrowDown" ? 1 : -1);
+}
+
+function isTextEntryTarget(target) {
+  if (!target) return false;
+  const tagName = target.tagName;
+  return target.isContentEditable || tagName === "INPUT" || tagName === "TEXTAREA" || tagName === "SELECT";
+}
+
+function preventWorkspacePageScroll(event) {
+  const scrollKeys = new Set(["ArrowDown", "ArrowUp", "PageDown", "PageUp", "Home", "End", " "]);
+  if (!scrollKeys.has(event.key) || isTextEntryTarget(event.target)) return;
+  if (state.activePage === "review" && (event.key === "ArrowDown" || event.key === "ArrowUp")) {
+    event.preventDefault();
+    event.stopPropagation();
+    selectAdjacentRecord(event.key === "ArrowDown" ? 1 : -1);
+    return;
+  }
+  if (state.activePage !== "review" && state.activePage !== "audit") return;
+  event.preventDefault();
+}
+
+function releasePdfKeyboardFocus(event) {
+  event.currentTarget.blur();
+  document.body.setAttribute("tabindex", "-1");
+  document.body.focus({ preventScroll: true });
+}
+
+function renderSelectedRecord() {
+  const record = state.selectedRecord;
+  $("saveBtn").disabled = false;
+  $("pdfTitle").textContent = record.source_file || "未找到文件";
+  $("pdfSubtitle").textContent = `${record.patient_name || ""} · ${record.inpatient_no || record.medical_record_no || ""} · ${record.primary_diagnosis || ""}`;
+  $("detailTitle").textContent = `${record.patient_name || "未命名"} · ${record.inpatient_no || record.medical_record_no || ""}`;
+  $("detailMeta").textContent = `${record.source_file || ""} · ${record.major_department || ""}`;
+  setPdfFrame("pdfFrame", record.pdf_url);
+  renderReviewStrip(record);
+  renderForm(record, {
+    formId: "detailForm",
+    targetId: "targetSummary",
+    noteId: "manualNote",
+    logCountId: "changeLogCount",
+    logTableId: "changeLogTable",
+  });
+  renderChangeLogs(record.review_logs || [], "changeLogCount", "changeLogTable");
+  setView(state.view);
+}
+
+async function refreshSelectedRecord() {
+  if (!state.selectedId) return;
+  try {
+    const data = await api(`/api/records/${state.selectedId}`);
+    if (state.selectedId !== data.record.id) return;
+    state.selectedRecord = data.record;
+    renderSelectedRecord();
+    const index = state.records.findIndex((record) => record.id === data.record.id);
+    if (index >= 0) state.records[index] = { ...state.records[index], ...compactRecord(data.record) };
+    renderRecords();
+  } catch (_) {
+    // keep the current UI if a background refresh loses the selected record
+  }
+}
+
+function reviewNotes(record) {
+  const notes = [];
+  ["review_notes", "quality_notes", "auto_corrections"].forEach((key) => {
+    const value = record[key];
+    if (Array.isArray(value)) {
+      value.forEach((item) => notes.push(typeof item === "string" ? item : JSON.stringify(item)));
+    } else if (value) {
+      notes.push(String(value));
+    }
+  });
+  return notes.filter(Boolean);
+}
+
+function reviewTargets(record) {
+  const notes = reviewNotes(record);
+  const text = notes.join("；");
+  const fieldAlerts = new Set();
+  const groupAlerts = new Set();
+  state.schema.forEach((group) => {
+    let groupHit = (GROUP_KEYWORDS[group.name] || []).some((keyword) => text.includes(keyword));
+    group.fields.forEach(([name, label]) => {
+      const keywords = [label, name, ...(FIELD_KEYWORDS[name] || [])].filter((keyword) => String(keyword).length > 1);
+      if (keywords.some((keyword) => text.includes(keyword))) {
+        fieldAlerts.add(name);
+        groupHit = true;
+      }
+    });
+    if (groupHit) groupAlerts.add(group.name);
+  });
+  return { notes, fieldAlerts, groupAlerts };
+}
+
+function renderReviewStrip(record) {
+  $("reviewStrip").innerHTML = `
+    
复核状态${statusTag(record.review_status)}

+    
+  `;
+}
+
+function renderForm(record, options = {}) {
+  const formId = options.formId || "detailForm";
+  const targetId = options.targetId || "targetSummary";
+  const noteId = options.noteId || "manualNote";
+  const form = $(formId);
+  const targets = reviewTargets(record);
+  const hasTarget = targets.groupAlerts.size > 0;
+  renderTargetSummary(targets, targetId, formId);
+  form.innerHTML = state.schema.map((group, index) => {
+    const groupAlert = targets.groupAlerts.has(group.name);
+    const fields = group.fields.map(([name, label, type, options]) => renderField(record, name, label, type, options, targets.fieldAlerts.has(name))).join("");
+    const open = groupAlert || (!hasTarget && index === 0);
+    return `
+      

+        ${escapeHtml(group.name)}${groupAlert ? "需核对" : ""}
+        
${fields}

+      

+    `;
+  }).join("");
+  if ($(noteId)) $(noteId).value = "";
+}
+
+function renderTargetSummary(targets, targetId = "targetSummary", formId = "detailForm") {
+  const groups = [...targets.groupAlerts];
+  const notesHtml = targets.notes.length
+    ? targets.notes.map((note) => `
  • ${escapeHtml(note)}
    • `).join("")
+    : "
  • 暂无复核提示
    • ";
+  $(targetId).innerHTML = `
+    
    
+      
+        复核定位
+        ${groups.length ? `${groups.length} 个模块` : "未定位"}
+      
+      
    
+        ${groups.length ? groups.map((group) => ``).join("") : "未定位到具体模块，先核对基本信息"}
+      
    
+      
      ${notesHtml}
    
+    
    
+  `;
+  $(targetId).querySelectorAll("[data-target-group]").forEach((button) => {
+    button.addEventListener("click", () => {
+      const group = $(formId).querySelector(`.field-group[data-group="${CSS.escape(button.dataset.targetGroup)}"]`);
+      if (group) {
+        group.open = true;
+        group.scrollIntoView({ block: "center", behavior: "smooth" });
+      }
+    });
+  });
+}
+
+function renderField(record, name, label, type, options, alerted = false) {
+  const value = record[name];
+  const wide = type === "json" || String(value || "").length > 28 ? " is-wide" : "";
+  const alertClass = alerted ? " is-alert" : "";
+  let control = "";
+  if (type === "select") {
+    control = ``;
+  } else if (type === "json") {
+    control = renderJsonEditor(name, value);
+  } else {
+    const inputType = type === "date" ? "date" : "text";
+    const placeholder = type === "datetime" ? "YYYY-MM-DD HH:MM:SS" : "";
+    control = ``;
+  }
+  return ``;
+}
+
+function preferredColumns(fieldName, value) {
+  if (fieldName === "operations") return OPERATION_COLUMNS;
+  if (fieldName === "discharge_diagnoses") return DIAGNOSIS_COLUMNS;
+  if (Array.isArray(value)) {
+    const keys = [];
+    value.forEach((row) => {
+      if (row && typeof row === "object" && !Array.isArray(row)) {
+        Object.keys(row).forEach((key) => {
+          if (!keys.includes(key)) keys.push(key);
+        });
+      }
+    });
+    return keys.length ? keys : ["值"];
+  }
+  if (value && typeof value === "object") return ["项目", "金额"];
+  return ["值"];
+}
+
+function renderJsonEditor(fieldName, value) {
+  const columns = preferredColumns(fieldName, value);
+  const rows = normalizeJsonRows(value, columns);
+  const header = columns.map((column) => `${escapeHtml(column)}`).join("");
+  const body = rows.map((row) => renderJsonRow(columns, row)).join("");
+  return `
+    
    
+      
    
+        
+        
+      
    
+      
    
+        
+          ${header}
+          ${body || renderJsonRow(columns, {})}
+        
    操作
    
+      
    
+    
    
+  `;
+}
+
+function normalizeJsonRows(value, columns) {
+  if (Array.isArray(value)) {
+    return value.map((row) => {
+      if (row && typeof row === "object" && !Array.isArray(row)) return row;
+      return { [columns[0] || "值"]: row ?? "" };
+    });
+  }
+  if (value && typeof value === "object") {
+    return Object.entries(value).map(([key, amount]) => ({ 项目: key, 金额: amount }));
+  }
+  if (value === null || value === undefined || value === "") return [];
+  return [{ [columns[0] || "值"]: value }];
+}
+
+function renderJsonRow(columns, row) {
+  const cells = columns.map((column) => {
+    const value = row[column] ?? "";
+    return ``;
+  }).join("");
+  return `${cells}`;
+}
+
+function collectFields(formId = "detailForm") {
+  const fields = {};
+  $(formId).querySelectorAll("input[data-field], select[data-field], textarea[data-field]").forEach((control) => {
+    fields[control.dataset.field] = control.value;
+  });
+  $(formId).querySelectorAll(".json-editor[data-json-field]").forEach((editor) => {
+    fields[editor.dataset.jsonField] = collectJsonEditor(editor);
+  });
+  return fields;
+}
+
+function collectJsonEditor(editor) {
+  const table = editor.querySelector("table");
+  const columns = JSON.parse(table.dataset.columns || "[]");
+  const rows = [...table.querySelectorAll("tbody tr")].map((tr) => {
+    const row = {};
+    tr.querySelectorAll("[data-json-key]").forEach((input) => {
+      row[input.dataset.jsonKey] = input.value;
+    });
+    return row;
+  }).filter((row) => Object.values(row).some((value) => String(value).trim() !== ""));
+
+  if (editor.dataset.kind === "object") {
+    const result = {};
+    rows.forEach((row) => {
+      const key = row["项目"];
+      if (key) result[key] = row["金额"] ?? "";
+    });
+    return result;
+  }
+  return rows;
+}
+
+async function advanceAfterSave(savedId, savedRecord) {
+  const savedSummary = compactRecord(savedRecord);
+  const listIndex = state.records.findIndex((record) => record.id === savedId);
+  let nextIndex = listIndex >= 0 ? listIndex : recordListIndex() + 1;
+
+  if (listIndex >= 0) {
+    if (recordMatchesActiveFilter(savedSummary)) {
+      state.records[listIndex] = { ...state.records[listIndex], ...savedSummary };
+      nextIndex = listIndex + 1;
+    } else {
+      state.records.splice(listIndex, 1);
+      state.recordOffset = Math.max(0, state.recordOffset - 1);
+      nextIndex = listIndex;
+    }
+  }
+
+  renderRecords();
+
+  if (nextIndex >= state.records.length && state.recordHasMore) {
+    await loadRecords({ append: true });
+  }
+
+  if (nextIndex < state.records.length) {
+    await selectRecord(state.records[nextIndex].id, { keepInView: true });
+    showMessage("已复核并保存，已自动跳到下一条", "ok");
+    return;
+  }
+
+  state.selectedId = savedId;
+  state.selectedRecord = savedRecord;
+  renderSelectedRecord();
+  renderRecords();
+  showMessage("已复核并保存，当前列表已没有下一条", "ok");
+}
+
+async function saveRecord() {
+  if (!state.selectedId) return;
+  $("saveBtn").disabled = true;
+  showMessage("正在保存...");
+  const savedId = state.selectedId;
+  try {
+    const payload = {
+      fields: collectFields(),
+      manual_note: $("manualNote").value.trim(),
+      note_prefix: "人工复核",
+    };
+    const data = await api(`/api/records/${state.selectedId}`, {
+      method: "POST",
+      body: JSON.stringify(payload),
+    });
+    state.selectedRecord = data.record;
+    renderChangeLogs(data.record.review_logs || [], "changeLogCount", "changeLogTable");
+    await advanceAfterSave(savedId, data.record);
+    loadStatus().catch(() => null);
+  } catch (error) {
+    showMessage(error.message, "error");
+  } finally {
+    $("saveBtn").disabled = false;
+  }
+}
+
+function compactRecord(record) {
+  return {
+    id: record.id,
+    source_file: record.source_file,
+    inpatient_no: record.inpatient_no,
+    medical_record_no: record.medical_record_no,
+    patient_name: record.patient_name,
+    review_status: record.review_status,
+    manual_corrected: record.manual_corrected,
+    major_department: record.major_department,
+    discharge_dept: record.discharge_dept,
+    primary_diagnosis: record.primary_diagnosis,
+    primary_diagnosis_code: record.primary_diagnosis_code,
+    contact_phone: record.contact_phone,
+    last_activity_at: record.last_activity_at,
+    has_pdf: Boolean(record.pdf_url),
+  };
+}
+
+function setView(view) {
+  state.view = "pdf";
+  $("pdfView").classList.remove("is-hidden");
+}
+
+function renderChangeLogs(logs, countId = "changeLogCount", tableId = "changeLogTable") {
+  $(countId).textContent = `${logs.length} 条`;
+  $(tableId).innerHTML = renderLogTable(logs);
+}
+
+function renderLogTable(logs) {
+  if (!logs.length) return `
    暂无修改记录
    `;
+  const rows = logs.map((log) => {
+    const fields = Array.isArray(log.changed_fields) ? log.changed_fields : [];
+    const summary = fields.length
+      ? fields.map((field) => `${field.label || field.field}: ${shortValue(field.old)} -> ${shortValue(field.new)}`).join("；")
+      : "仅记录人工备注或状态确认";
+    return `
+      
+        ${escapeHtml(formatTime(log.changed_at))}
+        ${escapeHtml(log.changed_by || "web")}
+        ${escapeHtml(summary)}
+        ${escapeHtml(log.manual_note || "")}
+      
+    `;
+  }).join("");
+  return `
+    
+      
+      ${rows}
+    
    修改时间来源修改内容备注
    
+  `;
+}
+
+function shortValue(value) {
+  if (value === null || value === undefined || value === "") return "空";
+  const text = typeof value === "string" ? value : JSON.stringify(value);
+  return text.length > 36 ? `${text.slice(0, 36)}...` : text;
+}
+
+async function sampleAudit() {
+  const source = encodeURIComponent($("auditSource").value);
+  const count = encodeURIComponent($("auditCount").value || "5");
+  showAuditMessage("正在抽取...");
+  try {
+    const data = await api(`/api/audit/sample?source=${source}&count=${count}`, { method: "POST" });
+    state.auditSamples = data.records.map((record) => ({
+      record,
+      audit_source: data.audit_source,
+      audit_status: "pending",
+      audit_notes: "",
+    }));
+    state.auditCurrentId = state.auditSamples[0]?.record.id || null;
+    renderAuditSamples();
+    if (state.auditCurrentId) renderAuditCurrent();
+    showAuditMessage(`已抽取 ${state.auditSamples.length} 条`, "ok");
+    await loadAuditLogs();
+    await loadOverview();
+  } catch (error) {
+    showAuditMessage(error.message, "error");
+  }
+}
+
+function renderAuditSamples() {
+  const box = $("auditSampleList");
+  if (!state.auditSamples.length) {
+    box.innerHTML = `
    暂无抽查样本
    `;
+    return;
+  }
+  box.innerHTML = state.auditSamples.map((item) => {
+    const { record } = item;
+    return `
+    
+  `;
+  }).join("");
+  box.querySelectorAll("[data-audit-record]").forEach((button) => {
+    button.addEventListener("click", () => {
+      state.auditCurrentId = Number(button.dataset.auditRecord);
+      renderAuditSamples();
+      renderAuditCurrent();
+    });
+  });
+}
+
+function renderAuditCurrent() {
+  const item = state.auditSamples.find((sample) => sample.record.id === state.auditCurrentId);
+  if (!item) {
+    $("auditPdfTitle").textContent = "未选择抽查记录";
+    $("auditPdfSubtitle").textContent = "";
+    setPdfFrame("auditPdfFrame", "");
+    $("auditDetailTitle").textContent = "抽查信息";
+    $("auditDetailMeta").textContent = "随机抽取后开始抽查";
+    $("auditDetailForm").innerHTML = "";
+    $("auditTargetSummary").innerHTML = "";
+    renderChangeLogs([], "auditChangeLogCount", "auditChangeLogTable");
+    return;
+  }
+  const record = item.record;
+  $("auditPdfTitle").textContent = record.source_file || "未找到文件";
+  $("auditPdfSubtitle").textContent = `${record.patient_name || ""} · ${record.inpatient_no || record.medical_record_no || ""} · ${record.primary_diagnosis || ""}`;
+  $("auditDetailTitle").textContent = `${record.patient_name || "未命名"} · ${record.inpatient_no || record.medical_record_no || ""}`;
+  $("auditDetailMeta").textContent = `${record.source_file || ""} · ${record.major_department || ""}`;
+  setPdfFrame("auditPdfFrame", record.pdf_url);
+  $("auditReviewStrip").innerHTML = `
+    
    复核状态${statusTag(record.review_status)}
    
+    
    抽查状态${auditTag(item.audit_status || "pending")}
    
+    
    主要诊断${escapeHtml(record.primary_diagnosis || "无")}
    
+    
    抽查来源${escapeHtml(item.audit_source || "")}
    
+  `;
+  renderForm(record, {
+    formId: "auditDetailForm",
+    targetId: "auditTargetSummary",
+    noteId: "auditNotes",
+  });
+  $("auditNotes").value = item.audit_notes || "";
+  renderChangeLogs(record.review_logs || [], "auditChangeLogCount", "auditChangeLogTable");
+  setAuditView(state.auditView);
+}
+
+async function saveAuditStatus(status) {
+  const item = state.auditSamples.find((sample) => sample.record.id === state.auditCurrentId);
+  if (!item) return;
+  try {
+    const data = await api("/api/audit/classify", {
+      method: "POST",
+      body: JSON.stringify({
+        record_id: item.record.id,
+        audit_source: item.audit_source || $("auditSource").value,
+        audit_status: status,
+        audit_notes: $("auditNotes").value.trim(),
+        fields: collectFields("auditDetailForm"),
+      }),
+    });
+    item.log = data.log;
+    item.audit_status = data.log.audit_status;
+    item.audit_notes = data.log.audit_notes || "";
+    item.record = data.record;
+    renderAuditSamples();
+    renderAuditCurrent();
+    showAuditMessage("抽查已归类并保存", "ok");
+    await loadAuditLogs();
+    await loadOverview();
+  } catch (error) {
+    showAuditMessage(error.message, "error");
+  }
+}
+
+function setAuditView(view) {
+  state.auditView = "pdf";
+  $("auditPdfView").classList.remove("is-hidden");
+}
+
+async function loadAuditLogs() {
+  const data = await api("/api/audit/logs?limit=200");
+  state.auditLogs = data.logs;
+  renderAuditHistory();
+}
+
+function renderAuditHistory() {
+  const box = $("auditHistoryTable");
+  if (!box) return;
+  if (!state.auditLogs.length) {
+    box.innerHTML = `
    暂无抽查记录
    `;
+    return;
+  }
+  const rows = state.auditLogs.map((log) => `
+    
+      ${escapeHtml(formatTime(log.updated_at || log.created_at))}
+      ${escapeHtml(log.patient_name || "")}
+      ${escapeHtml(log.inpatient_no || log.medical_record_no || "")}
+      ${escapeHtml(log.audit_source || "")}
+      ${auditTag(log.audit_status)}
+      ${escapeHtml(changeSummaryFromSnapshot(log.snapshot))}
+      ${escapeHtml(log.audit_notes || "")}
+    
+  `).join("");
+  box.innerHTML = `
+    
+      
+      ${rows}
+    
    时间姓名住院号来源状态修改内容人工备注
    
+  `;
+}
+
+function changeSummaryFromSnapshot(snapshot) {
+  const fields = Array.isArray(snapshot?.changed_fields) ? snapshot.changed_fields : [];
+  if (!fields.length) return "未修改字段";
+  return fields.map((field) => `${field.label || field.field}: ${shortValue(field.old)} -> ${shortValue(field.new)}`).join("；");
+}
+
+async function loadSettings() {
+  const data = await api("/api/settings");
+  state.settings = data;
+  state.ai = data.kimi || state.ai;
+  renderSettings();
+  renderAiActions();
+}
+
+async function loadAiConfig() {
+  const data = await api("/api/ai/config");
+  state.ai = data.kimi;
+  renderAiActions();
+}
+
+function aiAvailable() {
+  return Boolean(state.ai?.available);
+}
+
+function renderAiActions() {
+  const actions = $("aiActions");
+  if (!actions) return;
+  actions.classList.toggle("is-hidden", !aiAvailable());
+}
+
+function renderSettings() {
+  const status = state.status || {};
+  $("settingsStatus").innerHTML = `
+    
    数据库${escapeHtml(status.database || "未知")}
    
+    
    连接${escapeHtml(`${status.host || ""}:${status.port || ""}/${status.database_name || ""}`)}
    
+    
    表名${escapeHtml(status.table || "")}
    
+    
    PDF目录${escapeHtml(status.pdf_dir || "")}
    
+    
    上次检查${escapeHtml(status.checked_at ? formatTime(status.checked_at) : "尚未检查")}
    
+    
    下次检查${escapeHtml(status.next_check_at ? formatTime(status.next_check_at) : "--")}
    
+    
    检查结果${escapeHtml(status.message || "")}
    
+  `;
+  const settings = state.settings || { users: [], permission_labels: {} };
+  const system = settings.system || {};
+  const kimi = settings.kimi || {};
+  $("statusCheckTime").value = system.status_check_time || "03:00";
+  $("statusCheckMeta").textContent = `上次检查：${status.checked_at ? formatTime(status.checked_at) : "尚未检查"} · 下次：${status.next_check_at ? formatTime(status.next_check_at) : "--"}`;
+  $("kimiEnabled").checked = Boolean(kimi.enabled);
+  $("kimiModel").value = kimi.model || "";
+  $("kimiApiBase").value = kimi.api_base || "";
+  $("kimiConcurrency").value = kimi.concurrency || 3;
+  $("kimiMeta").textContent = `${kimi.api_key_configured ? "API Key 已从 .env 读取" : "未配置 API Key"} · ${kimi.available ? "AI核验已启用" : "AI核验已关闭"} · 并发 ${kimi.concurrency || 3}`;
+  $("settingsUserCount").textContent = `${settings.users.length} 个用户`;
+  renderPermissionControls("newUserPermissions", settings.permission_labels || {}, {});
+  $("userList").innerHTML = settings.users.map((user) => renderUserCard(user, settings.permission_labels || {})).join("");
+  $("userList").querySelectorAll("[data-save-user]").forEach((button) => {
+    button.addEventListener("click", () => saveUser(button.dataset.saveUser));
+  });
+  $("userList").querySelectorAll("[data-delete-user]").forEach((button) => {
+    button.addEventListener("click", () => deleteUser(button.dataset.deleteUser));
+  });
+}
+
+async function saveKimiSettings(event) {
+  event.preventDefault();
+  const button = $("saveKimiSettingsBtn");
+  button.disabled = true;
+  $("kimiMeta").textContent = "正在保存...";
+  try {
+    const data = await api("/api/settings/kimi", {
+      method: "POST",
+      body: JSON.stringify({
+        enabled: $("kimiEnabled").checked,
+        model: $("kimiModel").value.trim(),
+        api_base: $("kimiApiBase").value.trim(),
+        concurrency: Number($("kimiConcurrency").value) || 3,
+      }),
+    });
+    state.settings = data;
+    state.ai = data.kimi;
+    renderSettings();
+    renderAiActions();
+  } catch (error) {
+    $("kimiMeta").textContent = error.message;
+  } finally {
+    button.disabled = false;
+  }
+}
+
+function setAiButtonsDisabled(disabled) {
+  ["aiCurrentBtn", "aiFiveBtn", "aiAllBtn"].forEach((id) => {
+    if ($(id)) $(id).disabled = disabled;
+  });
+}
+
+async function pollAiReviewJob() {
+  const job = await api("/api/ai/review/status");
+  const total = job.total || 0;
+  const processed = job.processed || 0;
+  showMessage(`AI核验中：${processed}/${total}，并发 ${job.concurrency || 1}，无问题 ${job.ok || 0}，待确认 ${job.pending || 0}，失败 ${job.failed || 0}`, "");
+  if (job.running) {
+    state.aiPolling = setTimeout(() => pollAiReviewJob().catch((error) => showMessage(error.message, "error")), 2000);
+    return;
+  }
+  setAiButtonsDisabled(false);
+  showMessage(`AI核验完成：无问题 ${job.ok || 0}，待确认 ${job.pending || 0}，失败 ${job.failed || 0}`, job.failed ? "error" : "ok");
+  await loadStatus().catch(() => null);
+  await loadOverview().catch(() => null);
+  await loadRecords();
+  await refreshSelectedRecord();
+}
+
+async function runAiReview(scope) {
+  if (!aiAvailable()) {
+    showMessage("Kimi AI 核验未启用，请先在设置中开启", "error");
+    return;
+  }
+  if ((scope === "current" || scope === "five") && !state.selectedId) return;
+  if (scope === "all" && !window.confirm("确认对当前项之后的所有需复核记录执行 AI 核验？")) return;
+  if (state.aiPolling) clearTimeout(state.aiPolling);
+  setAiButtonsDisabled(true);
+  showMessage("正在启动 AI 核验...");
+  try {
+    const job = await api("/api/ai/review", {
+      method: "POST",
+      body: JSON.stringify({ scope, record_id: state.selectedId }),
+    });
+    showMessage(`AI核验已启动：0/${job.total || 0}`);
+    await pollAiReviewJob();
+  } catch (error) {
+    setAiButtonsDisabled(false);
+    showMessage(error.message, "error");
+  }
+}
+
+async function approveAiNoIssueRecords() {
+  if (!window.confirm("确认将所有“AI复核-无问题”批量标记为已人工复核？")) return;
+  const button = $("approveAiNoIssueBtn");
+  button.disabled = true;
+  showMessage("正在批量确认 AI 无问题记录...");
+  try {
+    const data = await api("/api/ai/review/approve-no-issue", { method: "POST" });
+    state.status = data.status_snapshot || state.status;
+    showMessage(`已批量确认 ${data.updated || 0} 条 AI 无问题记录`, "ok");
+    await loadStatus().catch(() => null);
+    await loadOverview().catch(() => null);
+    await loadRecords();
+  } catch (error) {
+    showMessage(error.message, "error");
+  } finally {
+    button.disabled = false;
+  }
+}
+
+async function submitReviewedRecords() {
+  if (!window.confirm("确认提交所有已人工复核项目？提交后这些项目将不再出现在复核工作台列表中。")) return;
+  const button = $("submitReviewedBtn");
+  button.disabled = true;
+  $("statusCheckMeta").textContent = "正在提交已人工复核项目...";
+  try {
+    const data = await api("/api/settings/submit-reviewed", { method: "POST" });
+    state.status = data.status_snapshot || state.status;
+    $("statusCheckMeta").textContent = `已提交 ${data.updated || 0} 条已人工复核项目`;
+    await loadStatus().catch(() => null);
+    await loadOverview().catch(() => null);
+    await loadRecords().catch(() => null);
+  } catch (error) {
+    $("statusCheckMeta").textContent = error.message;
+  } finally {
+    button.disabled = false;
+  }
+}
+
+async function saveSystemSettings(event) {
+  event.preventDefault();
+  const button = $("saveSystemSettingsBtn");
+  button.disabled = true;
+  $("statusCheckMeta").textContent = "正在保存...";
+  try {
+    const data = await api("/api/settings/system", {
+      method: "POST",
+      body: JSON.stringify({ status_check_time: $("statusCheckTime").value }),
+    });
+    state.settings = data;
+    state.status = data.status_snapshot || state.status;
+    renderSettings();
+    await loadStatus();
+  } catch (error) {
+    $("statusCheckMeta").textContent = error.message;
+  } finally {
+    button.disabled = false;
+  }
+}
+
+async function runStatusCheck() {
+  const button = $("statusCheckBtn");
+  button.disabled = true;
+  $("statusCheckMeta").textContent = "正在检查数据库与 PDF 目录...";
+  try {
+    const data = await api("/api/settings/status/check", { method: "POST" });
+    state.status = data.status_snapshot;
+    await loadSettings();
+    await loadStatus();
+  } catch (error) {
+    $("statusCheckMeta").textContent = error.message;
+  } finally {
+    button.disabled = false;
+  }
+}
+
+function renderPermissionControls(containerId, labels, permissions) {
+  const container = $(containerId);
+  container.innerHTML = Object.entries(labels).map(([key, label]) => `
+    
+  `).join("");
+}
+
+function renderUserCard(user, labels) {
+  const editable = user.source !== "env";
+  const permissionHtml = Object.entries(labels).map(([key, label]) => `
+    
+  `).join("");
+  return `
+    
    
+      
    
+        ${escapeHtml(user.username)}
+        ${user.source === "env" ? "内置管理员" : "本地配置用户"}${user.has_password ? " · 已设密码" : ""}
+      
    
+      
    
+        
+        
+      
    
+      
    ${permissionHtml}
    
+      ${editable ? `
+        
+      ` : ""}
+    
    
+  `;
+}
+
+function permissionsFromContainer(container) {
+  const result = {};
+  container.querySelectorAll("input[data-permission]").forEach((input) => {
+    result[input.dataset.permission] = input.checked;
+  });
+  return result;
+}
+
+function permissionsFromUserCard(card) {
+  const result = {};
+  card.querySelectorAll("input[data-user-permission]").forEach((input) => {
+    result[input.dataset.userPermission] = input.checked;
+  });
+  return result;
+}
+
+function cssEscape(value) {
+  if (window.CSS?.escape) return CSS.escape(value);
+  return String(value).replaceAll("\\", "\\\\").replaceAll('"', '\\"');
+}
+
+async function createUser(event) {
+  event.preventDefault();
+  const username = $("newUsername").value.trim();
+  const password = $("newPassword").value;
+  const permissions = permissionsFromContainer($("newUserPermissions"));
+  const data = await api("/api/settings/users", {
+    method: "POST",
+    body: JSON.stringify({ username, password, permissions }),
+  });
+  state.settings = data;
+  $("newUsername").value = "";
+  $("newPassword").value = "";
+  renderSettings();
+}
+
+async function saveUser(username) {
+  const card = document.querySelector(`[data-user-card="${cssEscape(username)}"]`);
+  if (!card) return;
+  const data = await api(`/api/settings/users/${encodeURIComponent(username)}`, {
+    method: "POST",
+    body: JSON.stringify({
+      username: card.querySelector("[data-user-name]").value.trim(),
+      password: card.querySelector("[data-user-password]").value,
+      permissions: permissionsFromUserCard(card),
+    }),
+  });
+  state.settings = data;
+  renderSettings();
+}
+
+async function deleteUser(username) {
+  if (!window.confirm(`确认删除用户 ${username}？`)) return;
+  const data = await api(`/api/settings/users/${encodeURIComponent(username)}`, { method: "DELETE" });
+  state.settings = data;
+  renderSettings();
+}
+
+async function login(event) {
+  event.preventDefault();
+  $("loginMessage").textContent = "";
+  $("loginBtn").disabled = true;
+  try {
+    const data = await api("/api/auth/login", {
+      method: "POST",
+      body: JSON.stringify({
+        username: $("loginUsername").value.trim(),
+        password: $("loginPassword").value,
+      }),
+    });
+    state.currentUser = data.user;
+    $("loginPassword").value = "";
+    showApp();
+    await loadWorkspace();
+  } catch (error) {
+    $("loginMessage").textContent = error.message;
+  } finally {
+    $("loginBtn").disabled = false;
+  }
+}
+
+async function logout() {
+  await api("/api/auth/logout", { method: "POST" }).catch(() => null);
+  state.currentUser = null;
+  state.records = [];
+  state.selectedId = null;
+  state.selectedRecord = null;
+  showLogin("已退出登录");
+}
+
+function wireJsonEditorButtons() {
+  document.querySelectorAll(".structured-form").forEach((form) => form.addEventListener("click", (event) => {
+    const target = event.target;
+    if (!(target instanceof HTMLElement)) return;
+    if (target.matches("[data-json-delete]")) {
+      const tbody = target.closest("tbody");
+      target.closest("tr")?.remove();
+      if (tbody && !tbody.querySelector("tr")) {
+        const table = tbody.closest("table");
+        const columns = JSON.parse(table.dataset.columns || "[]");
+        tbody.insertAdjacentHTML("beforeend", renderJsonRow(columns, {}));
+      }
+    }
+    if (target.matches("[data-json-add]")) {
+      const editor = target.closest(".json-editor");
+      const table = editor.querySelector("table");
+      const columns = JSON.parse(table.dataset.columns || "[]");
+      table.querySelector("tbody").insertAdjacentHTML("beforeend", renderJsonRow(columns, {}));
+    }
+    if (target.matches("[data-json-raw]")) {
+      const editor = target.closest(".json-editor");
+      const value = collectJsonEditor(editor);
+      navigator.clipboard?.writeText(JSON.stringify(value, null, 2));
+      showMessage("当前表格 JSON 已复制", "ok");
+    }
+  }));
+}
+
+function escapeHtml(value) {
+  return String(value)
+    .replaceAll("&", "&")
+    .replaceAll("<", "<")
+    .replaceAll(">", ">")
+    .replaceAll('"', """)
+    .replaceAll("'", "'");
+}
+
+function formatTime(value) {
+  if (!value) return "";
+  return String(value).replace("T", " ").slice(0, 19);
+}
+
+function debounce(fn, wait = 250) {
+  let timer = null;
+  return (...args) => {
+    clearTimeout(timer);
+    timer = setTimeout(() => fn(...args), wait);
+  };
+}
+
+async function boot() {
+  $("loginForm").addEventListener("submit", login);
+  $("logoutBtn").addEventListener("click", logout);
+  document.querySelectorAll(".nav-button").forEach((button) => {
+    button.addEventListener("click", () => switchPage(button.dataset.page));
+  });
+  $("overviewRefreshBtn").addEventListener("click", loadOverview);
+  $("searchInput").addEventListener("input", debounce(() => loadRecords()));
+  $("statusFilter").addEventListener("change", () => {
+    renderFilterActions();
+    loadRecords();
+  });
+  $("recordList").addEventListener("scroll", maybeLoadMoreRecords);
+  $("recordList").addEventListener("keydown", handleRecordListKeydown);
+  document.addEventListener("keydown", preventWorkspacePageScroll, { capture: true });
+  $("saveBtn").addEventListener("click", saveRecord);
+  $("aiCurrentBtn").addEventListener("click", () => runAiReview("current"));
+  $("aiFiveBtn").addEventListener("click", () => runAiReview("five"));
+  $("aiAllBtn").addEventListener("click", () => runAiReview("all"));
+  $("approveAiNoIssueBtn").addEventListener("click", approveAiNoIssueRecords);
+  $("pdfZoomOutBtn").addEventListener("click", () => changePdfZoom(-PDF_ZOOM_STEP));
+  $("pdfZoomInBtn").addEventListener("click", () => changePdfZoom(PDF_ZOOM_STEP));
+  $("pdfZoomResetBtn").addEventListener("click", () => setPdfZoom(100));
+  $("pdfFrame").addEventListener("focus", releasePdfKeyboardFocus);
+  $("auditPdfZoomOutBtn").addEventListener("click", () => changePdfZoom(-PDF_ZOOM_STEP));
+  $("auditPdfZoomInBtn").addEventListener("click", () => changePdfZoom(PDF_ZOOM_STEP));
+  $("auditPdfZoomResetBtn").addEventListener("click", () => setPdfZoom(100));
+  $("auditPdfFrame").addEventListener("focus", releasePdfKeyboardFocus);
+  updatePdfZoomLabels();
+  $("auditSampleBtn").addEventListener("click", sampleAudit);
+  $("auditPassBtn").addEventListener("click", () => saveAuditStatus("passed"));
+  $("auditFailBtn").addEventListener("click", () => saveAuditStatus("failed"));
+  $("auditUnsureBtn").addEventListener("click", () => saveAuditStatus("unsure"));
+  $("auditHistoryRefreshBtn").addEventListener("click", loadAuditLogs);
+  $("userForm").addEventListener("submit", createUser);
+  $("systemSettingsForm").addEventListener("submit", saveSystemSettings);
+  $("kimiSettingsForm").addEventListener("submit", saveKimiSettings);
+  $("submitReviewedBtn").addEventListener("click", submitReviewedRecords);
+  $("statusCheckBtn").addEventListener("click", runStatusCheck);
+  wireJsonEditorButtons();
+  try {
+    const session = await api("/api/auth/me");
+    if (!session.authenticated) {
+      showLogin();
+      return;
+    }
+    state.currentUser = session.user;
+    showApp();
+    await loadWorkspace();
+  } catch (error) {
+    showLogin(error.message);
+  }
+}
+
+async function loadWorkspace() {
+  await loadStatus();
+  await loadSchema();
+  await loadAiConfig().catch(() => null);
+  const page = canOpenPage(state.activePage) ? state.activePage : firstAllowedPage();
+  switchPage(page);
+}
+
+boot();
diff --git a/患者首页处理/数据可视化网页端/app/static/index.html b/患者首页处理/数据可视化网页端/app/static/index.html
new file mode 100644
index 0000000..13df763
--- /dev/null
+++ b/患者首页处理/数据可视化网页端/app/static/index.html
@@ -0,0 +1,273 @@
+
+
+
+  
+  
+  患者首页复核工作台
+  
+
+
+  
+
+  
+
+  
+
+
diff --git a/患者首页处理/数据可视化网页端/app/static/styles.css b/患者首页处理/数据可视化网页端/app/static/styles.css
new file mode 100644
index 0000000..1c9ef2c
--- /dev/null
+++ b/患者首页处理/数据可视化网页端/app/static/styles.css
@@ -0,0 +1,1193 @@
+:root {
+  --bg: #eef1f4;
+  --panel: #ffffff;
+  --ink: #15191f;
+  --muted: #66717f;
+  --line: #d5dbe2;
+  --line-dark: #b5bec8;
+  --accent: #1d5f8f;
+  --accent-dark: #17486c;
+  --ok: #1d7b52;
+  --warn: #9a6500;
+  --bad: #bd2d20;
+  --bad-bg: #fff1ef;
+  --warn-bg: #fff7df;
+  --ok-bg: #eef8f3;
+  --focus-bg: #fff7f5;
+  --shadow: 0 10px 28px rgba(28, 35, 45, 0.10);
+}
+
+* {
+  box-sizing: border-box;
+}
+
+html {
+  height: 100%;
+  overflow: hidden;
+}
+
+body {
+  margin: 0;
+  height: 100vh;
+  min-height: 100vh;
+  overflow: hidden;
+  background: var(--bg);
+  color: var(--ink);
+  font-family: "Noto Sans CJK SC", "Microsoft YaHei", sans-serif;
+  font-size: 14px;
+  letter-spacing: 0;
+}
+
+.login-view {
+  min-height: 100vh;
+  display: grid;
+  place-items: center;
+  padding: 24px;
+  background:
+    linear-gradient(135deg, rgba(29, 95, 143, .10), transparent 42%),
+    linear-gradient(315deg, rgba(29, 123, 82, .12), transparent 38%),
+    var(--bg);
+}
+
+.login-panel {
+  width: min(430px, 100%);
+  display: grid;
+  gap: 12px;
+  border: 1px solid var(--line-dark);
+  border-radius: 8px;
+  background: #fff;
+  box-shadow: 0 18px 48px rgba(28, 35, 45, .16);
+  padding: 24px;
+}
+
+.login-mark {
+  width: 42px;
+  height: 42px;
+  border: 2px solid var(--accent);
+  border-radius: 8px;
+  background:
+    linear-gradient(90deg, transparent 46%, rgba(29, 95, 143, .18) 47%, rgba(29, 95, 143, .18) 53%, transparent 54%),
+    linear-gradient(0deg, transparent 46%, rgba(29, 95, 143, .18) 47%, rgba(29, 95, 143, .18) 53%, transparent 54%),
+    #fff;
+}
+
+.login-panel p {
+  margin: -4px 0 4px;
+  color: var(--muted);
+}
+
+.login-panel label {
+  display: grid;
+  gap: 6px;
+  color: var(--muted);
+  font-size: 12px;
+}
+
+.login-panel button {
+  min-height: 40px;
+}
+
+.login-message {
+  min-height: 18px;
+  color: var(--bad);
+}
+
+.app-shell {
+  height: 100vh;
+  overflow: hidden;
+}
+
+button,
+input,
+select,
+textarea {
+  font: inherit;
+}
+
+button {
+  cursor: pointer;
+}
+
+h1,
+h2 {
+  margin: 0;
+  letter-spacing: 0;
+}
+
+h1 {
+  font-size: 18px;
+}
+
+h2 {
+  font-size: 16px;
+}
+
+.topbar {
+  min-height: 76px;
+  display: grid;
+  grid-template-columns: minmax(250px, 320px) auto minmax(720px, 1fr) auto;
+  align-items: center;
+  gap: 14px;
+  padding: 10px 14px;
+  border-bottom: 1px solid var(--line-dark);
+  background: #f9fafb;
+}
+
+.session-box {
+  display: inline-flex;
+  align-items: center;
+  justify-content: flex-end;
+  gap: 8px;
+  min-width: 160px;
+}
+
+.session-box span {
+  color: var(--muted);
+  white-space: nowrap;
+}
+
+.session-box button {
+  min-height: 34px;
+  padding: 6px 12px;
+}
+
+.brand {
+  display: flex;
+  align-items: center;
+  gap: 12px;
+  min-width: 0;
+}
+
+.brand p,
+.detail-head p {
+  margin: 3px 0 0;
+  color: var(--muted);
+  font-size: 12px;
+}
+
+.mark {
+  width: 34px;
+  height: 34px;
+  border: 2px solid var(--accent);
+  border-radius: 7px;
+  background:
+    linear-gradient(90deg, transparent 46%, rgba(29, 95, 143, .16) 47%, rgba(29, 95, 143, .16) 53%, transparent 54%),
+    linear-gradient(0deg, transparent 46%, rgba(29, 95, 143, .16) 47%, rgba(29, 95, 143, .16) 53%, transparent 54%),
+    #fff;
+  flex: 0 0 auto;
+}
+
+.top-nav {
+  display: inline-flex;
+  gap: 4px;
+  padding: 4px;
+  border: 1px solid var(--line);
+  border-radius: 8px;
+  background: #fff;
+  overflow: auto;
+}
+
+.nav-button {
+  min-height: 34px;
+  border: 0;
+  border-radius: 6px;
+  background: transparent;
+  color: var(--muted);
+  padding: 6px 10px;
+  white-space: nowrap;
+}
+
+.nav-button.is-active {
+  background: var(--accent);
+  color: #fff;
+}
+
+.status-grid {
+  display: grid;
+  grid-template-columns: repeat(6, minmax(104px, 1fr));
+  gap: 8px;
+}
+
+.status-card,
+.metric-card {
+  min-height: 48px;
+  padding: 7px 9px;
+  border: 1px solid var(--line);
+  border-radius: 7px;
+  background: #fff;
+}
+
+.status-card span,
+.metric-card span {
+  display: block;
+  color: var(--muted);
+  font-size: 12px;
+}
+
+.status-card strong,
+.metric-card strong {
+  display: block;
+  margin-top: 2px;
+  font-size: 18px;
+  white-space: nowrap;
+  overflow: hidden;
+  text-overflow: ellipsis;
+}
+
+.status-online {
+  color: var(--ok);
+}
+
+.status-offline {
+  color: var(--bad);
+}
+
+.status-muted {
+  color: var(--muted);
+}
+
+.page-view {
+  height: calc(100vh - 76px);
+  min-height: 0;
+  overflow: hidden;
+}
+
+.is-hidden {
+  display: none !important;
+}
+
+.overview-page,
+.audit-page,
+.audit-history-page,
+.settings-page {
+  padding: 12px;
+  overflow: auto;
+}
+
+.metric-grid {
+  display: grid;
+  grid-template-columns: repeat(auto-fit, minmax(150px, 1fr));
+  gap: 10px;
+  margin-bottom: 12px;
+}
+
+.page-band {
+  border: 1px solid var(--line);
+  border-radius: 8px;
+  background: var(--panel);
+  box-shadow: var(--shadow);
+  padding: 12px;
+  margin-bottom: 12px;
+}
+
+.section-head {
+  display: flex;
+  align-items: center;
+  justify-content: space-between;
+  gap: 12px;
+  margin-bottom: 10px;
+}
+
+.section-head.compact {
+  margin-bottom: 6px;
+}
+
+.section-head span {
+  color: var(--muted);
+  font-size: 12px;
+}
+
+.queue-summary,
+.settings-list {
+  display: grid;
+  gap: 8px;
+}
+
+.queue-line,
+.settings-list > div {
+  display: flex;
+  align-items: center;
+  justify-content: space-between;
+  gap: 12px;
+  min-height: 38px;
+  border-bottom: 1px solid var(--line);
+  padding: 6px 0;
+}
+
+.queue-line:last-child,
+.settings-list > div:last-child {
+  border-bottom: 0;
+}
+
+.queue-line span,
+.settings-list span {
+  color: var(--muted);
+}
+
+.workspace {
+  height: calc(100vh - 76px);
+  display: grid;
+  grid-template-columns: 300px minmax(720px, 1.25fr) minmax(390px, .85fr);
+  gap: 10px;
+  padding: 10px;
+}
+
+.record-panel,
+.pdf-panel,
+.detail-panel {
+  min-height: 0;
+  border: 1px solid var(--line);
+  border-radius: 8px;
+  background: var(--panel);
+  box-shadow: var(--shadow);
+  overflow: hidden;
+}
+
+.record-panel,
+.detail-panel {
+  display: flex;
+  flex-direction: column;
+}
+
+.panel-tools {
+  display: grid;
+  gap: 8px;
+  padding: 10px;
+  border-bottom: 1px solid var(--line);
+  background: #fbfcfd;
+}
+
+.bulk-action {
+  width: 100%;
+  min-height: 34px;
+}
+
+input,
+select,
+textarea {
+  width: 100%;
+  border: 1px solid var(--line-dark);
+  border-radius: 6px;
+  background: #fff;
+  color: var(--ink);
+}
+
+input,
+select {
+  min-height: 34px;
+  padding: 6px 8px;
+}
+
+textarea {
+  padding: 8px;
+  resize: vertical;
+}
+
+.record-list,
+.audit-list {
+  overflow: auto;
+  padding: 6px;
+}
+
+.record-item {
+  width: 100%;
+  display: block;
+  border: 1px solid transparent;
+  border-radius: 7px;
+  background: transparent;
+  text-align: left;
+  padding: 9px;
+}
+
+.record-item:hover,
+.record-item.is-active {
+  border-color: var(--line-dark);
+  background: #f3f6f8;
+}
+
+.record-main {
+  display: flex;
+  align-items: center;
+  justify-content: space-between;
+  gap: 8px;
+}
+
+.record-name {
+  font-weight: 700;
+  overflow: hidden;
+  text-overflow: ellipsis;
+  white-space: nowrap;
+}
+
+.tag {
+  display: inline-flex;
+  align-items: center;
+  flex: 0 0 auto;
+  padding: 2px 6px;
+  border-radius: 5px;
+  font-size: 12px;
+  border: 1px solid var(--line);
+  color: var(--muted);
+  background: #fff;
+  white-space: nowrap;
+}
+
+.tag.needs_review,
+.tag.audit-failed {
+  color: var(--bad);
+  border-color: #efb6ae;
+  background: var(--bad-bg);
+}
+
+.tag.auto_corrected,
+.tag.audit-unsure,
+.tag.ai_pending {
+  color: var(--warn);
+  border-color: #efd28b;
+  background: var(--warn-bg);
+}
+
+.tag.reviewed,
+.tag.auto_pass,
+.tag.ai_passed,
+.tag.audit-passed {
+  color: var(--ok);
+  border-color: #b6dec9;
+  background: var(--ok-bg);
+}
+
+.tag.submitted {
+  color: var(--muted);
+  border-color: var(--line-dark);
+  background: #f3f5f7;
+}
+
+.record-sub {
+  margin-top: 5px;
+  color: var(--muted);
+  font-size: 12px;
+  line-height: 1.45;
+}
+
+.record-time {
+  display: block;
+  margin-top: 2px;
+  color: #7b8794;
+}
+
+.pdf-panel {
+  display: flex;
+  flex-direction: column;
+}
+
+.pdf-toolbar,
+.detail-head {
+  min-height: 54px;
+  display: flex;
+  align-items: center;
+  justify-content: space-between;
+  gap: 10px;
+  padding: 10px 12px;
+  border-bottom: 1px solid var(--line);
+  background: #fbfcfd;
+}
+
+.pdf-toolbar > div:first-child {
+  min-width: 0;
+}
+
+.detail-actions {
+  display: flex;
+  align-items: center;
+  justify-content: flex-end;
+  gap: 8px;
+  flex-wrap: wrap;
+}
+
+.ai-actions {
+  display: grid;
+  grid-template-columns: repeat(3, minmax(0, 1fr));
+  gap: 6px;
+}
+
+.ai-actions button {
+  min-height: 32px;
+  border-color: #6f8f35;
+  background: #eef7df;
+  color: #304b12;
+}
+
+.ai-actions button:disabled {
+  cursor: wait;
+  color: var(--muted);
+  border-color: var(--line-dark);
+  background: #eef1f4;
+}
+
+#pdfSubtitle {
+  display: block;
+  margin-top: 3px;
+  color: var(--muted);
+  font-size: 12px;
+}
+
+.pdf-zoom-controls {
+  display: inline-grid;
+  grid-template-columns: 32px 50px 32px 42px;
+  align-items: center;
+  gap: 4px;
+  flex: 0 0 auto;
+}
+
+.pdf-zoom-controls button {
+  width: 100%;
+  height: 30px;
+  padding: 0;
+  border-radius: 6px;
+  font-weight: 700;
+  line-height: 1;
+}
+
+.pdf-zoom-controls span {
+  display: block;
+  color: var(--muted);
+  font-size: 12px;
+  font-weight: 700;
+  text-align: center;
+  font-variant-numeric: tabular-nums;
+}
+
+.view-switch {
+  display: inline-flex;
+  border: 1px solid var(--line-dark);
+  border-radius: 7px;
+  overflow: hidden;
+  background: #fff;
+}
+
+.view-switch button {
+  min-width: 52px;
+  border: 0;
+  border-right: 1px solid var(--line);
+  background: #fff;
+  padding: 6px 8px;
+}
+
+.view-switch button:last-child {
+  border-right: 0;
+}
+
+.view-switch button.is-active {
+  background: var(--accent);
+  color: #fff;
+}
+
+.viewer {
+  flex: 1;
+  min-height: 0;
+  background: #dfe4ea;
+}
+
+.review-strip {
+  display: grid;
+  grid-template-columns: 150px minmax(260px, 1fr);
+  gap: 8px;
+  padding: 8px 10px;
+  border-top: 1px solid var(--line);
+  background: #fbfcfd;
+}
+
+.strip-item {
+  min-width: 0;
+  padding: 6px 8px;
+  border: 1px solid var(--line);
+  border-radius: 7px;
+  background: #fff;
+}
+
+.strip-item span {
+  display: block;
+  color: var(--muted);
+  font-size: 12px;
+  margin-bottom: 3px;
+}
+
+.strip-item strong {
+  display: block;
+  overflow: hidden;
+  text-overflow: ellipsis;
+  white-space: nowrap;
+}
+
+.strip-note {
+  display: grid;
+  grid-template-columns: 96px minmax(0, 1fr);
+  align-items: center;
+  gap: 8px;
+}
+
+.strip-note span {
+  margin: 0;
+}
+
+.strip-note input {
+  min-height: 30px;
+}
+
+#pdfFrame,
+#auditPdfFrame {
+  display: block;
+  width: 100%;
+  height: 100%;
+  border: 0;
+  background: #fff;
+}
+
+.target-summary {
+  border-bottom: 1px solid var(--line);
+  background: #fffafa;
+  padding: 0;
+}
+
+.collapsible-panel > summary {
+  cursor: pointer;
+  list-style: none;
+}
+
+.collapsible-panel > summary::-webkit-details-marker {
+  display: none;
+}
+
+.target-panel > summary,
+.review-log-panel > summary {
+  display: flex;
+  align-items: center;
+  justify-content: space-between;
+  gap: 10px;
+  min-height: 40px;
+  padding: 8px 10px;
+}
+
+.target-panel > summary::before,
+.review-log-panel > summary::before {
+  content: "▸";
+  color: var(--muted);
+  font-size: 12px;
+}
+
+.target-panel[open] > summary::before,
+.review-log-panel[open] > summary::before {
+  content: "▾";
+}
+
+.target-panel > summary strong,
+.review-log-panel > summary h2 {
+  flex: 1;
+}
+
+.target-panel > summary span {
+  color: var(--muted);
+  font-size: 12px;
+}
+
+.target-chips {
+  display: flex;
+  flex-wrap: wrap;
+  gap: 6px;
+  padding: 0 10px 6px;
+}
+
+.target-chips button,
+.target-chips span {
+  border: 1px solid #e7aea7;
+  border-radius: 6px;
+  background: #fff;
+  color: var(--bad);
+  padding: 4px 7px;
+  font-size: 12px;
+}
+
+.target-summary ul {
+  margin: 0;
+  padding: 0 10px 10px 28px;
+  color: #6e332d;
+  font-size: 12px;
+  line-height: 1.5;
+}
+
+#saveBtn,
+.login-panel button,
+.session-box button,
+.audit-save-actions button,
+.page-band button,
+.bulk-action,
+.button-row button {
+  min-height: 34px;
+  border: 1px solid var(--accent-dark);
+  border-radius: 6px;
+  background: var(--accent);
+  color: #fff;
+  padding: 6px 10px;
+  white-space: nowrap;
+}
+
+#saveBtn:disabled {
+  cursor: not-allowed;
+  background: #b9c4ce;
+  border-color: #a9b4bf;
+}
+
+.button-row {
+  display: flex;
+  gap: 8px;
+  flex-wrap: wrap;
+}
+
+.button-row button.danger,
+.audit-save-actions button.danger,
+.page-band button.danger {
+  background: var(--bad);
+  border-color: #9f2118;
+}
+
+.audit-save-actions {
+  display: flex;
+  gap: 6px;
+  flex-wrap: wrap;
+  justify-content: flex-end;
+}
+
+.audit-save-actions button {
+  min-height: 30px;
+  padding: 5px 8px;
+  font-size: 12px;
+}
+
+.detail-form {
+  flex: 1;
+  min-height: 0;
+  overflow: auto;
+  padding: 10px;
+}
+
+.field-group {
+  border: 1px solid var(--line);
+  border-radius: 8px;
+  margin-bottom: 10px;
+  overflow: hidden;
+}
+
+.field-group.is-alert {
+  border-color: #df8e84;
+  box-shadow: inset 3px 0 0 var(--bad);
+}
+
+.field-group summary {
+  cursor: pointer;
+  padding: 8px 10px;
+  background: #eef2f5;
+  font-weight: 700;
+}
+
+.field-group.is-alert summary {
+  background: var(--bad-bg);
+  color: #6e241e;
+}
+
+.field-group summary span {
+  float: right;
+  font-size: 12px;
+  color: var(--bad);
+}
+
+.field-grid {
+  display: grid;
+  grid-template-columns: 1fr 1fr;
+  gap: 9px;
+  padding: 10px;
+}
+
+.field {
+  min-width: 0;
+}
+
+.field.is-wide {
+  grid-column: 1 / -1;
+}
+
+.field > span {
+  display: flex;
+  align-items: center;
+  justify-content: space-between;
+  gap: 8px;
+  margin-bottom: 4px;
+  color: var(--muted);
+  font-size: 12px;
+}
+
+.field > span em {
+  color: var(--bad);
+  font-style: normal;
+}
+
+.field.is-alert input,
+.field.is-alert select,
+.field.is-alert textarea,
+.field.is-alert .json-editor {
+  border-color: #d9796f;
+  background: var(--focus-bg);
+}
+
+.json-editor {
+  border: 1px solid var(--line);
+  border-radius: 7px;
+  overflow: hidden;
+  background: #fff;
+}
+
+.json-toolbar {
+  display: flex;
+  gap: 6px;
+  padding: 6px;
+  border-bottom: 1px solid var(--line);
+  background: #f8fafb;
+}
+
+.json-toolbar button,
+.json-table button {
+  border: 1px solid var(--line-dark);
+  border-radius: 5px;
+  background: #fff;
+  padding: 4px 7px;
+  color: var(--ink);
+}
+
+.json-table-wrap,
+.data-table-wrap {
+  overflow: auto;
+}
+
+.json-table-wrap {
+  max-height: 360px;
+}
+
+.json-table,
+.data-table {
+  width: max-content;
+  min-width: 100%;
+  border-collapse: collapse;
+}
+
+.json-table th,
+.json-table td,
+.data-table th,
+.data-table td {
+  border-bottom: 1px solid var(--line);
+  border-right: 1px solid var(--line);
+  padding: 6px;
+  vertical-align: top;
+}
+
+.json-table th,
+.data-table th {
+  position: sticky;
+  top: 0;
+  z-index: 1;
+  background: #eef2f5;
+  color: #303844;
+  font-size: 12px;
+  white-space: nowrap;
+}
+
+.json-table input {
+  min-width: 120px;
+  border-color: #c8d0d8;
+  background: #fff;
+}
+
+.json-table td:nth-child(4) input,
+.json-table td:nth-child(11) input {
+  min-width: 260px;
+}
+
+.manual-note,
+.review-log-panel {
+  padding: 10px;
+  border-top: 1px solid var(--line);
+  background: #fbfcfd;
+}
+
+.manual-note label,
+.audit-toolbar label,
+.audit-side label {
+  display: grid;
+  gap: 5px;
+  color: var(--muted);
+  font-size: 12px;
+}
+
+.review-log-panel {
+  max-height: 230px;
+  overflow: auto;
+  padding: 0;
+}
+
+.review-log-panel .data-table-wrap {
+  max-height: 184px;
+}
+
+.save-message,
+.inline-message {
+  min-height: 28px;
+  padding: 0 10px 10px;
+  color: var(--muted);
+}
+
+.inline-message {
+  padding: 0;
+}
+
+.save-message.is-ok,
+.inline-message.is-ok {
+  color: var(--ok);
+}
+
+.save-message.is-error,
+.inline-message.is-error {
+  color: var(--bad);
+}
+
+.audit-toolbar {
+  display: grid;
+  grid-template-columns: minmax(180px, 260px) 120px auto 1fr;
+  align-items: end;
+  gap: 10px;
+}
+
+.audit-layout {
+  height: calc(100vh - 170px);
+  display: grid;
+  grid-template-columns: 300px minmax(760px, 1.35fr) minmax(390px, .9fr);
+  gap: 10px;
+}
+
+.audit-list,
+.audit-record,
+.audit-side {
+  min-height: 0;
+  border: 1px solid var(--line);
+  border-radius: 8px;
+  background: #fff;
+  box-shadow: var(--shadow);
+  overflow: auto;
+}
+
+.audit-record {
+  padding: 10px;
+}
+
+.audit-side {
+  padding: 12px;
+  display: grid;
+  align-content: start;
+  gap: 12px;
+}
+
+.audit-title {
+  display: flex;
+  align-items: center;
+  justify-content: space-between;
+  gap: 10px;
+  margin-bottom: 4px;
+}
+
+.audit-meta {
+  color: var(--muted);
+  font-size: 12px;
+  margin-bottom: 8px;
+}
+
+.audit-pdf {
+  width: 100%;
+  height: calc(100vh - 250px);
+  min-height: 420px;
+  border: 1px solid var(--line);
+  background: #fff;
+}
+
+.settings-grid {
+  display: grid;
+  grid-template-columns: minmax(420px, 1fr) minmax(420px, 1fr);
+  gap: 12px;
+}
+
+.settings-wide {
+  grid-column: 1 / -1;
+}
+
+.settings-form {
+  display: grid;
+  grid-template-columns: minmax(180px, 1fr) minmax(180px, 1fr) minmax(360px, 2fr) auto;
+  gap: 10px;
+  align-items: end;
+  margin-bottom: 12px;
+}
+
+.status-check-form {
+  display: grid;
+  grid-template-columns: minmax(74px, 100px) minmax(120px, 1fr) minmax(180px, 1.4fr) 82px auto;
+  gap: 8px;
+  align-items: center;
+}
+
+.status-check-form input {
+  min-height: 34px;
+}
+
+.toggle-field {
+  min-height: 34px;
+  display: inline-flex;
+  align-items: center;
+  gap: 6px;
+  color: var(--ink);
+  white-space: nowrap;
+}
+
+.toggle-field input {
+  width: auto;
+  min-height: 0;
+}
+
+.permission-grid {
+  display: grid;
+  grid-template-columns: repeat(auto-fit, minmax(130px, 1fr));
+  gap: 8px;
+}
+
+.permission-item {
+  display: flex;
+  align-items: center;
+  gap: 7px;
+  min-height: 34px;
+  border: 1px solid var(--line);
+  border-radius: 6px;
+  background: #fbfcfd;
+  padding: 5px 8px;
+}
+
+.permission-item input {
+  width: auto;
+  min-height: auto;
+}
+
+.user-list {
+  display: grid;
+  gap: 10px;
+}
+
+.user-card {
+  border: 1px solid var(--line);
+  border-radius: 8px;
+  padding: 10px;
+  background: #fff;
+}
+
+.user-card-head {
+  display: flex;
+  align-items: center;
+  justify-content: space-between;
+  gap: 10px;
+  margin-bottom: 9px;
+}
+
+.user-card-head span {
+  color: var(--muted);
+  font-size: 12px;
+}
+
+.user-edit-grid {
+  display: grid;
+  grid-template-columns: repeat(2, minmax(180px, 1fr));
+  gap: 10px;
+  margin-bottom: 10px;
+}
+
+.user-edit-grid label {
+  display: grid;
+  gap: 5px;
+  color: var(--muted);
+  font-size: 12px;
+}
+
+.user-actions {
+  margin-top: 10px;
+}
+
+.empty {
+  padding: 18px;
+  color: var(--muted);
+}
+
+.empty.small {
+  padding: 8px;
+  font-size: 12px;
+}
+
+.list-note {
+  position: sticky;
+  top: 0;
+  z-index: 1;
+  padding: 8px 10px;
+  border-bottom: 1px solid var(--line);
+  background: #fffdf4;
+  color: var(--warn);
+  font-size: 12px;
+}
+
+.list-loading,
+.list-end {
+  padding: 10px;
+  color: var(--muted);
+  font-size: 12px;
+  text-align: center;
+}
+
+@media (max-width: 1280px) {
+  .topbar {
+    grid-template-columns: 1fr;
+  }
+
+  .status-grid {
+    width: 100%;
+  }
+
+  .workspace,
+  .audit-layout,
+  .settings-grid,
+  .settings-form {
+    height: auto;
+    min-height: calc(100vh - 130px);
+    grid-template-columns: 1fr;
+  }
+
+  .settings-form {
+    min-height: 0;
+  }
+
+  .status-check-form {
+    grid-template-columns: 1fr;
+  }
+
+  .record-panel,
+  .pdf-panel,
+  .detail-panel,
+  .audit-list,
+  .audit-record,
+  .audit-side {
+    min-height: 420px;
+  }
+
+  .review-strip {
+    grid-template-columns: 1fr 1fr;
+  }
+}
diff --git a/患者首页处理/数据可视化网页端/docker-compose.yml b/患者首页处理/数据可视化网页端/docker-compose.yml
new file mode 100644
index 0000000..7015d70
--- /dev/null
+++ b/患者首页处理/数据可视化网页端/docker-compose.yml
@@ -0,0 +1,20 @@
+name: patient-frontpage-review
+
+services:
+  patient-frontpage-review:
+    build:
+      context: .
+    container_name: patient-frontpage-review
+    env_file:
+      - ../.env
+    environment:
+      REVIEW_SETTINGS_PATH: /app/settings/review_settings.local.json
+    ports:
+      - "8501:8501"
+    volumes:
+      - ../已处理-患者首页PDF/2026_5_25_处理:/data/pdfs:ro
+      - review-settings:/app/settings
+    restart: unless-stopped
+
+volumes:
+  review-settings:
diff --git a/患者首页处理/数据处理工作区/01_配置规则/01_科室分类规则.json b/患者首页处理/数据处理工作区/01_配置规则/01_科室分类规则.json
new file mode 100644
index 0000000..110c8e6
--- /dev/null
+++ b/患者首页处理/数据处理工作区/01_配置规则/01_科室分类规则.json
@@ -0,0 +1,245 @@
+{
+  "说明": "每个具体子科室只归属一个大科室；aliases 用于把图片文件夹名归一到标准子科室名。",
+  "大科室列表": [
+    {
+      "大科室": "肝胆外科及肝移植相关",
+      "子科室": [
+        "肝胆外科1",
+        "肝胆外科2",
+        "肝胆外科3",
+        "肝胆外科4",
+        "肝胆外科5D",
+        "肝胆外科手术室",
+        "肝胆特种病区L",
+        "肝移植内",
+        "肝移植内N"
+      ]
+    },
+    {
+      "大科室": "普通外科及腹部外科",
+      "子科室": [
+        "普外科1",
+        "普外科2",
+        "普外科3",
+        "普外科4D",
+        "腹部外科L",
+        "普通腺体外科"
+      ]
+    },
+    {
+      "大科室": "急诊医学科",
+      "子科室": [
+        "急诊中心",
+        "急诊中心L",
+        "急诊重症医学科L"
+      ]
+    },
+    {
+      "大科室": "重症医学科",
+      "子科室": [
+        "※重症医学1病区",
+        "重症医学2N病区",
+        "重症医学3病区",
+        "重症医学4D病区",
+        "外科ICU",
+        "肝胆ICU",
+        "肿瘤外科重症",
+        "感染重症"
+      ]
+    },
+    {
+      "大科室": "骨科",
+      "子科室": [
+        "骨科",
+        "骨科1",
+        "骨科2"
+      ]
+    },
+    {
+      "大科室": "泌尿外科",
+      "子科室": [
+        "泌尿外1",
+        "泌尿外2"
+      ]
+    },
+    {
+      "大科室": "呼吸内科",
+      "子科室": [
+        "呼吸内科3病区"
+      ]
+    },
+    {
+      "大科室": "耳鼻喉头颈外科",
+      "子科室": [
+        "耳鼻喉D",
+        "耳鼻喉头颈外科",
+        "耳鼻喉头颈外科L",
+        "耳鼻喉2病区"
+      ]
+    },
+    {
+      "大科室": "日间诊疗中心",
+      "子科室": [
+        "日间手术中心",
+        "日间中心病房L"
+      ]
+    },
+    {
+      "大科室": "乳腺外科",
+      "子科室": [
+        "乳腺外科1N",
+        "乳腺外科2N",
+        "乳腺外科2D"
+      ]
+    },
+    {
+      "大科室": "胸外科",
+      "子科室": [
+        "胸外1",
+        "胸外2",
+        "胸外3",
+        "胸外4D"
+      ]
+    },
+    {
+      "大科室": "肝病内科",
+      "子科室": [
+        "肝病内科"
+      ]
+    },
+    {
+      "大科室": "感染科",
+      "子科室": [
+        "感染1",
+        "感染2",
+        "感染3"
+      ]
+    },
+    {
+      "大科室": "肿瘤放疗科",
+      "子科室": [
+        "肿瘤放疗1",
+        "肿瘤放疗2",
+        "肿瘤放疗3",
+        "肿瘤放疗病区L",
+        "肿瘤放疗中心L【无人】",
+        "肿瘤放疗日间【无人】"
+      ]
+    },
+    {
+      "大科室": "特需/涉外病房",
+      "子科室": [
+        "健苑五涉外病房"
+      ]
+    },
+    {
+      "大科室": "老年外科",
+      "子科室": [
+        "老年外科"
+      ]
+    }
+  ],
+  "aliases": {
+    "呼吸内科3": "呼吸内科3病区",
+    "外科ICU": "外科ICU",
+    "急诊中心病房": "急诊中心",
+    "急诊中心": "急诊中心",
+    "急诊中心病房L": "急诊中心L",
+    "急诊中心L": "急诊中心L",
+    "急诊重症医学L": "急诊重症医学科L",
+    "急诊重症L": "急诊重症医学科L",
+    "急诊重症医学科L": "急诊重症医学科L",
+    "感染科1": "感染1",
+    "感染科2": "感染2",
+    "感染科3": "感染3",
+    "感染1": "感染1",
+    "感染2": "感染2",
+    "感染3": "感染3",
+    "感染科重症": "感染重症",
+    "感染重症": "感染重症",
+    "肝病内科": "肝病内科",
+    "肝病内科N": "肝病内科",
+    "肝胆外科1": "肝胆外科1",
+    "肝胆外科2": "肝胆外科2",
+    "肝胆外科3": "肝胆外科3",
+    "肝胆外科4": "肝胆外科4",
+    "肝胆1": "肝胆外科1",
+    "肝胆2": "肝胆外科2",
+    "肝胆3": "肝胆外科3",
+    "肝胆4": "肝胆外科4",
+    "肝胆5D": "肝胆外科5D",
+    "肝胆外科5病房D": "肝胆外科5D",
+    "肝胆特种L": "肝胆特种病区L",
+    "肝胆特种病区L": "肝胆特种病区L",
+    "肝胆移植内": "肝移植内",
+    "肝移植内": "肝移植内",
+    "肝移植内科": "肝移植内",
+    "肝移植内N": "肝移植内N",
+    "肝移植内科N": "肝移植内N",
+    "肿瘤放疗3": "肿瘤放疗3",
+    "肿放3": "肿瘤放疗3",
+    "肿瘤放疗L": "肿瘤放疗病区L",
+    "肿放L": "肿瘤放疗病区L",
+    "重症医学1": "※重症医学1病区",
+    "重症医学科1": "※重症医学1病区",
+    "重症1": "※重症医学1病区",
+    "重症医学2": "重症医学2N病区",
+    "重症医学科2": "重症医学2N病区",
+    "重症2": "重症医学2N病区",
+    "重症2N": "重症医学2N病区",
+    "重症医学3": "重症医学3病区",
+    "重症医学科3": "重症医学3病区",
+    "重症3": "重症医学3病区",
+    "重症医学4": "重症医学4D病区",
+    "重症医学科4": "重症医学4D病区",
+    "重症4": "重症医学4D病区",
+    "重症4D": "重症医学4D病区",
+    "胸外1": "胸外1",
+    "胸外2": "胸外2",
+    "胸外3": "胸外3",
+    "胸外4D": "胸外4D",
+    "普外1": "普外科1",
+    "普外2": "普外科2",
+    "普外3": "普外科3",
+    "普外4D": "普外科4D",
+    "普通外科1": "普外科1",
+    "普通外科2": "普外科2",
+    "普通外科3": "普外科3",
+    "普通外科4": "普外科4D",
+    "普通外科4D": "普外科4D",
+    "普通腺体外科": "普通腺体外科",
+    "腹部外科L": "腹部外科L",
+    "肝胆ICU": "肝胆ICU",
+    "日间手术": "日间手术中心",
+    "日间手术中心": "日间手术中心",
+    "日间手术L": "日间中心病房L",
+    "日间中心病房L": "日间中心病房L",
+    "耳鼻喉": "耳鼻喉头颈外科",
+    "耳鼻喉D": "耳鼻喉D",
+    "耳鼻喉L": "耳鼻喉头颈外科L",
+    "耳鼻喉头颈外科": "耳鼻喉头颈外科",
+    "耳鼻喉头颈外科L": "耳鼻喉头颈外科L",
+    "耳鼻咽喉": "耳鼻喉头颈外科",
+    "耳鼻咽喉头颈外科": "耳鼻喉头颈外科",
+    "耳鼻喉2": "耳鼻喉2病区",
+    "耳鼻喉2D": "耳鼻喉2病区",
+    "耳鼻喉头颈2": "耳鼻喉2病区",
+    "乳腺外科": "乳腺外科1N",
+    "乳腺外科1": "乳腺外科1N",
+    "乳腺外科1N": "乳腺外科1N",
+    "乳腺外科2": "乳腺外科2N",
+    "乳腺外科2N": "乳腺外科2N",
+    "乳腺外科2D": "乳腺外科2D",
+    "乳腺1N": "乳腺外科1N",
+    "乳腺2N": "乳腺外科2N",
+    "乳腺2D": "乳腺外科2D",
+    "骨科": "骨科",
+    "骨科1": "骨科1",
+    "骨科2": "骨科2",
+    "泌尿1": "泌尿外1",
+    "泌尿2": "泌尿外2",
+    "健苑五涉外": "健苑五涉外病房",
+    "健苑五涉外病房": "健苑五涉外病房",
+    "老年外科": "老年外科"
+  }
+}
diff --git a/患者首页处理/数据处理工作区/02_解析入库/02_患者首页PDF解析与入库.py b/患者首页处理/数据处理工作区/02_解析入库/02_患者首页PDF解析与入库.py
new file mode 100755
index 0000000..de1f146
--- /dev/null
+++ b/患者首页处理/数据处理工作区/02_解析入库/02_患者首页PDF解析与入库.py
@@ -0,0 +1,2181 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+"""
+批量处理住院病案首页 PDF。
+
+默认从项目根目录的“待处理-患者首页PDF”读取 PDF，并把 CSV/JSONL/单份 JSON
+写入“数据处理结果区”。脚本只依赖 Python 标准库和系统命令 pdftotext。
+"""
+
+from __future__ import annotations
+
+import argparse
+import csv
+import json
+import os
+import re
+import shutil
+import subprocess
+import sys
+import tempfile
+from datetime import datetime
+from pathlib import Path
+from typing import Any
+
+
+PROJECT_ROOT = Path(__file__).resolve().parents[2]
+DEFAULT_INPUT_DIR = PROJECT_ROOT / "待处理-患者首页PDF"
+DEFAULT_OUTPUT_DIR = PROJECT_ROOT / "数据处理结果区"
+DEFAULT_PG_TABLE = "Patient_FrontPages"
+DEFAULT_DEPARTMENT_RULE_PATH = PROJECT_ROOT / "数据处理工作区" / "01_配置规则" / "01_科室分类规则.json"
+DEFAULT_MINERU_CLIENT_PATH = PROJECT_ROOT / "数据处理工作区" / "05_备用读取" / "05_备用PDF转Markdown_Mineru.py"
+DEFAULT_MINERU_MD_DIR = DEFAULT_OUTPUT_DIR / "06_Mineru_MD"
+DEFAULT_MINERU_URL = os.environ.get("MINERU_URL", "http://10.168.1.103:4000/extract")
+
+
+CSV_COLUMNS = [
+    "住院号",
+    "源文件",
+    "病案号",
+    "首页病案号",
+    "姓名",
+    "性别",
+    "出生日期",
+    "年龄",
+    "身份证号",
+    "新生儿年龄（月）",
+    "新生儿出生体重（克）",
+    "新生儿入院体重（克）",
+    "医疗机构",
+    "组织机构代码",
+    "医疗付费方式",
+    "住院次数",
+    "入院时间",
+    "入院科别",
+    "入院病房",
+    "转科科别",
+    "转科时间",
+    "出院时间",
+    "出院科别",
+    "出院病房",
+    "大科室",
+    "实际住院天数",
+    "门急诊诊断",
+    "门急诊诊断编码",
+    "主要诊断",
+    "主要诊断编码",
+    "主要诊断入院病情",
+    "出院诊断",
+    "手术操作",
+    "病理诊断",
+    "病理诊断编码",
+    "病理号",
+    "药物过敏代码",
+    "过敏药物",
+    "死亡患者尸检代码",
+    "血型代码",
+    "Rh代码",
+    "科主任",
+    "主任副主任医师",
+    "主治医师",
+    "住院医师",
+    "责任护士",
+    "进修医师",
+    "实习医师",
+    "规培医师",
+    "编码员",
+    "病案质量代码",
+    "质控医师",
+    "质控护士",
+    "质控日期",
+    "离院方式代码",
+    "出院31天内再住院计划代码",
+    "总费用",
+    "自付金额",
+    "质控状态",
+    "质控提示",
+    "复核状态",
+    "复核备注",
+    "文本抽取方式",
+    "自动修正",
+    "人工修正",
+]
+
+
+PG_COLUMNS: list[tuple[str, str]] = [
+    ("source_file", "TEXT NOT NULL"),
+    ("inpatient_no", "TEXT"),
+    ("medical_record_no", "TEXT"),
+    ("front_page_medical_record_no", "TEXT"),
+    ("patient_name", "TEXT"),
+    ("gender", "TEXT"),
+    ("birth_date", "DATE"),
+    ("age", "TEXT"),
+    ("nationality", "TEXT"),
+    ("id_card_no", "TEXT"),
+    ("neonatal_age_months", "INTEGER"),
+    ("newborn_birth_weight_g", "INTEGER"),
+    ("newborn_admission_weight_g", "INTEGER"),
+    ("hospital_name", "TEXT"),
+    ("organization_code", "TEXT"),
+    ("payment_method", "TEXT"),
+    ("health_card_no", "TEXT"),
+    ("admission_count", "INTEGER"),
+    ("birthplace", "TEXT"),
+    ("native_place", "TEXT"),
+    ("ethnicity", "TEXT"),
+    ("occupation", "TEXT"),
+    ("marital_status_code", "TEXT"),
+    ("current_address", "TEXT"),
+    ("current_address_phone", "TEXT"),
+    ("current_address_postcode", "TEXT"),
+    ("household_address", "TEXT"),
+    ("household_postcode", "TEXT"),
+    ("employer_address", "TEXT"),
+    ("employer_phone", "TEXT"),
+    ("employer_postcode", "TEXT"),
+    ("contact_name", "TEXT"),
+    ("contact_relationship", "TEXT"),
+    ("contact_address", "TEXT"),
+    ("contact_phone", "TEXT"),
+    ("admission_path_code", "TEXT"),
+    ("admission_time", "TIMESTAMP"),
+    ("admission_dept", "TEXT"),
+    ("admission_ward", "TEXT"),
+    ("transfer_dept", "TEXT"),
+    ("transfer_time", "TEXT"),
+    ("discharge_time", "TIMESTAMP"),
+    ("discharge_dept", "TEXT"),
+    ("discharge_ward", "TEXT"),
+    ("major_department", "TEXT"),
+    ("hospital_days", "INTEGER"),
+    ("outpatient_diagnosis", "TEXT"),
+    ("outpatient_diagnosis_code", "TEXT"),
+    ("primary_diagnosis", "TEXT"),
+    ("primary_diagnosis_code", "TEXT"),
+    ("primary_admission_condition", "TEXT"),
+    ("discharge_diagnoses", "JSONB"),
+    ("operations", "JSONB"),
+    ("injury_poisoning_external_cause", "TEXT"),
+    ("injury_poisoning_code", "TEXT"),
+    ("pathology_diagnosis", "TEXT"),
+    ("pathology_diagnosis_code", "TEXT"),
+    ("pathology_no", "TEXT"),
+    ("drug_allergy_code", "TEXT"),
+    ("allergy_drug", "TEXT"),
+    ("autopsy_code", "TEXT"),
+    ("blood_type_code", "TEXT"),
+    ("rh_code", "TEXT"),
+    ("department_director", "TEXT"),
+    ("chief_physician", "TEXT"),
+    ("attending_physician", "TEXT"),
+    ("resident_physician", "TEXT"),
+    ("responsible_nurse", "TEXT"),
+    ("refresher_physician", "TEXT"),
+    ("intern_physician", "TEXT"),
+    ("standardized_resident_physician", "TEXT"),
+    ("coder", "TEXT"),
+    ("record_quality_code", "TEXT"),
+    ("quality_control_physician", "TEXT"),
+    ("quality_control_nurse", "TEXT"),
+    ("quality_control_date", "DATE"),
+    ("discharge_disposition_code", "TEXT"),
+    ("receiving_org_name", "TEXT"),
+    ("readmission_plan_code", "TEXT"),
+    ("readmission_plan_purpose", "TEXT"),
+    ("coma_before_days", "INTEGER"),
+    ("coma_before_hours", "INTEGER"),
+    ("coma_before_minutes", "INTEGER"),
+    ("coma_after_days", "INTEGER"),
+    ("coma_after_hours", "INTEGER"),
+    ("coma_after_minutes", "INTEGER"),
+    ("total_cost", "NUMERIC(12,2)"),
+    ("self_pay_amount", "NUMERIC(12,2)"),
+    ("fee_details", "JSONB"),
+    ("quality_status", "TEXT"),
+    ("quality_notes", "JSONB"),
+    ("review_status", "TEXT NOT NULL DEFAULT 'pending'"),
+    ("review_notes", "JSONB NOT NULL DEFAULT '[]'::jsonb"),
+    ("manual_corrected", "BOOLEAN NOT NULL DEFAULT false"),
+    ("auto_corrections", "JSONB NOT NULL DEFAULT '[]'::jsonb"),
+    ("text_extraction_method", "TEXT"),
+    ("mineru_markdown_dir", "TEXT"),
+    ("raw_text", "TEXT"),
+]
+
+
+PG_COLUMN_COMMENTS: dict[str, str] = {
+    "id": "自增主键，仅用于数据库内部定位记录。",
+    "source_file": "来源PDF文件名；重复入库时以住院号为准更新。",
+    "inpatient_no": "患者号/住院号，作为首页与患者列表联动唯一键；不能为空，格式由患者目录核验端处理。",
+    "medical_record_no": "病案号，统一保存为10位文本，保留前导0。",
+    "front_page_medical_record_no": "PDF首页病案号，统一保存为10位文本，保留前导0。",
+    "patient_name": "患者姓名。",
+    "gender": "患者性别。",
+    "birth_date": "出生日期。",
+    "age": "首页记录的住院年龄。",
+    "nationality": "国籍。",
+    "id_card_no": "居民身份证号。",
+    "neonatal_age_months": "年龄不足1周岁患儿的年龄（月）。",
+    "newborn_birth_weight_g": "新生儿出生体重（克）。",
+    "newborn_admission_weight_g": "新生儿入院体重（克）。",
+    "hospital_name": "医疗机构名称。",
+    "organization_code": "医疗机构组织机构代码。",
+    "payment_method": "医疗付费方式。",
+    "health_card_no": "健康卡号。",
+    "admission_count": "本机构第几次住院。",
+    "birthplace": "出生地。",
+    "native_place": "籍贯。",
+    "ethnicity": "民族。",
+    "occupation": "职业。",
+    "marital_status_code": "婚姻状况代码：1未婚、2已婚、3丧偶、4离婚、9其他。",
+    "current_address": "现住址。",
+    "current_address_phone": "现住址联系电话。",
+    "current_address_postcode": "现住址邮编。",
+    "household_address": "户口地址。",
+    "household_postcode": "户口地址邮编。",
+    "employer_address": "工作单位及地址。",
+    "employer_phone": "单位电话。",
+    "employer_postcode": "单位邮编。",
+    "contact_name": "联系人姓名，位于首页第一面“联系人姓名”栏。",
+    "contact_relationship": "联系人与患者关系，位于首页第一面“关系”栏。",
+    "contact_address": "联系人地址，位于首页第一面“联系人姓名/关系/地址/电话”这一行的“地址”栏；不是入院途径选项。",
+    "contact_phone": "联系人电话，位于首页第一面“电话”栏。",
+    "admission_path_code": "入院途径代码：1急诊、2门诊、3其他医疗机构转入、9其他。",
+    "admission_time": "入院时间。",
+    "admission_dept": "入院科别。",
+    "admission_ward": "入院病房。",
+    "transfer_dept": "转科科别。",
+    "transfer_time": "转科时间；首页该项常为空或为横线。",
+    "discharge_time": "出院时间。",
+    "discharge_dept": "出院科别。",
+    "discharge_ward": "出院病房。",
+    "major_department": "大科室分类，由科室分类规则根据出院科别优先、入院科别兜底映射。",
+    "hospital_days": "实际住院天数。",
+    "outpatient_diagnosis": "门（急）诊诊断。",
+    "outpatient_diagnosis_code": "门（急）诊诊断疾病编码。",
+    "primary_diagnosis": "主要出院诊断名称。",
+    "primary_diagnosis_code": "主要出院诊断疾病编码。",
+    "primary_admission_condition": "主要诊断入院病情代码。",
+    "discharge_diagnoses": "出院诊断明细JSON数组，包含主要诊断和其他诊断。",
+    "operations": "手术及操作明细JSON数组。",
+    "injury_poisoning_external_cause": "损伤、中毒的外部原因。",
+    "injury_poisoning_code": "损伤、中毒外部原因疾病编码。",
+    "pathology_diagnosis": "病理诊断。",
+    "pathology_diagnosis_code": "病理诊断疾病编码。",
+    "pathology_no": "病理号。",
+    "drug_allergy_code": "药物过敏代码：1无、2有。",
+    "allergy_drug": "过敏药物名称。",
+    "autopsy_code": "死亡患者尸检代码：1是、2否、3-。",
+    "blood_type_code": "ABO血型代码：1A、2B、3O、4AB、5不详、6未查。",
+    "rh_code": "Rh血型代码：1阴、2阳、3不详、4未查。",
+    "department_director": "科主任。",
+    "chief_physician": "主任（副主任）医师。",
+    "attending_physician": "主治医师。",
+    "resident_physician": "住院医师。",
+    "responsible_nurse": "责任护士。",
+    "refresher_physician": "进修医师。",
+    "intern_physician": "实习医师。",
+    "standardized_resident_physician": "规培医师。",
+    "coder": "病案首页编码员。",
+    "record_quality_code": "病案质量代码：1甲、2乙、3丙。",
+    "quality_control_physician": "质控医师。",
+    "quality_control_nurse": "质控护士。",
+    "quality_control_date": "质控日期。",
+    "discharge_disposition_code": "离院方式代码：1医嘱离院、2医嘱转院、3医嘱转社区/乡镇卫生院、4非医嘱离院、5死亡、9其他。",
+    "receiving_org_name": "拟接收医疗机构名称。",
+    "readmission_plan_code": "是否有出院31天内再住院计划代码：1无、2有。",
+    "readmission_plan_purpose": "出院31天内再住院计划目的。",
+    "coma_before_days": "颅脑损伤患者昏迷时间：入院前天数。",
+    "coma_before_hours": "颅脑损伤患者昏迷时间：入院前小时数。",
+    "coma_before_minutes": "颅脑损伤患者昏迷时间：入院前分钟数。",
+    "coma_after_days": "颅脑损伤患者昏迷时间：入院后天数。",
+    "coma_after_hours": "颅脑损伤患者昏迷时间：入院后小时数。",
+    "coma_after_minutes": "颅脑损伤患者昏迷时间：入院后分钟数。",
+    "total_cost": "住院总费用。",
+    "self_pay_amount": "自付金额。",
+    "fee_details": "住院费用分类明细JSON对象。",
+    "quality_status": "程序质控状态。",
+    "quality_notes": "程序质控提示JSON数组。",
+    "review_status": "复核状态：auto_pass自动通过、auto_corrected已自动修正、needs_review需复核、reviewed已人工复核。",
+    "review_notes": "人工或程序复核备注JSON数组。",
+    "manual_corrected": "是否经过人工修正。",
+    "auto_corrections": "程序自动修正记录JSON数组。",
+    "text_extraction_method": "本次解析使用的文本抽取方式：pdftotext或mineru_markdown。",
+    "mineru_markdown_dir": "Mineru Markdown输出目录；未使用Mineru时为空。",
+    "raw_text": "PDF抽取出的首页原始文本，可能来自pdftotext或Mineru Markdown，用于追溯和人工核对。",
+}
+
+
+def normalize_spaces(text: str) -> str:
+    return re.sub(r"[ \t]+", " ", text.strip())
+
+
+def clean_value(value: str | None) -> str:
+    if value is None:
+        return ""
+    value = normalize_spaces(value)
+    return "" if value in {"-", "—", "无"} else value
+
+
+def clean_int_value(value: str | None) -> str:
+    value = clean_value(value)
+    return value if re.fullmatch(r"\d+", value) else ""
+
+
+def first_match(pattern: str, text: str, group: int = 1, flags: int = 0) -> str:
+    match = re.search(pattern, text, flags)
+    if not match:
+        return ""
+    return clean_value(match.group(group))
+
+
+def filename_medical_record_no(pdf_path: Path) -> str:
+    match = re.match(r"^ZY\d{2}(\d{10})", pdf_path.stem, flags=re.IGNORECASE)
+    return match.group(1) if match else ""
+
+
+def filename_admission_count(pdf_path: Path) -> str:
+    match = re.match(r"^ZY(\d{2})\d{10}", pdf_path.stem, flags=re.IGNORECASE)
+    return match.group(1) if match else ""
+
+
+def filename_inpatient_no(pdf_path: Path) -> str:
+    match = re.match(r"^(ZY\d{12})", pdf_path.stem, flags=re.IGNORECASE)
+    return match.group(1).upper() if match else ""
+
+
+def normalize_digits(value: Any, width: int) -> str:
+    digits = re.sub(r"\D", "", clean_value(str(value)) if value is not None else "")
+    if not digits:
+        return ""
+    return digits[-width:].zfill(width)
+
+
+def build_inpatient_no(
+    admission_count: Any,
+    front_page_medical_record_no: Any,
+    medical_record_no: Any,
+    pdf_path: Path,
+) -> tuple[str, list[str], list[str]]:
+    corrections: list[str] = []
+    notes: list[str] = []
+    filename_no = filename_inpatient_no(pdf_path)
+
+    admission = normalize_digits(admission_count, 2) or filename_admission_count(pdf_path)
+    page_no = normalize_digits(front_page_medical_record_no, 10) or normalize_digits(medical_record_no, 10) or filename_medical_record_no(pdf_path)
+    if admission and page_no:
+        inpatient_no = f"ZY{admission}{page_no}"
+        if filename_no and filename_no != inpatient_no:
+            notes.append(f"住院号{inpatient_no}与文件名住院号{filename_no}不一致，请核对住院次数和首页病案号")
+        if clean_value(str(front_page_medical_record_no)) and normalize_digits(front_page_medical_record_no, 10) != clean_value(str(front_page_medical_record_no)):
+            corrections.append(f"首页病案号用于住院号时补齐为{page_no}")
+        return inpatient_no, corrections, notes
+
+    if filename_no:
+        corrections.append(f"住院号由文件名补充为{filename_no}")
+        return filename_no, corrections, notes
+
+    notes.append("住院号无法生成：缺少住院次数或首页病案号")
+    return "", corrections, notes
+
+
+def normalize_medical_record_no(raw_no: str, pdf_path: Path) -> tuple[str, str, list[str], list[str]]:
+    original = clean_value(raw_no)
+    from_filename = filename_medical_record_no(pdf_path)
+    corrections: list[str] = []
+    notes: list[str] = []
+
+    if from_filename and original != from_filename:
+        if original and from_filename.endswith(original):
+            corrections.append(f"病案号由PDF值{original}按文件名补齐为{from_filename}")
+        elif original:
+            notes.append(f"PDF病案号{original}与文件名病案号{from_filename}不一致，已优先采用文件名")
+        else:
+            corrections.append(f"病案号由文件名补充为{from_filename}")
+        return from_filename, original, corrections, notes
+
+    if re.fullmatch(r"\d{1,9}", original):
+        normalized = original.zfill(10)
+        corrections.append(f"病案号由{original}补齐为{normalized}")
+        return normalized, original, corrections, notes
+
+    return original, original, corrections, notes
+
+
+def normalize_department_key(value: str) -> str:
+    return re.sub(r"\s+", "", clean_value(value))
+
+
+def department_lookup_candidates(value: str) -> list[str]:
+    base = normalize_department_key(value)
+    candidates: list[str] = []
+
+    def add(candidate: str) -> None:
+        candidate = normalize_department_key(candidate)
+        if candidate and candidate not in candidates:
+            candidates.append(candidate)
+
+    add(base)
+    add(base.replace("病房", ""))
+    add(base.replace("病区", ""))
+    add(base.replace("病房", "").replace("病区", ""))
+
+    for candidate in list(candidates):
+        add(candidate.replace("胸外科", "胸外"))
+        add(candidate.replace("泌尿外科", "泌尿外"))
+        add(candidate.replace("感染科", "感染"))
+        add(candidate.replace("普通外科", "普外科"))
+
+    return candidates
+
+
+def load_department_rules(rule_path: Path) -> dict[str, dict[str, str]]:
+    if not rule_path.exists():
+        return {}
+    data = json.loads(rule_path.read_text(encoding="utf-8"))
+    aliases = {
+        normalize_department_key(alias): normalize_department_key(standard)
+        for alias, standard in data.get("aliases", {}).items()
+    }
+    standard_to_major: dict[str, str] = {}
+    for group in data.get("大科室列表", []):
+        major = clean_value(group.get("大科室", ""))
+        for department in group.get("子科室", []):
+            standard = normalize_department_key(department)
+            standard_to_major[standard] = major
+            aliases.setdefault(standard, standard)
+    return {"aliases": aliases, "standard_to_major": standard_to_major}
+
+
+def classify_major_department(record: dict[str, Any], rules: dict[str, dict[str, str]]) -> tuple[str, str]:
+    if not rules:
+        return "", ""
+    aliases = rules.get("aliases", {})
+    standard_to_major = rules.get("standard_to_major", {})
+    for source_key in ["出院科别", "入院科别"]:
+        for candidate in department_lookup_candidates(str(record.get(source_key, ""))):
+            standard = aliases.get(candidate, candidate)
+            major = standard_to_major.get(standard)
+            if major:
+                return major, standard
+    return "", ""
+
+
+def extract_text_with_pdftotext(pdf_path: Path) -> str:
+    if shutil.which("pdftotext") is None:
+        raise RuntimeError("未找到 pdftotext。请先安装 poppler-utils 后再运行。")
+
+    command = ["pdftotext", "-layout", "-enc", "UTF-8", str(pdf_path), "-"]
+    completed = subprocess.run(
+        command,
+        check=False,
+        stdout=subprocess.PIPE,
+        stderr=subprocess.PIPE,
+        text=True,
+        encoding="utf-8",
+        errors="replace",
+    )
+    if completed.returncode != 0:
+        raise RuntimeError(completed.stderr.strip() or "pdftotext 转换失败")
+    return completed.stdout.replace("\r\n", "\n").replace("\r", "\n")
+
+
+def text_needs_mineru_fallback(text: str) -> bool:
+    stripped = text.strip()
+    if len(stripped) < 200:
+        return True
+    required_markers = ["病案号", "医疗机构", "主要诊断"]
+    return sum(1 for marker in required_markers if marker in stripped) < 2
+
+
+def markdown_files_for_pdf(pdf_path: Path, md_dir: Path) -> list[Path]:
+    folder = md_dir / pdf_path.stem
+    if not folder.exists():
+        return []
+    return sorted([*folder.rglob("*.md"), *folder.rglob("*.markdown"), *folder.rglob("*.txt")])
+
+
+def read_mineru_markdown(pdf_path: Path, md_dir: Path) -> str:
+    parts: list[str] = []
+    for file_path in markdown_files_for_pdf(pdf_path, md_dir):
+        parts.append(file_path.read_text(encoding="utf-8", errors="replace"))
+    return "\n\n".join(parts).replace("\r\n", "\n").replace("\r", "\n")
+
+
+def ensure_mineru_markdown(pdf_path: Path, args: argparse.Namespace) -> None:
+    md_dir = args.mineru_md_dir.resolve()
+    if markdown_files_for_pdf(pdf_path, md_dir):
+        return
+    client_path = args.mineru_client.resolve()
+    if not client_path.exists():
+        raise RuntimeError(f"未找到 Mineru 客户端：{client_path}")
+
+    command = [
+        sys.executable,
+        str(client_path),
+        "-s",
+        str(pdf_path.parent.resolve()),
+        "-t",
+        str(md_dir),
+        "-u",
+        args.mineru_url,
+    ]
+    if args.mineru_sync:
+        command.append("--sync")
+    completed = subprocess.run(
+        command,
+        check=False,
+        stdout=subprocess.PIPE,
+        stderr=subprocess.PIPE,
+        text=True,
+        encoding="utf-8",
+        errors="replace",
+    )
+    if completed.returncode != 0:
+        raise RuntimeError(completed.stderr.strip() or completed.stdout.strip() or "Mineru PDF转Markdown失败")
+
+
+def extract_record_text(pdf_path: Path, args: argparse.Namespace) -> tuple[str, str, str]:
+    text_source = args.text_source
+    if text_source == "pdftotext":
+        return extract_text_with_pdftotext(pdf_path), "pdftotext", ""
+
+    if text_source == "mineru":
+        ensure_mineru_markdown(pdf_path, args)
+        markdown_text = read_mineru_markdown(pdf_path, args.mineru_md_dir.resolve())
+        if not markdown_text.strip():
+            raise RuntimeError(f"Mineru未生成可读取的Markdown：{pdf_path.name}")
+        return markdown_text, "mineru_markdown", str((args.mineru_md_dir.resolve() / pdf_path.stem))
+
+    try:
+        pdftotext_text = extract_text_with_pdftotext(pdf_path)
+    except Exception:
+        pdftotext_text = ""
+    if pdftotext_text and not text_needs_mineru_fallback(pdftotext_text):
+        return pdftotext_text, "pdftotext", ""
+
+    try:
+        ensure_mineru_markdown(pdf_path, args)
+        markdown_text = read_mineru_markdown(pdf_path, args.mineru_md_dir.resolve())
+        if markdown_text.strip() and not text_needs_mineru_fallback(markdown_text):
+            return markdown_text, "mineru_markdown", str((args.mineru_md_dir.resolve() / pdf_path.stem))
+    except Exception as exc:
+        if pdftotext_text:
+            return pdftotext_text, f"pdftotext_mineru_failed:{exc}", ""
+        raise
+
+    if pdftotext_text:
+        return pdftotext_text, "pdftotext_suspicious_mineru_unusable", ""
+    return "", "empty", ""
+
+
+def meaningful_lines(text: str) -> list[str]:
+    lines: list[str] = []
+    for raw in text.splitlines():
+        line = raw.strip()
+        if line:
+            lines.append(line)
+    return lines
+
+
+def parse_header(text: str, lines: list[str]) -> dict[str, Any]:
+    result: dict[str, Any] = {
+        "组织机构代码": first_match(r"组织机构代码:\s*([^)）]+)", text),
+        "医疗机构": first_match(r"医疗机构:\s*(.+)", text),
+        "医疗付费方式": first_match(r"医疗付费方式:\s*(.+)", text),
+        "健康卡号": first_match(r"健康卡号:\s*(.*?)\s*第", text),
+        "住院次数": first_match(r"第\s*(\d+)\s*次住院", text),
+        "病案号": first_match(r"病案号:\s*([A-Za-z0-9]+)", text),
+    }
+
+    for idx, line in enumerate(lines):
+        if "病案号:" not in line or idx + 1 >= len(lines):
+            continue
+        patient_line = lines[idx + 1]
+        match = re.search(
+            r"^\s*(?P\S+)\s+(?P男|女)\s+"
+            r"(?P\d{4})\s*年\s*(?P\d{1,2})\s*月\s*"
+            r"(?P\d{1,2})\s*日\s+(?P[\d.]+)\s+(?P\S+)",
+            patient_line,
+        )
+        if match:
+            result.update(
+                {
+                    "姓名": clean_value(match.group("name")),
+                    "性别": match.group("sex"),
+                    "出生日期": f"{int(match.group('year')):04d}-{int(match.group('month')):02d}-{int(match.group('day')):02d}",
+                    "年龄": clean_value(match.group("age")),
+                    "国籍": clean_value(match.group("nation")),
+                }
+            )
+        break
+
+    return result
+
+
+def parse_newborn_info(lines: list[str]) -> dict[str, str]:
+    result = {
+        "新生儿年龄（月）": "",
+        "新生儿出生体重（克）": "",
+        "新生儿入院体重（克）": "",
+    }
+    for idx, line in enumerate(lines):
+        if "年龄不足1周岁" not in line:
+            continue
+        joined = line + " " + (lines[idx + 1] if idx + 1 < len(lines) else "")
+        result["新生儿年龄（月）"] = clean_int_value(first_match(r"）\s*([-\d]+)\s*月", joined))
+        weights = [clean_int_value(item) for item in re.findall(r"([-\d]+)\s*克", joined)]
+        weights = [item for item in weights if item]
+        if weights:
+            result["新生儿出生体重（克）"] = weights[0]
+            result["新生儿入院体重（克）"] = weights[-1]
+        break
+    return result
+
+
+def parse_basic_lines(lines: list[str]) -> dict[str, Any]:
+    result: dict[str, Any] = {}
+    id_index = -1
+
+    for idx, line in enumerate(lines):
+        match = re.match(r"^([1-9]\d{5}(?:18|19|20)\d{2}\d{2}\d{2}\d{3}[\dXx]|-)\s+(.+?)\s+([1-9])\s+1\.未婚", line)
+        if match:
+            id_index = idx
+            result["身份证号"] = clean_value(match.group(1)).upper()
+            result["职业"] = clean_value(match.group(2))
+            result["婚姻代码"] = clean_value(match.group(3))
+            if idx > 0:
+                parts = re.split(r"\s{2,}", lines[idx - 1].strip())
+                if len(parts) >= 3:
+                    result["出生地"] = clean_value(parts[0])
+                    result["籍贯"] = clean_value(parts[1])
+                    result["民族"] = clean_value(parts[-1])
+            break
+
+    if id_index >= 0:
+        address_names = [
+            ("现住址", "现住址电话", "现住址邮编"),
+            ("户口地址", "", "户口地址邮编"),
+            ("工作单位及地址", "单位电话", "单位邮编"),
+        ]
+        for offset, names in enumerate(address_names, start=1):
+            if id_index + offset >= len(lines):
+                continue
+            address, phone, postcode = split_address_phone_postcode(lines[id_index + offset])
+            result[names[0]] = address
+            if names[1]:
+                result[names[1]] = phone
+            result[names[2]] = postcode
+
+        contact_line = find_contact_line(lines, id_index + 4)
+        if contact_line:
+            contact = parse_contact_line(contact_line)
+            result.update(contact)
+
+    return result
+
+
+def split_address_phone_postcode(line: str) -> tuple[str, str, str]:
+    pieces = re.split(r"\s{2,}", line.strip())
+    if not pieces:
+        return "", "", ""
+    postcode = ""
+    phone = ""
+    if pieces and re.fullmatch(r"[\d-]{5,}|-", pieces[-1]):
+        postcode = clean_value(pieces.pop())
+    if pieces and re.fullmatch(r"[\d-]{7,}|-", pieces[-1]):
+        phone = clean_value(pieces.pop())
+    return clean_value(" ".join(pieces)), phone, postcode
+
+
+def parse_contact_line(line: str) -> dict[str, str]:
+    if is_admission_path_line(line):
+        return {
+            "联系人姓名": "",
+            "联系人关系": "",
+            "联系人地址": "",
+            "联系人电话": "",
+        }
+    pieces = [clean_value(p) for p in re.split(r"\s{2,}", line.strip()) if clean_value(p)]
+    result = {
+        "联系人姓名": "",
+        "联系人关系": "",
+        "联系人地址": "",
+        "联系人电话": "",
+    }
+    if len(pieces) >= 1:
+        result["联系人姓名"] = pieces[0]
+    if len(pieces) >= 2:
+        result["联系人关系"] = pieces[1]
+    if len(pieces) >= 3:
+        if re.fullmatch(r"\d{7,}", pieces[-1]):
+            result["联系人电话"] = pieces[-1]
+            result["联系人地址"] = clean_value(" ".join(pieces[2:-1]))
+        else:
+            result["联系人地址"] = clean_value(" ".join(pieces[2:]))
+    return result
+
+
+def is_admission_path_line(line: str) -> bool:
+    return bool(re.search(r"1\.急诊\s+2\.门诊", line)) or "其他医疗机构转入" in line
+
+
+def find_contact_line(lines: list[str], start_index: int) -> str:
+    for idx in range(start_index, min(start_index + 3, len(lines))):
+        line = lines[idx]
+        if is_admission_path_line(line):
+            return ""
+        if re.search(r"\d{4}\s*年\s*\d{1,2}\s*月", line):
+            return ""
+        pieces = [clean_value(p) for p in re.split(r"\s{2,}", line.strip()) if clean_value(p)]
+        if len(pieces) >= 2:
+            return line
+    return ""
+
+
+def parse_admission_discharge(lines: list[str]) -> dict[str, Any]:
+    result: dict[str, Any] = {}
+    date_line_indexes: list[int] = []
+
+    for idx, line in enumerate(lines):
+        if re.search(r"\d{4}\s*年\s*\d{1,2}\s*月\s*\d{1,2}\s*日\s*\d{1,2}\s*时", line):
+            date_line_indexes.append(idx)
+
+    if date_line_indexes:
+        result.update(parse_visit_line(lines[date_line_indexes[0]], prefix="入院"))
+        if date_line_indexes[0] + 1 < len(lines):
+            result["转科科别"] = clean_value(lines[date_line_indexes[0] + 1])
+        if date_line_indexes[0] + 2 < len(lines):
+            result["转科时间"] = clean_value(lines[date_line_indexes[0] + 2])
+    if len(date_line_indexes) >= 2:
+        result.update(parse_visit_line(lines[date_line_indexes[1]], prefix="出院"))
+
+    for idx, line in enumerate(lines):
+        if re.search(r"1\.急诊\s+2\.门诊", line):
+            result["入院途径代码"] = first_match(r"^([1-9-])\s+1\.急诊", line)
+        if "天" in line and idx in date_line_indexes:
+            result["实际住院天数"] = first_match(r"([0-9]+)\s*天\s*$", line)
+
+    if len(date_line_indexes) >= 2:
+        discharge_line = lines[date_line_indexes[1]]
+        result["实际住院天数"] = first_match(r"([0-9]+)\s*天\s*$", discharge_line) or result.get("实际住院天数", "")
+        if date_line_indexes[1] + 1 < len(lines):
+            diagnosis_line = lines[date_line_indexes[1] + 1]
+            diagnosis_match = re.match(r"(.+?)\s+([A-Z]\d{2}[\w.]*\d*)\s*$", diagnosis_line)
+            if diagnosis_match:
+                result["门急诊诊断"] = clean_value(diagnosis_match.group(1))
+                result["门急诊诊断编码"] = clean_value(diagnosis_match.group(2))
+
+    return result
+
+
+def parse_visit_line(line: str, prefix: str) -> dict[str, str]:
+    pattern = (
+        r"(?P\d{4})\s*年\s*(?P\d{1,2})\s*月\s*"
+        r"(?P\d{1,2})\s*日\s*(?P\d{1,2})\s*时\s+"
+        r"(?P.+?)\s{2,}(?P\S+)"
+    )
+    match = re.search(pattern, line)
+    if not match:
+        return {}
+    return {
+        f"{prefix}时间": (
+            f"{int(match.group('year')):04d}-{int(match.group('month')):02d}-"
+            f"{int(match.group('day')):02d} {int(match.group('hour')):02d}:00"
+        ),
+        f"{prefix}科别": clean_value(match.group("dept")),
+        f"{prefix}病房": clean_value(match.group("ward")),
+    }
+
+
+def parse_diagnoses(lines: list[str]) -> dict[str, Any]:
+    section = diagnosis_section(lines)
+    diagnoses: list[dict[str, str]] = []
+    if not section:
+        return {"出院诊断": diagnoses, "主要诊断": {}}
+
+    main_index = next((idx for idx, line in enumerate(section) if "主要诊断" in line), -1)
+    if main_index >= 0:
+        main_line = section[main_index]
+        normalized_main_line = normalize_spaces(main_line)
+        main_match = re.search(r"主要诊断\s*(.*?)\s+([A-Z]\d{2}[\w.]*\d*)\s+([0-9-])", normalized_main_line)
+        if main_match:
+            name = clean_value(main_match.group(1))
+            if not name and main_index > 0:
+                name = clean_value(section[main_index - 1])
+                if main_index + 1 < len(section) and not re.search(r"[A-Z]\d{2}[\w.]*\d*", section[main_index + 1]):
+                    name = clean_value(name + section[main_index + 1])
+            diagnoses.append(
+                {
+                    "诊断类别": "主要诊断",
+                    "出院诊断": name,
+                    "疾病编码": clean_value(main_match.group(2)),
+                    "入院病情": clean_value(main_match.group(3)),
+                }
+            )
+        else:
+            no_code_match = re.search(r"主要诊断\s*(.*?)\s+([0-9-])(?:\s+其他诊断)?$", normalized_main_line)
+            if no_code_match:
+                name = clean_value(no_code_match.group(1))
+                if not name:
+                    nearby_name_parts: list[str] = []
+                    if main_index > 0:
+                        previous_line = clean_value(section[main_index - 1])
+                        if previous_line and "其他诊断" not in previous_line and not re.search(r"[A-Z]\d{2}[\w.]*\d*", previous_line):
+                            nearby_name_parts.append(previous_line)
+                    if main_index + 1 < len(section):
+                        next_line = clean_value(section[main_index + 1])
+                        if next_line and "其他诊断" not in next_line and not re.search(r"[A-Z]\d{2}[\w.]*\d*", next_line):
+                            nearby_name_parts.append(next_line)
+                    name = clean_value("".join(nearby_name_parts))
+                if name:
+                    diagnoses.append(
+                        {
+                            "诊断类别": "主要诊断",
+                            "出院诊断": name,
+                            "疾病编码": "",
+                            "入院病情": clean_value(no_code_match.group(2)),
+                        }
+                    )
+
+    other_started = False
+    for line in section:
+        if "其他诊断" in line:
+            other_started = True
+            line = re.sub(r"^.*?其他诊断\s*", "", line).strip()
+            if not line:
+                continue
+        elif not other_started:
+            continue
+        parsed = parse_other_diagnosis_line(line)
+        if parsed:
+            diagnoses.append(parsed)
+
+    main = next((item for item in diagnoses if item["诊断类别"] == "主要诊断"), {})
+    return {"出院诊断": diagnoses, "主要诊断": main}
+
+
+def diagnosis_section(lines: list[str]) -> list[str]:
+    start = -1
+    for idx, line in enumerate(lines):
+        if "主要诊断" in line:
+            start = max(0, idx - 1)
+            break
+    if start < 0:
+        return []
+
+    section: list[str] = []
+    for line in lines[start:]:
+        if "B20" in line or re.match(r"^-+\s+-+$", line) or "1.无 2.有" in line:
+            break
+        section.append(line)
+    return section
+
+
+def parse_other_diagnosis_line(line: str) -> dict[str, str] | None:
+    line = normalize_spaces(line)
+    match = re.match(r"^(?:其他诊断\s+)?(.+?)\s+([A-Z]\d{2}[\w.]*\d*)\s+([0-9-])$", line)
+    if match:
+        return {
+            "诊断类别": "其他诊断",
+            "出院诊断": clean_value(match.group(1)),
+            "疾病编码": clean_value(match.group(2)),
+            "入院病情": clean_value(match.group(3)),
+        }
+    no_code_match = re.match(r"^(?:其他诊断\s+)?(.+?)\s+([0-9-])$", line)
+    if no_code_match:
+        return {
+            "诊断类别": "其他诊断",
+            "出院诊断": clean_value(no_code_match.group(1)),
+            "疾病编码": "",
+            "入院病情": clean_value(no_code_match.group(2)),
+        }
+    return None
+
+
+def parse_operations(lines: list[str]) -> list[dict[str, str]]:
+    operation_lines: list[str] = []
+    after_quality_line = False
+    for raw in lines:
+        line = raw.replace("\f", "").strip()
+        if not line:
+            continue
+        if "1.甲" in line and "2.乙" in line and "3.丙" in line:
+            after_quality_line = True
+            continue
+        if not after_quality_line and re.match(r"^[A-Z0-9]\d?\.\d", line):
+            after_quality_line = True
+        if after_quality_line and "1.医嘱离院" in line:
+            break
+        if after_quality_line:
+            operation_lines.append(line)
+
+    combined: list[str] = []
+    pending_prefix = ""
+    for line in operation_lines:
+        has_date = bool(re.search(r"\d{4}-\d{1,2}-\d{1,2}", line))
+        starts_with_code = bool(re.match(r"^[A-Z0-9]\d?\.\d", line))
+        if starts_with_code and not has_date:
+            pending_prefix = normalize_spaces(line)
+            continue
+        if has_date:
+            if starts_with_code:
+                combined.append(line)
+            elif pending_prefix:
+                combined.append(normalize_spaces(pending_prefix + " " + line))
+                pending_prefix = ""
+            else:
+                combined.append(line)
+        elif combined:
+            combined[-1] = normalize_spaces(combined[-1] + " " + line)
+
+    operations: list[dict[str, str]] = []
+    for line in combined:
+        match = re.match(
+            r"^(?:(?P[A-Z0-9.]+[a-zA-Z]*)\s+)?(?:(?P.*?)\s+)?"
+            r"(?P\d{4}-\d{1,2}-\d{1,2})\s+(?P.+)$",
+            line,
+        )
+        if not match:
+            continue
+        operation = parse_operation_columns(
+            code=clean_value(match.group("code")),
+            date=normalize_date(match.group("date")),
+            predate=clean_value(match.group("predate") or ""),
+            rest=clean_value(match.group("rest")),
+            raw_line=clean_value(line),
+        )
+        operations.append(operation)
+    return operations
+
+
+def parse_operation_columns(code: str, date: str, predate: str, rest: str, raw_line: str) -> dict[str, str]:
+    tokens = [clean_operation_token(token) for token in rest.split() if clean_operation_token(token)]
+    anesthesia_hint = predate if is_anesthesia_text(predate) else ""
+    name_prefix = "" if anesthesia_hint else predate
+
+    level = ""
+    if tokens and is_operation_level_token(tokens[0]):
+        raw_level = tokens.pop(0)
+        level = normalize_operation_level(raw_level)
+    if level == "诊断性操作" and tokens and tokens[-1] == "作":
+        level = "诊断性操作"
+        tokens.pop()
+
+    incision_index = next((idx for idx, token in enumerate(tokens) if is_incision_healing_token(token)), -1)
+    incision_healing = ""
+    if incision_index >= 0:
+        before_incision = tokens[:incision_index]
+        incision_healing = tokens[incision_index]
+        after_incision = tokens[incision_index + 1 :]
+    else:
+        before_incision = list(tokens)
+        after_incision = []
+
+    if incision_index < 0:
+        before_incision, anesthesiologist = split_no_incision_tail(before_incision)
+        anesthesia_method = anesthesia_hint
+        name_suffix = ""
+    else:
+        anesthesia_method, anesthesiologist, name_suffix = split_anesthesia_tail(after_incision, anesthesia_hint)
+
+    procedure_name, surgeon, assistant_1, assistant_2 = split_procedure_and_doctors(before_incision)
+    procedure_name = clean_value(" ".join(part for part in [name_prefix, procedure_name, name_suffix] if part))
+
+    return {
+        "手术操作编码": code,
+        "手术操作日期": date,
+        "手术级别": level,
+        "手术操作名称": procedure_name,
+        "术者": surgeon,
+        "I助": assistant_1,
+        "II助": assistant_2,
+        "切口愈合等级": incision_healing,
+        "麻醉方式": anesthesia_method,
+        "麻醉医师": anesthesiologist,
+        "原始内容": raw_line,
+    }
+
+
+def clean_operation_token(token: str) -> str:
+    token = token.strip().strip("()（）[]【】")
+    token = token.replace("Ⅰ", "I").replace("Ⅱ", "II").replace("Ⅲ", "III").replace("Ⅳ", "IV")
+    token = token.replace("／", "/")
+    return clean_value(token)
+
+
+def is_operation_level_token(token: str) -> bool:
+    return bool(token and (token in {"一级", "二级", "三级", "四级", "诊断性", "诊断性操作", "性操"} or token.endswith("级")))
+
+
+def normalize_operation_level(token: str) -> str:
+    return "诊断性操作" if token in {"性操", "诊断性"} else token
+
+
+def is_incision_healing_token(token: str) -> bool:
+    return bool(re.fullmatch(r"(?:0|I|II|III|IV|V|[一二三四五])/[甲乙丙-]", token))
+
+
+def is_anesthesia_text(text: str) -> bool:
+    return bool(text and re.search(r"麻醉|浸润|静吸|全凭|硬膜|局部", text))
+
+
+def has_procedure_keyword(token: str) -> bool:
+    return bool(re.search(r"术|切除|穿刺|栓塞|成型|松解|活检|操作|置入|修补|吻合", token))
+
+
+def looks_like_person_name(token: str) -> bool:
+    if not re.fullmatch(r"[\u4e00-\u9fa5]{2,4}", token):
+        return False
+    if has_procedure_keyword(token):
+        return False
+    if re.search(r"腹腔|经皮|经导管|CT|胆|肝|肺|肠|阑尾|粘连|总管|局部|静吸|全凭|脉麻醉", token):
+        return False
+    return True
+
+
+def split_procedure_and_doctors(tokens: list[str]) -> tuple[str, str, str, str]:
+    doctor_tokens: list[str] = []
+    remaining = list(tokens)
+    while remaining and len(doctor_tokens) < 3 and looks_like_person_name(remaining[-1]):
+        doctor_tokens.insert(0, remaining.pop())
+    procedure_name = clean_value(" ".join(remaining))
+    doctor_tokens = doctor_tokens[-3:]
+    surgeon = doctor_tokens[0] if len(doctor_tokens) >= 1 else ""
+    assistant_1 = doctor_tokens[1] if len(doctor_tokens) >= 2 else ""
+    assistant_2 = doctor_tokens[2] if len(doctor_tokens) >= 3 else ""
+    return procedure_name, surgeon, assistant_1, assistant_2
+
+
+def split_anesthesia_tail(tokens: list[str], anesthesia_hint: str = "") -> tuple[str, str, str]:
+    clean_tokens = [token for token in (clean_operation_token(item) for item in tokens) if token]
+    clean_tokens = [token for token in clean_tokens if not re.fullmatch(r"\d{2,4}", token) and token != "麻醉"]
+    doctor_index = next((idx for idx, token in enumerate(clean_tokens) if looks_like_person_name(token)), -1)
+    if doctor_index < 0:
+        method = clean_value("".join([anesthesia_hint, *clean_tokens]))
+        return method, "", ""
+
+    anesthesiologist = clean_tokens[doctor_index]
+    method_parts = clean_tokens[:doctor_index]
+    suffix_parts: list[str] = []
+    for token in clean_tokens[doctor_index + 1 :]:
+        if has_procedure_keyword(token):
+            suffix_parts.append(token)
+        else:
+            method_parts.append(token)
+    method = clean_value("".join([anesthesia_hint, *method_parts]))
+    name_suffix = clean_value(" ".join(suffix_parts))
+    return method, anesthesiologist, name_suffix
+
+
+def split_no_incision_tail(tokens: list[str]) -> tuple[list[str], str]:
+    remaining = [token for token in tokens if token != "麻醉" and not re.fullmatch(r"\d{2,4}", token)]
+    if remaining and looks_like_person_name(remaining[-1]):
+        return remaining[:-1], remaining[-1]
+    return remaining, ""
+
+
+def normalize_date(value: str) -> str:
+    match = re.match(r"(\d{4})-(\d{1,2})-(\d{1,2})", value)
+    if not match:
+        return value
+    return f"{int(match.group(1)):04d}-{int(match.group(2)):02d}-{int(match.group(3)):02d}"
+
+
+def parse_fees(text: str) -> dict[str, Any]:
+    result: dict[str, Any] = {
+        "总费用": first_match(r"\n\s*([0-9]+\.[0-9]{2})\s*\(\s*自付金额", text),
+        "自付金额": first_match(r"自付金额\s*([0-9]+\.[0-9]{2})", text),
+        "费用明细": {},
+    }
+    for index, name, amount in re.findall(r"\((\d+)\)\s*([^:：()]+?)\s*[:：]\s*([0-9]+(?:\.[0-9]+)?)", text):
+        key = f"{int(index):02d}_{clean_value(name)}"
+        result["费用明细"][key] = amount
+    for name, amount in re.findall(r"(? dict[str, str]:
+    result = {
+        "损伤中毒外部原因": "",
+        "损伤中毒疾病编码": "",
+        "病理诊断": "",
+        "病理诊断编码": "",
+        "病理号": "",
+        "药物过敏代码": "",
+        "过敏药物": "",
+        "死亡患者尸检代码": "",
+        "血型代码": "",
+        "Rh代码": "",
+        "科主任": "",
+        "主任副主任医师": "",
+        "主治医师": "",
+        "住院医师": "",
+        "责任护士": "",
+        "进修医师": "",
+        "实习医师": "",
+        "规培医师": "",
+        "编码员": "",
+        "病案质量代码": "",
+        "质控医师": "",
+        "质控护士": "",
+        "质控日期": "",
+    }
+
+    pathology_no_idx = next((idx for idx, line in enumerate(lines) if re.match(r"^B\d+", line)), -1)
+    if pathology_no_idx >= 0:
+        result["病理号"] = clean_value(lines[pathology_no_idx])
+        if pathology_no_idx > 0:
+            pathology_cols = split_layout_columns(lines[pathology_no_idx - 1])
+            if len(pathology_cols) >= 2:
+                result["病理诊断"] = clean_value(" ".join(pathology_cols[:-1]))
+                result["病理诊断编码"] = clean_value(pathology_cols[-1])
+        if pathology_no_idx > 2:
+            injury_cols = split_layout_columns(lines[pathology_no_idx - 2])
+            if len(injury_cols) >= 2:
+                result["损伤中毒外部原因"] = clean_value(injury_cols[0])
+                result["损伤中毒疾病编码"] = clean_value(injury_cols[-1])
+
+    allergy_idx = next((idx for idx, line in enumerate(lines) if "1.无 2.有" in line and "1.是 2.否" in line), -1)
+    if allergy_idx >= 0:
+        match = re.search(r"^\s*([1-9-])\s+1\.无 2\.有\s+(.*?)\s+([1-9-])\s+1\.是 2\.否", lines[allergy_idx])
+        if match:
+            result["药物过敏代码"] = clean_value(match.group(1))
+            result["过敏药物"] = clean_value(match.group(2))
+            result["死亡患者尸检代码"] = clean_value(match.group(3))
+
+    blood_idx = next((idx for idx, line in enumerate(lines) if "1.A" in line and "1.阴" in line), -1)
+    if blood_idx >= 0:
+        result["血型代码"] = first_match(r"^\s*([1-9-])\s+1\.A", lines[blood_idx])
+        result["Rh代码"] = first_match(r"6\.未查\s+([1-9-])\s+1\.阴", lines[blood_idx])
+        doctor_rows = [split_layout_columns_keep_positions(lines[i]) for i in range(blood_idx + 1, min(blood_idx + 3, len(lines)))]
+        if doctor_rows:
+            for key, value in zip(["科主任", "主任副主任医师", "主治医师", "住院医师"], doctor_rows[0]):
+                result[key] = clean_value(value)
+        if len(doctor_rows) >= 2:
+            for key, value in zip(["责任护士", "进修医师", "实习医师", "规培医师", "编码员"], doctor_rows[1]):
+                result[key] = clean_value(value)
+
+    quality_idx = next((idx for idx, line in enumerate(lines) if "1.甲" in line and "2.乙" in line and "3.丙" in line), -1)
+    if quality_idx >= 0:
+        line = lines[quality_idx]
+        result["病案质量代码"] = first_match(r"^\s*([1-9-])\s+1\.甲", line)
+        date_match = re.search(r"(\d{4})\s*年\s*(\d{1,2})\s*月\s*(\d{1,2})\s*日", line)
+        if date_match:
+            result["质控日期"] = f"{int(date_match.group(1)):04d}-{int(date_match.group(2)):02d}-{int(date_match.group(3)):02d}"
+            line = line[: date_match.start()]
+        line = re.sub(r"^\s*[1-9-]\s+1\.甲 2\.乙 3\.丙\s*", "", line)
+        cols = split_layout_columns(line)
+        if len(cols) >= 1:
+            result["质控医师"] = clean_value(cols[-2] if len(cols) >= 2 else cols[0])
+        if len(cols) >= 2:
+            result["质控护士"] = clean_value(cols[-1])
+
+    return result
+
+
+def split_layout_columns(line: str) -> list[str]:
+    return [clean_value(part) for part in re.split(r"\s{2,}", line.strip()) if clean_value(part)]
+
+
+def split_layout_columns_keep_positions(line: str) -> list[str]:
+    return [clean_value(normalize_spaces(part)) for part in re.split(r"\s{2,}", line.strip()) if part.strip()]
+
+
+def parse_discharge_followup(lines: list[str]) -> dict[str, str]:
+    result = {
+        "离院方式代码": "",
+        "拟接收医疗机构名称": "",
+        "出院31天内再住院计划代码": "",
+        "再住院计划目的": "",
+        "入院前昏迷天数": "",
+        "入院前昏迷小时": "",
+        "入院前昏迷分钟": "",
+        "入院后昏迷天数": "",
+        "入院后昏迷小时": "",
+        "入院后昏迷分钟": "",
+    }
+    for line in lines:
+        if "1.医嘱离院" in line:
+            result["离院方式代码"] = first_match(r"^\s*([1-9-])\s+1\.医嘱离院", line)
+            result["拟接收医疗机构名称"] = first_match(r"拟接收医疗机构名称:\s*(.*?)\s*$", line)
+        elif "是否有出院31天内再住院计划" in line:
+            result["出院31天内再住院计划代码"] = first_match(r"是否有出院31天内再住院计划\s*([1-9-])", line)
+            result["再住院计划目的"] = first_match(r"目的:\s*(.*?)\s*$", line)
+        elif "入院前" in line and "入院后" in line and "分钟" in line:
+            match = re.search(
+                r"入院前\s*([-\d]+)\s*天\s*([-\d]+)\s*小时\s*([-\d]+)\s*分钟\s*"
+                r"入院后\s*([-\d]+)\s*天\s*([-\d]+)\s*小时\s*([-\d]+)\s*分钟",
+                line,
+            )
+            if match:
+                result["入院前昏迷天数"] = clean_int_value(match.group(1))
+                result["入院前昏迷小时"] = clean_int_value(match.group(2))
+                result["入院前昏迷分钟"] = clean_int_value(match.group(3))
+                result["入院后昏迷天数"] = clean_int_value(match.group(4))
+                result["入院后昏迷小时"] = clean_int_value(match.group(5))
+                result["入院后昏迷分钟"] = clean_int_value(match.group(6))
+    return result
+
+
+def build_quality_warnings(record: dict[str, Any], text: str) -> list[str]:
+    warnings: list[str] = []
+    required = ["病案号", "姓名", "性别", "入院时间", "出院时间", "主要诊断"]
+    for key in required:
+        value = record.get(key)
+        if isinstance(value, dict):
+            if not value:
+                warnings.append(f"缺少{key}")
+        elif not value:
+            warnings.append(f"缺少{key}")
+    if len(text.strip()) < 200:
+        warnings.append("提取文本过短，可能是扫描件或加密PDF")
+    return warnings
+
+
+def validate_record(record: dict[str, Any], department_rules_loaded: bool = False) -> list[str]:
+    notes: list[str] = []
+    if not clean_value(str(record.get("住院号", ""))):
+        notes.append("患者号/住院号为空")
+    if not re.fullmatch(r"\d{10}", str(record.get("病案号", ""))):
+        notes.append("病案号不是10位数字")
+    if record.get("性别") and record["性别"] not in {"男", "女"}:
+        notes.append("性别不是男/女")
+    if record.get("身份证号") and not re.fullmatch(r"\d{17}[\dX]", record["身份证号"]):
+        notes.append("身份证号格式异常")
+    if record.get("出生日期") and not re.fullmatch(r"\d{4}-\d{2}-\d{2}", record["出生日期"]):
+        notes.append("出生日期格式异常")
+    if record.get("入院时间") and record.get("出院时间") and record["入院时间"] > record["出院时间"]:
+        notes.append("出院时间早于入院时间")
+    if record.get("实际住院天数") and not re.fullmatch(r"\d+", str(record["实际住院天数"])):
+        notes.append("实际住院天数不是整数")
+
+    main = record.get("主要诊断") if isinstance(record.get("主要诊断"), dict) else {}
+    if main and not re.match(r"^[A-Z]\d{2}", main.get("疾病编码", "")):
+        notes.append("主要诊断编码格式异常")
+    other_diagnoses = [diagnosis for diagnosis in record.get("出院诊断", []) if diagnosis.get("诊断类别") == "其他诊断"]
+    for index, diagnosis in enumerate(other_diagnoses, start=1):
+        if not re.match(r"^[A-Z]\d{2}", diagnosis.get("疾病编码", "")):
+            notes.append(f"其他诊断{index}编码格式异常")
+    suspicious_phrases = ["1.急诊", "2.门诊", "其他医疗机构转入", "1.医嘱离院"]
+    for key in ["联系人姓名", "联系人关系", "联系人地址", "联系人电话", "现住址", "户口地址", "工作单位及地址"]:
+        value = str(record.get(key, ""))
+        if any(phrase in value for phrase in suspicious_phrases):
+            notes.append(f"{key}疑似串入版式选项文本")
+    for index, operation in enumerate(record.get("手术操作", []), start=1):
+        if not operation.get("手术操作日期") or not operation.get("手术操作名称"):
+            notes.append(f"手术操作{index}缺少日期或名称")
+        if not operation.get("手术操作编码"):
+            notes.append(f"手术操作{index}缺少手术及操作编码")
+
+    for key in ["总费用", "自付金额"]:
+        if record.get(key) and not re.fullmatch(r"\d+(\.\d{1,2})?", str(record[key])):
+            notes.append(f"{key}不是金额格式")
+    if department_rules_loaded and (record.get("出院科别") or record.get("入院科别")) and not record.get("大科室"):
+        notes.append("科别未匹配到大科室分类")
+    return notes
+
+
+def apply_review_fields(record: dict[str, Any], validation_notes: list[str]) -> None:
+    auto_corrections = [
+        note for note in record.get("自动修正", [])
+        if not ("病案号由PDF值" in note and "按文件名补齐" in note)
+    ]
+    quality_notes = record.get("质控提示", [])
+    review_notes = list(dict.fromkeys([*quality_notes, *validation_notes, *record.get("复核备注", [])]))
+    record["自动修正"] = auto_corrections
+    record["复核备注"] = review_notes
+    record["人工修正"] = False
+    record["manual_corrected"] = False
+    if review_notes:
+        record["复核状态"] = "needs_review"
+    elif auto_corrections:
+        record["复核状态"] = "auto_corrected"
+    else:
+        record["复核状态"] = "auto_pass"
+
+
+def parse_record(
+    pdf_path: Path,
+    department_rules: dict[str, dict[str, str]] | None = None,
+    args: argparse.Namespace | None = None,
+) -> tuple[dict[str, Any], str]:
+    if args is None:
+        args = build_parser().parse_args([])
+    text, text_method, mineru_markdown_dir = extract_record_text(pdf_path, args)
+    lines = meaningful_lines(text)
+
+    diagnoses = parse_diagnoses(lines)
+    fees = parse_fees(text)
+    header = parse_header(text, lines)
+    normalized_no, original_no, corrections, no_notes = normalize_medical_record_no(header.get("病案号", ""), pdf_path)
+    header["病案号"] = normalized_no
+    header["首页病案号"] = normalize_digits(original_no or normalized_no, 10)
+    inpatient_no, inpatient_corrections, inpatient_notes = build_inpatient_no(
+        header.get("住院次数", ""),
+        header.get("首页病案号", ""),
+        header.get("病案号", ""),
+        pdf_path,
+    )
+    record: dict[str, Any] = {
+        "源文件": pdf_path.name,
+        "住院号": inpatient_no,
+        "解析时间": datetime.now().isoformat(timespec="seconds"),
+        "解析器版本": "patient-front-page-local-v1",
+        "文本抽取方式": text_method,
+        "Mineru Markdown目录": mineru_markdown_dir,
+        "自动修正": [*corrections, *inpatient_corrections],
+        "复核备注": [*no_notes, *inpatient_notes],
+    }
+    record.update(header)
+    record.update(parse_newborn_info(lines))
+    record.update(parse_basic_lines(lines))
+    record.update(parse_admission_discharge(lines))
+    department_rules = department_rules or {}
+    major_department, standard_department = classify_major_department(record, department_rules)
+    record["大科室"] = major_department
+    record["标准子科室"] = standard_department
+    record.update(diagnoses)
+    record["手术操作"] = parse_operations(lines)
+    record.update(parse_front_page_middle(lines))
+    record.update(parse_discharge_followup(lines))
+    record.update(fees)
+    record["原始文本"] = text
+
+    main = diagnoses.get("主要诊断") or {}
+    if main:
+        record["主要诊断名称"] = main.get("出院诊断", "")
+        record["主要诊断编码"] = main.get("疾病编码", "")
+        record["主要诊断入院病情"] = main.get("入院病情", "")
+
+    warnings = build_quality_warnings(record, text)
+    validation_notes = validate_record(record, department_rules_loaded=bool(department_rules))
+    record["质控状态"] = "需复核" if warnings else "通过"
+    record["质控提示"] = warnings
+    apply_review_fields(record, validation_notes)
+    return record, text
+
+
+def flatten_for_csv(record: dict[str, Any]) -> dict[str, str]:
+    main = record.get("主要诊断") if isinstance(record.get("主要诊断"), dict) else {}
+    return {
+        "住院号": record.get("住院号", ""),
+        "源文件": record.get("源文件", ""),
+        "病案号": record.get("病案号", ""),
+        "首页病案号": record.get("首页病案号", ""),
+        "姓名": record.get("姓名", ""),
+        "性别": record.get("性别", ""),
+        "出生日期": record.get("出生日期", ""),
+        "年龄": record.get("年龄", ""),
+        "身份证号": record.get("身份证号", ""),
+        "新生儿年龄（月）": record.get("新生儿年龄（月）", ""),
+        "新生儿出生体重（克）": record.get("新生儿出生体重（克）", ""),
+        "新生儿入院体重（克）": record.get("新生儿入院体重（克）", ""),
+        "医疗机构": record.get("医疗机构", ""),
+        "组织机构代码": record.get("组织机构代码", ""),
+        "医疗付费方式": record.get("医疗付费方式", ""),
+        "住院次数": record.get("住院次数", ""),
+        "入院时间": record.get("入院时间", ""),
+        "入院科别": record.get("入院科别", ""),
+        "入院病房": record.get("入院病房", ""),
+        "转科科别": record.get("转科科别", ""),
+        "转科时间": record.get("转科时间", ""),
+        "出院时间": record.get("出院时间", ""),
+        "出院科别": record.get("出院科别", ""),
+        "出院病房": record.get("出院病房", ""),
+        "大科室": record.get("大科室", ""),
+        "实际住院天数": record.get("实际住院天数", ""),
+        "门急诊诊断": record.get("门急诊诊断", ""),
+        "门急诊诊断编码": record.get("门急诊诊断编码", ""),
+        "主要诊断": main.get("出院诊断", ""),
+        "主要诊断编码": main.get("疾病编码", ""),
+        "主要诊断入院病情": main.get("入院病情", ""),
+        "出院诊断": json.dumps(record.get("出院诊断", []), ensure_ascii=False),
+        "手术操作": json.dumps(record.get("手术操作", []), ensure_ascii=False),
+        "病理诊断": record.get("病理诊断", ""),
+        "病理诊断编码": record.get("病理诊断编码", ""),
+        "病理号": record.get("病理号", ""),
+        "药物过敏代码": record.get("药物过敏代码", ""),
+        "过敏药物": record.get("过敏药物", ""),
+        "死亡患者尸检代码": record.get("死亡患者尸检代码", ""),
+        "血型代码": record.get("血型代码", ""),
+        "Rh代码": record.get("Rh代码", ""),
+        "科主任": record.get("科主任", ""),
+        "主任副主任医师": record.get("主任副主任医师", ""),
+        "主治医师": record.get("主治医师", ""),
+        "住院医师": record.get("住院医师", ""),
+        "责任护士": record.get("责任护士", ""),
+        "进修医师": record.get("进修医师", ""),
+        "实习医师": record.get("实习医师", ""),
+        "规培医师": record.get("规培医师", ""),
+        "编码员": record.get("编码员", ""),
+        "病案质量代码": record.get("病案质量代码", ""),
+        "质控医师": record.get("质控医师", ""),
+        "质控护士": record.get("质控护士", ""),
+        "质控日期": record.get("质控日期", ""),
+        "离院方式代码": record.get("离院方式代码", ""),
+        "出院31天内再住院计划代码": record.get("出院31天内再住院计划代码", ""),
+        "总费用": record.get("总费用", ""),
+        "自付金额": record.get("自付金额", ""),
+        "质控状态": record.get("质控状态", ""),
+        "质控提示": "；".join(record.get("质控提示", [])),
+        "复核状态": record.get("复核状态", ""),
+        "复核备注": "；".join(record.get("复核备注", [])),
+        "文本抽取方式": record.get("文本抽取方式", ""),
+        "自动修正": "；".join(record.get("自动修正", [])),
+        "人工修正": str(record.get("人工修正", False)),
+    }
+
+
+def write_outputs(records: list[dict[str, Any]], output_dir: Path, save_text: bool) -> None:
+    output_dir.mkdir(parents=True, exist_ok=True)
+    structured_dir = output_dir / "01_结构化结果"
+    per_record_dir = output_dir / "02_单份JSON"
+    text_dir = output_dir / "03_提取文本"
+    review_dir = output_dir / "04_复核与人工校正"
+    structured_dir.mkdir(exist_ok=True)
+    per_record_dir.mkdir(exist_ok=True)
+    review_dir.mkdir(exist_ok=True)
+    if save_text:
+        text_dir.mkdir(exist_ok=True)
+
+    jsonl_path = structured_dir / "患者首页_结构化结果.jsonl"
+    csv_path = structured_dir / "患者首页_结构化结果.csv"
+    review_path = review_dir / "患者首页_复核清单.csv"
+
+    with jsonl_path.open("w", encoding="utf-8") as fp:
+        for record in records:
+            fp.write(json.dumps(record, ensure_ascii=False) + "\n")
+
+    with csv_path.open("w", encoding="utf-8-sig", newline="") as fp:
+        writer = csv.DictWriter(fp, fieldnames=CSV_COLUMNS)
+        writer.writeheader()
+        for record in records:
+            writer.writerow(flatten_for_csv(record))
+
+    with review_path.open("w", encoding="utf-8-sig", newline="") as fp:
+        writer = csv.DictWriter(fp, fieldnames=["住院号", "源文件", "病案号", "首页病案号", "姓名", "复核状态", "复核备注", "自动修正", "人工修正"])
+        writer.writeheader()
+        for record in records:
+            if record.get("复核状态") != "auto_pass":
+                writer.writerow(
+                    {
+                        "住院号": record.get("住院号", ""),
+                        "源文件": record.get("源文件", ""),
+                        "病案号": record.get("病案号", ""),
+                        "首页病案号": record.get("首页病案号", ""),
+                        "姓名": record.get("姓名", ""),
+                        "复核状态": record.get("复核状态", ""),
+                        "复核备注": "；".join(record.get("复核备注", [])),
+                        "自动修正": "；".join(record.get("自动修正", [])),
+                        "人工修正": str(record.get("人工修正", False)),
+                    }
+                )
+
+    for record in records:
+        stem = Path(record["源文件"]).stem
+        json_path = per_record_dir / f"{stem}.json"
+        with json_path.open("w", encoding="utf-8") as fp:
+            json.dump(record, fp, ensure_ascii=False, indent=2)
+        if save_text:
+            (text_dir / f"{stem}.txt").write_text(record.get("原始文本", ""), encoding="utf-8")
+
+
+def quote_pg_identifier(identifier: str) -> str:
+    if not re.fullmatch(r"[A-Za-z_][A-Za-z0-9_]*", identifier):
+        raise ValueError(f"非法 PostgreSQL 表名：{identifier}")
+    return '"' + identifier.replace('"', '""') + '"'
+
+
+def pg_literal(value: str) -> str:
+    return "'" + value.replace("'", "''") + "'"
+
+
+def int_or_empty(value: Any) -> str:
+    value = clean_value(str(value)) if value is not None else ""
+    return value if re.fullmatch(r"\d+", value) else ""
+
+
+def decimal_or_empty(value: Any) -> str:
+    value = clean_value(str(value)) if value is not None else ""
+    return value if re.fullmatch(r"\d+(\.\d{1,2})?", value) else ""
+
+
+def pg_json(value: Any) -> str:
+    return json.dumps(value if value is not None else [], ensure_ascii=False)
+
+
+def record_to_pg_row(record: dict[str, Any]) -> dict[str, str]:
+    main = record.get("主要诊断") if isinstance(record.get("主要诊断"), dict) else {}
+    return {
+        "source_file": record.get("源文件", ""),
+        "inpatient_no": record.get("住院号", ""),
+        "medical_record_no": record.get("病案号", ""),
+        "front_page_medical_record_no": record.get("首页病案号", ""),
+        "patient_name": record.get("姓名", ""),
+        "gender": record.get("性别", ""),
+        "birth_date": record.get("出生日期", ""),
+        "age": record.get("年龄", ""),
+        "nationality": record.get("国籍", ""),
+        "id_card_no": record.get("身份证号", ""),
+        "neonatal_age_months": int_or_empty(record.get("新生儿年龄（月）", "")),
+        "newborn_birth_weight_g": int_or_empty(record.get("新生儿出生体重（克）", "")),
+        "newborn_admission_weight_g": int_or_empty(record.get("新生儿入院体重（克）", "")),
+        "hospital_name": record.get("医疗机构", ""),
+        "organization_code": record.get("组织机构代码", ""),
+        "payment_method": record.get("医疗付费方式", ""),
+        "health_card_no": record.get("健康卡号", ""),
+        "admission_count": int_or_empty(record.get("住院次数", "")),
+        "birthplace": record.get("出生地", ""),
+        "native_place": record.get("籍贯", ""),
+        "ethnicity": record.get("民族", ""),
+        "occupation": record.get("职业", ""),
+        "marital_status_code": record.get("婚姻代码", ""),
+        "current_address": record.get("现住址", ""),
+        "current_address_phone": record.get("现住址电话", ""),
+        "current_address_postcode": record.get("现住址邮编", ""),
+        "household_address": record.get("户口地址", ""),
+        "household_postcode": record.get("户口地址邮编", ""),
+        "employer_address": record.get("工作单位及地址", ""),
+        "employer_phone": record.get("单位电话", ""),
+        "employer_postcode": record.get("单位邮编", ""),
+        "contact_name": record.get("联系人姓名", ""),
+        "contact_relationship": record.get("联系人关系", ""),
+        "contact_address": record.get("联系人地址", ""),
+        "contact_phone": record.get("联系人电话", ""),
+        "admission_path_code": record.get("入院途径代码", ""),
+        "admission_time": record.get("入院时间", ""),
+        "admission_dept": record.get("入院科别", ""),
+        "admission_ward": record.get("入院病房", ""),
+        "transfer_dept": record.get("转科科别", ""),
+        "transfer_time": record.get("转科时间", ""),
+        "discharge_time": record.get("出院时间", ""),
+        "discharge_dept": record.get("出院科别", ""),
+        "discharge_ward": record.get("出院病房", ""),
+        "major_department": record.get("大科室", ""),
+        "hospital_days": int_or_empty(record.get("实际住院天数", "")),
+        "outpatient_diagnosis": record.get("门急诊诊断", ""),
+        "outpatient_diagnosis_code": record.get("门急诊诊断编码", ""),
+        "primary_diagnosis": main.get("出院诊断", ""),
+        "primary_diagnosis_code": main.get("疾病编码", ""),
+        "primary_admission_condition": main.get("入院病情", ""),
+        "discharge_diagnoses": pg_json(record.get("出院诊断", [])),
+        "operations": pg_json(record.get("手术操作", [])),
+        "injury_poisoning_external_cause": record.get("损伤中毒外部原因", ""),
+        "injury_poisoning_code": record.get("损伤中毒疾病编码", ""),
+        "pathology_diagnosis": record.get("病理诊断", ""),
+        "pathology_diagnosis_code": record.get("病理诊断编码", ""),
+        "pathology_no": record.get("病理号", ""),
+        "drug_allergy_code": record.get("药物过敏代码", ""),
+        "allergy_drug": record.get("过敏药物", ""),
+        "autopsy_code": record.get("死亡患者尸检代码", ""),
+        "blood_type_code": record.get("血型代码", ""),
+        "rh_code": record.get("Rh代码", ""),
+        "department_director": record.get("科主任", ""),
+        "chief_physician": record.get("主任副主任医师", ""),
+        "attending_physician": record.get("主治医师", ""),
+        "resident_physician": record.get("住院医师", ""),
+        "responsible_nurse": record.get("责任护士", ""),
+        "refresher_physician": record.get("进修医师", ""),
+        "intern_physician": record.get("实习医师", ""),
+        "standardized_resident_physician": record.get("规培医师", ""),
+        "coder": record.get("编码员", ""),
+        "record_quality_code": record.get("病案质量代码", ""),
+        "quality_control_physician": record.get("质控医师", ""),
+        "quality_control_nurse": record.get("质控护士", ""),
+        "quality_control_date": record.get("质控日期", ""),
+        "discharge_disposition_code": record.get("离院方式代码", ""),
+        "receiving_org_name": record.get("拟接收医疗机构名称", ""),
+        "readmission_plan_code": record.get("出院31天内再住院计划代码", ""),
+        "readmission_plan_purpose": record.get("再住院计划目的", ""),
+        "coma_before_days": int_or_empty(record.get("入院前昏迷天数", "")),
+        "coma_before_hours": int_or_empty(record.get("入院前昏迷小时", "")),
+        "coma_before_minutes": int_or_empty(record.get("入院前昏迷分钟", "")),
+        "coma_after_days": int_or_empty(record.get("入院后昏迷天数", "")),
+        "coma_after_hours": int_or_empty(record.get("入院后昏迷小时", "")),
+        "coma_after_minutes": int_or_empty(record.get("入院后昏迷分钟", "")),
+        "total_cost": decimal_or_empty(record.get("总费用", "")),
+        "self_pay_amount": decimal_or_empty(record.get("自付金额", "")),
+        "fee_details": pg_json(record.get("费用明细", {})),
+        "quality_status": record.get("质控状态", ""),
+        "quality_notes": pg_json(record.get("质控提示", [])),
+        "review_status": record.get("复核状态", "pending"),
+        "review_notes": pg_json(record.get("复核备注", [])),
+        "manual_corrected": "true" if record.get("manual_corrected") or record.get("人工修正") else "false",
+        "auto_corrections": pg_json(record.get("自动修正", [])),
+        "text_extraction_method": record.get("文本抽取方式", ""),
+        "mineru_markdown_dir": record.get("Mineru Markdown目录", ""),
+        "raw_text": record.get("原始文本", ""),
+    }
+
+
+def pg_copy_cast(column_name: str, column_type: str) -> str:
+    if column_name == "inpatient_no":
+        return "NULLIF(BTRIM(inpatient_no), '')"
+    if column_type in {"INTEGER"}:
+        return f"NULLIF({column_name}, '')::{column_type}"
+    if column_type.startswith("NUMERIC"):
+        return f"NULLIF({column_name}, '')::{column_type}"
+    if column_type in {"DATE", "TIMESTAMP"}:
+        return f"NULLIF({column_name}, '')::{column_type}"
+    if column_type.startswith("JSONB"):
+        return f"COALESCE(NULLIF({column_name}, ''), 'null')::jsonb"
+    if column_type.startswith("BOOLEAN"):
+        return f"COALESCE(NULLIF({column_name}, '')::boolean, false)"
+    return f"NULLIF({column_name}, '')"
+
+
+def write_postgres(records: list[dict[str, Any]], args: argparse.Namespace) -> None:
+    if not records:
+        return
+    if shutil.which("psql") is None:
+        raise RuntimeError("未找到 psql。请先安装 PostgreSQL 客户端，或取消 --write-postgres。")
+
+    table_name = quote_pg_identifier(args.pg_table)
+    list_table_name = quote_pg_identifier("Patient_Lists")
+    trigger_function_name = quote_pg_identifier(f"{args.pg_table}_sync_patient_lists_trigger_fn")
+    trigger_name = quote_pg_identifier(f"trg_{args.pg_table}_sync_patient_lists")
+    dedupe_trigger_function_name = quote_pg_identifier(f"{args.pg_table}_dedupe_inpatient_no_trigger_fn")
+    dedupe_trigger_name = quote_pg_identifier(f"trg_{args.pg_table}_dedupe_inpatient_no")
+    source_file_constraint = quote_pg_identifier(f"{args.pg_table}_source_file_key")
+    inpatient_no_constraint = quote_pg_identifier(f"{args.pg_table}_inpatient_no_key")
+    inpatient_no_check_constraints = [
+        quote_pg_identifier(f"ck_{args.pg_table}_inpatient_no_format"),
+        quote_pg_identifier(f"ck_{args.pg_table.lower()}_inpatient_no_format"),
+        quote_pg_identifier(f"ck_{args.pg_table}_inpatient_no_required"),
+    ]
+    inpatient_no_required_constraint = quote_pg_identifier(f"ck_{args.pg_table}_inpatient_no_required")
+    list_inpatient_no_check_constraints = [
+        quote_pg_identifier("ck_Patient_Lists_inpatient_no_format"),
+        quote_pg_identifier("ck_patient_lists_inpatient_no_format"),
+        quote_pg_identifier("ck_patient_lists_inpatient_no_required"),
+    ]
+    list_inpatient_no_required_constraint = quote_pg_identifier("ck_patient_lists_inpatient_no_required")
+    connection_env = os.environ.copy()
+    connection_env.update(
+        {
+            "PGHOST": args.pg_host or os.environ.get("PGHOST", ""),
+            "PGPORT": str(args.pg_port or os.environ.get("PGPORT", "")),
+            "PGDATABASE": args.pg_database or os.environ.get("PGDATABASE", ""),
+            "PGUSER": args.pg_user or os.environ.get("PGUSER", ""),
+            "PGPASSWORD": args.pg_password or os.environ.get("PGPASSWORD", ""),
+        }
+    )
+    missing = [name for name in ["PGHOST", "PGPORT", "PGDATABASE", "PGUSER", "PGPASSWORD"] if not connection_env.get(name)]
+    if missing:
+        raise RuntimeError("缺少 PostgreSQL 连接配置：" + "、".join(missing))
+
+    with tempfile.TemporaryDirectory(prefix="patient_front_pages_") as tmpdir:
+        csv_path = Path(tmpdir) / "records.csv"
+        sql_path = Path(tmpdir) / "import.sql"
+        with csv_path.open("w", encoding="utf-8", newline="") as fp:
+            fieldnames = [name for name, _type in PG_COLUMNS]
+            writer = csv.DictWriter(fp, fieldnames=fieldnames)
+            writer.writeheader()
+            for record in records:
+                writer.writerow(record_to_pg_row(record))
+
+        escaped_csv = str(csv_path).replace("'", "''")
+        column_defs = ",\n    ".join(
+            ["id BIGSERIAL PRIMARY KEY", *[f"{name} {column_type}" for name, column_type in PG_COLUMNS], "UNIQUE (inpatient_no)"]
+        )
+        alter_add_columns = "\n".join(
+            [f"ALTER TABLE {table_name} ADD COLUMN IF NOT EXISTS {name} {column_type};" for name, column_type in PG_COLUMNS]
+        )
+        table_comments = "\n".join(
+            [
+                f"COMMENT ON COLUMN {table_name}.{quote_pg_identifier(column_name)} IS {pg_literal(comment)};"
+                for column_name, comment in PG_COLUMN_COMMENTS.items()
+            ]
+        )
+        import_column_defs = ",\n    ".join([f"{name} TEXT" for name, _type in PG_COLUMNS])
+        import_columns = ", ".join([name for name, _type in PG_COLUMNS])
+        select_values = ",\n    ".join([pg_copy_cast(name, column_type) for name, column_type in PG_COLUMNS])
+        update_values = ",\n    ".join([f"{name} = EXCLUDED.{name}" for name, _type in PG_COLUMNS if name != "inpatient_no"])
+        dedupe_update_values = ",\n        ".join([f"{name} = NEW.{name}" for name, _type in PG_COLUMNS if name != "inpatient_no"])
+        drop_inpatient_no_checks = "\n".join(
+            [f"ALTER TABLE {table_name} DROP CONSTRAINT IF EXISTS {constraint};" for constraint in inpatient_no_check_constraints]
+        )
+        drop_list_inpatient_no_checks = "\n".join(
+            [
+                f"ALTER TABLE {list_table_name} DROP CONSTRAINT IF EXISTS {constraint};"
+                for constraint in list_inpatient_no_check_constraints
+            ]
+        )
+        sql_path.write_text(
+            f"""
+\\set ON_ERROR_STOP on
+CREATE TABLE IF NOT EXISTS {table_name} (
+    {column_defs}
+);
+
+ALTER TABLE {table_name} DROP COLUMN IF EXISTS payload;
+ALTER TABLE {table_name} DROP COLUMN IF EXISTS parsed_at;
+ALTER TABLE {table_name} DROP COLUMN IF EXISTS created_at;
+ALTER TABLE {table_name} DROP COLUMN IF EXISTS updated_at;
+DO $$
+BEGIN
+    IF EXISTS (
+        SELECT 1 FROM information_schema.columns
+        WHERE table_schema = 'public' AND table_name = {args.pg_table!r} AND column_name = 'original_medical_record_no'
+    ) AND NOT EXISTS (
+        SELECT 1 FROM information_schema.columns
+        WHERE table_schema = 'public' AND table_name = {args.pg_table!r} AND column_name = 'front_page_medical_record_no'
+    ) THEN
+        ALTER TABLE {table_name} RENAME COLUMN original_medical_record_no TO front_page_medical_record_no;
+    END IF;
+
+    IF EXISTS (
+        SELECT 1 FROM information_schema.columns
+        WHERE table_schema = 'public' AND table_name = {args.pg_table!r} AND column_name = 'diagnoses'
+    ) AND NOT EXISTS (
+        SELECT 1 FROM information_schema.columns
+        WHERE table_schema = 'public' AND table_name = {args.pg_table!r} AND column_name = 'discharge_diagnoses'
+    ) THEN
+        ALTER TABLE {table_name} RENAME COLUMN diagnoses TO discharge_diagnoses;
+    END IF;
+END $$;
+ALTER TABLE {table_name} DROP COLUMN IF EXISTS other_diagnoses;
+{alter_add_columns}
+UPDATE {table_name}
+SET front_page_medical_record_no = RIGHT(LPAD(regexp_replace(front_page_medical_record_no, '\\D', '', 'g'), 10, '0'), 10)
+WHERE front_page_medical_record_no IS NOT NULL
+  AND front_page_medical_record_no <> ''
+  AND front_page_medical_record_no !~ '^\\d{{10}}$'
+  AND regexp_replace(front_page_medical_record_no, '\\D', '', 'g') <> '';
+UPDATE {table_name}
+SET inpatient_no =
+    'ZY'
+    || COALESCE(
+        LPAD(admission_count::text, 2, '0'),
+        substring(source_file from '^ZY([0-9]{{2}})[0-9]{{10}}')
+    )
+    || RIGHT(
+        LPAD(
+            COALESCE(
+                NULLIF(regexp_replace(COALESCE(front_page_medical_record_no, ''), '\\D', '', 'g'), ''),
+                NULLIF(regexp_replace(COALESCE(medical_record_no, ''), '\\D', '', 'g'), ''),
+                substring(source_file from '^ZY[0-9]{{2}}([0-9]{{10}})')
+            ),
+            10,
+            '0'
+        ),
+        10
+    )
+WHERE (inpatient_no IS NULL OR BTRIM(inpatient_no) = '')
+  AND COALESCE(
+        LPAD(admission_count::text, 2, '0'),
+        substring(source_file from '^ZY([0-9]{{2}})[0-9]{{10}}')
+      ) IS NOT NULL
+  AND COALESCE(
+        NULLIF(regexp_replace(COALESCE(front_page_medical_record_no, ''), '\\D', '', 'g'), ''),
+        NULLIF(regexp_replace(COALESCE(medical_record_no, ''), '\\D', '', 'g'), ''),
+        substring(source_file from '^ZY[0-9]{{2}}([0-9]{{10}})')
+      ) IS NOT NULL;
+ALTER TABLE {table_name} DROP CONSTRAINT IF EXISTS {source_file_constraint};
+{drop_inpatient_no_checks}
+DELETE FROM {table_name}
+WHERE NULLIF(BTRIM(inpatient_no), '') IS NULL;
+UPDATE {table_name}
+SET inpatient_no = BTRIM(inpatient_no)
+WHERE inpatient_no <> BTRIM(inpatient_no);
+WITH ranked AS (
+    SELECT
+        id,
+        ROW_NUMBER() OVER (PARTITION BY BTRIM(inpatient_no) ORDER BY id DESC) AS duplicate_rank
+    FROM {table_name}
+    WHERE NULLIF(BTRIM(inpatient_no), '') IS NOT NULL
+)
+DELETE FROM {table_name} p
+USING ranked
+WHERE p.id = ranked.id
+  AND ranked.duplicate_rank > 1;
+ALTER TABLE {table_name} ALTER COLUMN inpatient_no SET NOT NULL;
+ALTER TABLE {table_name} DROP CONSTRAINT IF EXISTS {inpatient_no_constraint};
+ALTER TABLE {table_name} ADD CONSTRAINT {inpatient_no_constraint} UNIQUE (inpatient_no);
+DO $$
+BEGIN
+    IF NOT EXISTS (SELECT 1 FROM pg_constraint WHERE conname = {pg_literal(f"ck_{args.pg_table}_inpatient_no_required")}) THEN
+        ALTER TABLE {table_name} ADD CONSTRAINT {inpatient_no_required_constraint} CHECK (NULLIF(BTRIM(inpatient_no), '') IS NOT NULL);
+    END IF;
+END $$;
+COMMENT ON TABLE {table_name} IS '患者住院病案首页结构化宽表，由PDF首页解析程序生成并支持人工复核。';
+{table_comments}
+
+CREATE TABLE IF NOT EXISTS {list_table_name} (
+    record_id BIGSERIAL PRIMARY KEY,
+    batch_name TEXT NOT NULL DEFAULT 'Patient_FrontPages',
+    major_department TEXT NOT NULL DEFAULT '',
+    sub_department TEXT NOT NULL DEFAULT '',
+    source_folder TEXT NOT NULL DEFAULT 'Patient_FrontPages',
+    image_path TEXT NOT NULL DEFAULT '',
+    image_name TEXT NOT NULL DEFAULT '',
+    image_row_no INTEGER NOT NULL DEFAULT 0,
+    patient_name TEXT NOT NULL DEFAULT '',
+    gender TEXT,
+    age TEXT,
+    inpatient_no TEXT NOT NULL,
+    diagnosis TEXT,
+    admission_time TEXT,
+    last_write_time TEXT,
+    hospital_days INTEGER,
+    discharge_time TEXT,
+    postoperative_days TEXT,
+    review_status TEXT NOT NULL DEFAULT '首页自动关联',
+    review_notes TEXT,
+    manual_corrected BOOLEAN NOT NULL DEFAULT false,
+    imported_at TIMESTAMPTZ NOT NULL DEFAULT now()
+);
+ALTER TABLE {list_table_name} ADD COLUMN IF NOT EXISTS has_front_page BOOLEAN NOT NULL DEFAULT false;
+ALTER TABLE {list_table_name} ADD COLUMN IF NOT EXISTS front_page_id BIGINT;
+ALTER TABLE {list_table_name} ADD COLUMN IF NOT EXISTS front_page_source_file TEXT;
+{drop_list_inpatient_no_checks}
+DELETE FROM {list_table_name}
+WHERE NULLIF(BTRIM(inpatient_no), '') IS NULL;
+UPDATE {list_table_name}
+SET inpatient_no = BTRIM(inpatient_no)
+WHERE inpatient_no <> BTRIM(inpatient_no);
+WITH ranked AS (
+    SELECT
+        record_id,
+        ROW_NUMBER() OVER (PARTITION BY BTRIM(inpatient_no) ORDER BY record_id DESC) AS duplicate_rank
+    FROM {list_table_name}
+    WHERE NULLIF(BTRIM(inpatient_no), '') IS NOT NULL
+)
+DELETE FROM {list_table_name} pl
+USING ranked
+WHERE pl.record_id = ranked.record_id
+  AND ranked.duplicate_rank > 1;
+ALTER TABLE {list_table_name} ALTER COLUMN inpatient_no SET NOT NULL;
+DO $$
+BEGIN
+    IF NOT EXISTS (SELECT 1 FROM pg_constraint WHERE conname = 'ck_patient_lists_inpatient_no_required') THEN
+        ALTER TABLE {list_table_name} ADD CONSTRAINT {list_inpatient_no_required_constraint} CHECK (NULLIF(BTRIM(inpatient_no), '') IS NOT NULL);
+    END IF;
+END $$;
+COMMENT ON COLUMN {list_table_name}.has_front_page IS '是否有患者首页：由Patient_FrontPages按住院号自动联动。';
+COMMENT ON COLUMN {list_table_name}.front_page_id IS '关联的Patient_FrontPages.id。';
+COMMENT ON COLUMN {list_table_name}.front_page_source_file IS '关联患者首页PDF文件名。';
+CREATE UNIQUE INDEX IF NOT EXISTS uq_patient_lists_inpatient_no ON {list_table_name}(inpatient_no);
+
+CREATE OR REPLACE FUNCTION {dedupe_trigger_function_name}()
+RETURNS trigger
+LANGUAGE plpgsql
+AS $dedupe_trigger$
+DECLARE
+    existing_id BIGINT;
+BEGIN
+    NEW.inpatient_no := BTRIM(NEW.inpatient_no);
+    IF NULLIF(NEW.inpatient_no, '') IS NULL THEN
+        RETURN NEW;
+    END IF;
+
+    IF TG_OP = 'INSERT' THEN
+        SELECT id
+        INTO existing_id
+        FROM {table_name}
+        WHERE BTRIM(inpatient_no) = NEW.inpatient_no
+        ORDER BY id DESC
+        LIMIT 1;
+
+        IF existing_id IS NOT NULL THEN
+            DELETE FROM {table_name}
+            WHERE BTRIM(inpatient_no) = NEW.inpatient_no
+              AND id <> existing_id;
+
+            UPDATE {table_name}
+            SET {dedupe_update_values}
+            WHERE id = existing_id;
+
+            RETURN NULL;
+        END IF;
+    END IF;
+
+    DELETE FROM {table_name}
+    WHERE BTRIM(inpatient_no) = NEW.inpatient_no
+      AND id <> NEW.id;
+
+    RETURN NEW;
+END;
+$dedupe_trigger$;
+
+DROP TRIGGER IF EXISTS {dedupe_trigger_name} ON {table_name};
+
+CREATE OR REPLACE FUNCTION {trigger_function_name}()
+RETURNS trigger
+LANGUAGE plpgsql
+AS $trigger$
+BEGIN
+    IF TG_OP = 'DELETE' THEN
+        IF NULLIF(BTRIM(OLD.inpatient_no), '') IS NOT NULL THEN
+            UPDATE {list_table_name} AS pl
+            SET has_front_page = false,
+                front_page_id = NULL,
+                front_page_source_file = NULL,
+                imported_at = now()
+            WHERE pl.inpatient_no = BTRIM(OLD.inpatient_no)
+              AND NOT EXISTS (
+                  SELECT 1 FROM {table_name} fp
+                  WHERE BTRIM(fp.inpatient_no) = BTRIM(OLD.inpatient_no)
+              );
+        END IF;
+        RETURN OLD;
+    END IF;
+
+    IF TG_OP = 'UPDATE'
+       AND NULLIF(BTRIM(OLD.inpatient_no), '') IS NOT NULL
+       AND BTRIM(OLD.inpatient_no) IS DISTINCT FROM BTRIM(NEW.inpatient_no) THEN
+        UPDATE {list_table_name} AS pl
+        SET has_front_page = false,
+            front_page_id = NULL,
+            front_page_source_file = NULL,
+            imported_at = now()
+        WHERE pl.inpatient_no = BTRIM(OLD.inpatient_no)
+          AND NOT EXISTS (
+              SELECT 1 FROM {table_name} fp
+              WHERE BTRIM(fp.inpatient_no) = BTRIM(OLD.inpatient_no)
+          );
+    END IF;
+
+    IF NULLIF(BTRIM(NEW.inpatient_no), '') IS NOT NULL THEN
+        INSERT INTO {list_table_name} (
+            batch_name, major_department, sub_department, source_folder, image_path, image_name,
+            image_row_no, patient_name, gender, age, inpatient_no, diagnosis, admission_time,
+            hospital_days, discharge_time, review_status, review_notes, manual_corrected,
+            has_front_page, front_page_id, front_page_source_file, imported_at
+        )
+        VALUES (
+            'Patient_FrontPages',
+            COALESCE(NEW.major_department, ''),
+            COALESCE(NEW.discharge_dept, NEW.admission_dept, ''),
+            'Patient_FrontPages',
+            COALESCE(NEW.source_file, ''),
+            COALESCE(NEW.source_file, ''),
+            0,
+            COALESCE(NEW.patient_name, ''),
+            NEW.gender,
+            NEW.age,
+            BTRIM(NEW.inpatient_no),
+            NEW.primary_diagnosis,
+            to_char(NEW.admission_time, 'YYYY-MM-DD HH24:MI:SS'),
+            NEW.hospital_days,
+            to_char(NEW.discharge_time, 'YYYY-MM-DD HH24:MI:SS'),
+            '首页自动关联',
+            '由Patient_FrontPages触发器按住院号自动关联',
+            COALESCE(NEW.manual_corrected, false),
+            true,
+            NEW.id,
+            NEW.source_file,
+            now()
+        )
+        ON CONFLICT (inpatient_no) DO UPDATE SET
+            has_front_page = true,
+            front_page_id = EXCLUDED.front_page_id,
+            front_page_source_file = EXCLUDED.front_page_source_file,
+            patient_name = COALESCE(NULLIF(EXCLUDED.patient_name, ''), {list_table_name}.patient_name),
+            gender = EXCLUDED.gender,
+            age = EXCLUDED.age,
+            major_department = EXCLUDED.major_department,
+            sub_department = EXCLUDED.sub_department,
+            manual_corrected = EXCLUDED.manual_corrected,
+            imported_at = now();
+    END IF;
+
+    RETURN NEW;
+END;
+$trigger$;
+
+DROP TRIGGER IF EXISTS {trigger_name} ON {table_name};
+
+CREATE TEMP TABLE patient_front_pages_import (
+    {import_column_defs}
+);
+
+\\copy patient_front_pages_import({import_columns}) FROM '{escaped_csv}' WITH (FORMAT csv, HEADER true)
+
+DELETE FROM patient_front_pages_import
+WHERE NULLIF(BTRIM(inpatient_no), '') IS NULL;
+
+INSERT INTO {table_name} (
+    {import_columns}
+)
+SELECT
+    {select_values}
+FROM (
+    SELECT DISTINCT ON (BTRIM(inpatient_no)) *
+    FROM patient_front_pages_import
+    WHERE NULLIF(BTRIM(inpatient_no), '') IS NOT NULL
+    ORDER BY BTRIM(inpatient_no), ctid DESC
+) patient_front_pages_import
+ON CONFLICT (inpatient_no) DO UPDATE SET
+    {update_values};
+
+WITH front_pages AS (
+    SELECT DISTINCT ON (BTRIM(inpatient_no))
+        id,
+        BTRIM(inpatient_no) AS inpatient_no,
+        source_file,
+        COALESCE(patient_name, '') AS patient_name,
+        gender,
+        age,
+        COALESCE(major_department, '') AS major_department,
+        COALESCE(discharge_dept, admission_dept, '') AS sub_department,
+        primary_diagnosis,
+        admission_time,
+        discharge_time,
+        hospital_days,
+        manual_corrected
+    FROM {table_name}
+    WHERE NULLIF(BTRIM(inpatient_no), '') IS NOT NULL
+    ORDER BY BTRIM(inpatient_no), id DESC
+)
+INSERT INTO {list_table_name} (
+    batch_name, major_department, sub_department, source_folder, image_path, image_name,
+    image_row_no, patient_name, gender, age, inpatient_no, diagnosis, admission_time,
+    hospital_days, discharge_time, review_status, review_notes, manual_corrected,
+    has_front_page, front_page_id, front_page_source_file, imported_at
+)
+SELECT
+    'Patient_FrontPages',
+    major_department,
+    sub_department,
+    'Patient_FrontPages',
+    source_file,
+    source_file,
+    0,
+    patient_name,
+    gender,
+    age,
+    inpatient_no,
+    primary_diagnosis,
+    to_char(admission_time, 'YYYY-MM-DD HH24:MI:SS'),
+    hospital_days,
+    to_char(discharge_time, 'YYYY-MM-DD HH24:MI:SS'),
+    '首页自动关联',
+    '由Patient_FrontPages按住院号自动关联',
+    manual_corrected,
+    true,
+    id,
+    source_file,
+    now()
+FROM front_pages
+ON CONFLICT (inpatient_no) DO UPDATE SET
+    has_front_page = true,
+    front_page_id = EXCLUDED.front_page_id,
+    front_page_source_file = EXCLUDED.front_page_source_file,
+    patient_name = COALESCE(NULLIF(EXCLUDED.patient_name, ''), {list_table_name}.patient_name),
+    gender = EXCLUDED.gender,
+    age = EXCLUDED.age,
+    major_department = EXCLUDED.major_department,
+    sub_department = EXCLUDED.sub_department,
+    manual_corrected = EXCLUDED.manual_corrected,
+    imported_at = now();
+
+UPDATE {list_table_name} AS pl
+SET has_front_page = false,
+    front_page_id = NULL,
+    front_page_source_file = NULL,
+    imported_at = now()
+WHERE has_front_page IS TRUE
+  AND NOT EXISTS (
+      SELECT 1 FROM {table_name} fp
+      WHERE BTRIM(fp.inpatient_no) = pl.inpatient_no
+  );
+
+CREATE TRIGGER {dedupe_trigger_name}
+BEFORE INSERT OR UPDATE OF inpatient_no ON {table_name}
+FOR EACH ROW EXECUTE FUNCTION {dedupe_trigger_function_name}();
+
+CREATE TRIGGER {trigger_name}
+AFTER INSERT OR UPDATE OR DELETE ON {table_name}
+FOR EACH ROW EXECUTE FUNCTION {trigger_function_name}();
+""".strip()
+            + "\n",
+            encoding="utf-8",
+        )
+
+        completed = subprocess.run(
+            ["psql", "--no-password", "--file", str(sql_path)],
+            check=False,
+            stdout=subprocess.PIPE,
+            stderr=subprocess.PIPE,
+            text=True,
+            encoding="utf-8",
+            errors="replace",
+            env=connection_env,
+        )
+    if completed.returncode != 0:
+        raise RuntimeError(completed.stderr.strip() or completed.stdout.strip() or "PostgreSQL 写入失败")
+    print(f"PostgreSQL 写入完成：{len(records)} 条 -> {connection_env['PGHOST']}:{connection_env['PGPORT']}/{connection_env['PGDATABASE']}.{args.pg_table}")
+
+
+def process(input_dir: Path, output_dir: Path, save_text: bool, args: argparse.Namespace) -> int:
+    pdf_files = sorted(input_dir.glob("*.pdf"))
+    if not pdf_files:
+        print(f"未找到 PDF：{input_dir}", file=sys.stderr)
+        return 2
+
+    department_rules = load_department_rules(args.department_rules.resolve()) if args.department_rules else {}
+    if department_rules:
+        print(f"已加载科室分类规则：{args.department_rules}")
+
+    records: list[dict[str, Any]] = []
+    failures: list[dict[str, str]] = []
+    for index, pdf_path in enumerate(pdf_files, start=1):
+        print(f"[{index}/{len(pdf_files)}] 处理 {pdf_path.name}")
+        try:
+            record, _text = parse_record(pdf_path, department_rules=department_rules, args=args)
+            records.append(record)
+        except Exception as exc:  # noqa: BLE001 - 批处理需要继续处理下一份
+            failures.append({"源文件": pdf_path.name, "错误": str(exc)})
+
+    write_outputs(records, output_dir, save_text=save_text)
+    if args.write_postgres:
+        write_postgres(records, args)
+
+    if failures:
+        fail_path = output_dir / "04_复核与人工校正" / "患者首页_处理失败.csv"
+        with fail_path.open("w", encoding="utf-8-sig", newline="") as fp:
+            writer = csv.DictWriter(fp, fieldnames=["源文件", "错误"])
+            writer.writeheader()
+            writer.writerows(failures)
+
+    print(f"完成：成功 {len(records)} 份，失败 {len(failures)} 份。结果目录：{output_dir}")
+    return 1 if failures else 0
+
+
+def build_parser() -> argparse.ArgumentParser:
+    parser = argparse.ArgumentParser(description="批量解析患者住院病案首页 PDF，输出 JSONL/CSV/单份 JSON。")
+    parser.add_argument("-i", "--input-dir", type=Path, default=DEFAULT_INPUT_DIR, help="PDF 输入目录")
+    parser.add_argument("-o", "--output-dir", type=Path, default=DEFAULT_OUTPUT_DIR, help="结果输出目录")
+    parser.add_argument("--no-text", action="store_true", help="不额外保存 pdftotext 抽取出的 txt")
+    parser.add_argument("--write-postgres", action="store_true", help="同时写入 PostgreSQL 表")
+    parser.add_argument("--pg-host", default=os.environ.get("PGHOST", ""), help="PostgreSQL 主机；也可用 PGHOST")
+    parser.add_argument("--pg-port", default=os.environ.get("PGPORT", "5432"), help="PostgreSQL 端口；也可用 PGPORT")
+    parser.add_argument("--pg-database", default=os.environ.get("PGDATABASE", ""), help="PostgreSQL 数据库；也可用 PGDATABASE")
+    parser.add_argument("--pg-user", default=os.environ.get("PGUSER", ""), help="PostgreSQL 用户名；也可用 PGUSER")
+    parser.add_argument("--pg-password", default=os.environ.get("PGPASSWORD", ""), help="PostgreSQL 密码；也可用 PGPASSWORD")
+    parser.add_argument("--pg-table", default=os.environ.get("PG_PATIENT_TABLE", DEFAULT_PG_TABLE), help="PostgreSQL 表名")
+    parser.add_argument("--department-rules", type=Path, default=DEFAULT_DEPARTMENT_RULE_PATH, help="科室到大科室分类规则 JSON")
+    parser.add_argument(
+        "--text-source",
+        choices=["auto", "pdftotext", "mineru"],
+        default=os.environ.get("PATIENT_FRONT_TEXT_SOURCE", "auto"),
+        help="PDF文本抽取方式：auto先pdftotext异常时再Mineru，pdftotext只用本地转换，mineru强制PDF转Markdown",
+    )
+    parser.add_argument("--mineru-client", type=Path, default=DEFAULT_MINERU_CLIENT_PATH, help="Mineru PDF转Markdown客户端脚本路径")
+    parser.add_argument("--mineru-md-dir", type=Path, default=DEFAULT_MINERU_MD_DIR, help="Mineru Markdown输出目录")
+    parser.add_argument("--mineru-url", default=DEFAULT_MINERU_URL, help="Mineru API服务地址；也可用 MINERU_URL")
+    parser.add_argument("--mineru-sync", action="store_true", help="调用Mineru时开启同步清理")
+    return parser
+
+
+def main() -> int:
+    args = build_parser().parse_args()
+    return process(args.input_dir.resolve(), args.output_dir.resolve(), save_text=not args.no_text, args=args)
+
+
+if __name__ == "__main__":
+    raise SystemExit(main())
diff --git a/患者首页处理/数据处理工作区/03_人工复核/03_人工复核导出与回写.py b/患者首页处理/数据处理工作区/03_人工复核/03_人工复核导出与回写.py
new file mode 100755
index 0000000..0e4406a
--- /dev/null
+++ b/患者首页处理/数据处理工作区/03_人工复核/03_人工复核导出与回写.py
@@ -0,0 +1,224 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+"""
+从 PostgreSQL 导出患者首页复核表，并把人工复核结果回写到 Patient_FrontPages。
+
+依赖系统命令 psql；数据库密码通过 PGPASSWORD 环境变量或 --pg-password 传入。
+"""
+
+from __future__ import annotations
+
+import argparse
+import os
+import re
+import shutil
+import subprocess
+import sys
+import tempfile
+from pathlib import Path
+
+
+PROJECT_ROOT = Path(__file__).resolve().parents[2]
+DEFAULT_OUTPUT = PROJECT_ROOT / "数据处理结果区" / "04_复核与人工校正" / "患者首页_人工复核表.csv"
+DEFAULT_TABLE = "Patient_FrontPages"
+
+
+def quote_pg_identifier(identifier: str) -> str:
+    if not re.fullmatch(r"[A-Za-z_][A-Za-z0-9_]*", identifier):
+        raise ValueError(f"非法 PostgreSQL 表名：{identifier}")
+    return '"' + identifier.replace('"', '""') + '"'
+
+
+def connection_env(args: argparse.Namespace) -> dict[str, str]:
+    env = os.environ.copy()
+    env.update(
+        {
+            "PGHOST": args.pg_host or os.environ.get("PGHOST", ""),
+            "PGPORT": str(args.pg_port or os.environ.get("PGPORT", "")),
+            "PGDATABASE": args.pg_database or os.environ.get("PGDATABASE", ""),
+            "PGUSER": args.pg_user or os.environ.get("PGUSER", ""),
+            "PGPASSWORD": args.pg_password or os.environ.get("PGPASSWORD", ""),
+        }
+    )
+    missing = [name for name in ["PGHOST", "PGPORT", "PGDATABASE", "PGUSER", "PGPASSWORD"] if not env.get(name)]
+    if missing:
+        raise RuntimeError("缺少 PostgreSQL 连接配置：" + "、".join(missing))
+    return env
+
+
+def run_psql(sql: str, args: argparse.Namespace) -> None:
+    if shutil.which("psql") is None:
+        raise RuntimeError("未找到 psql。请先安装 PostgreSQL 客户端。")
+    with tempfile.TemporaryDirectory(prefix="patient_front_review_") as tmpdir:
+        sql_path = Path(tmpdir) / "review.sql"
+        sql_path.write_text(sql, encoding="utf-8")
+        completed = subprocess.run(
+            ["psql", "--no-password", "--file", str(sql_path)],
+            check=False,
+            stdout=subprocess.PIPE,
+            stderr=subprocess.PIPE,
+            text=True,
+            encoding="utf-8",
+            errors="replace",
+            env=connection_env(args),
+        )
+    if completed.returncode != 0:
+        raise RuntimeError(completed.stderr.strip() or completed.stdout.strip() or "psql 执行失败")
+    if completed.stdout.strip():
+        print(completed.stdout.strip())
+
+
+def export_review(args: argparse.Namespace) -> None:
+    output = args.output.resolve()
+    output.parent.mkdir(parents=True, exist_ok=True)
+    escaped_output = str(output).replace("'", "''")
+    table = quote_pg_identifier(args.pg_table)
+    where_sql = "" if args.all else "WHERE review_status <> 'auto_pass' OR manual_corrected IS TRUE"
+    select_sql = (
+        "SELECT inpatient_no, source_file, medical_record_no, front_page_medical_record_no, patient_name, "
+        "gender, birth_date, discharge_dept, major_department, primary_diagnosis, "
+        "primary_diagnosis_code, total_cost, self_pay_amount, review_status, "
+        "review_notes::text AS review_notes, auto_corrections::text AS auto_corrections, "
+        "manual_corrected, contact_address, contact_phone, discharge_diagnoses::text AS discharge_diagnoses, "
+        "operations::text AS operations, pathology_diagnosis, pathology_diagnosis_code, pathology_no, "
+        "discharge_disposition_code, readmission_plan_code, quality_status, quality_notes::text AS quality_notes, "
+        "'' AS manual_review_notes, '' AS corrected_medical_record_no, "
+        "'' AS corrected_patient_name, '' AS corrected_major_department, "
+        "'' AS corrected_primary_diagnosis, '' AS corrected_primary_diagnosis_code, "
+        "'' AS corrected_total_cost, '' AS corrected_self_pay_amount, '' AS mark_review_status "
+        f"FROM {table} {where_sql} ORDER BY source_file"
+    )
+    sql = f"\\set ON_ERROR_STOP on\n\\copy ({select_sql}) TO '{escaped_output}' WITH (FORMAT csv, HEADER true, ENCODING 'UTF8')\n"
+    run_psql(sql, args)
+    print(f"已导出复核表：{output}")
+
+
+def import_review(args: argparse.Namespace) -> None:
+    input_path = args.input.resolve()
+    if not input_path.exists():
+        raise RuntimeError(f"复核表不存在：{input_path}")
+    escaped_input = str(input_path).replace("'", "''")
+    table = quote_pg_identifier(args.pg_table)
+    sql = f"""
+\\set ON_ERROR_STOP on
+CREATE TEMP TABLE patient_front_review_import (
+    inpatient_no TEXT,
+    source_file TEXT,
+    medical_record_no TEXT,
+    front_page_medical_record_no TEXT,
+    patient_name TEXT,
+    gender TEXT,
+    birth_date TEXT,
+    discharge_dept TEXT,
+    major_department TEXT,
+    primary_diagnosis TEXT,
+    primary_diagnosis_code TEXT,
+    total_cost TEXT,
+    self_pay_amount TEXT,
+    review_status TEXT,
+    review_notes TEXT,
+    auto_corrections TEXT,
+    manual_corrected TEXT,
+    contact_address TEXT,
+    contact_phone TEXT,
+    discharge_diagnoses TEXT,
+    operations TEXT,
+    pathology_diagnosis TEXT,
+    pathology_diagnosis_code TEXT,
+    pathology_no TEXT,
+    discharge_disposition_code TEXT,
+    readmission_plan_code TEXT,
+    quality_status TEXT,
+    quality_notes TEXT,
+    manual_review_notes TEXT,
+    corrected_medical_record_no TEXT,
+    corrected_patient_name TEXT,
+    corrected_major_department TEXT,
+    corrected_primary_diagnosis TEXT,
+    corrected_primary_diagnosis_code TEXT,
+    corrected_total_cost TEXT,
+    corrected_self_pay_amount TEXT,
+    mark_review_status TEXT
+);
+
+\\copy patient_front_review_import FROM '{escaped_input}' WITH (FORMAT csv, HEADER true, ENCODING 'UTF8')
+
+UPDATE {table} AS t
+SET
+    medical_record_no = COALESCE(NULLIF(i.corrected_medical_record_no, ''), t.medical_record_no),
+    patient_name = COALESCE(NULLIF(i.corrected_patient_name, ''), t.patient_name),
+    major_department = COALESCE(NULLIF(i.corrected_major_department, ''), t.major_department),
+    primary_diagnosis = COALESCE(NULLIF(i.corrected_primary_diagnosis, ''), t.primary_diagnosis),
+    primary_diagnosis_code = COALESCE(NULLIF(i.corrected_primary_diagnosis_code, ''), t.primary_diagnosis_code),
+    total_cost = COALESCE(NULLIF(i.corrected_total_cost, '')::numeric, t.total_cost),
+    self_pay_amount = COALESCE(NULLIF(i.corrected_self_pay_amount, '')::numeric, t.self_pay_amount),
+    review_status = COALESCE(NULLIF(i.mark_review_status, ''), 'reviewed'),
+    review_notes = CASE
+        WHEN NULLIF(i.manual_review_notes, '') IS NULL THEN t.review_notes
+        ELSE t.review_notes || jsonb_build_array(i.manual_review_notes)
+    END,
+    manual_corrected = CASE
+        WHEN NULLIF(i.corrected_medical_record_no, '') IS NOT NULL
+          OR NULLIF(i.corrected_patient_name, '') IS NOT NULL
+          OR NULLIF(i.corrected_major_department, '') IS NOT NULL
+          OR NULLIF(i.corrected_primary_diagnosis, '') IS NOT NULL
+          OR NULLIF(i.corrected_primary_diagnosis_code, '') IS NOT NULL
+          OR NULLIF(i.corrected_total_cost, '') IS NOT NULL
+          OR NULLIF(i.corrected_self_pay_amount, '') IS NOT NULL
+          OR lower(COALESCE(i.manual_corrected, '')) IN ('true', 't', '1', 'yes', 'y')
+        THEN true
+        ELSE t.manual_corrected
+    END
+FROM patient_front_review_import AS i
+WHERE (
+    NULLIF(i.inpatient_no, '') IS NOT NULL
+    AND t.inpatient_no = i.inpatient_no
+) OR (
+    NULLIF(i.inpatient_no, '') IS NULL
+    AND t.source_file = i.source_file
+);
+""".strip()
+    run_psql(sql, args)
+    print(f"已回写人工复核表：{input_path}")
+
+
+def build_parser() -> argparse.ArgumentParser:
+    parser = argparse.ArgumentParser(description="导出/回写患者首页 PostgreSQL 人工复核表。")
+    subparsers = parser.add_subparsers(dest="command", required=True)
+
+    def add_pg_options(subparser: argparse.ArgumentParser) -> None:
+        subparser.add_argument("--pg-host", default=os.environ.get("PGHOST", ""), help="PostgreSQL 主机；也可用 PGHOST")
+        subparser.add_argument("--pg-port", default=os.environ.get("PGPORT", "5432"), help="PostgreSQL 端口；也可用 PGPORT")
+        subparser.add_argument("--pg-database", default=os.environ.get("PGDATABASE", ""), help="PostgreSQL 数据库；也可用 PGDATABASE")
+        subparser.add_argument("--pg-user", default=os.environ.get("PGUSER", ""), help="PostgreSQL 用户名；也可用 PGUSER")
+        subparser.add_argument("--pg-password", default=os.environ.get("PGPASSWORD", ""), help="PostgreSQL 密码；也可用 PGPASSWORD")
+        subparser.add_argument("--pg-table", default=os.environ.get("PG_PATIENT_TABLE", DEFAULT_TABLE), help="PostgreSQL 表名")
+
+    export_parser = subparsers.add_parser("export", help="从 PostgreSQL 导出待复核 CSV")
+    add_pg_options(export_parser)
+    export_parser.add_argument("-o", "--output", type=Path, default=DEFAULT_OUTPUT, help="复核 CSV 输出路径")
+    export_parser.add_argument("--all", action="store_true", help="导出全部记录，而不仅是需复核/已人工修正记录")
+
+    import_parser = subparsers.add_parser("import", help="把人工复核 CSV 回写 PostgreSQL")
+    add_pg_options(import_parser)
+    import_parser.add_argument("-i", "--input", type=Path, default=DEFAULT_OUTPUT, help="人工复核 CSV 路径")
+    return parser
+
+
+def main() -> int:
+    args = build_parser().parse_args()
+    if args.command == "export":
+        export_review(args)
+    elif args.command == "import":
+        import_review(args)
+    else:
+        raise RuntimeError(f"未知命令：{args.command}")
+    return 0
+
+
+if __name__ == "__main__":
+    try:
+        raise SystemExit(main())
+    except Exception as exc:  # noqa: BLE001 - 命令行工具需要清楚输出错误
+        print(f"错误：{exc}", file=sys.stderr)
+        raise SystemExit(1)
diff --git a/患者首页处理/数据处理工作区/04_质量体检/04_字段核验与数据库体检.py b/患者首页处理/数据处理工作区/04_质量体检/04_字段核验与数据库体检.py
new file mode 100755
index 0000000..65c308f
--- /dev/null
+++ b/患者首页处理/数据处理工作区/04_质量体检/04_字段核验与数据库体检.py
@@ -0,0 +1,175 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+"""
+对 Patient_FrontPages 做字段级体检。
+
+输出：
+- 数据处理结果区/05_质量体检/患者首页_数据库体检报告.txt
+- 数据处理结果区/05_质量体检/患者首页_字段空值统计.csv
+- 数据处理结果区/05_质量体检/患者首页_疑似异常记录.csv
+"""
+
+from __future__ import annotations
+
+import argparse
+import os
+import re
+import shutil
+import subprocess
+import sys
+import tempfile
+from pathlib import Path
+
+
+PROJECT_ROOT = Path(__file__).resolve().parents[2]
+DEFAULT_OUTPUT_DIR = PROJECT_ROOT / "数据处理结果区" / "05_质量体检"
+DEFAULT_TABLE = "Patient_FrontPages"
+
+
+def quote_pg_identifier(identifier: str) -> str:
+    if not re.fullmatch(r"[A-Za-z_][A-Za-z0-9_]*", identifier):
+        raise ValueError(f"非法 PostgreSQL 表名：{identifier}")
+    return '"' + identifier.replace('"', '""') + '"'
+
+
+def connection_env(args: argparse.Namespace) -> dict[str, str]:
+    env = os.environ.copy()
+    env.update(
+        {
+            "PGHOST": args.pg_host or os.environ.get("PGHOST", ""),
+            "PGPORT": str(args.pg_port or os.environ.get("PGPORT", "")),
+            "PGDATABASE": args.pg_database or os.environ.get("PGDATABASE", ""),
+            "PGUSER": args.pg_user or os.environ.get("PGUSER", ""),
+            "PGPASSWORD": args.pg_password or os.environ.get("PGPASSWORD", ""),
+        }
+    )
+    missing = [name for name in ["PGHOST", "PGPORT", "PGDATABASE", "PGUSER", "PGPASSWORD"] if not env.get(name)]
+    if missing:
+        raise RuntimeError("缺少 PostgreSQL 连接配置：" + "、".join(missing))
+    return env
+
+
+def run_psql(sql: str, args: argparse.Namespace) -> str:
+    if shutil.which("psql") is None:
+        raise RuntimeError("未找到 psql。请先安装 PostgreSQL 客户端。")
+    with tempfile.TemporaryDirectory(prefix="patient_front_audit_") as tmpdir:
+        sql_path = Path(tmpdir) / "audit.sql"
+        sql_path.write_text(sql, encoding="utf-8")
+        completed = subprocess.run(
+            ["psql", "--no-password", "--file", str(sql_path)],
+            check=False,
+            stdout=subprocess.PIPE,
+            stderr=subprocess.PIPE,
+            text=True,
+            encoding="utf-8",
+            errors="replace",
+            env=connection_env(args),
+        )
+    if completed.returncode != 0:
+        raise RuntimeError(completed.stderr.strip() or completed.stdout.strip() or "psql 执行失败")
+    return completed.stdout
+
+
+def build_empty_stat_sql(table: str) -> str:
+    monitored_columns = [
+        "inpatient_no",
+        "medical_record_no",
+        "patient_name",
+        "gender",
+        "birth_date",
+        "id_card_no",
+        "contact_address",
+        "admission_time",
+        "discharge_time",
+        "major_department",
+        "primary_diagnosis",
+        "primary_diagnosis_code",
+        "discharge_diagnoses",
+        "operations",
+        "quality_status",
+        "review_status",
+    ]
+    selects = []
+    for column in monitored_columns:
+        identifier = quote_pg_identifier(column)
+        if column in {"discharge_diagnoses", "operations"}:
+            empty_expr = f"{identifier} IS NULL OR jsonb_array_length({identifier}) = 0"
+        else:
+            empty_expr = f"{identifier} IS NULL OR {identifier}::text = ''"
+        selects.append(f"SELECT '{column}' AS column_name, count(*) FILTER (WHERE {empty_expr}) AS empty_count, count(*) AS total_count FROM {table}")
+    return " UNION ALL ".join(selects)
+
+
+def audit(args: argparse.Namespace) -> None:
+    output_dir = args.output_dir.resolve()
+    output_dir.mkdir(parents=True, exist_ok=True)
+    table = quote_pg_identifier(args.pg_table)
+    empty_csv = output_dir / "患者首页_字段空值统计.csv"
+    suspicious_csv = output_dir / "患者首页_疑似异常记录.csv"
+    report_path = output_dir / "患者首页_数据库体检报告.txt"
+    empty_stat_sql = build_empty_stat_sql(table)
+    suspicious_sql = (
+        "SELECT source_file, inpatient_no, medical_record_no, patient_name, review_status, "
+        "review_notes::text AS review_notes, contact_address, primary_diagnosis, primary_diagnosis_code "
+        f"FROM {table} "
+        "WHERE review_status <> 'auto_pass' "
+        "OR inpatient_no !~ '^ZY\\d{12}$' "
+        "OR medical_record_no !~ '^\\d{10}$' "
+        "OR contact_address ~ '急诊|门诊|其他医疗机构转入|医嘱离院' "
+        "OR primary_diagnosis_code IS NULL "
+        "OR primary_diagnosis_code = '' "
+        "ORDER BY source_file"
+    )
+
+    sql = f"""
+\\set ON_ERROR_STOP on
+\\pset pager off
+\\o {str(report_path).replace("'", "''")}
+SELECT '总记录数' AS item, count(*)::text AS value FROM {table}
+UNION ALL
+SELECT '需复核记录数', count(*)::text FROM {table} WHERE review_status <> 'auto_pass'
+UNION ALL
+SELECT '缺字段注释数', count(*)::text
+FROM pg_class c
+JOIN pg_namespace n ON n.oid = c.relnamespace
+JOIN pg_attribute a ON a.attrelid = c.oid
+WHERE n.nspname = 'public'
+  AND c.relname = {args.pg_table!r}
+  AND a.attnum > 0
+  AND NOT a.attisdropped
+  AND col_description(c.oid, a.attnum) IS NULL;
+\\o
+\\copy ({empty_stat_sql}) TO '{str(empty_csv).replace("'", "''")}' WITH (FORMAT csv, HEADER true, ENCODING 'UTF8')
+\\copy ({suspicious_sql}) TO '{str(suspicious_csv).replace("'", "''")}' WITH (FORMAT csv, HEADER true, ENCODING 'UTF8')
+""".strip()
+    stdout = run_psql(sql, args)
+    if stdout.strip():
+        print(stdout.strip())
+    print(f"体检报告：{report_path}")
+    print(f"字段空值统计：{empty_csv}")
+    print(f"疑似异常记录：{suspicious_csv}")
+
+
+def build_parser() -> argparse.ArgumentParser:
+    parser = argparse.ArgumentParser(description="核验 Patient_FrontPages 字段注释、空值和疑似错位记录。")
+    parser.add_argument("--pg-host", default=os.environ.get("PGHOST", ""), help="PostgreSQL 主机；也可用 PGHOST")
+    parser.add_argument("--pg-port", default=os.environ.get("PGPORT", "5432"), help="PostgreSQL 端口；也可用 PGPORT")
+    parser.add_argument("--pg-database", default=os.environ.get("PGDATABASE", ""), help="PostgreSQL 数据库；也可用 PGDATABASE")
+    parser.add_argument("--pg-user", default=os.environ.get("PGUSER", ""), help="PostgreSQL 用户名；也可用 PGUSER")
+    parser.add_argument("--pg-password", default=os.environ.get("PGPASSWORD", ""), help="PostgreSQL 密码；也可用 PGPASSWORD")
+    parser.add_argument("--pg-table", default=os.environ.get("PG_PATIENT_TABLE", DEFAULT_TABLE), help="PostgreSQL 表名")
+    parser.add_argument("-o", "--output-dir", type=Path, default=DEFAULT_OUTPUT_DIR, help="体检结果输出目录")
+    return parser
+
+
+def main() -> int:
+    audit(build_parser().parse_args())
+    return 0
+
+
+if __name__ == "__main__":
+    try:
+        raise SystemExit(main())
+    except Exception as exc:  # noqa: BLE001
+        print(f"错误：{exc}", file=sys.stderr)
+        raise SystemExit(1)
diff --git a/患者首页处理/数据处理工作区/05_备用读取/05_备用PDF转Markdown_Mineru.py b/患者首页处理/数据处理工作区/05_备用读取/05_备用PDF转Markdown_Mineru.py
new file mode 100644
index 0000000..ad111f0
--- /dev/null
+++ b/患者首页处理/数据处理工作区/05_备用读取/05_备用PDF转Markdown_Mineru.py
@@ -0,0 +1,130 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+
+import os
+import requests
+import zipfile
+import io
+import shutil  # 【新增】用于删除非空文件夹
+import argparse  # <--- 1. 确保这里导入了
+
+def main():
+    # 1. 配置命令行参数解析
+    parser = argparse.ArgumentParser(description="Mineru PDF 转 Markdown 批量处理工具")
+    
+    # 添加参数
+    parser.add_argument("-s", "--source", type=str, default="./Papers/ORI_PDF",
+                        help="源 PDF 文件夹路径 (默认: ./Papers/ORI_PDF)")
+    parser.add_argument("-t", "--target", type=str, default="./Papers/ORI_MD",
+                        help="目标 Markdown 文件夹路径 (默认: ./Papers/ORI_MD)")
+    parser.add_argument("-u", "--url", type=str, default="http://10.168.1.103:4000/extract",
+                        help="API 服务端地址 (默认: http://10.168.1.103:4000/extract)")
+    parser.add_argument("--sync", action="store_true", 
+                        help="任务完成后是否开启反向同步清理（删除多余的输出文件夹）")
+
+    args = parser.parse_args()
+    
+    # 使用解析后的参数
+    url = args.url
+    source_dir = args.source
+    target_dir = args.target
+
+    # 2. 确保源目录存在，避免报错
+    if not os.path.exists(source_dir):
+        print(f"错误: 找不到源文件夹 '{source_dir}'")
+        exit(1)
+
+    # 确保目标主文件夹存在
+    os.makedirs(target_dir, exist_ok=True)
+
+    # 3. 遍历源文件夹下的所有文件进行上传处理
+    for filename in os.listdir(source_dir):
+        # 只处理 PDF 文件
+        if not filename.lower().endswith(".pdf"):
+            continue
+            
+        file_path = os.path.join(source_dir, filename)
+        # 获取去掉 .pdf 后缀的文件名
+        pdf_name_no_ext = os.path.splitext(filename)[0]
+        
+        # 对应的输出文件夹路径
+        output_folder = os.path.join(target_dir, pdf_name_no_ext)
+        
+        # ==========================================
+        # 检查是否已经处理过
+        # 如果输出文件夹已存在，且内部有文件，则视为已转换，直接跳过
+        # ==========================================
+        if os.path.exists(output_folder) and len(os.listdir(output_folder)) > 0:
+            print(f"⏩ 已存在转换结果，跳过处理: {filename}")
+            continue
+        
+        print(f"正在上传并处理 {filename}...")
+        
+        try:
+            # 4. 发送请求
+            with open(file_path, "rb") as f:
+                files = {"file": (filename, f, "application/pdf")}
+                response = requests.post(url, files=files)
+
+            # 5. 处理响应结果
+            if response.status_code == 200:
+                # 检查服务端是否返回了包含报错信息的 JSON
+                if response.headers.get('content-type') == 'application/json':
+                    error_msg = response.json()
+                    print(f"❌ 服务端处理失败 ({filename})：{error_msg.get('message', '未知错误')}")
+                    continue # 跳过解压，处理下一个
+
+                # 确保该 PDF 专属的输出文件夹存在
+                os.makedirs(output_folder, exist_ok=True)
+                
+                # 核心：使用 io.BytesIO 直接在内存中读取 zip 内容并解压
+                try:
+                    with zipfile.ZipFile(io.BytesIO(response.content)) as zip_ref:
+                        zip_ref.extractall(output_folder)
+                    print(f"✅ 成功！已解压并保存至文件夹: {output_folder}")
+                except zipfile.BadZipFile:
+                    print(f"❌ 失败！{filename} 返回的内容不是有效的 ZIP 格式。")
+            else:
+                print(f"❌ 失败！{filename} 状态码: {response.status_code}, 报错: {response.text}")
+                
+        except requests.exceptions.RequestException as e:
+            print(f"❌ 网络请求异常 ({filename}): {e}")
+        except Exception as e:
+            print(f"❌ 处理 {filename} 时发生未知错误: {e}")
+
+    print("-" * 30)
+    print("转换任务结束，开始进行文件夹同步清理...")
+
+    # ==========================================
+    # 【新增逻辑】：反向比对，清理多余的 MD 文件夹
+    # ==========================================
+    # 1. 收集当前 ORI_PDF 中所有有效的 PDF 名字（不含后缀）
+    valid_pdf_names = set()
+    for filename in os.listdir(source_dir):
+        if filename.lower().endswith(".pdf"):
+            valid_pdf_names.add(os.path.splitext(filename)[0])
+
+    # 2. 遍历 ORI_MD 文件夹
+    if os.path.exists(target_dir):
+        for folder_name in os.listdir(target_dir):
+            folder_path = os.path.join(target_dir, folder_name)
+            
+            # 仅处理文件夹（以防里面有意外的独立文件）
+            if os.path.isdir(folder_path):
+                # 3. 如果这个文件夹的名字不在有效的 PDF 列表中，说明是被删掉的“孤儿”
+                if folder_name not in valid_pdf_names:
+                    print(f"🧹 发现多余文件，正在清理: {folder_name}")
+                    try:
+                        # 使用 shutil.rmtree 删除整个文件夹及其内部所有内容
+                        shutil.rmtree(folder_path)
+                    except Exception as e:
+                        print(f"❌ 删除 {folder_name} 时发生错误: {e}")
+
+    print("-" * 30)
+    print("所有批量任务彻底完成！")
+
+if __name__ == "__main__":
+    try:
+        main()
+    except KeyboardInterrupt:
+        print("\n检测到用户中断，程序退出。")
diff --git a/患者首页处理/数据处理工作区/06_图片对照核验/06_PDF转图片与对照核验.py b/患者首页处理/数据处理工作区/06_图片对照核验/06_PDF转图片与对照核验.py
new file mode 100644
index 0000000..aea2167
--- /dev/null
+++ b/患者首页处理/数据处理工作区/06_图片对照核验/06_PDF转图片与对照核验.py
@@ -0,0 +1,697 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+"""
+把患者首页 PDF 转为图片，并生成图片-结构化字段对照核验清单。
+
+依赖系统命令 pdftoppm，通常由 poppler-utils 提供。
+"""
+
+from __future__ import annotations
+
+import argparse
+import csv
+import html
+import json
+import re
+import shutil
+import subprocess
+from dataclasses import dataclass
+from pathlib import Path
+from string import Template
+from typing import Any
+
+
+PROJECT_ROOT = Path(__file__).resolve().parents[2]
+DEFAULT_INPUT_DIR = PROJECT_ROOT / "待处理-患者首页PDF"
+DEFAULT_RESULT_DIR = PROJECT_ROOT / "数据处理结果区"
+DEFAULT_IMAGE_REVIEW_DIR = DEFAULT_RESULT_DIR / "06_PDF图片对照"
+
+
+@dataclass(frozen=True)
+class FieldCheck:
+    group: str
+    key: str
+    level: str
+    location_hint: str
+
+
+FIELD_CHECKS = [
+    FieldCheck("基本信息", "医疗机构", "recommended", "首页抬头及组织机构代码附近"),
+    FieldCheck("基本信息", "医疗付费方式", "recommended", "首页左上付费方式"),
+    FieldCheck("基本信息", "健康卡号", "recommended", "首页左上健康卡号"),
+    FieldCheck("基本信息", "住院次数", "recommended", "首页左上住院次数"),
+    FieldCheck("基本信息", "病案号", "required", "首页左上病案号"),
+    FieldCheck("基本信息", "姓名", "required", "基本信息行"),
+    FieldCheck("基本信息", "性别", "required", "基本信息行"),
+    FieldCheck("基本信息", "出生日期", "required", "基本信息行"),
+    FieldCheck("基本信息", "年龄", "required", "基本信息行"),
+    FieldCheck("基本信息", "国籍", "recommended", "基本信息行"),
+    FieldCheck("基本信息", "身份证号", "required", "基本信息行"),
+    FieldCheck("基本信息", "职业", "recommended", "基本信息行"),
+    FieldCheck("基本信息", "婚姻代码", "recommended", "基本信息行"),
+    FieldCheck("基本信息", "出生地", "recommended", "地址信息区"),
+    FieldCheck("基本信息", "籍贯", "recommended", "地址信息区"),
+    FieldCheck("基本信息", "民族", "recommended", "地址信息区"),
+    FieldCheck("地址联系人", "现住址", "required", "现住址行"),
+    FieldCheck("地址联系人", "现住址电话", "recommended", "现住址行"),
+    FieldCheck("地址联系人", "现住址邮编", "recommended", "现住址行"),
+    FieldCheck("地址联系人", "户口地址", "recommended", "户口地址行"),
+    FieldCheck("地址联系人", "户口地址邮编", "recommended", "户口地址行"),
+    FieldCheck("地址联系人", "工作单位及地址", "recommended", "工作单位及地址行"),
+    FieldCheck("地址联系人", "单位电话", "recommended", "工作单位及地址行"),
+    FieldCheck("地址联系人", "单位邮编", "recommended", "工作单位及地址行"),
+    FieldCheck("地址联系人", "联系人姓名", "required", "联系人信息行"),
+    FieldCheck("地址联系人", "联系人关系", "recommended", "联系人信息行"),
+    FieldCheck("地址联系人", "联系人地址", "recommended", "联系人信息行"),
+    FieldCheck("地址联系人", "联系人电话", "required", "联系人信息行"),
+    FieldCheck("入出院", "入院途径代码", "recommended", "入院途径勾选项"),
+    FieldCheck("入出院", "入院时间", "required", "入院记录行"),
+    FieldCheck("入出院", "入院科别", "required", "入院记录行"),
+    FieldCheck("入出院", "入院病房", "recommended", "入院记录行"),
+    FieldCheck("入出院", "转科科别", "optional", "转科科别行"),
+    FieldCheck("入出院", "出院时间", "required", "出院记录行"),
+    FieldCheck("入出院", "出院科别", "required", "出院记录行"),
+    FieldCheck("入出院", "出院病房", "recommended", "出院记录行"),
+    FieldCheck("入出院", "实际住院天数", "required", "出院记录行"),
+    FieldCheck("入出院", "大科室", "required", "由入院/出院科别映射"),
+    FieldCheck("诊断手术", "门急诊诊断", "recommended", "诊断区顶部"),
+    FieldCheck("诊断手术", "门急诊诊断编码", "recommended", "诊断区顶部"),
+    FieldCheck("诊断手术", "主要诊断名称", "required", "出院诊断表第一行"),
+    FieldCheck("诊断手术", "主要诊断编码", "required", "出院诊断表第一行"),
+    FieldCheck("诊断手术", "主要诊断入院病情", "required", "出院诊断表第一行"),
+    FieldCheck("诊断手术", "出院诊断", "recommended", "出院诊断表全部行"),
+    FieldCheck("诊断手术", "手术操作", "recommended", "手术及操作表"),
+    FieldCheck("诊断手术", "损伤中毒外部原因", "optional", "损伤中毒外部原因行"),
+    FieldCheck("诊断手术", "损伤中毒疾病编码", "optional", "损伤中毒外部原因行"),
+    FieldCheck("诊断手术", "病理诊断", "optional", "病理诊断行"),
+    FieldCheck("诊断手术", "病理诊断编码", "optional", "病理诊断行"),
+    FieldCheck("诊断手术", "病理号", "optional", "病理号"),
+    FieldCheck("质控信息", "药物过敏代码", "recommended", "药物过敏勾选项"),
+    FieldCheck("质控信息", "过敏药物", "optional", "过敏药物填写处"),
+    FieldCheck("质控信息", "死亡患者尸检代码", "recommended", "死亡患者尸检勾选项"),
+    FieldCheck("质控信息", "血型代码", "recommended", "血型勾选项"),
+    FieldCheck("质控信息", "Rh代码", "recommended", "Rh勾选项"),
+    FieldCheck("质控信息", "科主任", "recommended", "医师签名区"),
+    FieldCheck("质控信息", "主任副主任医师", "recommended", "医师签名区"),
+    FieldCheck("质控信息", "主治医师", "recommended", "医师签名区"),
+    FieldCheck("质控信息", "住院医师", "recommended", "医师签名区"),
+    FieldCheck("质控信息", "责任护士", "recommended", "护理签名区"),
+    FieldCheck("质控信息", "编码员", "recommended", "编码员签名区"),
+    FieldCheck("质控信息", "病案质量代码", "recommended", "病案质量勾选项"),
+    FieldCheck("质控信息", "质控医师", "recommended", "质控签名区"),
+    FieldCheck("质控信息", "质控护士", "recommended", "质控签名区"),
+    FieldCheck("质控信息", "质控日期", "recommended", "质控日期"),
+    FieldCheck("离院费用", "离院方式代码", "recommended", "离院方式勾选项"),
+    FieldCheck("离院费用", "出院31天内再住院计划代码", "recommended", "再住院计划勾选项"),
+    FieldCheck("离院费用", "再住院计划目的", "optional", "再住院计划目的"),
+    FieldCheck("离院费用", "入院前昏迷天数", "optional", "昏迷时间区"),
+    FieldCheck("离院费用", "入院后昏迷天数", "optional", "昏迷时间区"),
+    FieldCheck("离院费用", "总费用", "optional", "费用区"),
+    FieldCheck("离院费用", "自付金额", "optional", "费用区"),
+    FieldCheck("离院费用", "费用明细", "optional", "费用明细区"),
+]
+
+
+def parse_args() -> argparse.Namespace:
+    parser = argparse.ArgumentParser(description="患者首页 PDF 转图片并生成字段对照核验报告")
+    parser.add_argument("-i", "--input-dir", type=Path, default=DEFAULT_INPUT_DIR, help="PDF 输入目录")
+    parser.add_argument("-r", "--result-dir", type=Path, default=DEFAULT_RESULT_DIR, help="解析结果根目录")
+    parser.add_argument("-o", "--output-dir", type=Path, default=DEFAULT_IMAGE_REVIEW_DIR, help="图片对照输出目录")
+    parser.add_argument("--dpi", type=int, default=180, help="图片分辨率，默认 180")
+    parser.add_argument("--format", choices=["png", "jpeg"], default="png", help="图片格式，默认 png")
+    parser.add_argument("--force", action="store_true", help="重新生成已存在图片")
+    parser.add_argument("--strict-recommended", action="store_true", help="把建议字段缺项也写入对照建议")
+    return parser.parse_args()
+
+
+def run(cmd: list[str]) -> subprocess.CompletedProcess[str]:
+    return subprocess.run(cmd, check=True, text=True, capture_output=True)
+
+
+def numeric_page_key(path: Path) -> tuple[int, str]:
+    match = re.search(r"-(\d+)\.(png|jpe?g)$", path.name, flags=re.IGNORECASE)
+    if not match:
+        return (10**9, path.name)
+    return (int(match.group(1)), path.name)
+
+
+def convert_pdf_to_images(pdf_path: Path, output_dir: Path, dpi: int, image_format: str, force: bool) -> tuple[list[Path], str]:
+    pdftoppm = shutil.which("pdftoppm")
+    if not pdftoppm:
+        return [], "未找到 pdftoppm，请安装 poppler-utils。"
+
+    output_dir.mkdir(parents=True, exist_ok=True)
+    extension = "jpg" if image_format == "jpeg" else "png"
+    existing = sorted(output_dir.glob(f"page-*.{extension}"), key=numeric_page_key)
+    if existing and not force:
+        return existing, ""
+
+    if force:
+        for old_image in output_dir.glob("page-*.*"):
+            old_image.unlink()
+
+    prefix = output_dir / "page"
+    cmd = [pdftoppm, "-r", str(dpi), f"-{image_format}", str(pdf_path), str(prefix)]
+    try:
+        run(cmd)
+    except subprocess.CalledProcessError as exc:
+        detail = exc.stderr.strip() or exc.stdout.strip() or str(exc)
+        return [], f"图片转换失败：{detail}"
+
+    images = sorted(output_dir.glob(f"page-*.{extension}"), key=numeric_page_key)
+    normalized_images: list[Path] = []
+    for index, image_path in enumerate(images, start=1):
+        target = output_dir / f"page-{index:03d}.{extension}"
+        if image_path != target:
+            if target.exists():
+                target.unlink()
+            image_path.rename(target)
+        normalized_images.append(target)
+    return normalized_images, "" if normalized_images else "图片转换后未找到输出文件。"
+
+
+def load_record(result_dir: Path, pdf_path: Path) -> tuple[dict[str, Any], Path | None, str]:
+    json_path = result_dir / "02_单份JSON" / f"{pdf_path.stem}.json"
+    if not json_path.exists():
+        return {}, None, "未找到结构化 JSON，请先运行 02_解析入库 步骤。"
+    try:
+        return json.loads(json_path.read_text(encoding="utf-8")), json_path, ""
+    except json.JSONDecodeError as exc:
+        return {}, json_path, f"结构化 JSON 读取失败：{exc}"
+
+
+def is_blank(value: Any) -> bool:
+    if value is None:
+        return True
+    if isinstance(value, str):
+        text = value.strip()
+        return text == "" or text in {"-", "--", "null", "None", "[]", "{}"}
+    if isinstance(value, (list, tuple, set, dict)):
+        return len(value) == 0
+    return False
+
+
+def value_preview(value: Any, max_length: int = 120) -> str:
+    if value is None:
+        return ""
+    if isinstance(value, (list, dict)):
+        text = json.dumps(value, ensure_ascii=False)
+    else:
+        text = str(value)
+    text = re.sub(r"\s+", " ", text).strip()
+    if len(text) > max_length:
+        return text[: max_length - 1] + "…"
+    return text
+
+
+def build_field_rows(record: dict[str, Any]) -> tuple[list[dict[str, str]], list[str], list[str]]:
+    rows: list[dict[str, str]] = []
+    missing_required: list[str] = []
+    missing_recommended: list[str] = []
+
+    for check in FIELD_CHECKS:
+        value = record.get(check.key)
+        missing = is_blank(value)
+        if missing and check.level == "required":
+            missing_required.append(check.key)
+        elif missing and check.level == "recommended":
+            missing_recommended.append(check.key)
+        rows.append(
+            {
+                "group": check.group,
+                "key": check.key,
+                "level": check.level,
+                "location_hint": check.location_hint,
+                "value": value_preview(value),
+                "missing": "是" if missing else "否",
+            }
+        )
+
+    return rows, missing_required, missing_recommended
+
+
+def as_csv_text(value: Any) -> str:
+    if value is None:
+        return ""
+    if isinstance(value, (list, dict)):
+        return json.dumps(value, ensure_ascii=False)
+    return str(value)
+
+
+def relative_to(path: Path, start: Path) -> str:
+    try:
+        return path.relative_to(start).as_posix()
+    except ValueError:
+        return path.as_posix()
+
+
+def html_escape(value: Any) -> str:
+    return html.escape(str(value), quote=True)
+
+
+def make_case_summary(
+    pdf_path: Path,
+    images: list[Path],
+    image_error: str,
+    record: dict[str, Any],
+    json_path: Path | None,
+    json_error: str,
+    output_dir: Path,
+    strict_recommended: bool,
+) -> dict[str, Any]:
+    field_rows, missing_required, missing_recommended = build_field_rows(record) if record else ([], [], [])
+    review_notes = record.get("复核备注", []) if record else []
+    if not isinstance(review_notes, list):
+        review_notes = [review_notes]
+
+    suggestions: list[str] = []
+    if image_error:
+        suggestions.append(image_error)
+    if json_error:
+        suggestions.append(json_error)
+    if missing_required:
+        suggestions.append("核心字段缺项，需对照图片补齐：" + "、".join(missing_required))
+    if strict_recommended and missing_recommended:
+        suggestions.append("建议字段缺项，抽样确认是否首页未填写：" + "、".join(missing_recommended))
+    if record.get("复核状态") and record.get("复核状态") != "auto_pass":
+        suggestions.append("该记录已有复核状态：" + str(record.get("复核状态")))
+    if review_notes:
+        suggestions.append("已有复核备注：" + "；".join(value_preview(note, 80) for note in review_notes))
+    if not suggestions:
+        suggestions.append("图片与结构化字段抽查一致即可。")
+
+    return {
+        "pdf_path": pdf_path,
+        "source_file": pdf_path.name,
+        "images": images,
+        "image_dir": images[0].parent if images else output_dir / "图片" / pdf_path.stem,
+        "first_image": images[0] if images else None,
+        "page_count": len(images),
+        "image_error": image_error,
+        "record": record,
+        "json_path": json_path,
+        "json_error": json_error,
+        "field_rows": field_rows,
+        "missing_required": missing_required,
+        "missing_recommended": missing_recommended,
+        "suggestions": suggestions,
+    }
+
+
+def write_index_csv(cases: list[dict[str, Any]], output_dir: Path) -> Path:
+    csv_path = output_dir / "患者首页_PDF图片对照索引.csv"
+    fieldnames = [
+        "源文件",
+        "页数",
+        "图片目录",
+        "首页图片",
+        "结构化JSON",
+        "病案号",
+        "姓名",
+        "复核状态",
+        "复核备注",
+        "核心缺项",
+        "建议核对缺项",
+        "对照建议",
+    ]
+    with csv_path.open("w", encoding="utf-8-sig", newline="") as fp:
+        writer = csv.DictWriter(fp, fieldnames=fieldnames)
+        writer.writeheader()
+        for case in cases:
+            record = case["record"]
+            writer.writerow(
+                {
+                    "源文件": case["source_file"],
+                    "页数": case["page_count"],
+                    "图片目录": relative_to(case["image_dir"], output_dir),
+                    "首页图片": relative_to(case["first_image"], output_dir) if case["first_image"] else "",
+                    "结构化JSON": relative_to(case["json_path"], output_dir) if case["json_path"] else "",
+                    "病案号": record.get("病案号", ""),
+                    "姓名": record.get("姓名", ""),
+                    "复核状态": record.get("复核状态", ""),
+                    "复核备注": as_csv_text(record.get("复核备注", [])),
+                    "核心缺项": "、".join(case["missing_required"]),
+                    "建议核对缺项": "、".join(case["missing_recommended"]),
+                    "对照建议": "；".join(case["suggestions"]),
+                }
+            )
+    return csv_path
+
+
+def render_field_table(field_rows: list[dict[str, str]]) -> str:
+    if not field_rows:
+        return '
    未生成字段核验表。
    '
+
+    parts = [
+        '',
+        "",
+        "",
+    ]
+    for row in field_rows:
+        missing_class = " is-missing" if row["missing"] == "是" and row["level"] != "optional" else ""
+        parts.append(
+            ""
+            ""
+            "".format(
+                missing_class=missing_class.strip(),
+                group=html_escape(row["group"]),
+                key=html_escape(row["key"]),
+                level=html_escape(row["level"]),
+                location=html_escape(row["location_hint"]),
+                value=html_escape(row["value"]),
+                missing=html_escape(row["missing"]),
+            )
+        )
+    parts.append("
    组字段级别首页位置值缺项
    {group}{key}{level}{location}{value}{missing}
    ")
+    return "\n".join(parts)
+
+
+def render_html(cases: list[dict[str, Any]], output_dir: Path) -> Path:
+    html_path = output_dir / "患者首页_PDF图片对照.html"
+    total = len(cases)
+    with_required_missing = sum(1 for case in cases if case["missing_required"])
+    with_review_status = sum(1 for case in cases if case["record"].get("复核状态") not in {"", "auto_pass", None})
+    failed_images = sum(1 for case in cases if case["image_error"])
+
+    nav_rows = []
+    case_sections = []
+    for index, case in enumerate(cases, start=1):
+        anchor = f"case-{index}"
+        record = case["record"]
+        status = record.get("复核状态", "未解析")
+        nav_rows.append(
+            "{source}{name}{mrn}{status}{required}".format(
+                anchor=anchor,
+                source=html_escape(case["source_file"]),
+                name=html_escape(record.get("姓名", "")),
+                mrn=html_escape(record.get("病案号", "")),
+                status=html_escape(status),
+                required=html_escape("、".join(case["missing_required"]) or "无"),
+            )
+        )
+
+        images_html = []
+        for image_path in case["images"]:
+            rel_image = relative_to(image_path, output_dir)
+            images_html.append(
+                '
    {alt}
    {caption}
    '.format(
+                    src=html_escape(rel_image),
+                    alt=html_escape(f"{case['source_file']} {image_path.name}"),
+                    caption=html_escape(image_path.name),
+                )
+            )
+        if not images_html:
+            images_html.append('
    未生成图片。
    ')
+
+        suggestion_items = "\n".join(f"
  • {html_escape(note)}
    • " for note in case["suggestions"])
+        summary_items = [
+            ("病案号", record.get("病案号", "")),
+            ("姓名", record.get("姓名", "")),
+            ("大科室", record.get("大科室", "")),
+            ("出院科别", record.get("出院科别", "")),
+            ("主要诊断", record.get("主要诊断名称", "")),
+            ("主要诊断编码", record.get("主要诊断编码", "")),
+            ("复核状态", status),
+        ]
+        summary_html = "\n".join(
+            f"
    {html_escape(label)}
    {html_escape(value_preview(value, 160))}
    " for label, value in summary_items
+        )
+        case_sections.append(
+            """
+            
    
+              
    
+                
    
+                  
    {source}
    
+                  
    {name} · {mrn} · {status}
    
+                
    
+                返回顶部
+              
    
+              
    
+                
    {images}
    
+                
    
+                  
    {summary}
    
+                  
    对照建议
      {suggestions}
    
+                  {field_table}
+                
    
+              
    
+            
    
+            """.format(
+                anchor=anchor,
+                source=html_escape(case["source_file"]),
+                name=html_escape(record.get("姓名", "")),
+                mrn=html_escape(record.get("病案号", "")),
+                status=html_escape(status),
+                images="\n".join(images_html),
+                summary=summary_html,
+                suggestions=suggestion_items,
+                field_table=render_field_table(case["field_rows"]),
+            )
+        )
+
+    page_template = Template(
+        """
+
+
+  
+  
+  患者首页 PDF 图片对照核验
+  
+
+
+  
    
+    
    患者首页 PDF 图片对照核验
    
+    
    
+      PDF：$total
+      核心缺项：$with_required_missing
+      需复核状态：$with_review_status
+      图片失败：$failed_images
+    
    
+  
    
+  
    
+    
    
+      
+        
+        $nav_rows
+      
    源文件姓名病案号复核状态核心缺项
    
+    
    
+    $case_sections
+  
    
+
+
+"""
+    )
+    html_path.write_text(
+        page_template.substitute(
+            total=total,
+            with_required_missing=with_required_missing,
+            with_review_status=with_review_status,
+            failed_images=failed_images,
+            nav_rows="\n".join(nav_rows),
+            case_sections="\n".join(case_sections),
+        ),
+        encoding="utf-8",
+    )
+    return html_path
+
+
+def main() -> int:
+    args = parse_args()
+    input_dir = args.input_dir.resolve()
+    result_dir = args.result_dir.resolve()
+    output_dir = args.output_dir.resolve()
+    image_root = output_dir / "图片"
+    output_dir.mkdir(parents=True, exist_ok=True)
+
+    pdf_files = sorted(input_dir.glob("*.pdf"))
+    if not pdf_files:
+        raise SystemExit(f"未找到 PDF：{input_dir}")
+
+    cases: list[dict[str, Any]] = []
+    for pdf_path in pdf_files:
+        case_image_dir = image_root / pdf_path.stem
+        images, image_error = convert_pdf_to_images(pdf_path, case_image_dir, args.dpi, args.format, args.force)
+        record, json_path, json_error = load_record(result_dir, pdf_path)
+        cases.append(
+            make_case_summary(
+                pdf_path=pdf_path,
+                images=images,
+                image_error=image_error,
+                record=record,
+                json_path=json_path,
+                json_error=json_error,
+                output_dir=output_dir,
+                strict_recommended=args.strict_recommended,
+            )
+        )
+
+    index_csv = write_index_csv(cases, output_dir)
+    html_path = render_html(cases, output_dir)
+    required_missing_count = sum(1 for case in cases if case["missing_required"])
+    print(f"PDF 数量：{len(cases)}")
+    print(f"图片输出：{image_root}")
+    print(f"对照索引：{index_csv}")
+    print(f"HTML 对照页：{html_path}")
+    print(f"核心缺项病例数：{required_missing_count}")
+    return 0
+
+
+if __name__ == "__main__":
+    raise SystemExit(main())
diff --git a/患者首页处理/数据处理工作区/07_Kimi视觉兜底/07_Kimi图片识别辅助.py b/患者首页处理/数据处理工作区/07_Kimi视觉兜底/07_Kimi图片识别辅助.py
new file mode 100644
index 0000000..0f62c8c
--- /dev/null
+++ b/患者首页处理/数据处理工作区/07_Kimi视觉兜底/07_Kimi图片识别辅助.py
@@ -0,0 +1,205 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+"""
+Kimi 视觉兜底识别：用于 pdftotext/Mineru 对局部字段解析不稳定时，人工指定 PDF 图片页进行辅助抽取。
+
+默认只输出 JSON 文件，不直接覆盖 PostgreSQL。
+"""
+
+from __future__ import annotations
+
+import argparse
+import base64
+import json
+import mimetypes
+import os
+import subprocess
+import sys
+import tempfile
+import urllib.error
+import urllib.request
+from datetime import datetime
+from pathlib import Path
+from typing import Any
+
+
+PROJECT_ROOT = Path(__file__).resolve().parents[2]
+DEFAULT_IMAGE_ROOT = PROJECT_ROOT / "数据处理结果区" / "06_PDF图片对照" / "图片"
+DEFAULT_PDF_ROOT = PROJECT_ROOT / "已处理-患者首页PDF" / "2026_5_25_处理"
+DEFAULT_OUTPUT_DIR = PROJECT_ROOT / "数据处理结果区" / "07_Kimi视觉识别"
+DEFAULT_API_BASE = "https://api.moonshot.cn/v1"
+DEFAULT_MODEL = os.environ.get("KIMI_MODEL", "kimi-k2.6")
+
+
+PROMPT = """请读取这张住院病案首页图片，仅返回 JSON。
+优先核对这些区域：基本信息、地址联系人、入出院信息、诊断、手术及操作、质控签名、离院方式、费用。
+如果是手术及操作区域，请按以下列拆分：
+手术操作编码、手术操作日期、手术级别、手术操作名称、术者、I助、II助、切口愈合等级、麻醉方式、麻醉医师。
+无法确认的字段填空字符串，不要编造。输出结构：
+{
+  "page_summary": "",
+  "suspected_missing_fields": [],
+  "fields": {},
+  "discharge_diagnoses": [],
+  "operations": [],
+  "fees": {},
+  "notes": []
+}
+"""
+
+
+def parse_args() -> argparse.Namespace:
+    parser = argparse.ArgumentParser(description="使用 Kimi 视觉模型辅助识别患者首页图片")
+    parser.add_argument("--image-root", type=Path, default=DEFAULT_IMAGE_ROOT, help="PDF 转图片根目录")
+    parser.add_argument("--pdf-root", type=Path, default=DEFAULT_PDF_ROOT, help="没有图片目录时，从该 PDF 目录临时渲染页面")
+    parser.add_argument("-o", "--output-dir", type=Path, default=DEFAULT_OUTPUT_DIR, help="Kimi 识别结果输出目录")
+    parser.add_argument("--case", help="只处理指定病例目录名，例如 ZY010001672803_ori")
+    parser.add_argument("--page", help="只处理指定图片名，例如 page-001.png")
+    parser.add_argument("--pdf-page", type=int, help="从 PDF 临时渲染时，只处理指定页码，从 1 开始")
+    parser.add_argument("--dpi", type=int, default=160, help="PDF 临时渲染分辨率")
+    parser.add_argument("--model", default=DEFAULT_MODEL, help="Kimi 模型名")
+    parser.add_argument("--api-base", default=os.environ.get("MOONSHOT_API_BASE", DEFAULT_API_BASE), help="Moonshot API base URL")
+    parser.add_argument("--api-key", default=os.environ.get("MOONSHOT_API_KEY") or os.environ.get("KIMI_API_KEY"), help="API Key，也可用 MOONSHOT_API_KEY")
+    return parser.parse_args()
+
+
+def image_to_data_url(path: Path) -> str:
+    mime_type, _ = mimetypes.guess_type(path.name)
+    if mime_type not in {"image/png", "image/jpeg", "image/webp", "image/gif"}:
+        mime_type = "image/png"
+    encoded = base64.b64encode(path.read_bytes()).decode("utf-8")
+    return f"data:{mime_type};base64,{encoded}"
+
+
+def call_kimi(image_path: Path, args: argparse.Namespace) -> dict[str, Any]:
+    if not args.api_key:
+        raise RuntimeError("未设置 MOONSHOT_API_KEY/KIMI_API_KEY，无法调用 Kimi API。")
+    payload = {
+        "model": args.model,
+        "messages": [
+            {"role": "system", "content": "你是严谨的病案首页结构化抽取助手，只输出 JSON。"},
+            {
+                "role": "user",
+                "content": [
+                    {"type": "image_url", "image_url": {"url": image_to_data_url(image_path)}},
+                    {"type": "text", "text": PROMPT},
+                ],
+            },
+        ],
+    }
+    request = urllib.request.Request(
+        f"{args.api_base.rstrip('/')}/chat/completions",
+        data=json.dumps(payload, ensure_ascii=False).encode("utf-8"),
+        headers={
+            "Content-Type": "application/json",
+            "Authorization": f"Bearer {args.api_key}",
+        },
+        method="POST",
+    )
+    try:
+        with urllib.request.urlopen(request, timeout=180) as response:
+            data = json.loads(response.read().decode("utf-8"))
+    except urllib.error.HTTPError as exc:
+        detail = exc.read().decode("utf-8", errors="replace")
+        raise RuntimeError(f"Kimi API 返回错误 {exc.code}: {detail}") from exc
+
+    content = data["choices"][0]["message"]["content"]
+    parsed = parse_json_content(content)
+    return {
+        "image": str(image_path),
+        "model": args.model,
+        "recognized_at": datetime.now().isoformat(timespec="seconds"),
+        "raw_response": content,
+        "parsed": parsed,
+    }
+
+
+def parse_json_content(content: str) -> Any:
+    text = content.strip()
+    if text.startswith("```"):
+        text = text.strip("`")
+        if text.startswith("json"):
+            text = text[4:].strip()
+    try:
+        return json.loads(text)
+    except json.JSONDecodeError:
+        start = text.find("{")
+        end = text.rfind("}")
+        if start >= 0 and end > start:
+            return json.loads(text[start : end + 1])
+        return {"notes": ["Kimi 返回内容不是合法 JSON"], "text": content}
+
+
+def iter_images(args: argparse.Namespace) -> list[Path]:
+    root = args.image_root
+    if args.case:
+        case_dir = root / args.case
+        if args.page:
+            return [case_dir / args.page]
+        return sorted(case_dir.glob("page-*.*"))
+    images: list[Path] = []
+    for case_dir in sorted(path for path in root.iterdir() if path.is_dir()):
+        images.extend(sorted(case_dir.glob("page-*.*")))
+    return images
+
+
+def find_case_pdf(args: argparse.Namespace) -> Path | None:
+    if not args.case:
+        return None
+    stem = Path(args.case).stem
+    candidates = [
+        args.pdf_root / f"{stem}.pdf",
+        args.pdf_root / args.case,
+    ]
+    return next((path for path in candidates if path.exists() and path.is_file()), None)
+
+
+def render_pdf_pages(pdf_path: Path, output_root: Path, args: argparse.Namespace) -> list[Path]:
+    case_dir = output_root / pdf_path.stem
+    case_dir.mkdir(parents=True, exist_ok=True)
+    if args.pdf_page:
+        prefix = case_dir / f"page-{args.pdf_page:03d}"
+        command = [
+            "pdftoppm",
+            "-png",
+            "-r",
+            str(args.dpi),
+            "-f",
+            str(args.pdf_page),
+            "-singlefile",
+            str(pdf_path),
+            str(prefix),
+        ]
+    else:
+        command = ["pdftoppm", "-png", "-r", str(args.dpi), str(pdf_path), str(case_dir / "page")]
+    subprocess.run(command, check=True)
+    return sorted(case_dir.glob("page-*.png")) or sorted(case_dir.glob("page.png"))
+
+
+def main() -> int:
+    args = parse_args()
+    args.output_dir.mkdir(parents=True, exist_ok=True)
+    images = [path for path in iter_images(args) if path.exists() and path.is_file()]
+    with tempfile.TemporaryDirectory(prefix="frontpage_kimi_") as temp_dir:
+        if not images:
+            pdf_path = find_case_pdf(args)
+            if pdf_path:
+                images = render_pdf_pages(pdf_path, Path(temp_dir), args)
+        if not images:
+            raise SystemExit("未找到待识别图片，也未找到可临时渲染的 PDF。")
+
+        for image_path in images:
+            print(f"识别：{image_path}")
+            result = call_kimi(image_path, args)
+            out_path = args.output_dir / f"{image_path.parent.name}_{image_path.stem}.json"
+            out_path.write_text(json.dumps(result, ensure_ascii=False, indent=2), encoding="utf-8")
+            print(f"输出：{out_path}")
+    return 0
+
+
+if __name__ == "__main__":
+    try:
+        raise SystemExit(main())
+    except Exception as exc:  # noqa: BLE001
+        print(f"错误：{exc}", file=sys.stderr)
+        raise SystemExit(1)