lami/lami-base-v1/scripts/extract_qiwei_api.py

import re
import json
from pathlib import Path

ROOT = Path(__file__).resolve().parents[1]
MD_DIR = ROOT / "doc" / "文档" / "QiWe开放平台文档" / "md"
README = ROOT / "doc" / "文档" / "QiWe开放平台文档" / "README.md"

API_FILES = {
    "API-01": [],
    "API-02": [
        "创建设备（步骤1）.md",
        "二维码-获取（步骤2）.md",
        "二维码状态-检测（步骤3）.md",
        "二维码-code验证（步骤4）.md",
    ],
    "API-03": ["设置回调地址.md"],
    "API-04": ["回调结构说明.md"],
    "API-05": ["用户状态.md"],
    "API-06": ["同步历史消息分页.md"],
    "API-07": ["群分页.md"],
    "API-08": ["群详情-批量.md"],
    "API-09": ["群成员变动查询.md"],
    "API-10": ["创建群.md"],
    "API-11": ["修改群公告.md"],
    "API-12": ["群消息置顶-列表.md"],
    "API-13": ["群消息置顶-添加.md"],
    "API-14": ["发送纯文本消息.md"],
    "API-15": ["群发消息.md"],
    "API-16": ["群发消息-状态查询.md"],
    "API-17": ["外部联系人分页.md"],
    "API-18": ["联系人详情-批量.md"],
    "API-19": ["客户标签-增删.md"],
    "API-20": ["添加群成员好友.md"],
    "API-21": ["企微文件下载.md"],
}


def extract(path: Path) -> dict:
    if not path.exists():
        return {}
    text = path.read_text(encoding="utf-8")
    out: dict = {}
    mm = re.search(r'"method"\s*:\s*"([^"]+)"', text)
    if mm:
        out["method"] = mm.group(1)
    blocks = re.findall(r"```\s*\n\s*(\{[\s\S]*?\})\s*\n\s*```", text)
    for b in blocks:
        if '"method"' in b and "params" in b:
            out["request"] = b.strip()
            break
    for b in blocks:
        if '"code"' in b and "data" in b and "method" not in b:
            out["response"] = b.strip()
            break
  # notes from bullet lines
    notes = [
        ln.strip("•").strip()
        for ln in text.splitlines()
        if ln.strip().startswith("•") or ln.strip().startswith("-")
    ]
    if notes:
        out["notes"] = notes[:5]
    return out


def main():
    index = {}
    for line in README.read_text(encoding="utf-8").splitlines():
        m = re.match(r"- \[(.+?)\]\(md/(.+?)\) - (https://.+)", line)
        if m:
            index[m.group(2)] = {"title": m.group(1), "url": m.group(3)}

    result = {}
    for api, files in API_FILES.items():
        result[api] = {"files": []}
        for f in files:
            p = MD_DIR / f
            info = extract(p)
            info["file"] = f
            info["local"] = f"./文档/QiWe开放平台文档/md/{f}"
            if f in index:
                info["title"] = index[f]["title"]
                info["url"] = index[f]["url"]
            result[api]["files"].append(info)
            if info.get("method") and "method" not in result[api]:
                result[api]["method"] = info["method"]

    out = ROOT / "doc" / "文档" / "qiwei_api_extracted.json"
    out.write_text(json.dumps(result, ensure_ascii=False, indent=2), encoding="utf-8")
    print("wrote", out)
    for api, data in result.items():
        print(api, data.get("method", "-"), len(data["files"]))


if __name__ == "__main__":
    main()