sophhub 0.2.4 → 0.4.0

package/skills/notes-hub-assistant/src/scripts/openclaw_meeting_minutes.py

@@ -0,0 +1,473 @@
+#!/usr/bin/env python3
+"""OpenClaw downstream meeting-minutes orchestrator.
+
+This script implements the reusable half of the OpenClaw flow:
+
+minute_token -> lark-cli minutes/vc/docs -> Markdown -> JSON result
+
+It intentionally does not upload audio into Feishu Minutes. That step must be
+handled by an upstream ingest adapter which returns a minute_token.
+"""
+
+from __future__ import annotations
+
+import argparse
+import json
+import os
+import subprocess
+import sys
+from pathlib import Path
+from typing import Any, Dict, Optional
+
+
+REQUIRED_SCOPES = (
+    "vc:note:read "
+    "minutes:minutes:readonly "
+    "minutes:minutes.artifacts:read "
+    "minutes:minutes.transcript:export "
+    "docx:document:create"
+)
+
+
+class CLIError(Exception):
+    def __init__(self, message: str, code: str = "cli_error", detail: Any = None):
+        super().__init__(message)
+        self.code = code
+        self.detail = detail
+
+
+def parse_args() -> argparse.Namespace:
+    parser = argparse.ArgumentParser(
+        description="Build meeting minutes Markdown from an existing minute_token."
+    )
+    parser.add_argument("--minute-token", required=True, help="Feishu minute_token")
+    parser.add_argument("--job-id", required=True, help="OpenClaw job ID")
+    parser.add_argument("--title", help="Override meeting title")
+    parser.add_argument("--folder-token", help="Target Feishu folder token")
+    parser.add_argument(
+        "--output-dir",
+        default=".openclaw-meeting-minutes",
+        help="Directory for transcript/artifact files",
+    )
+    parser.add_argument(
+        "--config-dir",
+        help="Override LARKSUITE_CLI_CONFIG_DIR for per-user isolation",
+    )
+    parser.add_argument(
+        "--lark-cli-prefix",
+        action="append",
+        default=[],
+        help="repeatable argv prefix, e.g. --lark-cli-prefix npx --lark-cli-prefix -y ...",
+    )
+    parser.add_argument(
+        "--lark-cli-bin",
+        default="lark-cli",
+        help="(legacy) Path to the lark-cli executable; ignored if --lark-cli-prefix is provided",
+    )
+    return parser.parse_args()
+
+
+def build_env(config_dir: Optional[str]) -> Dict[str, str]:
+    env = os.environ.copy()
+    if config_dir:
+        env["LARKSUITE_CLI_CONFIG_DIR"] = config_dir
+    return env
+
+
+def run_cli(
+    cli_prefix: list[str],
+    env: Dict[str, str],
+    args: list[str],
+    check: bool = True,
+    cwd: Optional[Path] = None,
+) -> Any:
+    proc = subprocess.run(
+        cli_prefix + args,
+        capture_output=True,
+        text=True,
+        env=env,
+        cwd=str(cwd) if cwd else None,
+    )
+    stdout = proc.stdout.strip()
+    stderr = proc.stderr.strip()
+
+    parsed: Any = None
+    if stdout:
+        try:
+            parsed = json.loads(stdout)
+        except json.JSONDecodeError as exc:
+            raise CLIError(
+                f"lark-cli returned non-JSON stdout: {stdout[:400]}",
+                code="invalid_cli_output",
+                detail={"stderr": stderr, "args": args},
+            ) from exc
+
+    if check and proc.returncode != 0:
+        error_detail = parsed if isinstance(parsed, dict) else {"stderr": stderr}
+        raise CLIError(
+            f"lark-cli command failed: {' '.join(args)}",
+            code="lark_cli_failed",
+            detail=error_detail,
+        )
+
+    return parsed
+
+
+def unwrap_payload(payload: Any) -> Any:
+    if isinstance(payload, dict) and "ok" in payload:
+        if not payload.get("ok", False):
+            error = payload.get("error") or {}
+            raise CLIError(
+                error.get("message", "lark-cli returned an error envelope"),
+                code=error.get("type", "lark_error"),
+                detail=payload,
+            )
+        return payload.get("data")
+    return payload
+
+
+def ensure_auth(cli_prefix: list[str], env: Dict[str, str]) -> None:
+    run_cli(cli_prefix, env, ["auth", "check", "--scope", REQUIRED_SCOPES], check=True)
+
+
+def minute_info(
+    cli_prefix: list[str], env: Dict[str, str], minute_token: str
+) -> Dict[str, Any]:
+    payload = run_cli(
+        cli_prefix,
+        env,
+        [
+            "api",
+            "GET",
+            f"/open-apis/minutes/v1/minutes/{minute_token}",
+            "--as",
+            "bot",
+            "--format",
+            "json",
+        ],
+    )
+    data = unwrap_payload(payload)
+    if isinstance(data, dict) and "data" in data and isinstance(data["data"], dict):
+        data = data["data"]
+    if not isinstance(data, dict):
+        raise CLIError("unexpected minutes.get payload", code="invalid_minutes_payload")
+    minute = data.get("minute")
+    if not isinstance(minute, dict):
+        raise CLIError("minutes.get missing minute object", code="missing_minute")
+    return minute
+
+
+def notes_info(
+    cli_prefix: list[str], env: Dict[str, str], minute_token: str, output_dir: Path
+) -> Dict[str, Any]:
+    output_dir = output_dir.resolve()
+    payload = run_cli(
+        cli_prefix,
+        env,
+        [
+            "vc",
+            "+notes",
+            "--minute-tokens",
+            minute_token,
+            "--output-dir",
+            "vc-notes-artifacts",
+            "--format",
+            "json",
+        ],
+        cwd=output_dir,
+    )
+    data = unwrap_payload(payload)
+    if not isinstance(data, dict):
+        raise CLIError("unexpected vc +notes payload", code="invalid_notes_payload")
+    notes = data.get("notes")
+    if not isinstance(notes, list) or not notes:
+        raise CLIError("vc +notes returned no notes", code="missing_notes")
+    first = notes[0]
+    if not isinstance(first, dict):
+        raise CLIError("invalid note item shape", code="invalid_note_item")
+    if first.get("error"):
+        raise CLIError(str(first["error"]), code="note_query_failed", detail=first)
+    artifacts = first.get("artifacts")
+    if isinstance(artifacts, dict):
+        transcript_file = artifacts.get("transcript_file")
+        if isinstance(transcript_file, str) and transcript_file:
+            transcript_path = Path(transcript_file)
+            if not transcript_path.is_absolute():
+                artifacts["transcript_file"] = str(
+                    (output_dir / transcript_path).resolve()
+                )
+    return first
+
+
+def read_text(path: Optional[str]) -> str:
+    if not path:
+        return ""
+    p = Path(path)
+    if not p.exists():
+        return ""
+    return p.read_text(encoding="utf-8").strip()
+
+
+def transcript_excerpt(text: str, limit: int = 2000) -> str:
+    if not text:
+        return ""
+    if len(text) <= limit:
+        return text
+    return text[:limit].rstrip() + "\n..."
+
+
+def format_inline(value: Any) -> str:
+    if value in (None, "", [], {}):
+        return "暂无 AI 产物"
+    if isinstance(value, str):
+        return value.strip() or "暂无 AI 产物"
+    if isinstance(value, list):
+        items = [format_list_item(item) for item in value]
+        items = [item for item in items if item]
+        return "\n".join(items) if items else "暂无 AI 产物"
+    if isinstance(value, dict):
+        return "```json\n" + json.dumps(value, ensure_ascii=False, indent=2) + "\n```"
+    return str(value)
+
+
+def format_list_item(item: Any) -> str:
+    if isinstance(item, str):
+        return f"- {item}"
+    if isinstance(item, dict):
+        preferred = [
+            item.get("title"),
+            item.get("summary"),
+            item.get("content"),
+            item.get("text"),
+            item.get("todo"),
+            item.get("speaker"),
+        ]
+        text = next((str(v).strip() for v in preferred if v), "")
+        if text:
+            extras: list[str] = []
+            for key in ("assignee", "owner", "deadline", "timestamp", "time"):
+                if item.get(key):
+                    extras.append(f"{key}: {item[key]}")
+            if extras:
+                return f"- {text} ({', '.join(extras)})"
+            return f"- {text}"
+        return "- " + json.dumps(item, ensure_ascii=False, sort_keys=True)
+    return f"- {item}"
+
+
+def build_markdown(
+    title: str,
+    minute_token: str,
+    minutes_url: str,
+    note: Dict[str, Any],
+    transcript_text: str,
+    doc_url: str = "",
+) -> str:
+    artifacts = note.get("artifacts") or {}
+    if not isinstance(artifacts, dict):
+        artifacts = {}
+    lines = [
+        f"# {title}",
+        "",
+        "## 会议信息",
+        "",
+        f"- 标题：{title}",
+        f"- minute_token：`{minute_token}`",
+        f"- 妙记链接：{minutes_url or '暂无'}",
+        f"- 纪要文档 Token：`{note.get('note_doc_token') or '暂无'}`",
+        f"- 逐字稿文档 Token：`{note.get('verbatim_doc_token') or '暂无'}`",
+        f"- 逐字稿文件：{artifacts.get('transcript_file') or '暂无'}",
+        "",
+        "## 摘要",
+        "",
+        format_inline(artifacts.get("summary")),
+        "",
+        "## 待办",
+        "",
+        format_inline(artifacts.get("todos")),
+        "",
+        "## 章节",
+        "",
+        format_inline(artifacts.get("chapters")),
+        "",
+        "## 逐字稿",
+        "",
+        transcript_text or "暂无 AI 产物",
+        "",
+        "## 原始飞书产物",
+        "",
+        f"- 妙记链接：{minutes_url or '暂无'}",
+        f"- 飞书文档链接：{doc_url or '待创建'}",
+        f"- 纪要文档 Token：`{note.get('note_doc_token') or '暂无'}`",
+        f"- 逐字稿文档 Token：`{note.get('verbatim_doc_token') or '暂无'}`",
+    ]
+    return "\n".join(lines).strip() + "\n"
+
+
+def create_doc(
+    cli_prefix: list[str],
+    env: Dict[str, str],
+    title: str,
+    markdown: str,
+    folder_token: Optional[str],
+) -> Dict[str, Any]:
+    args = [
+        "docs",
+        "+create",
+        "--as",
+        "bot",
+        "--title",
+        title,
+        "--markdown",
+        markdown,
+    ]
+    if folder_token:
+        args.extend(["--folder-token", folder_token])
+    payload = run_cli(cli_prefix, env, args)
+    data = unwrap_payload(payload)
+    if not isinstance(data, dict):
+        raise CLIError("unexpected docs +create payload", code="invalid_doc_payload")
+    return data
+
+
+def partial_result(
+    job_id: str,
+    title: str,
+    minute_token: str,
+    minutes_url: str,
+    markdown: str,
+    transcript_file: str,
+    transcript_text: str,
+    error: CLIError,
+) -> Dict[str, Any]:
+    return {
+        "job_id": job_id,
+        "status": "partial",
+        "title": title,
+        "minute_token": minute_token,
+        "minutes_url": minutes_url,
+        "markdown": markdown,
+        "feishu_doc_url": "",
+        "feishu_doc_token": "",
+        "transcript_file": transcript_file,
+        "transcript_excerpt": transcript_excerpt(transcript_text),
+        "error_code": error.code,
+        "error_message": str(error),
+    }
+
+
+def failed_result(job_id: str, minute_token: str, error: CLIError) -> Dict[str, Any]:
+    return {
+        "job_id": job_id,
+        "status": "failed",
+        "title": "",
+        "minute_token": minute_token,
+        "minutes_url": "",
+        "markdown": "",
+        "feishu_doc_url": "",
+        "feishu_doc_token": "",
+        "transcript_file": "",
+        "transcript_excerpt": "",
+        "error_code": error.code,
+        "error_message": str(error),
+    }
+
+
+def main() -> int:
+    args = parse_args()
+    env = build_env(args.config_dir)
+    from _resolve_lark_cli import as_prefix
+
+    cli_prefix = as_prefix(args.lark_cli_bin, args.lark_cli_prefix)
+    base_output_dir = Path(args.output_dir).resolve()
+    job_dir = base_output_dir / args.job_id
+    job_dir.mkdir(parents=True, exist_ok=True)
+
+    try:
+        ensure_auth(cli_prefix, env)
+        minute = minute_info(cli_prefix, env, args.minute_token)
+        note = notes_info(cli_prefix, env, args.minute_token, job_dir)
+    except CLIError as exc:
+        json.dump(
+            failed_result(args.job_id, args.minute_token, exc),
+            sys.stdout,
+            ensure_ascii=False,
+            indent=2,
+        )
+        sys.stdout.write("\n")
+        return 0
+
+    title = args.title or str(minute.get("title") or "会议纪要")
+    minutes_url = str(minute.get("url") or "")
+    artifacts = note.get("artifacts") or {}
+    transcript_file = ""
+    if isinstance(artifacts, dict):
+        transcript_file = str(artifacts.get("transcript_file") or "")
+    transcript_text = read_text(transcript_file)
+
+    markdown = build_markdown(
+        title=title,
+        minute_token=args.minute_token,
+        minutes_url=minutes_url,
+        note=note,
+        transcript_text=transcript_text,
+    )
+
+    try:
+        doc = create_doc(
+            cli_prefix,
+            env,
+            title,
+            markdown,
+            args.folder_token,
+        )
+    except CLIError as exc:
+        json.dump(
+            partial_result(
+                job_id=args.job_id,
+                title=title,
+                minute_token=args.minute_token,
+                minutes_url=minutes_url,
+                markdown=markdown,
+                transcript_file=transcript_file,
+                transcript_text=transcript_text,
+                error=exc,
+            ),
+            sys.stdout,
+            ensure_ascii=False,
+            indent=2,
+        )
+        sys.stdout.write("\n")
+        return 0
+
+    doc_url = str(doc.get("doc_url") or "")
+    doc_token = str(doc.get("doc_id") or "")
+    final_markdown = build_markdown(
+        title=title,
+        minute_token=args.minute_token,
+        minutes_url=minutes_url,
+        note=note,
+        transcript_text=transcript_text,
+        doc_url=doc_url,
+    )
+    result = {
+        "job_id": args.job_id,
+        "status": "succeeded",
+        "title": title,
+        "minute_token": args.minute_token,
+        "minutes_url": minutes_url,
+        "markdown": final_markdown,
+        "feishu_doc_url": doc_url,
+        "feishu_doc_token": doc_token,
+        "transcript_file": transcript_file,
+        "transcript_excerpt": transcript_excerpt(transcript_text),
+        "error_code": "",
+        "error_message": "",
+    }
+    json.dump(result, sys.stdout, ensure_ascii=False, indent=2)
+    sys.stdout.write("\n")
+    return 0
+
+
+if __name__ == "__main__":
+    raise SystemExit(main())