fix:优化数据

2026-03-15 16:38:59 +08:00
parent a609f81a36
commit 3aa1a586e5
43 changed files with 14565 additions and 294 deletions
--- a/backend/app/services/cloud_doc_service.py
+++ b/backend/app/services/cloud_doc_service.py
@@ -0,0 +1,315 @@
+"""
+云文档集成：飞书、语雀、腾讯文档的文档创建/更新。
+统一以 Markdown 为中间格式，由各平台 API 写入。
+
+扩展建议：可增加「月度财务明细表」自动导出——每月在飞书/腾讯文档生成表格，
+插入当月发票等附件预览链接，供财务查看（需对接财务记录与附件列表）。
+"""
+from typing import Any, Dict, List, Tuple
+
+import httpx
+
+FEISHU_BASE = "https://open.feishu.cn"
+YUQUE_BASE = "https://www.yuque.com/api/v2"
+
+
+async def get_feishu_tenant_token(app_id: str, app_secret: str) -> str:
+    """获取飞书 tenant_access_token。"""
+    async with httpx.AsyncClient() as client:
+        r = await client.post(
+            f"{FEISHU_BASE}/open-apis/auth/v3/tenant_access_token/internal",
+            json={"app_id": app_id, "app_secret": app_secret},
+            timeout=10.0,
+        )
+        r.raise_for_status()
+        data = r.json()
+        if data.get("code") != 0:
+            raise RuntimeError(data.get("msg", "飞书鉴权失败"))
+        return data["tenant_access_token"]
+
+
+def _feishu_text_block_elements(md: str) -> List[Dict[str, Any]]:
+    """将 Markdown 转为飞书文本块 elements（按行拆成 textRun，简单实现）。"""
+    elements: List[Dict[str, Any]] = []
+    for line in md.split("\n"):
+        line = line.rstrip()
+        if not line:
+            elements.append({"type": "textRun", "text_run": {"text": "\n"}})
+        else:
+            elements.append({"type": "textRun", "text_run": {"text": line + "\n"}})
+    if not elements:
+        elements.append({"type": "textRun", "text_run": {"text": " "}})
+    return elements
+
+
+async def feishu_create_doc(
+    token: str, title: str, body_md: str, folder_token: str = ""
+) -> Tuple[str, str]:
+    """
+    创建飞书文档并写入内容。返回 (document_id, url)。
+    使用 docx/v1：创建文档后向根块下添加子块写入 Markdown 文本。
+    """
+    async with httpx.AsyncClient() as client:
+        headers = {"Authorization": f"Bearer {token}"}
+        # 1. 创建文档
+        create_body: Dict[str, Any] = {"title": title[:50] or "未命名文档"}
+        if folder_token:
+            create_body["folder_token"] = folder_token
+        r = await client.post(
+            f"{FEISHU_BASE}/open-apis/docx/v1/documents",
+            headers=headers,
+            json=create_body,
+            timeout=15.0,
+        )
+        r.raise_for_status()
+        data = r.json()
+        if data.get("code") != 0:
+            raise RuntimeError(data.get("msg", "飞书创建文档失败"))
+        doc = data.get("data", {})
+        document_id = doc.get("document", {}).get("document_id")
+        if not document_id:
+            raise RuntimeError("飞书未返回 document_id")
+        url = doc.get("document", {}).get("url", "")
+        # 2. 根块 ID 即 document_id（飞书约定）
+        block_id = document_id
+        # 3. 添加子块（内容）
+        elements = _feishu_text_block_elements(body_md)
+        # 单块有长度限制，分批写入多块
+        chunk_size = 3000
+        for i in range(0, len(elements), chunk_size):
+            chunk = elements[i : i + chunk_size]
+            body_json = {"children": [{"block_type": "text", "text": {"elements": chunk}}], "index": -1}
+            r3 = await client.post(
+                f"{FEISHU_BASE}/open-apis/docx/v1/documents/{document_id}/blocks/{block_id}/children",
+                headers=headers,
+                json=body_json,
+                timeout=15.0,
+            )
+            r3.raise_for_status()
+            res = r3.json()
+            if res.get("code") != 0:
+                raise RuntimeError(res.get("msg", "飞书写入块失败"))
+            # 下一批挂在刚创建的块下
+            new_items = res.get("data", {}).get("children", [])
+            if new_items:
+                block_id = new_items[0].get("block_id", block_id)
+        return document_id, url or f"https://feishu.cn/docx/{document_id}"
+
+
+async def feishu_update_doc(token: str, document_id: str, body_md: str) -> str:
+    """
+    更新飞书文档内容：获取现有块并批量更新首个文本块，或追加新块。
+    返回文档 URL。
+    """
+    async with httpx.AsyncClient() as client:
+        headers = {"Authorization": f"Bearer {token}"}
+        r = await client.get(
+            f"{FEISHU_BASE}/open-apis/docx/v1/documents/{document_id}/blocks",
+            headers=headers,
+            params={"document_id": document_id},
+            timeout=10.0,
+        )
+        r.raise_for_status()
+        data = r.json()
+        if data.get("code") != 0:
+            raise RuntimeError(data.get("msg", "飞书获取块失败"))
+        items = data.get("data", {}).get("items", [])
+        elements = _feishu_text_block_elements(body_md)
+        if items:
+            first_id = items[0].get("block_id")
+            if first_id:
+                # 批量更新：只更新第一个块的内容
+                update_body = {
+                    "requests": [
+                        {
+                            "request_type": "blockUpdate",
+                            "block_id": first_id,
+                            "update_text": {"elements": elements},
+                        }
+                    ]
+                }
+                r2 = await client.patch(
+                    f"{FEISHU_BASE}/open-apis/docx/v1/documents/{document_id}/blocks/batch_update",
+                    headers=headers,
+                    json=update_body,
+                    timeout=15.0,
+                )
+                r2.raise_for_status()
+                if r2.json().get("code") != 0:
+                    # 若 PATCH 不支持该块类型，则追加新块
+                    pass
+                else:
+                    return f"https://feishu.cn/docx/{document_id}"
+        # 无块或更新失败：在根下追加子块
+        block_id = document_id
+        for i in range(0, len(elements), 3000):
+            chunk = elements[i : i + 3000]
+            body_json = {"children": [{"block_type": "text", "text": {"elements": chunk}}], "index": -1}
+            r3 = await client.post(
+                f"{FEISHU_BASE}/open-apis/docx/v1/documents/{document_id}/blocks/{block_id}/children",
+                headers=headers,
+                json=body_json,
+                timeout=15.0,
+            )
+            r3.raise_for_status()
+            res = r3.json()
+            if res.get("data", {}).get("children"):
+                block_id = res["data"]["children"][0].get("block_id", block_id)
+        return f"https://feishu.cn/docx/{document_id}"
+
+
+# --------------- 语雀 ---------------
+
+
+async def yuque_create_doc(
+    token: str, repo_id_or_namespace: str, title: str, body_md: str
+) -> Tuple[str, str]:
+    """
+    在语雀知识库创建文档。repo_id_or_namespace 可为 repo_id 或 namespace（如 user/repo）。
+    返回 (doc_id, url)。
+    """
+    async with httpx.AsyncClient() as client:
+        headers = {
+            "X-Auth-Token": token,
+            "Content-Type": "application/json",
+            "User-Agent": "OpsCore-CloudDoc/1.0",
+        }
+        # 若为 namespace 需先解析为 repo_id（语雀 API 创建文档用 repo_id）
+        repo_id = repo_id_or_namespace
+        if "/" in repo_id_or_namespace:
+            r_repo = await client.get(
+                f"{YUQUE_BASE}/repos/{repo_id_or_namespace}",
+                headers=headers,
+                timeout=10.0,
+            )
+            if r_repo.status_code == 200 and r_repo.json().get("data"):
+                repo_id = str(r_repo.json()["data"]["id"])
+        r = await client.post(
+            f"{YUQUE_BASE}/repos/{repo_id}/docs",
+            headers=headers,
+            json={
+                "title": title[:100] or "未命名",
+                "body": body_md,
+                "format": "markdown",
+            },
+            timeout=15.0,
+        )
+        r.raise_for_status()
+        data = r.json()
+        doc = data.get("data", {})
+        doc_id = str(doc.get("id", ""))
+        url = doc.get("url", "")
+        if not url and doc.get("slug"):
+            url = f"https://www.yuque.com/{doc.get('namespace', '').replace('/', '/')}/{doc.get('slug', '')}"
+        return doc_id, url or ""
+
+
+async def yuque_update_doc(
+    token: str, repo_id_or_namespace: str, doc_id: str, title: str, body_md: str
+) -> str:
+    """更新语雀文档。返回文档 URL。"""
+    async with httpx.AsyncClient() as client:
+        headers = {
+            "X-Auth-Token": token,
+            "Content-Type": "application/json",
+            "User-Agent": "OpsCore-CloudDoc/1.0",
+        }
+        r = await client.put(
+            f"{YUQUE_BASE}/repos/{repo_id_or_namespace}/docs/{doc_id}",
+            headers=headers,
+            json={
+                "title": title[:100] or "未命名",
+                "body": body_md,
+                "format": "markdown",
+            },
+            timeout=15.0,
+        )
+        r.raise_for_status()
+        data = r.json()
+        doc = data.get("data", {})
+        return doc.get("url", "") or f"https://www.yuque.com/docs/{doc_id}"
+
+
+async def yuque_list_docs(token: str, repo_id_or_namespace: str) -> List[Dict[str, Any]]:
+    """获取知识库文档列表。"""
+    async with httpx.AsyncClient() as client:
+        headers = {
+            "X-Auth-Token": token,
+            "User-Agent": "OpsCore-CloudDoc/1.0",
+        }
+        r = await client.get(
+            f"{YUQUE_BASE}/repos/{repo_id_or_namespace}/docs",
+            headers=headers,
+            timeout=10.0,
+        )
+        r.raise_for_status()
+        data = r.json()
+        return data.get("data", [])
+
+
+# --------------- 腾讯文档（占位） ---------------
+
+
+async def tencent_create_doc(client_id: str, client_secret: str, title: str, body_md: str) -> Tuple[str, str]:
+    """
+    腾讯文档需 OAuth 用户授权与文件创建 API，此处返回占位。
+    正式接入需在腾讯开放平台创建应用并走 OAuth 流程。
+    """
+    raise RuntimeError(
+        "腾讯文档 Open API 需在开放平台配置 OAuth 并获取用户授权；当前版本请先用飞书或语雀推送。"
+    )
+
+
+# --------------- 统一入口 ---------------
+
+
+class CloudDocManager:
+    """统一封装：读取配置并执行创建/更新，支持增量（有 cloud_doc_id 则更新）。"""
+
+    def __init__(self, credentials: Dict[str, Dict[str, str]]):
+        self.credentials = credentials
+
+    async def push_markdown(
+        self,
+        platform: str,
+        title: str,
+        body_md: str,
+        existing_doc_id: str | None = None,
+        extra: Dict[str, str] | None = None,
+    ) -> Tuple[str, str]:
+        """
+        将 Markdown 推送到指定平台。若 existing_doc_id 存在则更新，否则创建。
+        返回 (cloud_doc_id, url)。
+        extra: 平台相关参数，如 yuque 的 default_repo。
+        """
+        extra = extra or {}
+        if platform == "feishu":
+            cred = self.credentials.get("feishu") or {}
+            app_id = (cred.get("app_id") or "").strip()
+            app_secret = (cred.get("app_secret") or "").strip()
+            if not app_id or not app_secret:
+                raise RuntimeError("请先在设置中配置飞书 App ID 与 App Secret")
+            token = await get_feishu_tenant_token(app_id, app_secret)
+            if existing_doc_id:
+                url = await feishu_update_doc(token, existing_doc_id, body_md)
+                return existing_doc_id, url
+            return await feishu_create_doc(token, title, body_md)
+
+        if platform == "yuque":
+            cred = self.credentials.get("yuque") or {}
+            token = (cred.get("token") or "").strip()
+            default_repo = (cred.get("default_repo") or extra.get("repo") or "").strip()
+            if not token:
+                raise RuntimeError("请先在设置中配置语雀 Personal Access Token")
+            if not default_repo:
+                raise RuntimeError("请先在设置中配置语雀默认知识库（namespace，如 user/repo）")
+            if existing_doc_id:
+                url = await yuque_update_doc(token, default_repo, existing_doc_id, title, body_md)
+                return existing_doc_id, url
+            return await yuque_create_doc(token, default_repo, title, body_md)
+
+        if platform == "tencent":
+            await tencent_create_doc("", "", title, body_md)
+            return "", ""
+
+        raise RuntimeError(f"不支持的平台: {platform}")