Fix Feishu wiki child node retrieval

ai_knowledge_base_v2/apps/backend/app/services/feishu_service.py

@@ -80,9 +80,17 @@ def _feishu_get(path: str, params: dict | None = None, config: FeishuRetrievalCo
     url = f"{FEISHU_OPEN_API}{path}"
     headers = {"Authorization": f"Bearer {token}"}
 
-    resp = httpx.get(url, headers=headers, params=params, timeout=config.timeout_seconds)
-    resp.raise_for_status()
-    data = resp.json()
+    try:
+        resp = httpx.get(url, headers=headers, params=params, timeout=config.timeout_seconds)
+        resp.raise_for_status()
+        data = resp.json()
+    except httpx.HTTPStatusError as exc:
+        raise ExternalServiceError(
+            f"飞书HTTP请求失败 [{path}]：{exc.response.status_code} {_response_error_detail(exc.response)}",
+            provider="feishu",
+        ) from exc
+    except httpx.HTTPError as exc:
+        raise ExternalServiceError(f"飞书网络请求失败 [{path}]：{exc}", provider="feishu") from exc
 
     if data.get("code") != 0:
         raise ExternalServiceError(
@@ -100,9 +108,17 @@ def _feishu_post(path: str, payload: dict | None = None, config: FeishuRetrieval
     url = f"{FEISHU_OPEN_API}{path}"
     headers = {"Authorization": f"Bearer {token}"}
 
-    resp = httpx.post(url, headers=headers, json=payload, timeout=config.timeout_seconds)
-    resp.raise_for_status()
-    data = resp.json()
+    try:
+        resp = httpx.post(url, headers=headers, json=payload, timeout=config.timeout_seconds)
+        resp.raise_for_status()
+        data = resp.json()
+    except httpx.HTTPStatusError as exc:
+        raise ExternalServiceError(
+            f"飞书HTTP请求失败 [{path}]：{exc.response.status_code} {_response_error_detail(exc.response)}",
+            provider="feishu",
+        ) from exc
+    except httpx.HTTPError as exc:
+        raise ExternalServiceError(f"飞书网络请求失败 [{path}]：{exc}", provider="feishu") from exc
 
     if data.get("code") != 0:
         raise ExternalServiceError(
@@ -112,6 +128,15 @@ def _feishu_post(path: str, payload: dict | None = None, config: FeishuRetrieval
     return data
 
 
+def _response_error_detail(response: httpx.Response) -> str:
+    try:
+        body = response.json()
+    except ValueError:
+        return response.text[:200]
+    message = body.get("msg") or body.get("message") or body.get("error") or body
+    return str(message)[:300]
+
+
 def _extract_text(elements: list[dict]) -> str:
     """从 elements 列表中提取纯文本"""
     texts = []
@@ -208,8 +233,16 @@ def _get_wiki_node_info(node_token: str, config: FeishuRetrievalConfig) -> dict:
     }
 
 
-def _get_wiki_children(node_token: str, config: FeishuRetrievalConfig, page_size: int = 50) -> list[dict]:
+def _get_wiki_children(
+    space_id: str,
+    node_token: str,
+    config: FeishuRetrievalConfig,
+    page_size: int = 50,
+) -> list[dict]:
     """获取 Wiki 节点的子节点列表"""
+    if not space_id:
+        raise ExternalServiceError("飞书知识库缺少 SpaceID，无法读取子节点", provider="feishu")
+
     all_children = []
     page_token = ""
     while True:
@@ -220,7 +253,7 @@ def _get_wiki_children(node_token: str, config: FeishuRetrievalConfig, page_size
         if page_token:
             params["page_token"] = page_token
 
-        data = _feishu_get("/open-apis/wiki/v2/spaces/nodes/get_child_nodes", params=params, config=config)
+        data = _feishu_get(f"/open-apis/wiki/v2/spaces/{space_id}/nodes", params=params, config=config)
         children = data.get("data", {}).get("items", [])
         all_children.extend(children)
 
@@ -231,7 +264,7 @@ def _get_wiki_children(node_token: str, config: FeishuRetrievalConfig, page_size
     return all_children
 
 
-def _get_space_node_tree(space_id: str) -> list[dict]:
+def _get_space_node_tree(space_id: str, config: FeishuRetrievalConfig | None = None) -> list[dict]:
     """获取知识空间下所有根节点"""
     all_nodes = []
     page_token = ""
@@ -243,7 +276,7 @@ def _get_space_node_tree(space_id: str) -> list[dict]:
         if page_token:
             params["page_token"] = page_token
 
-        data = _feishu_get("/open-apis/wiki/v2/spaces/nodes", params=params)
+        data = _feishu_get(f"/open-apis/wiki/v2/spaces/{space_id}/nodes", params=params, config=config)
         items = data.get("data", {}).get("items", [])
         all_nodes.extend(items)
 
@@ -398,7 +431,7 @@ class FeishuKnowledgeService:
             return [(title, content, source_url)]
 
         # 如果节点是文件夹，获取子节点
-        children = cls._get_wiki_children_with_cache(node_id, config)
+        children = cls._get_wiki_children_with_cache(space_id, node_id, config)
 
         documents = []
         for child in children:
@@ -457,9 +490,14 @@ class FeishuKnowledgeService:
         return content
 
     @classmethod
-    def _get_wiki_children_with_cache(cls, node_token: str, config: FeishuRetrievalConfig) -> list[dict]:
+    def _get_wiki_children_with_cache(
+        cls,
+        space_id: str,
+        node_token: str,
+        config: FeishuRetrievalConfig,
+    ) -> list[dict]:
         """带缓存获取子节点"""
-        cache_key = f"{config.app_id}:children:{node_token}"
+        cache_key = f"{config.app_id}:children:{space_id}:{node_token}"
         now = time.time()
 
         if cache_key in cls._content_cache:
@@ -467,7 +505,7 @@ class FeishuKnowledgeService:
             if now - ts < cls._CACHE_TTL:
                 return eval(cached_data)  # noqa: S307
 
-        children = _get_wiki_children(node_token, config)
+        children = _get_wiki_children(space_id, node_token, config)
         cls._content_cache[cache_key] = (str(children), now)
         return children