6

backend/app/services/kb_public_service.py

@@ -1,14 +1,17 @@
 """公共知识库服务（Phase 10/11/12 统一数据来源）。
 
 HTML/MD/TXT/JSON/搜索全部通过此 Service 获取数据，不各自写查询逻辑。
+支持目录树结构和分类过滤。
 """
 
 from app.core.errors import NotFoundError
 from app.core.security import hash_token
 from app.models.document import Document
+from app.models.document_category import DocumentCategory
 from app.models.knowledge_base import KnowledgeBase
 from app.repositories.doc_repo import DocumentRepository
 from app.repositories.kb_repo import KnowledgeBaseRepository
+from sqlalchemy import func, select
 from sqlalchemy.orm import Session
 
 
@@ -26,12 +29,55 @@ class KbPublicService:
             raise NotFoundError("知识库不存在。")
         return kb
 
+    def get_category_tree(self, kb: KnowledgeBase) -> list[dict]:
+        """获取目录树（含文档数量）。"""
+        stmt = (
+            select(DocumentCategory)
+            .where(DocumentCategory.knowledge_base_id == kb.id)
+            .order_by(DocumentCategory.sort_order, DocumentCategory.name)
+        )
+        all_cats = list(self._session.scalars(stmt).all())
+
+        # 获取每个分类的文档数量
+        count_stmt = (
+            select(Document.category_id, func.count())
+            .where(
+                Document.knowledge_base_id == kb.id,
+                Document.status != "DELETED",
+            )
+            .group_by(Document.category_id)
+        )
+        doc_count_map = dict(self._session.execute(count_stmt).all())
+
+        # 构建树
+        cat_map = {}
+        for cat in all_cats:
+            cat_map[cat.id] = {
+                "id": cat.id,
+                "name": cat.name,
+                "path": cat.path or "/",
+                "is_folder": getattr(cat, 'is_folder', True),
+                "sort_order": cat.sort_order,
+                "doc_count": doc_count_map.get(cat.id, 0),
+                "children": [],
+            }
+
+        roots = []
+        for cat in all_cats:
+            node = cat_map[cat.id]
+            if cat.parent_id and cat.parent_id in cat_map:
+                cat_map[cat.parent_id]["children"].append(node)
+            else:
+                roots.append(node)
+
+        return roots
+
     def list_documents(
-        self, kb: KnowledgeBase, *, page: int = 1, page_size: int = 50
+        self, kb: KnowledgeBase, *, category_id: str | None = None, page: int = 1, page_size: int = 50
     ) -> tuple[list[Document], int]:
-        """获取知识库的文档列表（排除 DELETED）。"""
+        """获取知识库的文档列表（排除 DELETED）。支持按分类过滤。"""
         return self._doc_repo.list_by_knowledge_base(
-            kb.id, page=page, page_size=page_size
+            kb.id, category_id=category_id, page=page, page_size=page_size
         )
 
     def get_document_by_token(self, kb: KnowledgeBase, doc_token: str) -> Document:
@@ -42,34 +88,38 @@ class KbPublicService:
             raise NotFoundError("文档不存在。")
         return doc
 
-    def get_document_markdown(self, doc: Document) -> str:
-        """读取文档的 Markdown 内容。"""
-        from app.storage.local_storage import get_storage
+    def get_document_content(self, doc: Document) -> str:
+        """获取文档内容（优先直接内容，其次 Markdown 文件）。"""
+        # 直接文本内容
+        if doc.content:
+            return doc.content
 
-        if not doc.markdown_path:
-            return ""
-        storage = get_storage()
-        content = storage.read(doc.markdown_path)
-        return content.decode("utf-8")
+        # 从文件读取 Markdown
+        if doc.markdown_path:
+            from app.storage.local_storage import get_storage
+
+            storage = get_storage()
+            content = storage.read(doc.markdown_path)
+            return content.decode("utf-8")
+
+        return ""
 
     def search_documents(
         self, kb: KnowledgeBase, query: str, *, page: int = 1, page_size: int = 20
     ) -> tuple[list[dict], int]:
         """关键词搜索文档（Phase 12：SQLite LIKE / FTS5）。"""
-        # MVP 简化：使用 LIKE 搜索标题+描述+关键词
-        from sqlalchemy import func, or_, select
-
         conditions = [
             Document.knowledge_base_id == kb.id,
             Document.status == "READY",
         ]
 
         like_pattern = f"%{query}%"
-        search_condition = or_(
-            Document.title.like(like_pattern),
-            Document.description.like(like_pattern),
-            Document.keywords.like(like_pattern),
-            Document.content_summary.like(like_pattern),
+        search_condition = (
+            Document.title.like(like_pattern)
+            | Document.description.like(like_pattern)
+            | Document.keywords.like(like_pattern)
+            | Document.content_summary.like(like_pattern)
+            | Document.content.like(like_pattern)
         )
         conditions.append(search_condition)
 
@@ -87,6 +137,13 @@ class KbPublicService:
 
         results = []
         for doc in docs:
+            # 获取分类路径
+            category_path = ""
+            if doc.category_id:
+                cat = self._session.get(DocumentCategory, doc.category_id)
+                if cat:
+                    category_path = cat.path or ""
+
             results.append({
                 "id": doc.id,
                 "title": doc.title or doc.original_filename,
@@ -95,6 +152,7 @@ class KbPublicService:
                 "file_type": doc.file_ext,
                 "updated_at": doc.updated_at,
                 "url_hint": doc.doc_token_hint,
+                "category_path": category_path,
             })
 
         return results, total