docs(fix): 修正引用跳转文档错误及修复 fix_image_paths 脚本

- 修正 preview 接口路径（去掉多余的 /api 前缀）
- 修正响应示例中 chunk id 格式（文件名_序号，非 UUID）
- 补充 preview 响应中 status、version 字段
- 明确前端引用匹配方式为 chunk_id 查找而非数组下标
- fix_image_paths.py 改为遍历知识库子目录，避免生成空 chroma.sqlite3

scripts/fix_image_paths.py

@@ -17,86 +17,109 @@ import sys
 sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
 
 import chromadb
+from config import CHROMA_DB_PATH
 
 
 def fix_image_paths(dry_run: bool = False):
     """
     修复向量库中的图片路径格式
 
+    多向量库模式下，遍历 chroma/ 下的每个知识库子目录，
+    分别创建 PersistentClient 处理各自的集合。
+
     Args:
         dry_run: 仅预览，不实际修改
     """
-    chroma_path = "knowledge/vector_store/chroma"
-    client = chromadb.PersistentClient(path=chroma_path)
+    # 扫描知识库子目录（每个子目录是一个独立的向量库）
+    kb_dirs = []
+    if os.path.exists(CHROMA_DB_PATH):
+        for item in os.listdir(CHROMA_DB_PATH):
+            item_path = os.path.join(CHROMA_DB_PATH, item)
+            if os.path.isdir(item_path) and not item.startswith('.'):
+                if os.path.exists(os.path.join(item_path, 'chroma.sqlite3')):
+                    kb_dirs.append(item)
 
-    # 获取所有集合
-    collections = client.list_collections()
-    print(f"找到 {len(collections)} 个集合")
+    if not kb_dirs:
+        print("未找到任何知识库子目录")
+        return
 
-    for collection in collections:
-        print(f"\n处理集合: {collection.name}")
+    print(f"发现 {len(kb_dirs)} 个知识库: {kb_dirs}\n")
 
-        # 获取所有图片类型的切片
-        try:
-            results = collection.get(
-                where={"chunk_type": "image"},
-                include=["metadatas", "documents", "embeddings"]
-            )
+    for kb_name in kb_dirs:
+        kb_path = os.path.join(CHROMA_DB_PATH, kb_name)
+        print(f"处理知识库: {kb_name}")
+        client = chromadb.PersistentClient(path=kb_path)
 
-            if not results["ids"]:
-                print(f"  无图片类型切片")
-                continue
+        # 获取所有集合
+        collections = client.list_collections()
+        print(f"  找到 {len(collections)} 个集合")
 
-            print(f"  找到 {len(results['ids'])} 个图片切片")
+        for collection in collections:
+            print(f"\n  处理集合: {collection.name}")
 
-            # 检查需要修复的路径
-            needs_fix = []
-            for i, (id, meta) in enumerate(zip(results["ids"], results["metadatas"])):
-                image_path = meta.get("image_path", "")
-                if image_path and ("/" in image_path or "\\" in image_path):
-                    # 路径包含目录，需要修复
-                    new_path = os.path.basename(image_path)
-                    needs_fix.append({
-                        "id": id,
-                        "old_path": image_path,
-                        "new_path": new_path,
-                        "index": i
-                    })
+            # 获取所有图片类型的切片
+            try:
+                results = collection.get(
+                    where={"chunk_type": "image"},
+                    include=["metadatas", "documents", "embeddings"]
+                )
 
-            if not needs_fix:
-                print(f"  所有路径格式正确，无需修复")
-                continue
+                if not results["ids"]:
+                    print(f"    无图片类型切片")
+                    continue
 
-            print(f"  需要修复 {len(needs_fix)} 条记录:")
-            for item in needs_fix[:5]:  # 只显示前5条
-                print(f"    {item['old_path']} -> {item['new_path']}")
-            if len(needs_fix) > 5:
-                print(f"    ... 还有 {len(needs_fix) - 5} 条")
+                print(f"    找到 {len(results['ids'])} 个图片切片")
 
-            if dry_run:
-                print(f"  [DRY-RUN] 跳过实际修改")
-                continue
+                # 检查需要修复的路径
+                needs_fix = []
+                for i, (id, meta) in enumerate(zip(results["ids"], results["metadatas"])):
+                    image_path = meta.get("image_path", "")
+                    if image_path and ("/" in image_path or "\\" in image_path):
+                        # 路径包含目录，需要修复
+                        new_path = os.path.basename(image_path)
+                        needs_fix.append({
+                            "id": id,
+                            "old_path": image_path,
+                            "new_path": new_path,
+                            "index": i
+                        })
 
-            # 执行修复
-            ids_to_update = []
-            metadatas_to_update = []
+                if not needs_fix:
+                    print(f"    所有路径格式正确，无需修复")
+                    continue
 
-            for item in needs_fix:
-                # 更新元数据
-                new_meta = results["metadatas"][item["index"]].copy()
-                new_meta["image_path"] = item["new_path"]
-                ids_to_update.append(item["id"])
-                metadatas_to_update.append(new_meta)
+                print(f"    需要修复 {len(needs_fix)} 条记录:")
+                for item in needs_fix[:5]:  # 只显示前5条
+                    print(f"      {item['old_path']} -> {item['new_path']}")
+                if len(needs_fix) > 5:
+                    print(f"      ... 还有 {len(needs_fix) - 5} 条")
 
-            # 批量更新
-            collection.update(
-                ids=ids_to_update,
-                metadatas=metadatas_to_update
-            )
-            print(f"  ✓ 已修复 {len(ids_to_update)} 条记录")
+                if dry_run:
+                    print(f"    [DRY-RUN] 跳过实际修改")
+                    continue
 
-        except Exception as e:
-            print(f"  处理集合 {collection.name} 时出错: {e}")
+                # 执行修复
+                ids_to_update = []
+                metadatas_to_update = []
+
+                for item in needs_fix:
+                    # 更新元数据
+                    new_meta = results["metadatas"][item["index"]].copy()
+                    new_meta["image_path"] = item["new_path"]
+                    ids_to_update.append(item["id"])
+                    metadatas_to_update.append(new_meta)
+
+                # 批量更新
+                collection.update(
+                    ids=ids_to_update,
+                    metadatas=metadatas_to_update
+                )
+                print(f"    ✓ 已修复 {len(ids_to_update)} 条记录")
+
+            except Exception as e:
+                print(f"    处理集合 {collection.name} 时出错: {e}")
+
+        print()
 
 
 def main():