提取操作使用千问plus大模型

app/services/text_service.py

@@ -6,7 +6,7 @@ import docx
 from app.clients.llm.base import LLMClient
 from app.clients.storage.base import StorageClient
 from app.core.config import get_config
-from app.core.exceptions import UnsupportedFileTypeError
+from app.core.exceptions import StorageError, UnsupportedFileTypeError
 from app.core.json_utils import extract_json
 from app.core.logging import get_logger
 from app.models.text_models import (
@@ -61,7 +61,21 @@ async def extract_triples(
     bucket = cfg["storage"]["buckets"]["source_data"]
     model = req.model or cfg["models"]["default_text"]
 
-    data = await storage.download_bytes(bucket, req.file_path)
+    try:
+        data = await storage.download_bytes(bucket, req.file_path)
+        logger.info("文件下载成功", extra={
+            "file_name": req.file_name,
+            "size_bytes": len(data)
+        })
+    except Exception as e:
+        logger.error("文件下载失败", extra={
+            "file_name": req.file_name,
+            "file_path": req.file_path,
+            "bucket": bucket,
+            "error_type": type(e).__name__,
+            "error_message": str(e)
+        }, exc_info=True)
+        raise StorageError(f"下载文件失败: {str(e)}") from e
 
     if ext == ".txt":
         text = _parse_txt(data)