log add context

docker/Dockerfile

@@ -11,7 +11,7 @@ ENV HTTPS_PROXY=
 RUN mkdir -p /opt/bisheng-unstructured/bin
 COPY ./docker/entrypoint.sh /opt/bisheng-unstructured/bin/
 COPY config /opt/bisheng-unstructured/
-COPY ./src/bisheng_unstructured/ /opt/bisheng-unstructured/
+COPY ./src/ /opt/bisheng-unstructured/
 
 # FIX the high risk scan
 RUN rm -f /opt/texlive/2023/texmf-dist/scripts/tlcockpit/tlcockpit.jar

src/bisheng_unstructured/api/main.py

@@ -71,7 +71,7 @@ async def update_config(inp: ConfigInput):
 
     if inp.rt_ep is not None:
         # update environment
-        os.environ['rt_server'] = inp.rt_ep
+        os.environ["rt_server"] = inp.rt_ep
         pdf_model_params = {}
         for k, v in pdf_model_params_temp.items():
             pdf_model_params[k] = v.format(inp.rt_ep)
@@ -81,7 +81,7 @@ async def update_config(inp: ConfigInput):
         config_dict = inp.dict()
 
     # update persist data
-    with open(config_file, 'wb') as file:
+    with open(config_file, "wb") as file:
         file.write(json.loads(config_dict))
 
     pipeline.update_config(config_dict)
@@ -95,7 +95,6 @@ async def config():
 
 @app.post("/v1/etl4llm/predict", response_model=UnstructuredOutput)
 async def etl4_llm(inp: UnstructuredInput):
-
     filename = inp.filename
     b64_data = inp.b64_data
     file_type = filename.rsplit(".", 1)[1].lower()
@@ -114,8 +113,8 @@ async def etl4_llm(inp: UnstructuredInput):
                 with open(file_path, "wb") as fout:
                     fout.write(base64.b64decode(b64_data[0]))
             except Exception:
-                logger.error(f"b64_data is damaged filename=[{inp.filename}]")
-                return Exception(f"b64_data is damaged")
+                logger.error(f"b64_data is damaged filename=[{inp.filename}]", exc_info=True)
+                raise Exception(f"b64_data is damaged")
         else:
             headers = inp.parameters.get("headers", {})
             ssl_verify = inp.parameters.get("ssl_verify", True)

src/bisheng_unstructured/api/pipeline.py

@@ -57,9 +57,8 @@ def partition_image(filename, model_params, **kwargs):
 
 
 class Pipeline(object):
-
     def __init__(self, config_file: str):
-        ''' k8s 使用cm 创建环境变量'''
+        """k8s 使用cm 创建环境变量"""
         tmp_dict = json.load(open(config_file))
         rt_ep = os.getenv("rt_server")
         if rt_ep:
@@ -89,7 +88,7 @@ def to_pdf(self, inp: UnstructuredInput) -> UnstructuredOutput:
             result = UnstructuredOutput(b64_pdf=output)
             return result
         except Exception as e:
-            logger.error(f"error in topdf filename=[{inp.filename}] err=[{e}]")
+            logger.error(f"error in topdf filename=[{inp.filename}] err=[{e}]", exc_info=True)
             return UnstructuredOutput(status_code=400, status_message=str(e))
 
     def predict(self, inp: UnstructuredInput) -> UnstructuredOutput: