fix config hot update

fix config hot update

Author: hrfng

docker/Dockerfile

@@ -1,5 +1,5 @@
 FROM dataelement/bisheng-unstructured-runtime:0.0.1
-MAINTAINER "Dataelem inc."
+LABEL org.opencontainers.image.authors="Dataelem inc."
 
 ARG BISHENG_UNS_VER=0.0.2
 

src/bisheng_unstructured/api/main.py

@@ -4,7 +4,7 @@
 import tempfile
 
 import requests
-from fastapi import Depends, FastAPI, Header, HTTPException, Request, status
+from fastapi import FastAPI, HTTPException, Request
 from fastapi.middleware.cors import CORSMiddleware
 from fastapi.responses import ORJSONResponse
 
@@ -15,7 +15,6 @@
 
 logger = get_logger("BishengUns", "/app/log/bisheng-uns.log")
 
-
 # Fastapi App
 
 
@@ -71,6 +70,8 @@ async def update_config(inp: ConfigInput):
     }
 
     if inp.rt_ep is not None:
+        # update environment
+        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)
@@ -79,6 +80,10 @@ async def update_config(inp: ConfigInput):
     else:
         config_dict = inp.dict()
 
+    # update persist data
+    with open(config_file, 'wb') as file:
+        file.write(json.loads(config_dict))
+
     pipeline.update_config(config_dict)
     return {"status": "OK"}
 
@@ -90,6 +95,7 @@ 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()

src/bisheng_unstructured/api/pipeline.py

@@ -1,5 +1,5 @@
 import json
-from typing import Dict
+import os
 
 from bisheng_unstructured.common import get_logger
 from bisheng_unstructured.documents.html_utils import save_to_txt, visualize_html
@@ -57,8 +57,22 @@ def partition_image(filename, model_params, **kwargs):
 
 
 class Pipeline(object):
+
     def __init__(self, config_file: str):
-        self.config = json.load(open(config_file))
+        ''' 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:
+            pdf_model_params_temp = {
+                "layout_ep": f"http://{rt_ep}/v2.1/models/elem_layout_v1/infer",
+                "cell_model_ep": f"http://{rt_ep}/v2.1/models/elem_table_cell_detect_v1/infer",
+                "rowcol_model_ep": f"http://{rt_ep}/v2.1/models/elem_table_rowcol_detect_v1/infer",
+                "table_model_ep": f"http://{rt_ep}/v2.1/models/elem_table_detect_v1/infer",
+                "ocr_model_ep": f"http://{rt_ep}/v2.1/models/elem_ocr_collection_v3/infer",
+            }
+            self.config = {"pdf_model_params": pdf_model_params_temp}
+        else:
+            self.config = tmp_dict
         self.pdf_model_params = self.config.get("pdf_model_params")
         topdf_model_params = self.config.get("topdf_model_params", {})
         self.pdf_creator = Any2PdfCreator(topdf_model_params)