Update Azure Blob Service Config

services/embed_model/embed_model.py

@@ -1,6 +1,7 @@
 import os
 import dotenv
-dotenv.load_dotenv()
+from config.env_constant import EnvFilepath
+dotenv.load_dotenv(EnvFilepath.ENVPATH)
 from datetime import datetime
 from langchain_openai import AzureOpenAIEmbeddings
 

services/uploader/azure_blob_service.py

@@ -1,14 +1,15 @@
 import os
 import dotenv
-dotenv.load_dotenv()
+from config.env_constant import EnvFilepath
+dotenv.load_dotenv(EnvFilepath.ENVPATH)
 
+from config.constant import AzureBlobConstants
 from azure.identity import DefaultAzureCredential
 from azure.storage.blob.aio import BlobServiceClient, ContainerClient
 # from azure.storage.blob import BlobServiceClient, ContainerClient
 from fastapi import UploadFile
 
 
-from config.config import azure_blob_config
 from utils.logger import get_logger
 # from src.utils.decorator import trace_runtime
 
@@ -20,13 +21,32 @@ logger = get_logger("azure blob")
 # CONNECTION_STRING = os.getenv("AZURE_STORAGE_CONNECTION_STRING")
 # CONTAINER_NAME = "pdf-uploads"  # The name of your Azure Blob Storage container
 
-async def get_blob_service_client(url=os.environ.get('azureai__container__endpoint')) -> BlobServiceClient:
+# async def get_blob_service_client(url=os.environ.get('azureai__container__endpoint')) -> BlobServiceClient:
+#     try:
+#         default_credential = DefaultAzureCredential()
+#         blob_service_client = BlobServiceClient(url, credential=default_credential)
+#         return blob_service_client
+#     except Exception as E:
+#         logger.error(f'❌ Getting blob service client error, {E}')
+
+async def get_blob_service_client() -> BlobServiceClient:
     try:
-        default_credential = DefaultAzureCredential()
-        blob_service_client = BlobServiceClient(url, credential=default_credential)
+        account_url = os.environ.get("azureai__container__endpoint")
+        sas_token = os.environ.get("azureai__search__sas")
+
+        if not account_url or not sas_token:
+            raise ValueError("Missing AZURE_STORAGE_ACCOUNT_URL or AZURE_STORAGE_SAS_TOKEN")
+
+        blob_service_client = BlobServiceClient(
+            account_url=account_url,
+            credential=sas_token
+        )
+
         return blob_service_client
-    except Exception as E:
-        logger.error(f'❌ Getting blob service client error, {E}')
+
+    except Exception as e:
+        logger.error(f"❌ Getting blob service client error: {e}")
+        raise
 
 
 async def get_container_client(url=os.environ.get('azureai__container__endpoint'), container_name=os.environ.get("azureai__container__name")) -> ContainerClient:
@@ -47,7 +67,7 @@ class AzureBlobStorageService:
         # Initialize the BlobServiceClient for the entire application lifetime
         self.blob_service_client = get_blob_service_client()
         self.container_client = get_container_client()
-        self.prefix = azure_blob_config.blob_prefix
+        self.prefix = AzureBlobConstants.BLOB_PREFIX
         
         if not self.blob_service_client:
             raise ValueError("azure service client is not set.")

utils/utils.py

@@ -2,17 +2,17 @@ import os
 import sys
 import io
 import time
-import dotenv
 import PyPDF2
 import asyncio
-import pandas as pd
 # import fitz
 import pytesseract
-dotenv.load_dotenv()
+import dotenv
+from config.env_constant import EnvFilepath
+dotenv.load_dotenv(EnvFilepath.ENVPATH)
 from PyPDF2 import PdfReader
 from functools import wraps
 from typing import ByteString
-from pdf2image import convert_from_path, convert_from_bytes
+from pdf2image import convert_from_bytes
 
 
 def measure_runtime(func):