initialize logger in config before anything else

Author: dminnear-rh

.github/workflows/ci-pipeline.yaml

@@ -110,6 +110,7 @@ jobs:
       - name: Run embed job
         run: |
           docker run --rm --network host \
+            -e LOG_LEVEL=debug \
             -e DB_TYPE=${{ matrix.db }} \
             test-image:${{ needs.build.outputs.image_tag }}
 

config.py

@@ -1,4 +1,5 @@
 import json
+import logging
 import os
 from dataclasses import dataclass
 from typing import Dict, List
@@ -97,13 +98,28 @@ def load() -> "Config":
         load_dotenv()
         get = Config._get_required_env_var
 
+        # Initialize logger
+        log_level_name = get("LOG_LEVEL").lower()
+        log_levels = {
+            "debug": 10,
+            "info": 20,
+            "warning": 30,
+            "error": 40,
+            "critical": 50,
+        }
+        if log_level_name not in log_levels:
+            raise ValueError(
+                f"Invalid LOG_LEVEL: '{log_level_name}'. Must be one of: {', '.join(log_levels)}"
+            )
+        log_level = log_levels[log_level_name]
+        logging.basicConfig(level=log_level)
+        logger = logging.getLogger(__name__)
+        logger.debug("Logging initialized at level: %s", log_level_name.upper())
+
+        # Initialize db
         db_type = get("DB_TYPE")
         db_provider = Config._init_db_provider(db_type)
 
-        chunk_size = int(get("CHUNK_SIZE"))
-        chunk_overlap = int(get("CHUNK_OVERLAP"))
-        temp_dir = get("TEMP_DIR")
-
         # Web URLs
         web_sources_raw = get("WEB_SOURCES")
         try:
@@ -118,20 +134,10 @@ def load() -> "Config":
         except json.JSONDecodeError as e:
             raise ValueError(f"Invalid REPO_SOURCES JSON: {e}") from e
 
-        # Logging
-        log_level_name = get("LOG_LEVEL").lower()
-        log_levels = {
-            "debug": 10,
-            "info": 20,
-            "warning": 30,
-            "error": 40,
-            "critical": 50,
-        }
-        if log_level_name not in log_levels:
-            raise ValueError(
-                f"Invalid LOG_LEVEL: '{log_level_name}'. Must be one of: {', '.join(log_levels)}"
-            )
-        log_level = log_levels[log_level_name]
+        # Misc
+        chunk_size = int(get("CHUNK_SIZE"))
+        chunk_overlap = int(get("CHUNK_OVERLAP"))
+        temp_dir = get("TEMP_DIR")
 
         return Config(
             db_provider=db_provider,

embed_documents.py

@@ -11,11 +11,7 @@
 from loaders.pdf import PDFLoader
 from loaders.web import WebLoader
 
-# Load configuration
 config = Config.load()
-
-# Set up logging
-logging.basicConfig(level=config.log_level)
 logger = logging.getLogger(__name__)