Update logging config

Author: MathyouMB

scripts/ingest_webpages_list.py

@@ -1,3 +1,7 @@
+import logging
+
+logging.getLogger("sqlalchemy.engine.Engine").disabled = True
+
 from ingestion.tasks.ingest_webpages_list_task import ingest_webpages_list
 
 if __name__ == "__main__":

src/config/celery.py

@@ -1,3 +1,7 @@
+import logging
+
+logging.getLogger("sqlalchemy.engine.Engine").disabled = True
+
 from celery import Celery
 
 celery = Celery(

src/config/logger.py

@@ -2,13 +2,13 @@
 import json
 from typing import Optional, Dict, Any
 
-# ANSI escape codes
+# ANSI escape codes for colors
 LOG_COLORS = {
-    "DEBUG": "\033[94m",  # Blue
-    "INFO": "\033[92m",  # Green
-    "WARNING": "\033[93m",  # Yellow
-    "ERROR": "\033[91m",  # Red
-    "RESET": "\033[0m",  # Reset to default
+    "DEBUG": "\033[94m",
+    "INFO": "\033[92m",
+    "WARNING": "\033[93m",
+    "ERROR": "\033[91m",
+    "RESET": "\033[0m",
 }
 
 
@@ -23,14 +23,16 @@ def format(self, record: logging.LogRecord) -> str:
 class SemanticLogger:
     def __init__(self, name: Optional[str] = None):
         self.logger = logging.getLogger(name or __name__)
-        handler = logging.StreamHandler()
+        self.logger.propagate = False  # 🔒 Prevent double logging
 
-        formatter = ColorFormatter("%(asctime)s [%(levelname)s] %(message)s")
-        handler.setFormatter(formatter)
+        # Only add handler if none exist
+        if not self.logger.handlers:
+            handler = logging.StreamHandler()
+            formatter = ColorFormatter("%(asctime)s [%(levelname)s] %(message)s")
+            handler.setFormatter(formatter)
+            self.logger.addHandler(handler)
 
         self.logger.setLevel(logging.INFO)
-        if not self.logger.hasHandlers():
-            self.logger.addHandler(handler)
 
     def info(self, event: str, **data: Any):
         self.logger.info(self._format(event, data))
@@ -49,5 +51,4 @@ def _format(self, event: str, data: Dict[str, Any]) -> str:
         return json.dumps(structured)
 
 
-# Usage
-logger = SemanticLogger()
+logger = SemanticLogger(__name__)

src/ingestion/services/webpage_ingestion_service.py

@@ -3,6 +3,7 @@
 from infrastructure.web import fetch_webpage
 from infrastructure.db.source_repository import SourceRepository
 from ingestion.services.webpage_parser_service import WebpageParserService
+from config import logger
 
 # NOTE: Solves the follow error:
 # Cannot copy out of meta tensor; no data! Please use torch.nn.Module.to_empty() instead of torch.nn.Module.to() when moving module from meta to a different device.
@@ -16,10 +17,15 @@ def __init__(self, url: str, repository: SourceRepository):
 
     def process(self):
         try:
-            print(f"🌐 Fetching content from: {self.url}")
+            logger.info("webpage_ingestion_service.process", url=self.url)
             self.__ingest_webpage()
         except Exception as e:
-            print(f"⚠️ Failed to fetch {self.url}: {e}")
+            logger.error(
+                "webpage_ingestion_service.error",
+                url=self.url,
+                error=str(e),
+            )
+            raise e
 
     def __ingest_webpage(self):
         webpage = fetch_webpage(self.url)