fix(explorer): resolve notion sync, add debug logging, and fix UI display for industries v0.6.1

company-explorer/backend/scripts/sync_notion_industries.py

@@ -11,7 +11,7 @@ from backend.database import SessionLocal, Industry, RoboticsCategory, init_db
 from backend.config import settings
 
 # Setup Logging
-logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s')
+logging.basicConfig(level=logging.DEBUG, format='%(asctime)s - %(levelname)s - %(message)s')
 logger = logging.getLogger(__name__)
 
 NOTION_TOKEN_FILE = "/app/notion_token.txt"
@@ -71,6 +71,9 @@ def extract_number(prop):
 
 def sync_categories(token, session):
     logger.info("Syncing Robotics Categories...")
+    # session.query(RoboticsCategory).delete() # DANGEROUS - Reverted to Upsert
+    # session.commit()
+
     pages = query_notion_db(token, CATEGORIES_DB_ID)
     
     count = 0
@@ -98,16 +101,19 @@ def sync_categories(token, session):
         
         cat.name = name
         cat.description = description
-        # cat.reasoning_guide = ... ? Maybe 'Constrains'?
         cat.reasoning_guide = extract_rich_text(props.get("Constrains"))
         
         count += 1
     
     session.commit()
-    logger.info(f"Synced {count} categories.")
+    logger.info(f"Synced (Upsert) {count} categories.")
 
 def sync_industries(token, session):
     logger.info("Syncing Industries...")
+    logger.warning("DELETING all existing industries before sync...")
+    session.query(Industry).delete()
+    session.commit()
+
     pages = query_notion_db(token, INDUSTRIES_DB_ID)
     
     count = 0
@@ -115,21 +121,17 @@ def sync_industries(token, session):
         props = page.get("properties", {})
         
         notion_id = page["id"]
-        name = extract_title(props.get("Industry"))
+        # In Notion, the column is now 'Vertical' not 'Industry'
+        name = extract_title(props.get("Vertical"))
         if not name: continue
         
-        # Upsert Logic: Check ID -> Check Name -> Create
-        industry = session.query(Industry).filter(Industry.notion_id == notion_id).first()
-        if not industry:
-            industry = session.query(Industry).filter(Industry.name == name).first()
-            if industry:
-                logger.info(f"Linked existing industry '{name}' to Notion ID {notion_id}")
-                industry.notion_id = notion_id
-            else:
-                industry = Industry(notion_id=notion_id, name=name)
-                session.add(industry)
+        # Removed full Notion props debug log - no longer needed
+        
+        # Logic is now INSERT only
+        industry = Industry(notion_id=notion_id, name=name)
+        session.add(industry)
             
-        # Map Fields
+        # Map Fields from Notion Schema
         industry.name = name
         industry.description = extract_rich_text(props.get("Definition"))
         
@@ -137,18 +139,19 @@ def sync_industries(token, session):
         industry.status_notion = status
         industry.is_focus = (status == "P1 Focus Industry")
         
-        industry.industry_group = extract_rich_text(props.get("Industry-Group"))
-        industry.whale_threshold = extract_number(props.get("Whale Threshold"))
+        # New Schema Fields
+        industry.metric_type = extract_select(props.get("Metric Type"))
         industry.min_requirement = extract_number(props.get("Min. Requirement"))
+        industry.whale_threshold = extract_number(props.get("Whale Threshold"))
+        industry.proxy_factor = extract_number(props.get("Proxy Factor"))
+        industry.scraper_search_term = extract_select(props.get("Scraper Search Term")) # <-- FIXED HERE
         industry.scraper_keywords = extract_rich_text(props.get("Scraper Keywords"))
-        industry.core_unit = extract_select(props.get("Core Unit"))
-        industry.proxy_factor = extract_rich_text(props.get("Proxy Factor"))
-        
+        industry.standardization_logic = extract_rich_text(props.get("Stanardization Logic"))
+
         # Relation: Primary Product Category
         relation = props.get("Primary Product Category", {}).get("relation", [])
         if relation:
             related_id = relation[0]["id"]
-            # Find Category by notion_id
             cat = session.query(RoboticsCategory).filter(RoboticsCategory.notion_id == related_id).first()
             if cat:
                 industry.primary_category_id = cat.id
@@ -173,4 +176,4 @@ if __name__ == "__main__":
     except Exception as e:
         logger.error(f"Sync failed: {e}", exc_info=True)
     finally:
-        db.close()
+        db.close()