fix: [BUG-001] TypeError in nutrition_weekly endpoint

Problem:
- /api/nutrition/weekly crashed with 500 Internal Server Error
- TypeError: strptime() argument 1 must be str, not datetime.date

Root Cause:
- d['date'] from PostgreSQL is already datetime.date object
- datetime.strptime() expects string input
- Line 156: wk=datetime.strptime(d['date'],'%Y-%m-%d').strftime('%Y-W%V')

Solution:
- Added type check before strptime()
- If date already has strftime method → use directly
- Else → parse as string first
- Works with both datetime.date objects and strings

Tested:
- /nutrition page loads without error
- Weekly aggregation works correctly
- Chart displays nutrition data

Closes: BUG-001

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

backend/routers/nutrition.py

@@ -153,7 +153,9 @@ def nutrition_weekly(weeks: int=16, x_profile_id: Optional[str]=Header(default=N
     if not rows: return []
     wm={}
     for d in rows:
-        wk=datetime.strptime(d['date'],'%Y-%m-%d').strftime('%Y-W%V')
+        # Handle both datetime.date objects (from DB) and strings
+        date_obj = d['date'] if hasattr(d['date'], 'strftime') else datetime.strptime(d['date'],'%Y-%m-%d')
+        wk = date_obj.strftime('%Y-W%V')
         wm.setdefault(wk,[]).append(d)
     result=[]
     for wk in sorted(wm):