fix: Part B protein placeholders - aggregate by date

Fixes calculate_protein_g_per_kg and calculate_protein_days_in_target:

**Problem:**
Both functions were treating individual nutrition_log entries as days,
causing incorrect calculations when multiple entries exist per day
(e.g., from CSV imports: 233 entries across 7 days).

**Solution:**
1. calculate_protein_g_per_kg:
   - Added GROUP BY date, SUM(protein_g) to aggregate by day
   - Now averages daily totals, not individual entries
   - Correct: 7 days → 7 values, not 233 entries → 233 values

2. calculate_protein_days_in_target:
   - Added GROUP BY date, SUM(protein_g) to aggregate by day
   - Calculates target range in absolute grams (not g/kg per entry)
   - Counts unique DAYS in range, not entries
   - Correct format: "5/7" (5 of 7 days), not "150/233" (entries)

**Impact:**
- protein_g_per_kg: was returning "nicht verfügbar" → now returns correct value
- protein_days_in_target: was returning "nicht verfügbar" → now returns correct format

**Root Cause:**
Functions expected 7 unique dates but got 233 entries.
With export date 2026-04-02 and last data 2026-03-26,
the 7-day window had insufficient unique dates.

Issue reported by user: Part B placeholders not showing correct values
in extended export (registry metadata was correct, but computed values failed).

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

backend/data_layer/nutrition_metrics.py

@@ -575,22 +575,23 @@ def calculate_protein_g_per_kg(profile_id: str) -> Optional[float]:
 
         weight = float(weight_row['weight'])
 
-        # Get protein intake
+        # Get protein intake aggregated by day (SUM per day)
         cur.execute("""
-            SELECT protein_g
+            SELECT date, SUM(protein_g) as daily_protein
             FROM nutrition_log
             WHERE profile_id = %s
               AND date >= CURRENT_DATE - INTERVAL '7 days'
               AND protein_g IS NOT NULL
+            GROUP BY date
             ORDER BY date DESC
         """, (profile_id,))
 
-        protein_values = [row['protein_g'] for row in cur.fetchall()]
+        daily_protein = [float(row['daily_protein']) for row in cur.fetchall()]
 
-        if len(protein_values) < 4:
+        if len(daily_protein) < 4:  # At least 4 days with data
             return None
 
-        avg_protein = float(sum(protein_values) / len(protein_values))
+        avg_protein = sum(daily_protein) / len(daily_protein)
         protein_per_kg = avg_protein / weight
 
         return round(protein_per_kg, 2)
@@ -619,28 +620,33 @@ def calculate_protein_days_in_target(profile_id: str, target_low: float = 1.6, t
 
         weight = float(weight_row['weight'])
 
-        # Get protein intake last 7 days
+        # Calculate protein target range (absolute values)
+        target_low_g = target_low * weight
+        target_high_g = target_high * weight
+
+        # Get protein intake aggregated by day (SUM per day)
         cur.execute("""
-            SELECT protein_g, date
+            SELECT date, SUM(protein_g) as daily_protein
             FROM nutrition_log
             WHERE profile_id = %s
               AND date >= CURRENT_DATE - INTERVAL '7 days'
               AND protein_g IS NOT NULL
+            GROUP BY date
             ORDER BY date DESC
         """, (profile_id,))
 
-        protein_data = cur.fetchall()
+        daily_data = cur.fetchall()
 
-        if len(protein_data) < 4:
+        if len(daily_data) < 4:  # At least 4 days with data
             return None
 
         # Count days in target range
         days_in_target = 0
-        total_days = len(protein_data)
+        total_days = len(daily_data)
 
-        for row in protein_data:
-            protein_per_kg = float(row['protein_g']) / weight
-            if target_low <= protein_per_kg <= target_high:
+        for row in daily_data:
+            daily_protein = float(row['daily_protein'])
+            if target_low_g <= daily_protein <= target_high_g:
                 days_in_target += 1
 
         return f"{days_in_target}/{total_days}"