Normalize converter labels and fix card grid sizing

migrate.py

@@ -31,6 +31,62 @@ CATEGORY_KEYS = [
 
 CATEGORY_SET = set(CATEGORY_KEYS)
 
+# Lightweight label normalization to catch duplicate/identity conversions
+# that differ only by abbreviations (e.g., "cm" vs "centimeters").
+TOKEN_MAP = {
+    'cm': 'centimeter',
+    'centimeter': 'centimeter',
+    'centimetre': 'centimeter',
+    'centimetres': 'centimeter',
+    'centimeters': 'centimeter',
+    'mm': 'millimeter',
+    'millimeter': 'millimeter',
+    'millimeters': 'millimeter',
+    'millimetre': 'millimeter',
+    'millimetres': 'millimeter',
+    'm': 'meter',
+    'meter': 'meter',
+    'meters': 'meter',
+    'metre': 'meter',
+    'metres': 'meter',
+    'km': 'kilometer',
+    'kilometer': 'kilometer',
+    'kilometers': 'kilometer',
+    'kilometre': 'kilometer',
+    'kilometres': 'kilometer',
+    'in': 'inch',
+    'inch': 'inch',
+    'inches': 'inch',
+    'ft': 'foot',
+    'foot': 'foot',
+    'feet': 'foot',
+}
+
+
+def normalize_label(label: str) -> str:
+    """Canonicalize a unit label for duplicate detection.
+
+    - Lowercase
+    - Replace '/' with ' per ' to align fraction style with text style
+    - Strip punctuation into tokens
+    - Collapse common abbreviations/plurals via TOKEN_MAP and simple singularization
+    """
+    cleaned = label.lower().replace('/', ' per ')
+    tokens = re.split(r'[^a-z0-9]+', cleaned)
+
+    normalized_tokens = []
+    for tok in tokens:
+        if not tok:
+            continue
+        base = tok
+        # Drop a trailing 's' for simple plurals, but avoid short abbreviations like 'cms'
+        if base.endswith('s') and len(base) > 3:
+            base = base[:-1]
+        base = TOKEN_MAP.get(base, base)
+        normalized_tokens.append(base)
+
+    return ' '.join(normalized_tokens)
+
 def load_external_descriptions():
     # Placeholder for future enrichment sources.
     return {}
@@ -92,6 +148,7 @@ def process():
     calculators_ts_entries = []
     
     seen_slugs = set()
+    seen_norm_pairs = set()
     for raw_name, slug, category_raw, factor_raw in active_rows:
         if raw_name == 'Calculator Name' or not slug:
             continue
@@ -106,7 +163,21 @@ def process():
             in1, in2 = "From", "To"
 
         custom_labels = None
-            
+        
+        norm_in1 = normalize_label(in1)
+        norm_in2 = normalize_label(in2)
+
+        # Skip identity conversions that only differ by spelling/abbreviation
+        if norm_in1 == norm_in2:
+            print(f"Skipping identity converter {slug}: {in1} -> {in2}")
+            continue
+
+        pair_key = (norm_in1, norm_in2)
+        if pair_key in seen_norm_pairs:
+            print(f"Skipping duplicate converter {slug}: {in1} -> {in2}")
+            continue
+        seen_norm_pairs.add(pair_key)
+
         category = normalize_category(category_raw)
         if not category:
             raise ValueError(f'Category required for {display_name}')