Simplify data tag id to only contain descriptive identifier

Remove numeric value and unit from id attribute, keeping only the
context-derived identifier (e.g., "avgift", "ranta", "moms").

This allows tracking the same data point across law amendments,
since the id stays constant while only the value changes.

Before: id="avgift-1500-kr"
After:  id="avgift"

Author: claude

formatters/tag_swedish_amounts.py

@@ -73,60 +73,41 @@ def normalize_number(num_str: str) -> str:
     return normalized
 
 
-def generate_amount_slug(value: str, multiplier: Optional[str], currency: str, context: str) -> str:
+def generate_amount_slug(context: str) -> str:
     """
     Generate a descriptive slug for an amount based on context.
 
+    The slug identifies what the amount represents, not its value.
+    This allows tracking changes across law amendments.
+
     Args:
-        value: The numeric value (normalized)
-        multiplier: Optional multiplier like "miljoner", "miljarder", "tusen"
-        currency: The currency unit used
         context: Surrounding text for context extraction
 
     Returns:
-        A slug like "belopp-1000000-kr" or "avgift-500-kr"
+        A slug like "avgift" or "bidrag" that identifies the amount
     """
-    # Try to extract a descriptive word from context
+    # Extract a descriptive word from context
     prefix = _extract_context_word(context)
 
-    # Format the value with multiplier
-    if multiplier:
-        multiplier_lower = multiplier.lower()
-        if 'miljard' in multiplier_lower:
-            suffix = 'mdkr'
-        elif 'miljon' in multiplier_lower:
-            suffix = 'mkr'
-        elif 'tusen' in multiplier_lower:
-            suffix = 'tkr'
-        else:
-            suffix = 'kr'
-    else:
-        suffix = 'kr'
-
-    # Create slug
-    slug_parts = [prefix, value, suffix]
-    slug = '-'.join(filter(None, slug_parts))
-
-    return _slugify(slug)
+    return _slugify(prefix)
 
 
-def generate_percentage_slug(value: str, context: str) -> str:
+def generate_percentage_slug(context: str) -> str:
     """
     Generate a descriptive slug for a percentage based on context.
 
+    The slug identifies what the percentage represents, not its value.
+    This allows tracking changes across law amendments.
+
     Args:
-        value: The numeric value (normalized)
         context: Surrounding text for context extraction
 
     Returns:
-        A slug like "ranta-5-procent" or "andel-25-procent"
+        A slug like "ranta" or "moms" that identifies the percentage
     """
     prefix = _extract_context_word(context)
 
-    slug_parts = [prefix, value, 'procent']
-    slug = '-'.join(filter(None, slug_parts))
-
-    return _slugify(slug)
+    return _slugify(prefix)
 
 
 def _extract_context_word(context: str) -> str:
@@ -316,15 +297,13 @@ def _tag_amounts_in_line(line: str) -> str:
     def replace_amount_with_multiplier(match):
         full_match = match.group(0)
         number = match.group(1)
-        multiplier = match.group(3)
-        currency = match.group(4)
 
         # Get context (text before match)
         start_pos = match.start()
         context = line[:start_pos]
 
         normalized_value = normalize_number(number)
-        slug = generate_amount_slug(normalized_value, multiplier, currency, context)
+        slug = generate_amount_slug(context)
 
         return f'<data id="{slug}" type="amount" value="{normalized_value}">{full_match}</data>'
 
@@ -338,12 +317,11 @@ def replace_simple_amount(match):
             return full_match
 
         number = match.group(1)
-        currency = match.group(3)
 
         context = line[:start_pos]
 
         normalized_value = normalize_number(number)
-        slug = generate_amount_slug(normalized_value, None, currency, context)
+        slug = generate_amount_slug(context)
 
         return f'<data id="{slug}" type="amount" value="{normalized_value}">{full_match}</data>'
 
@@ -377,7 +355,7 @@ def replace_percentage(match):
         context = line[:start_pos]
 
         normalized_value = normalize_number(number)
-        slug = generate_percentage_slug(normalized_value, context)
+        slug = generate_percentage_slug(context)
 
         return f'<data id="{slug}" type="percentage" value="{normalized_value}">{full_match}</data>'
 

test/test_tag_swedish_amounts.py

@@ -121,27 +121,23 @@ class TestGenerateAmountSlug:
 
     def test_simple_amount(self):
         """Test generating slug for simple amount."""
-        slug = generate_amount_slug("1000", None, "kronor", "En avgift på ")
-        assert "1000" in slug
-        assert "kr" in slug
+        slug = generate_amount_slug("En avgift på ")
+        assert slug == "avgift"
 
-    def test_amount_with_miljoner(self):
-        """Test generating slug for amount with 'miljoner'."""
-        slug = generate_amount_slug("5", "miljoner", "kronor", "Kapitalet är ")
-        assert "5" in slug
-        assert "mkr" in slug
+    def test_amount_with_context(self):
+        """Test generating slug extracts context word."""
+        slug = generate_amount_slug("Kapitalet är ")
+        assert slug == "kapital"
 
-    def test_amount_with_miljarder(self):
-        """Test generating slug for amount with 'miljarder'."""
-        slug = generate_amount_slug("2", "miljarder", "kronor", "Omsättningen är ")
-        assert "2" in slug
-        assert "mdkr" in slug
+    def test_amount_with_omsattning(self):
+        """Test generating slug for 'omsättning'."""
+        slug = generate_amount_slug("Omsättningen är ")
+        assert slug == "omsattning"
 
-    def test_amount_with_tusen(self):
-        """Test generating slug for amount with 'tusen'."""
-        slug = generate_amount_slug("50", "tusen", "kronor", "Priset är ")
-        assert "50" in slug
-        assert "tkr" in slug
+    def test_amount_with_pris(self):
+        """Test generating slug for 'pris'."""
+        slug = generate_amount_slug("Priset är ")
+        assert slug == "pris"
 
 
 # ===========================================================================
@@ -154,15 +150,13 @@ class TestGeneratePercentageSlug:
 
     def test_simple_percentage(self):
         """Test generating slug for simple percentage."""
-        slug = generate_percentage_slug("5", "Räntan är ")
-        assert "5" in slug
-        assert "procent" in slug
+        slug = generate_percentage_slug("Räntan är ")
+        assert slug == "ranta"
 
     def test_percentage_with_context(self):
         """Test generating slug with context extraction."""
-        slug = generate_percentage_slug("25", "Momsen är ")
-        assert "25" in slug
-        assert "procent" in slug
+        slug = generate_percentage_slug("Momsen är ")
+        assert slug == "moms"
 
 
 # ===========================================================================
@@ -361,14 +355,22 @@ class TestTagSwedishAmountsContextSlugs:
     def test_avgift_context(self):
         """Test slug generation with 'avgift' context."""
         result = tag_swedish_amounts("Avgiften är 500 kronor.")
-        assert 'id="avgift-500-kr"' in result
+        assert 'id="avgift"' in result
 
     def test_ranta_context_percentage(self):
         """Test slug generation with 'ränta' context for percentage."""
         result = tag_swedish_amounts("Räntan är 5 procent.")
-        # Ränta should be extracted and slugified
-        assert 'id="' in result
-        assert 'procent"' in result
+        assert 'id="ranta"' in result
+
+    def test_same_id_different_values(self):
+        """Test that same context gives same id regardless of value."""
+        result1 = tag_swedish_amounts("Avgiften är 500 kronor.")
+        result2 = tag_swedish_amounts("Avgiften är 1000 kronor.")
+        # Both should have id="avgift" but different values
+        assert 'id="avgift"' in result1
+        assert 'id="avgift"' in result2
+        assert 'value="500"' in result1
+        assert 'value="1000"' in result2
 
 
 # ===========================================================================