Update parser.py

parser.py

@@ -50,6 +50,23 @@ def is_blank_or_comment(line):
     stripped = line.strip()
     return not stripped or stripped.startswith('#')
 
+def create_vector(category, level, location, total_lines, parent_path):
+    """Create a 6D vector optimized for role similarity, integrating variable roles into category_id."""
+    category_map = {
+        'import': 1, 'function': 2, 'async_function': 3, 'class': 4,
+        'if': 5, 'while': 6, 'for': 7, 'try': 8, 'expression': 9, 'spacer': 10,
+        'other': 11, 'elif': 12, 'else': 13, 'except': 14, 'finally': 15, 'return': 16,
+        'assigned_variable': 17, 'input_variable': 18, 'returned_variable': 19
+    }
+    category_id = category_map.get(category, 0)  # Default to 0 for unknown categories
+    start_line, end_line = location
+    span = (end_line - start_line + 1) / total_lines
+    center_pos = ((start_line + end_line) / 2) / total_lines
+    parent_depth = len(parent_path)
+    parent_weight = sum(category_map.get(parent.split('[')[0].lower(), 0) * (1 / (i + 1)) 
+                        for i, parent in enumerate(parent_path)) / max(1, len(category_map))
+    return [category_id, level, center_pos, span, parent_depth, parent_weight]
+
 
 def collect_variable_usage(tree):
     """Collect definitions and uses of variables, respecting scope."""