thuban 0.3.1 → 0.3.3

package/packages/scanner/copy-paste-detector.js

@@ -149,21 +149,34 @@ class CopyPasteDriftDetector {
   }
 
   _normalize(code) {
-    return code
+    // Two-pass normalization:
+    // Pass 1: collect all declared variable names
+    const declaredNames = [];
+    const declPattern = /(?:const|let|var|function)\s+([a-zA-Z_$][a-zA-Z0-9_$]*)/g;
+    let m;
+    while ((m = declPattern.exec(code)) !== null) {
+      if (!declaredNames.includes(m[1])) declaredNames.push(m[1]);
+    }
+
+    let normalized = code
       // Remove comments
       .replace(/\/\/.*$/gm, '')
       .replace(/\/\*[\s\S]*?\*\//g, '')
-      // Normalize whitespace
-      .replace(/\s+/g, ' ')
-      // Normalize variable names to placeholders
-      .replace(/(?:const|let|var)\s+([a-zA-Z_$]\w*)/g, 'const VAR')
       // Normalize string literals
       .replace(/'[^']*'/g, 'STR')
       .replace(/"[^"]*"/g, 'STR')
       .replace(/`[^`]*`/g, 'STR')
       // Normalize numbers
-      .replace(/\b\d+\.?\d*\b/g, 'NUM')
-      .trim();
+      .replace(/\b\d+\.?\d*\b/g, 'NUM');
+
+    // Pass 2: replace ALL occurrences of each declared name with VAR_N
+    declaredNames.forEach((name, idx) => {
+      const nameRegex = new RegExp(`\\b${name.replace(/[.*+?^${}()|[\]\\]/g, '\\$&')}\\b`, 'g');
+      normalized = normalized.replace(nameRegex, `VAR_${idx}`);
+    });
+
+    // Normalize whitespace
+    return normalized.replace(/\s+/g, ' ').trim();
   }
 
   _tokenize(normalized) {
@@ -218,10 +231,23 @@ class CopyPasteDriftDetector {
 
   _detectFunction(line) {
     const trimmed = line.trim();
+    // JavaScript/TypeScript: function/const/let/var
     let match = trimmed.match(/^(?:export\s+)?(?:async\s+)?function\s+([a-zA-Z_$][a-zA-Z0-9_$]*)/);
     if (match) return { name: match[1] };
     match = trimmed.match(/^(?:export\s+)?(?:const|let|var)\s+([a-zA-Z_$][a-zA-Z0-9_$]*)\s*=\s*(?:async\s+)?(?:function|\()/);
     if (match) return { name: match[1] };
+    // Kotlin: fun
+    match = trimmed.match(/^(?:private\s+|public\s+|internal\s+|protected\s+)?(?:suspend\s+)?fun\s+([a-zA-Z_][a-zA-Z0-9_]*)\s*[(<]/);
+    if (match) return { name: match[1] };
+    // Rust: fn
+    match = trimmed.match(/^(?:pub(?:\([^)]*\))?\s+)?(?:async\s+)?fn\s+([a-zA-Z_][a-zA-Z0-9_]*)\s*[(<]/);
+    if (match) return { name: match[1] };
+    // PHP: function
+    match = trimmed.match(/^(?:(?:public|private|protected|static)\s+)*function\s+([a-zA-Z_][a-zA-Z0-9_]*)\s*\(/);
+    if (match) return { name: match[1] };
+    // Ruby: def
+    match = trimmed.match(/^def\s+(?:self\.)?([a-zA-Z_][a-zA-Z0-9_?!]*)\s*[(\s]?/);
+    if (match) return { name: match[1] };
     return null;
   }