ctxman 3.0.0 → 3.1.0

package/lib/analyzers/duplicate-detector.js

@@ -0,0 +1,264 @@
+/**
+ * Duplicate Code Detector
+ * Detects duplicate and near-duplicate code blocks
+ * Part of FEAT-005: AI-Powered Context Suggestions
+ */
+
+import { getLogger } from '../utils/logger.js';
+
+const logger = getLogger('DuplicateDetector');
+
+export class DuplicateDetector {
+  constructor(options = {}) {
+    this.options = {
+      minLines: options.minLines || 5,
+      similarityThreshold: options.similarityThreshold || 0.8,
+      maxBlocks: options.maxBlocks || 100,
+      ...options,
+    };
+    this.duplicates = [];
+  }
+
+  /**
+   * Detect duplicate code blocks in files
+   * @param {Array} files - Array of file objects with content
+   * @returns {Array} Array of duplicate block groups
+   */
+  detect(files) {
+    logger.debug(`Analyzing ${files.length} files for duplicates`);
+
+    const blocks = this.extractBlocks(files);
+    const duplicateGroups = this.findDuplicates(blocks);
+
+    this.duplicates = duplicateGroups;
+    logger.info(`Found ${duplicateGroups.length} duplicate code groups`);
+
+    return duplicateGroups;
+  }
+
+  /**
+   * Extract code blocks from files
+   * @param {Array} files - Array of file objects
+   * @returns {Array} Array of code blocks
+   */
+  extractBlocks(files) {
+    const blocks = [];
+
+    for (const file of files) {
+      if (!file.content || !this.isCodeFile(file.path)) {
+        continue;
+      }
+
+      const lines = file.content.split('\n');
+      const minLines = this.options.minLines;
+
+      // Extract sliding windows of code blocks
+      for (let i = 0; i <= lines.length - minLines; i++) {
+        const blockLines = lines.slice(i, i + minLines);
+        const content = blockLines.join('\n');
+
+        // Skip empty or comment-only blocks
+        if (this.isSignificantBlock(content)) {
+          blocks.push({
+            file: file.path,
+            relativePath: file.relativePath || file.path,
+            startLine: i + 1,
+            endLine: i + minLines,
+            content,
+            hash: this.hashBlock(content),
+            normalizedHash: this.hashBlock(this.normalizeBlock(content)),
+          });
+        }
+      }
+    }
+
+    return blocks;
+  }
+
+  /**
+   * Find duplicate blocks by comparing hashes
+   * @param {Array} blocks - Array of code blocks
+   * @returns {Array} Array of duplicate groups
+   */
+  findDuplicates(blocks) {
+    const hashMap = new Map();
+    const normalizedHashMap = new Map();
+
+    // Group by exact hash
+    for (const block of blocks) {
+      // Exact duplicates
+      if (!hashMap.has(block.hash)) {
+        hashMap.set(block.hash, []);
+      }
+      hashMap.get(block.hash).push(block);
+
+      // Near duplicates (normalized)
+      if (!normalizedHashMap.has(block.normalizedHash)) {
+        normalizedHashMap.set(block.normalizedHash, []);
+      }
+      normalizedHashMap.get(block.normalizedHash).push(block);
+    }
+
+    // Find groups with more than one occurrence
+    const duplicateGroups = [];
+    const seen = new Set();
+
+    // Process exact duplicates
+    for (const [hash, group] of hashMap) {
+      if (group.length > 1) {
+        const key = `exact:${hash}`;
+        if (!seen.has(key)) {
+          seen.add(key);
+          duplicateGroups.push({
+            type: 'exact',
+            count: group.length,
+            files: [...new Set(group.map((b) => b.relativePath))],
+            locations: group.map((b) => ({
+              file: b.relativePath,
+              line: b.startLine,
+            })),
+            tokens: this.estimateTokens(group[0].content),
+            preview: this.getPreview(group[0].content),
+          });
+        }
+      }
+    }
+
+    // Process near duplicates (normalized)
+    for (const [hash, group] of normalizedHashMap) {
+      if (group.length > 2) {
+        // Higher threshold for normalized
+        const key = `normalized:${hash}`;
+        if (!seen.has(key)) {
+          seen.add(key);
+
+          // Calculate similarity within group
+          const uniqueFiles = [...new Set(group.map((b) => b.relativePath))];
+          if (uniqueFiles.length > 1) {
+            duplicateGroups.push({
+              type: 'similar',
+              count: group.length,
+              files: uniqueFiles,
+              locations: group.slice(0, 5).map((b) => ({
+                file: b.relativePath,
+                line: b.startLine,
+              })),
+              tokens: this.estimateTokens(group[0].content),
+              preview: this.getPreview(group[0].content),
+            });
+          }
+        }
+      }
+    }
+
+    // Sort by impact (count * tokens)
+    duplicateGroups.sort((a, b) => b.count * b.tokens - a.count * a.tokens);
+
+    return duplicateGroups.slice(0, this.options.maxBlocks);
+  }
+
+  /**
+   * Check if file is a code file
+   */
+  isCodeFile(filePath) {
+    const codeExtensions = [
+      '.js',
+      '.ts',
+      '.jsx',
+      '.tsx',
+      '.py',
+      '.java',
+      '.go',
+      '.rs',
+      '.cs',
+      '.php',
+      '.rb',
+    ];
+    const ext = filePath.substring(filePath.lastIndexOf('.')).toLowerCase();
+    return codeExtensions.includes(ext);
+  }
+
+  /**
+   * Check if block has significant code (not just comments/whitespace)
+   */
+  isSignificantBlock(content) {
+    const lines = content.split('\n');
+    let significantLines = 0;
+
+    for (const line of lines) {
+      const trimmed = line.trim();
+      // Skip empty lines and single-line comments
+      if (
+        trimmed &&
+        !trimmed.startsWith('//') &&
+        !trimmed.startsWith('#') &&
+        !trimmed.startsWith('/*')
+      ) {
+        significantLines++;
+      }
+    }
+
+    return significantLines >= 3;
+  }
+
+  /**
+   * Create a hash of a code block
+   */
+  hashBlock(content) {
+    let hash = 0;
+    const str = content;
+
+    for (let i = 0; i < str.length; i++) {
+      const char = str.charCodeAt(i);
+      hash = (hash << 5) - hash + char;
+      hash = hash & hash; // Convert to 32-bit integer
+    }
+
+    return hash.toString(16);
+  }
+
+  /**
+   * Normalize block for near-duplicate detection
+   */
+  normalizeBlock(content) {
+    return content
+      .replace(/\s+/g, ' ') // Normalize whitespace
+      .replace(/[a-zA-Z_][a-zA-Z0-9_]*/g, 'VAR') // Replace identifiers
+      .replace(/"[^"]*"/g, 'STR') // Replace strings
+      .replace(/'[^']*'/g, 'STR')
+      .replace(/\d+/g, 'NUM') // Replace numbers
+      .trim();
+  }
+
+  /**
+   * Estimate tokens for content
+   */
+  estimateTokens(content) {
+    return Math.ceil(content.length / 3.5);
+  }
+
+  /**
+   * Get preview of code block
+   */
+  getPreview(content, maxLength = 60) {
+    const firstLine = content.split('\n')[0].trim();
+    if (firstLine.length <= maxLength) {
+      return firstLine;
+    }
+    return firstLine.substring(0, maxLength - 3) + '...';
+  }
+
+  /**
+   * Get statistics about duplicates
+   */
+  getStats() {
+    return {
+      totalGroups: this.duplicates.length,
+      exactDuplicates: this.duplicates.filter((d) => d.type === 'exact').length,
+      similarCode: this.duplicates.filter((d) => d.type === 'similar').length,
+      wastedTokens: this.duplicates.reduce((sum, d) => sum + d.tokens * (d.count - 1), 0),
+    };
+  }
+}
+
+export default DuplicateDetector;

package/lib/analyzers/go-method-analyzer.js

@@ -6,121 +6,162 @@
 import path from 'path';
 
 class GoMethodAnalyzer {
-    extractMethods(content, filePath) {
-        const patterns = [
-            // Regular function: func FunctionName(params) returnType { ... }
-            { regex: /func\s+([A-Z_][a-zA-Z0-9_]*)\s*\(/g, type: 'function' },
-
-            // Method (receiver function): func (r *Receiver) MethodName(params) returnType { ... }
-            { regex: /func\s+\([^)]+\)\s+([A-Z_][a-zA-Z0-9_]*)\s*\(/g, type: 'method' },
-
-            // Interface method declarations: MethodName(params) returnType
-            { regex: /^\s+([A-Z_][a-zA-Z0-9_]*)\s*\([^)]*\)\s*(?:[a-zA-Z0-9_*\[\]]+)?/gm, type: 'interface' },
-        ];
-
-        const methodsMap = new Map();
-        const processedLines = new Map();
-
-        patterns.forEach(({ regex, type }) => {
-            let match;
-            while ((match = regex.exec(content)) !== null) {
-                const line = this.getLineNumber(content, match.index);
-                const methodName = match[1];
-
-                if (methodName && !this.isKeyword(methodName)) {
-                    const key = `${methodName}:${line}`;
-
-                    // Skip if this line was already processed
-                    if (processedLines.has(key)) {
-                        continue;
-                    }
-
-                    // Validate it's not inside a comment
-                    if (this.isInComment(content, match.index)) {
-                        continue;
-                    }
-
-                    methodsMap.set(key, {
-                        name: methodName,
-                        line: line,
-                        file: path.relative(process.cwd(), filePath),
-                        type: type
-                    });
-                    processedLines.set(key, true);
-                }
-            }
-        });
-
-        return Array.from(methodsMap.values());
-    }
-
-    getLineNumber(content, index) {
-        return content.substring(0, index).split('\n').length;
-    }
-
-    isInComment(content, index) {
-        // Check if position is inside a comment
-        const beforeContent = content.substring(0, index);
-        const lines = beforeContent.split('\n');
-        const currentLine = lines[lines.length - 1];
-
-        // Single-line comment check
-        if (currentLine.includes('//')) {
-            const commentPos = currentLine.indexOf('//');
-            const matchPos = currentLine.length;
-            if (commentPos < matchPos) {
-                return true;
-            }
+  extractMethods(content, filePath) {
+    const patterns = [
+      // Regular function: func FunctionName(params) returnType { ... }
+      { regex: /func\s+([A-Z_][a-zA-Z0-9_]*)\s*\(/g, type: 'function' },
+
+      // Method (receiver function): func (r *Receiver) MethodName(params) returnType { ... }
+      { regex: /func\s+\([^)]+\)\s+([A-Z_][a-zA-Z0-9_]*)\s*\(/g, type: 'method' },
+
+      // Interface method declarations: MethodName(params) returnType
+      {
+        regex: /^\s+([A-Z_][a-zA-Z0-9_]*)\s*\([^)]*\)\s*(?:[a-zA-Z0-9_*\[\]]+)?/gm,
+        type: 'interface',
+      },
+    ];
+
+    const methodsMap = new Map();
+    const processedLines = new Map();
+
+    patterns.forEach(({ regex, type }) => {
+      let match;
+      while ((match = regex.exec(content)) !== null) {
+        const line = this.getLineNumber(content, match.index);
+        const methodName = match[1];
+
+        if (methodName && !this.isKeyword(methodName)) {
+          const key = `${methodName}:${line}`;
+
+          // Skip if this line was already processed
+          if (processedLines.has(key)) {
+            continue;
+          }
+
+          // Validate it's not inside a comment
+          if (this.isInComment(content, match.index)) {
+            continue;
+          }
+
+          methodsMap.set(key, {
+            name: methodName,
+            line: line,
+            file: path.relative(process.cwd(), filePath),
+            type: type,
+          });
+          processedLines.set(key, true);
         }
-
-        // Multi-line comment check (simple heuristic)
-        const beforeText = content.substring(0, index);
-        const openComments = (beforeText.match(/\/\*/g) || []).length;
-        const closeComments = (beforeText.match(/\*\//g) || []).length;
-
-        return openComments > closeComments;
-    }
-
-    isKeyword(name) {
-        // Go keywords and common types
-        const keywords = new Set([
-            'break', 'case', 'chan', 'const', 'continue', 'default', 'defer',
-            'else', 'fallthrough', 'for', 'func', 'go', 'goto', 'if',
-            'import', 'interface', 'map', 'package', 'range', 'return',
-            'select', 'struct', 'switch', 'type', 'var',
-            // Common types
-            'bool', 'byte', 'complex64', 'complex128', 'error', 'float32',
-            'float64', 'int', 'int8', 'int16', 'int32', 'int64', 'rune',
-            'string', 'uint', 'uint8', 'uint16', 'uint32', 'uint64', 'uintptr'
-        ]);
-        return keywords.has(name);
+      }
+    });
+
+    return Array.from(methodsMap.values());
+  }
+
+  getLineNumber(content, index) {
+    return content.substring(0, index).split('\n').length;
+  }
+
+  isInComment(content, index) {
+    // Check if position is inside a comment
+    const beforeContent = content.substring(0, index);
+    const lines = beforeContent.split('\n');
+    const currentLine = lines[lines.length - 1];
+
+    // Single-line comment check
+    if (currentLine.includes('//')) {
+      const commentPos = currentLine.indexOf('//');
+      const matchPos = currentLine.length;
+      if (commentPos < matchPos) {
+        return true;
+      }
     }
 
-    extractMethodContent(content, methodName) {
-        // Extract function/method body
-        const patterns = [
-            new RegExp(`(func\\s+${methodName}\\s*\\([^)]*\\)[^{]*\\{)`, 'g'),
-            new RegExp(`(func\\s+\\([^)]+\\)\\s+${methodName}\\s*\\([^)]*\\)[^{]*\\{)`, 'g')
-        ];
-
-        for (const pattern of patterns) {
-            const match = pattern.exec(content);
-            if (match) {
-                // Find the matching closing brace
-                let braceCount = 1;
-                let pos = match.index + match[0].length;
-
-                while (braceCount > 0 && pos < content.length) {
-                    if (content[pos] === '{') braceCount++;
-                    if (content[pos] === '}') braceCount--;
-                    pos++;
-                }
-
-                return content.substring(match.index, pos);
-            }
+    // Multi-line comment check (simple heuristic)
+    const beforeText = content.substring(0, index);
+    const openComments = (beforeText.match(/\/\*/g) || []).length;
+    const closeComments = (beforeText.match(/\*\//g) || []).length;
+
+    return openComments > closeComments;
+  }
+
+  isKeyword(name) {
+    // Go keywords and common types
+    const keywords = new Set([
+      'break',
+      'case',
+      'chan',
+      'const',
+      'continue',
+      'default',
+      'defer',
+      'else',
+      'fallthrough',
+      'for',
+      'func',
+      'go',
+      'goto',
+      'if',
+      'import',
+      'interface',
+      'map',
+      'package',
+      'range',
+      'return',
+      'select',
+      'struct',
+      'switch',
+      'type',
+      'var',
+      // Common types
+      'bool',
+      'byte',
+      'complex64',
+      'complex128',
+      'error',
+      'float32',
+      'float64',
+      'int',
+      'int8',
+      'int16',
+      'int32',
+      'int64',
+      'rune',
+      'string',
+      'uint',
+      'uint8',
+      'uint16',
+      'uint32',
+      'uint64',
+      'uintptr',
+    ]);
+    return keywords.has(name);
+  }
+
+  extractMethodContent(content, methodName) {
+    // Extract function/method body
+    const patterns = [
+      new RegExp(`(func\\s+${methodName}\\s*\\([^)]*\\)[^{]*\\{)`, 'g'),
+      new RegExp(`(func\\s+\\([^)]+\\)\\s+${methodName}\\s*\\([^)]*\\)[^{]*\\{)`, 'g'),
+    ];
+
+    for (const pattern of patterns) {
+      const match = pattern.exec(content);
+      if (match) {
+        // Find the matching closing brace
+        let braceCount = 1;
+        let pos = match.index + match[0].length;
+
+        while (braceCount > 0 && pos < content.length) {
+          if (content[pos] === '{') braceCount++;
+          if (content[pos] === '}') braceCount--;
+          pos++;
         }
-        return null;
+
+        return content.substring(match.index, pos);
+      }
     }
+    return null;
+  }
 }
 
 export default GoMethodAnalyzer;