aiwf 0.3.0 → 0.3.2

package/lib/task-analyzer.js

@@ -0,0 +1,599 @@
+/**
+ * Task Analyzer
+ * Analyzes tasks to determine the optimal AI persona
+ */
+
+import fs from 'fs/promises';
+import path from 'path';
+import { EventEmitter } from 'events';
+import { execSync } from 'child_process';
+
+class TaskAnalyzer extends EventEmitter {
+  constructor(options = {}) {
+    super();
+    
+    this.options = {
+      recentFileLimit: options.recentFileLimit || 10,
+      keywordWeight: options.keywordWeight || 0.4,
+      patternWeight: options.patternWeight || 0.3,
+      contextWeight: options.contextWeight || 0.2,
+      historyWeight: options.historyWeight || 0.1,
+      ...options
+    };
+
+    // Pattern definitions for each persona
+    this.patterns = {
+      architect: {
+        keywords: [
+          'design', 'architect', 'structure', 'system', 'scalab',
+          'interface', 'api', 'contract', 'schema', 'pattern',
+          'module', 'component', 'integration', 'dependency'
+        ],
+        filePatterns: ['*.md', '*.yml', '*.yaml', 'architecture/**', 'design/**'],
+        contextIndicators: ['creating new feature', 'planning', 'designing']
+      },
+      
+      debugger: {
+        keywords: [
+          'bug', 'error', 'fix', 'issue', 'problem', 'debug',
+          'trace', 'stack', 'exception', 'fail', 'crash',
+          'investigate', 'diagnose', 'troubleshoot'
+        ],
+        filePatterns: ['*.log', '*.test.*', 'test/**', 'debug/**'],
+        contextIndicators: ['error in', 'not working', 'broken', 'failing']
+      },
+      
+      reviewer: {
+        keywords: [
+          'review', 'audit', 'check', 'quality', 'standard',
+          'security', 'vulnerability', 'smell', 'refactor',
+          'improve', 'optimize', 'clean', 'lint'
+        ],
+        filePatterns: ['*.js', '*.ts', '*.py', 'src/**', 'lib/**'],
+        contextIndicators: ['review code', 'check quality', 'pull request']
+      },
+      
+      documenter: {
+        keywords: [
+          'document', 'docs', 'readme', 'guide', 'tutorial',
+          'explain', 'describe', 'comment', 'annotation',
+          'example', 'usage', 'api', 'reference'
+        ],
+        filePatterns: ['*.md', 'README*', 'docs/**', 'examples/**'],
+        contextIndicators: ['write documentation', 'explain', 'document']
+      },
+      
+      optimizer: {
+        keywords: [
+          'optimiz', 'performance', 'speed', 'efficient',
+          'benchmark', 'profile', 'memory', 'token',
+          'slow', 'bottleneck', 'improve', 'fast'
+        ],
+        filePatterns: ['*.bench.*', 'benchmark/**', 'perf/**'],
+        contextIndicators: ['optimize', 'improve performance', 'too slow']
+      },
+      
+      developer: {
+        keywords: [
+          'implement', 'create', 'build', 'develop',
+          'add', 'feature', 'function', 'code',
+          'write', 'make', 'new', 'update'
+        ],
+        filePatterns: ['*.js', '*.ts', '*.py', 'src/**', 'lib/**'],
+        contextIndicators: ['implement', 'create', 'build', 'add feature']
+      }
+    };
+
+    // Compile regex patterns
+    this.compiledPatterns = {};
+    for (const [persona, config] of Object.entries(this.patterns)) {
+      this.compiledPatterns[persona] = {
+        keywordRegex: new RegExp(config.keywords.join('|'), 'i'),
+        contextRegex: new RegExp(config.contextIndicators.join('|'), 'i')
+      };
+    }
+  }
+
+  /**
+   * Analyze a task to determine characteristics
+   */
+  async analyzeTask(taskContext) {
+    const analysis = {
+      command: taskContext.command || '',
+      description: taskContext.description || '',
+      recentFiles: await this.getRecentFiles(),
+      keywords: this.extractKeywords(taskContext),
+      gitContext: await this.getGitContext(),
+      errorState: await this.checkErrorState(),
+      fileTypes: await this.analyzeFileTypes(),
+      scores: {},
+      timestamp: new Date()
+    };
+
+    // Calculate scores for each persona
+    for (const personaName of Object.keys(this.patterns)) {
+      analysis.scores[personaName] = await this.calculatePersonaScore(
+        personaName,
+        analysis
+      );
+    }
+
+    // Determine primary task type
+    analysis.primaryType = this.determinePrimaryType(analysis.scores);
+
+    this.emit('taskAnalyzed', analysis);
+    return analysis;
+  }
+
+  /**
+   * Extract keywords from task context
+   */
+  extractKeywords(taskContext) {
+    const text = `${taskContext.command || ''} ${taskContext.description || ''}`;
+    
+    // Split into words and filter
+    const words = text
+      .toLowerCase()
+      .split(/\s+/)
+      .filter(word => word.length > 2)
+      .filter(word => !this.isStopWord(word));
+
+    // Find matching keywords for any persona
+    const keywords = [];
+    for (const [persona, config] of Object.entries(this.patterns)) {
+      for (const keyword of config.keywords) {
+        if (words.some(word => word.includes(keyword))) {
+          keywords.push(keyword);
+        }
+      }
+    }
+
+    return [...new Set(keywords)]; // Remove duplicates
+  }
+
+  /**
+   * Check if word is a stop word
+   */
+  isStopWord(word) {
+    const stopWords = [
+      'the', 'is', 'at', 'which', 'on', 'and', 'a', 'an',
+      'as', 'are', 'been', 'by', 'for', 'from', 'has', 'have',
+      'in', 'it', 'of', 'or', 'that', 'to', 'was', 'will', 'with'
+    ];
+    return stopWords.includes(word);
+  }
+
+  /**
+   * Get recently modified files
+   */
+  async getRecentFiles() {
+    try {
+      // Use git to get recently modified files
+      const output = execSync(
+        `git log --name-only --pretty=format: -${this.options.recentFileLimit} | sort | uniq | grep -v '^$'`,
+        { encoding: 'utf-8', cwd: process.cwd() }
+      );
+      
+      return output
+        .split('\n')
+        .filter(file => file.trim())
+        .slice(0, this.options.recentFileLimit);
+    } catch (error) {
+      // Fallback to file system scan
+      return await this.scanRecentFiles();
+    }
+  }
+
+  /**
+   * Scan file system for recent files
+   */
+  async scanRecentFiles() {
+    const files = [];
+    const now = Date.now();
+    const maxAge = 7 * 24 * 60 * 60 * 1000; // 7 days
+
+    async function scan(dir, level = 0) {
+      if (level > 3) return; // Max depth
+      
+      try {
+        const entries = await fs.readdir(dir, { withFileTypes: true });
+        
+        for (const entry of entries) {
+          if (entry.name.startsWith('.') || entry.name === 'node_modules') {
+            continue;
+          }
+          
+          const fullPath = path.join(dir, entry.name);
+          
+          if (entry.isDirectory()) {
+            await scan(fullPath, level + 1);
+          } else {
+            const stats = await fs.stat(fullPath);
+            if (now - stats.mtime.getTime() < maxAge) {
+              files.push({
+                path: path.relative(process.cwd(), fullPath),
+                modified: stats.mtime
+              });
+            }
+          }
+        }
+      } catch (error) {
+        // Ignore errors
+      }
+    }
+
+    await scan(process.cwd());
+    
+    // Sort by modification time and return paths
+    return files
+      .sort((a, b) => b.modified - a.modified)
+      .slice(0, this.options.recentFileLimit)
+      .map(f => f.path);
+  }
+
+  /**
+   * Get git context
+   */
+  async getGitContext() {
+    const context = {
+      branch: null,
+      isPullRequest: false,
+      hasUncommittedChanges: false,
+      recentCommitTypes: []
+    };
+
+    try {
+      // Get current branch
+      context.branch = execSync('git branch --show-current', {
+        encoding: 'utf-8'
+      }).trim();
+
+      // Check if PR branch
+      context.isPullRequest = context.branch.includes('feature/') || 
+                             context.branch.includes('fix/') ||
+                             context.branch.includes('pr/');
+
+      // Check for uncommitted changes
+      const status = execSync('git status --porcelain', {
+        encoding: 'utf-8'
+      });
+      context.hasUncommittedChanges = status.trim().length > 0;
+
+      // Get recent commit types
+      const commits = execSync('git log --pretty=format:%s -10', {
+        encoding: 'utf-8'
+      }).split('\n');
+      
+      context.recentCommitTypes = commits
+        .map(msg => {
+          const match = msg.match(/^(\w+)(\(.*?\))?:/);
+          return match ? match[1] : null;
+        })
+        .filter(Boolean);
+    } catch (error) {
+      // Git not available or not a git repo
+    }
+
+    return context;
+  }
+
+  /**
+   * Check for error state
+   */
+  async checkErrorState() {
+    const errorState = {
+      hasErrors: false,
+      errorFiles: [],
+      testFailures: false
+    };
+
+    try {
+      // Check for error logs
+      const errorLogPatterns = ['error.log', 'errors.log', '*.error.log'];
+      for (const pattern of errorLogPatterns) {
+        try {
+          const files = await this.findFiles(pattern);
+          if (files.length > 0) {
+            errorState.hasErrors = true;
+            errorState.errorFiles.push(...files);
+          }
+        } catch (error) {
+          // Ignore
+        }
+      }
+
+      // Check test status
+      try {
+        execSync('npm test -- --passWithNoTests', {
+          encoding: 'utf-8',
+          stdio: 'pipe'
+        });
+      } catch (error) {
+        errorState.testFailures = true;
+        errorState.hasErrors = true;
+      }
+    } catch (error) {
+      // Ignore errors in error checking
+    }
+
+    return errorState;
+  }
+
+  /**
+   * Analyze file types in project
+   */
+  async analyzeFileTypes() {
+    const fileTypes = {
+      hasTests: false,
+      hasDocumentation: false,
+      hasConfiguration: false,
+      primaryLanguage: null,
+      frameworks: []
+    };
+
+    try {
+      // Check for test files
+      fileTypes.hasTests = await this.checkFileExists(['*.test.*', '*.spec.*', 'test/**']);
+
+      // Check for documentation
+      fileTypes.hasDocumentation = await this.checkFileExists(['*.md', 'docs/**']);
+
+      // Check for configuration
+      fileTypes.hasConfiguration = await this.checkFileExists(['*.config.*', '*.json', '*.yml']);
+
+      // Determine primary language
+      const languages = {
+        javascript: await this.countFiles(['*.js', '*.jsx']),
+        typescript: await this.countFiles(['*.ts', '*.tsx']),
+        python: await this.countFiles(['*.py']),
+        java: await this.countFiles(['*.java'])
+      };
+      
+      fileTypes.primaryLanguage = Object.entries(languages)
+        .sort(([,a], [,b]) => b - a)[0][0];
+
+      // Detect frameworks
+      if (await this.checkFileExists('package.json')) {
+        try {
+          const pkg = JSON.parse(
+            await fs.readFile(path.join(process.cwd(), 'package.json'), 'utf-8')
+          );
+          
+          const deps = { ...pkg.dependencies, ...pkg.devDependencies };
+          
+          if (deps.react) fileTypes.frameworks.push('react');
+          if (deps.vue) fileTypes.frameworks.push('vue');
+          if (deps.angular) fileTypes.frameworks.push('angular');
+          if (deps.express) fileTypes.frameworks.push('express');
+          if (deps.next) fileTypes.frameworks.push('nextjs');
+        } catch (error) {
+          // Ignore
+        }
+      }
+    } catch (error) {
+      // Ignore errors
+    }
+
+    return fileTypes;
+  }
+
+  /**
+   * Calculate score for a specific persona
+   */
+  async calculatePersonaScore(personaName, analysis) {
+    const config = this.patterns[personaName];
+    const compiled = this.compiledPatterns[personaName];
+    
+    let score = 0;
+
+    // Keyword matching in command/description
+    const text = `${analysis.command} ${analysis.description}`.toLowerCase();
+    const keywordMatches = (text.match(compiled.keywordRegex) || []).length;
+    score += keywordMatches * 10;
+
+    // Context indicator matching
+    if (compiled.contextRegex.test(text)) {
+      score += 20;
+    }
+
+    // File pattern matching
+    if (analysis.recentFiles?.length) {
+      const matchingFiles = analysis.recentFiles.filter(file =>
+        config.filePatterns.some(pattern => this.matchesFilePattern(file, pattern))
+      );
+      score += (matchingFiles.length / analysis.recentFiles.length) * 30;
+    }
+
+    // Special context bonuses
+    score += this.applyContextBonuses(personaName, analysis);
+
+    return Math.min(score, 100); // Cap at 100
+  }
+
+  /**
+   * Apply context-specific bonuses
+   */
+  applyContextBonuses(personaName, analysis) {
+    let bonus = 0;
+
+    switch (personaName) {
+      case 'debugger':
+        if (analysis.errorState?.hasErrors) bonus += 30;
+        if (analysis.errorState?.testFailures) bonus += 20;
+        if (analysis.gitContext?.recentCommitTypes.includes('fix')) bonus += 10;
+        break;
+
+      case 'reviewer':
+        if (analysis.gitContext?.isPullRequest) bonus += 25;
+        if (analysis.command?.includes('review')) bonus += 20;
+        break;
+
+      case 'architect':
+        if (analysis.command?.includes('design')) bonus += 20;
+        if (analysis.fileTypes?.hasConfiguration) bonus += 10;
+        if (!analysis.fileTypes?.hasTests) bonus += 15; // Early stage project
+        break;
+
+      case 'documenter':
+        if (analysis.fileTypes?.hasDocumentation) bonus += 15;
+        if (analysis.command?.includes('document')) bonus += 25;
+        break;
+
+      case 'optimizer':
+        if (analysis.command?.includes('slow')) bonus += 20;
+        if (analysis.command?.includes('performance')) bonus += 25;
+        break;
+    }
+
+    return bonus;
+  }
+
+  /**
+   * Determine primary task type from scores
+   */
+  determinePrimaryType(scores) {
+    const sorted = Object.entries(scores)
+      .sort(([,a], [,b]) => b - a);
+    
+    // If top score is significantly higher, return it
+    if (sorted[0][1] > sorted[1][1] * 1.5) {
+      return sorted[0][0];
+    }
+
+    // Otherwise, consider it mixed and return developer
+    return 'developer';
+  }
+
+  /**
+   * Calculate similarity between two tasks
+   */
+  calculateSimilarity(task1, task2) {
+    if (!task1 || !task2) return 0;
+
+    // Extract features from both tasks
+    const features1 = this.extractTaskFeatures(task1);
+    const features2 = this.extractTaskFeatures(task2);
+
+    // Calculate Jaccard similarity
+    const intersection = features1.filter(f => features2.includes(f));
+    const union = [...new Set([...features1, ...features2])];
+
+    return union.length > 0 ? intersection.length / union.length : 0;
+  }
+
+  /**
+   * Extract features from a task for similarity comparison
+   */
+  extractTaskFeatures(task) {
+    const features = [];
+
+    // Add keywords
+    if (task.keywords) {
+      features.push(...task.keywords);
+    }
+
+    // Add primary type
+    if (task.primaryType) {
+      features.push(`type:${task.primaryType}`);
+    }
+
+    // Add file types
+    if (task.fileTypes) {
+      if (task.fileTypes.primaryLanguage) {
+        features.push(`lang:${task.fileTypes.primaryLanguage}`);
+      }
+      if (task.fileTypes.frameworks) {
+        task.fileTypes.frameworks.forEach(fw => 
+          features.push(`framework:${fw}`)
+        );
+      }
+    }
+
+    return features;
+  }
+
+  /**
+   * Check if file exists matching patterns
+   */
+  async checkFileExists(patterns) {
+    for (const pattern of (Array.isArray(patterns) ? patterns : [patterns])) {
+      const files = await this.findFiles(pattern);
+      if (files.length > 0) return true;
+    }
+    return false;
+  }
+
+  /**
+   * Count files matching patterns
+   */
+  async countFiles(patterns) {
+    let count = 0;
+    for (const pattern of (Array.isArray(patterns) ? patterns : [patterns])) {
+      const files = await this.findFiles(pattern);
+      count += files.length;
+    }
+    return count;
+  }
+
+  /**
+   * Find files matching a pattern
+   */
+  async findFiles(pattern, dir = process.cwd(), results = []) {
+    try {
+      const entries = await fs.readdir(dir, { withFileTypes: true });
+      
+      for (const entry of entries) {
+        if (entry.name.startsWith('.') || entry.name === 'node_modules') {
+          continue;
+        }
+        
+        const fullPath = path.join(dir, entry.name);
+        
+        if (entry.isDirectory() && pattern.includes('**')) {
+          await this.findFiles(pattern, fullPath, results);
+        } else if (entry.isFile() && this.matchesFilePattern(entry.name, pattern)) {
+          results.push(path.relative(process.cwd(), fullPath));
+        }
+      }
+    } catch (error) {
+      // Ignore errors
+    }
+    
+    return results;
+  }
+
+  /**
+   * Check if file matches pattern
+   */
+  matchesFilePattern(file, pattern) {
+    // Simple glob pattern matching
+    const regex = pattern
+      .replace(/\*\*/g, '.*')
+      .replace(/\*/g, '[^/]*')
+      .replace(/\?/g, '.')
+      .replace(/\./g, '\\.');
+    
+    return new RegExp(regex).test(file);
+  }
+
+  /**
+   * Get task analysis for a command
+   */
+  async analyzeCommand(command) {
+    return await this.analyzeTask({
+      command,
+      description: '',
+      timestamp: new Date()
+    });
+  }
+
+  /**
+   * Cleanup resources
+   */
+  cleanup() {
+    this.removeAllListeners();
+  }
+}
+
+export default TaskAnalyzer;
+export { TaskAnalyzer };