cntx-ui 3.0.2 → 3.0.4

package/lib/file-system-manager.js

@@ -3,7 +3,7 @@
  * Handles file operations, pattern matching, and directory traversal
  */
 
-import { readdirSync, statSync, existsSync, watch } from 'fs';
+import { readdirSync, readFileSync, statSync, existsSync, watch } from 'fs';
 import { join, relative, extname, basename, dirname } from 'path';
 
 export default class FileSystemManager {

package/lib/heuristics-manager.js

@@ -127,7 +127,7 @@ export default class HeuristicsManager {
     const domains = new Set();
     const name = func.name.toLowerCase();
     const path = (func.pathParts || []).join('/').toLowerCase();
-    const imports = func.includes?.imports || [];
+    const imports = (func.includes?.imports || []).filter(i => typeof i === 'string');
 
     // Path-based domains
     if (path.includes('auth')) domains.add('authentication');
@@ -330,7 +330,9 @@ export default class HeuristicsManager {
     if (condition.includes('chunk.imports.includes(')) {
       const importMatch = condition.match(/chunk\.imports\.includes\(['"]([^'"]+)['"]\)/)
       if (importMatch && func.includes?.imports) {
-        return func.includes.imports.some(imp => imp.includes(importMatch[1]))
+        return func.includes.imports
+          .filter(i => typeof i === 'string')
+          .some(imp => imp.includes(importMatch[1]))
       }
     }
 

package/lib/semantic-splitter.js

@@ -6,7 +6,6 @@
 
 import { readFileSync, existsSync } from 'fs'
 import { join, extname } from 'path'
-import glob from 'glob'
 import Parser from 'tree-sitter'
 import JavaScript from 'tree-sitter-javascript'
 import TypeScript from 'tree-sitter-typescript'
@@ -45,18 +44,14 @@ export default class SemanticSplitter {
 
   /**
    * Main entry point - extract semantic chunks from project
+   * Now accepts a pre-filtered list of files from FileSystemManager
    */
-  async extractSemanticChunks(projectPath, patterns = ['**/*.{js,jsx,ts,tsx,mjs}'], bundleConfig = null) {
-    console.log('🔪 Starting surgical semantic splitting via tree-sitter...');
-    console.log(`📂 Project path: ${projectPath}`);
-    console.log(`📋 Patterns: ${patterns}`);
+  async extractSemanticChunks(projectPath, files = [], bundleConfig = null) {
+    console.log('🔪 Starting surgical semantic splitting via tree-sitter...')
+    console.log(`📂 Project path: ${projectPath}`)
     
-    this.bundleConfig = bundleConfig;
-    const files = this.findFiles(projectPath, patterns);
-    console.log(`📁 Found ${files.length} files to split`);
-    if (files.length > 0) {
-      console.log('📄 Sample files:', files.slice(0, 5));
-    }
+    this.bundleConfig = bundleConfig
+    console.log(`📁 Processing ${files.length} filtered files`)
     
     const allChunks = []
     
@@ -74,48 +69,32 @@ export default class SemanticSplitter {
       summary: {
         totalFiles: files.length,
         totalChunks: allChunks.length,
-        averageSize: allChunks.reduce((sum, c) => sum + c.code.length, 0) / allChunks.length
+        averageSize: allChunks.length > 0 ? allChunks.reduce((sum, c) => sum + c.code.length, 0) / allChunks.length : 0
       },
       chunks: allChunks
     }
   }
 
-  findFiles(projectPath, patterns) {
-    const files = []
-    for (const pattern of patterns) {
-      const matches = glob.sync(pattern, {
-        cwd: projectPath,
-        ignore: ['node_modules/**', 'dist/**', '.git/**', '*.test.*', '*.spec.*']
-      })
-      files.push(...matches)
-    }
-    return [...new Set(files)]
-  }
-
   processFile(relativePath, projectPath) {
-    const fullPath = join(projectPath, relativePath);
-    if (!existsSync(fullPath)) return [];
+    const fullPath = join(projectPath, relativePath)
+    if (!existsSync(fullPath)) return []
     
-    // console.log(`  📄 Processing: ${relativePath}`);
-    const content = readFileSync(fullPath, 'utf8');
-    const parser = this.getParser(relativePath);
-    const tree = parser.parse(content);
-    const root = tree.rootNode;
+    const content = readFileSync(fullPath, 'utf8')
+    const parser = this.getParser(relativePath)
+    const tree = parser.parse(content)
+    const root = tree.rootNode
     
     const elements = {
       functions: [],
       types: [],
       imports: this.extractImports(root, content, relativePath)
-    };
+    }
 
     // Traverse AST for functions and types
-    this.traverse(root, content, relativePath, elements);
+    this.traverse(root, content, relativePath, elements)
 
-    const chunks = this.createChunks(elements, content, relativePath);
-    // if (chunks.length > 0) {
-    //   console.log(`    ✅ Found ${chunks.length} chunks in ${relativePath}`);
-    // }
-    return chunks;
+    // Create chunks from elements
+    return this.createChunks(elements, content, relativePath)
   }
 
   traverse(node, content, filePath, elements) {
@@ -235,10 +214,6 @@ export default class SemanticSplitter {
       const technicalPatterns = this.heuristicsManager.inferTechnicalPatterns(heuristicContext);
       const tags = this.generateTags(func);
 
-      // if (businessDomain.length > 0) {
-      //   console.log(`  💎 ${func.name}: Domains: [${businessDomain}], Patterns: [${technicalPatterns}]`);
-      // }
-
       let chunkCode = '';
       if (this.options.includeContext) {
         const relevantImports = elements.imports

package/lib/simple-vector-store.js

@@ -54,20 +54,30 @@ export default class SimpleVectorStore {
     const queryEmbedding = await this.generateEmbedding(query);
     
     // Load all embeddings from DB
-    // Optimization: In a huge codebase, we'd use a real vector DB or FAISS
-    // For now, SQLite + Manual Cosine Similarity is fine for local repos
     const rows = this.db.db.prepare('SELECT chunk_id, embedding FROM vector_embeddings WHERE model_name = ?').all(this.modelName);
     
     const results = [];
-    for (const row of rows) {
-      const embedding = new Float32Array(row.embedding.buffer, row.embedding.byteOffset, row.embedding.byteLength / 4);
-      const similarity = this.cosineSimilarity(queryEmbedding, embedding);
+    const batchSize = 100;
+    
+    // Process in batches to prevent blocking the event loop
+    for (let i = 0; i < rows.length; i += batchSize) {
+      const batch = rows.slice(i, i + batchSize);
+      
+      for (const row of batch) {
+        const embedding = new Float32Array(row.embedding.buffer, row.embedding.byteOffset, row.embedding.byteLength / 4);
+        const similarity = this.cosineSimilarity(queryEmbedding, embedding);
+        
+        if (similarity >= threshold) {
+          results.push({
+            chunkId: row.chunk_id,
+            similarity
+          });
+        }
+      }
       
-      if (similarity >= threshold) {
-        results.push({
-          chunkId: row.chunk_id,
-          similarity
-        });
+      // Give other tasks a chance to run
+      if (i + batchSize < rows.length) {
+        await new Promise(resolve => setImmediate(resolve));
       }
     }
 

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "cntx-ui",
   "type": "module",
-  "version": "3.0.2",
+  "version": "3.0.4",
   "description": "Autonomous Repository Intelligence engine with web UI and MCP server. Unified semantic code understanding, local RAG, and agent working memory.",
   "keywords": [
     "repository-intelligence",

package/server.js

@@ -4,7 +4,7 @@
  */
 
 import { createServer } from 'http';
-import { join, dirname } from 'path';
+import { join, dirname, relative, extname } from 'path';
 import { fileURLToPath } from 'url';
 import { existsSync, mkdirSync, readFileSync, writeFileSync, copyFileSync, cpSync } from 'fs';
 import * as fs from 'fs';
@@ -435,13 +435,17 @@ export class CntxServer {
 
     // 2. Perform fresh analysis if DB is empty
     try {
-      const patterns = ['**/*.{js,jsx,ts,tsx,mjs}'];
+      const supportedExtensions = ['.js', '.jsx', '.ts', '.tsx', '.mjs'];
+      const files = this.fileSystemManager.getAllFiles()
+        .filter(f => supportedExtensions.includes(extname(f).toLowerCase()))
+        .map(f => relative(this.CWD, f));
+      
       let bundleConfig = null;
       if (existsSync(this.configManager.CONFIG_FILE)) {
         bundleConfig = JSON.parse(readFileSync(this.configManager.CONFIG_FILE, 'utf8'));
       }
 
-      this.semanticCache = await this.semanticSplitter.extractSemanticChunks(this.CWD, patterns, bundleConfig);
+      this.semanticCache = await this.semanticSplitter.extractSemanticChunks(this.CWD, files, bundleConfig);
       this.lastSemanticAnalysis = Date.now();
 
       // 3. Persist chunks to SQLite immediately