scai 0.1.24 → 0.1.26

package/dist/commands/DaemonCmd.js

@@ -4,25 +4,29 @@ import { log } from '../utils/log.js';
 import { spawn } from 'child_process';
 import { fileURLToPath } from 'url';
 import path from 'path';
-// 🚀 Ensure daemon starts in the background
 export async function startDaemon() {
+    // If there's a PID file, check if the process is still running
     if (fsSync.existsSync(PID_PATH)) {
-        log(`⚠️ Daemon already running (PID file found at ${PID_PATH}). Skipping launch.`);
-    }
-    else {
-        log('🚀 Starting summarizer daemon in background mode...');
-        log(`📝 Logs will be saved to: ${LOG_PATH}`);
-        // Before starting the background process, set the environment variable
-        process.env.BACKGROUND_MODE = 'true'; // Set the mode to background
-        // Compute absolute path to the background worker (adjust path if needed)
-        const __filename = fileURLToPath(import.meta.url);
-        const __dirname = path.dirname(__filename);
-        const daemonWorkerPath = path.join(__dirname, '../daemon/daemonWorker.js');
-        // Spawn the daemonWorker.js file in the background
-        const child = spawn(process.execPath, [daemonWorkerPath], {
-            detached: true, // Detach the process so it runs independently
-            stdio: ['ignore', 'ignore', 'ignore'], // Suppress the output
-        });
-        child.unref(); // Allow the parent process to exit without waiting for the child
+        const pid = parseInt(fsSync.readFileSync(PID_PATH, 'utf8'));
+        try {
+            process.kill(pid, 0); // Check if process exists
+            log(`⚠️ Daemon already running with PID ${pid}.`);
+            return;
+        }
+        catch {
+            log(`⚠️ Stale PID file found. Removing and restarting daemon...`);
+            fsSync.unlinkSync(PID_PATH);
+        }
     }
+    log('🚀 Starting summarizer daemon in background mode...');
+    log(`📝 Logs will be saved to: ${LOG_PATH}`);
+    process.env.BACKGROUND_MODE = 'true';
+    const __filename = fileURLToPath(import.meta.url);
+    const __dirname = path.dirname(__filename);
+    const daemonWorkerPath = path.join(__dirname, '../daemon/daemonWorker.js');
+    const child = spawn(process.execPath, [daemonWorkerPath], {
+        detached: true,
+        stdio: ['ignore', 'ignore', 'ignore'],
+    });
+    child.unref();
 }

package/dist/commands/StopDaemonCmd.js

@@ -9,15 +9,26 @@ export async function runStopDaemonCommand() {
     }
     const pid = parseInt(fs.readFileSync(PID_PATH, 'utf-8'), 10);
     if (isNaN(pid)) {
-        console.error('⚠️ Invalid PID file.');
+        console.error('⚠️ Invalid PID file. Removing it...');
+        fs.unlinkSync(PID_PATH);
+        return;
+    }
+    try {
+        // Check if process exists
+        process.kill(pid, 0);
+    }
+    catch {
+        console.warn(`⚠️ No running process with PID ${pid}. Removing stale PID file.`);
+        fs.unlinkSync(PID_PATH);
         return;
     }
     try {
-        process.kill(pid);
+        // Attempt to terminate the process
+        process.kill(pid, 'SIGTERM');
         fs.unlinkSync(PID_PATH);
         console.log(`✅ Daemon process ${pid} stopped.`);
     }
     catch (err) {
-        console.error(`❌ Failed to stop process ${pid}:`, err instanceof Error ? err.message : err);
+        console.error(`❌ Failed to stop daemon process ${pid}:`, err instanceof Error ? err.message : err);
     }
 }

package/dist/config/IgnoredExtensions.js

@@ -1,46 +1,68 @@
-// src/config/IgnoredExtensions.ts
 export const IGNORED_EXTENSIONS = [
-    // 🖼 Media
+    // Media
     '.png', '.jpg', '.jpeg', '.gif', '.webp', '.svg', '.ico',
     '.mp4', '.mp3', '.mov', '.avi', '.mkv', '.flv', '.wav', '.flac',
     '.aac', '.m4a', '.wma', '.3gp', '.webm', '.ogg', '.aiff', '.au',
-    // 📦 Archives & install packages
+    // Archives & install packages
     '.zip', '.tar', '.gz', '.bz2', '.xz', '.rar', '.7z',
     '.jar', '.war', '.ear', '.deb', '.rpm', '.pkg', '.msi', '.dmg', '.cab', '.apk',
     '.tar.gz', '.tar.bz2', '.tar.xz', '.tar.lzma', '.tar.zst',
-    // 🧱 Binaries & executables
+    // Binaries & executables
     '.exe', '.dll', '.bin', '.so', '.dylib', '.a', '.lib',
     '.iso', '.img', '.elf', '.o', '.obj', '.msm', '.vbs', '.jscript',
     '.cmd', '.bat', '.ps1', '.sh', '.bash', '.run',
-    // 🧪 Runtime / build / cache
+    // Runtime / build / cache
     '.log', '.tmp', '.map',
     '.db', '.sqlite', '.pkl', '.sav', '.rdb', '.ldb',
     '.pyc', '.class', '.tsbuildinfo', '.coverage', '.eslintcache',
     '.yarn', '.webpack', '.babel', '.compilercache',
-    // 🔤 Fonts & styles
+    // Fonts & styles
     '.woff', '.woff2', '.ttf', '.eot', '.otf', '.css.map',
     '.scss', '.sass', '.less', '.styl',
-    // 🔐 Certs, keys, credentials
+    // Certs, keys, credentials
     '.crt', '.key', '.pem', '.pub', '.asc', '.gpg', '.p12', '.csr', '.der', '.pfx',
-    // ♻️ Backups / temp
+    // Backups / temp
     '.bak', '.old', '.swp', '.swo', '.orig',
     '.sublime-workspace', '.sublime-project', '.db-shm', '.db-wal',
-    // 🌐 System/config folders (still ignored by path, not extension)
+    // System/config folders (still ignored by path, not extension)
     '.DS_Store', '.bundle', '.npmrc',
-    // 🗺️ GIS / Geospatial
+    // GIS / Geospatial
     '.shp', '.shx', '.dbf', '.prj', '.qix', '.sbn', '.sbx', '.shp.xml', '.cpg', '.gpkg', '.mif', '.mid',
-    // 📊 Enterprise BI / Reporting
+    // Enterprise BI / Reporting
     '.pbix', '.rdl', '.rpt', '.bqy', '.iqy',
-    // 🧪 ETL / DWH / Modeling
+    // ETL / DWH / Modeling
     '.abf', '.dtsx', '.bim', '.xmi',
-    // 🏗️ CAD / Engineering
+    // CAD / Engineering
     '.dwg', '.dxf', '.step', '.stp', '.sldprt', '.sldasm',
     '.iges', '.igs', '.3ds', '.fbx',
-    // 🧾 Forms / Print / Publishing
+    // Forms / Print / Publishing
     '.xps', '.afpub', '.pub', '.indd', '.qxd', '.frm', '.frx', '.frl',
-    // 💰 ERP / Finance / Legacy DB
+    // ERP / Finance / Legacy DB
     '.mbd', '.fdb', '.nav', '.accdb', '.mdb', '.gdb',
     '.sap', '.sappkg', '.qbw', '.qbb',
-    // 🔒 Lock files (but NOT lock *configs*)
+    // Lock files (but NOT lock *configs*)
     '.lck', '.lockfile', '.db-lock', '.pid', '.socket',
+    // Documentation and Readme Files
+    '.md', // Markdown files for documentation
+    '.rst', // reStructuredText files for documentation
+    '.txt', // Plain text files (can be used for documentation or notes)
+    // Configuration files (JSON/YAML/XML)
+    '.json', // General JSON config files
+    '.yaml', // YAML configuration files
+    '.yml', // YAML configuration files
+    '.xml', // XML configuration files
+    // Miscellaneous Important Files
+    '.properties', // Java properties files (configuration)
+    '.xml.bck', // Backup XML files
+    '.bak', // General backup files
+    '.swp', // Swap files (e.g., vim/sublime temp files)
+    '.swo', // Swap files
+    '.orig', // Original backup files
+    '.old', // Old version files
+    '.~', // Temp files generated by some editors (e.g., vim)
+    '.md5', // MD5 checksums
+    '.sha1', // SHA1 checksums
+    '.sha256', // SHA256 checksums
+    '.tmp', // Temporary files
+    '.bak', // Backup file (for various applications)
 ];

package/dist/config/StopWords.js

@@ -1,15 +1,28 @@
-// src/config/StopWords.ts
 /**
  * These common words are ignored from search queries
  * to reduce noise and improve FTS and embedding match quality.
  */
 export const STOP_WORDS = new Set([
-    'a', 'an', 'and', 'are', 'as', 'at', 'be', 'but', 'by',
-    'for', 'if', 'in', 'into', 'is', 'it', 'no', 'not',
-    'of', 'on', 'or', 'such', 'that', 'the', 'their',
-    'then', 'there', 'these', 'they', 'this', 'to', 'was',
-    'will', 'with', 'what', 'which', 'who', 'whom', 'where',
-    'when', 'why', 'how', 'from', 'all', 'any', 'can',
-    'did', 'do', 'has', 'have', 'i', 'me', 'my', 'you',
-    'your', 'we', 'us', 'our'
+    // Articles & conjunctions
+    'a', 'an', 'and', 'but', 'or', 'nor',
+    // Prepositions
+    'at', 'by', 'for', 'from', 'in', 'into', 'of', 'on', 'to', 'with', 'about', 'above', 'below', 'under', 'over', 'through',
+    // Pronouns
+    'i', 'me', 'my', 'mine', 'you', 'your', 'yours', 'he', 'him', 'his', 'she', 'her', 'hers', 'it', 'its',
+    'we', 'us', 'our', 'ours', 'they', 'them', 'their', 'theirs',
+    // Determiners
+    'this', 'that', 'these', 'those', 'some', 'any', 'each', 'every', 'either', 'neither',
+    // Auxiliary and modal verbs
+    'am', 'is', 'are', 'was', 'were', 'be', 'been', 'being',
+    'do', 'does', 'did',
+    'have', 'has', 'had',
+    'can', 'could', 'shall', 'should', 'will', 'would', 'may', 'might', 'must',
+    // Wh-words and generic question words
+    'what', 'which', 'who', 'whom', 'whose', 'where', 'when', 'why', 'how',
+    // Misc common functional words
+    'as', 'if', 'than', 'then', 'there', 'because', 'so', 'just', 'only', 'not', 'no',
+    // Very common verbs to strip from noisy queries
+    'use', 'get', 'make', 'need', 'want', 'let', 'help', 'work', 'see', 'look', 'like', 'know',
+    // Other frequent noise
+    'all', 'more', 'most', 'many', 'much', 'such', 'also', 'again'
 ]);

package/dist/utils/removeIgnoredFiles.js

@@ -0,0 +1,43 @@
+import Database from 'better-sqlite3';
+import path from 'path';
+import os from 'os';
+import { IGNORED_EXTENSIONS } from '../config/IgnoredExtensions.js';
+import { specificFileExceptions } from './specificFileExceptions.js';
+// THIS FILE IS MEANT TO BE RUN AS A NODE JS SCRIPT. node dist/src/utilsremoveIgnoredFiles.js
+// It removes wrongly indexed files that don't add value to the model. 
+const DB_PATH = path.join(os.homedir(), '.scai', 'db.sqlite');
+const db = new Database(DB_PATH);
+console.log('🧹 Removing files with ignored extensions from the database...');
+// === Remove Files with Ignored Extensions, Excluding Specific Exceptions ===
+IGNORED_EXTENSIONS.forEach(ext => {
+    try {
+        const filesToDelete = db.prepare(`
+      SELECT path FROM files WHERE path LIKE ?
+    `).all(`%${ext}`);
+        ;
+        let deletedCount = 0;
+        filesToDelete.forEach(file => {
+            // Check if the file is in the exception list
+            if (!specificFileExceptions.includes(file.path)) {
+                // Delete the file from the database
+                const deleted = db.prepare(`DELETE FROM files WHERE path = ?`).run(file.path);
+                if (deleted.changes > 0) {
+                    deletedCount++;
+                }
+            }
+            else {
+                console.log(`⚠️ Skipped file (exception): ${file.path}`);
+            }
+        });
+        if (deletedCount > 0) {
+            console.log(`✅ Removed ${deletedCount} files with extension: ${ext}`);
+        }
+        else {
+            console.log(`⚠️ No deletions for files with extension: ${ext}`);
+        }
+    }
+    catch (err) {
+        console.error("❌ Failed to remove files with extension ${ext}:", err instanceof Error ? err.message : err);
+    }
+});
+console.log('🧹 Finished removing ignored files.');

package/dist/utils/shouldIgnoreFiles.js

@@ -1,6 +1,14 @@
 import path from 'path';
 import { IGNORED_EXTENSIONS } from '../config/IgnoredExtensions.js';
+import { specificFileExceptions } from '../utils/specificFileExceptions.js';
 export function shouldIgnoreFile(filePath) {
+    // Get file extension
     const ext = path.extname(filePath).toLowerCase();
+    // Check if the file is explicitly listed in the exceptions
+    const fileName = path.basename(filePath);
+    if (specificFileExceptions.includes(fileName)) {
+        return false; // Don't ignore if it's in the exceptions list
+    }
+    // If not in exceptions, check against ignored extensions
     return IGNORED_EXTENSIONS.includes(ext);
 }

package/dist/utils/specificFileExceptions.js

@@ -0,0 +1,90 @@
+export const specificFileExceptions = [
+    // 🧑‍💻 Project Configuration Files
+    'package.json', // Keep package.json for NPM/Yarn dependency management
+    'package-lock.json', // Keep package-lock.json for npm lockfile
+    'yarn.lock', // Keep yarn.lock for Yarn dependency lockfile
+    'pnpm-lock.yaml', // Keep pnpm-lock.yaml for pnpm lockfile
+    'tsconfig.json', // Keep TypeScript configuration file
+    'tsconfig.build.json', // Keep build-specific tsconfig file
+    'tsconfig.prod.json', // Keep production-specific tsconfig file
+    'tsconfig.dev.json', // Keep development-specific tsconfig file
+    'jsconfig.json', // Keep jsconfig.json for JavaScript projects
+    'eslint.json', // Keep eslint configuration
+    'eslint.config.js', // Keep eslint config file
+    'babel.config.js', // Keep Babel configuration
+    'webpack.config.js', // Keep Webpack configuration
+    'webpack.dev.config.js', // Keep development-specific Webpack config
+    'webpack.prod.config.js', // Keep production-specific Webpack config
+    'rollup.config.js', // Keep Rollup configuration file
+    'gulpfile.js', // Keep Gulp task runner file
+    'Makefile', // Keep Makefile for project builds
+    // 🧪 Docker & CI/CD
+    'Dockerfile', // Keep Dockerfile
+    'Dockerfile.dev', // Keep Dockerfile for development
+    'docker-compose.yaml', // Keep docker-compose.yaml for container orchestration
+    'docker-compose.yml', // Keep docker-compose.yml (common variation)
+    'ci.yml', // Keep CI configuration file (e.g., GitHub Actions)
+    'gitlab-ci.yml', // Keep GitLab CI configuration file
+    'Jenkinsfile', // Keep Jenkins pipeline file
+    'circleci/config.yml', // Keep CircleCI configuration file
+    // 📜 Documentation and Readme Files
+    'README.md', // Keep README file for project documentation
+    'README.rst', // Keep README in reStructuredText format
+    'CONTRIBUTING.md', // Keep contributing guidelines
+    'CHANGELOG.md', // Keep changelog for tracking project history
+    'LICENSE', // Keep project license
+    'LICENSE.txt', // Keep license in text format
+    'LICENSE.md', // Keep license in markdown format
+    'NOTICE.txt', // Keep NOTICE file
+    'INSTALL.md', // Keep installation instructions
+    // 🛠️ Build and Deployment Configuration Files
+    'build.gradle', // Keep Gradle build file
+    'pom.xml', // Keep Maven Project Object Model (POM) file
+    'settings.gradle', // Keep Gradle settings file
+    'build.sh', // Keep shell script for building the project
+    'build.bash', // Keep bash build script
+    'deploy.sh', // Keep shell script for deployment
+    'ci.sh', // Keep shell script for CI
+    // 🔧 Other Project Files
+    'Makefile.am', // Keep Automake Makefile
+    'config.yaml', // Keep general config file in YAML format
+    'config.json', // Keep general config file in JSON format
+    'config.toml', // Keep TOML configuration file
+    'settings.json', // Keep settings configuration file
+    'settings.yml', // Keep settings configuration file in YAML format
+    'secrets.json', // Keep secrets (make sure they are handled securely)
+    // 📂 Web Development & Frontend
+    'index.html', // Keep main HTML file
+    'index.php', // Keep main PHP file
+    'app.js', // Keep main JavaScript entry file
+    'app.ts', // Keep main TypeScript entry file
+    'styles.css', // Keep main CSS file
+    'main.scss', // Keep main SCSS file
+    'main.less', // Keep main LESS file
+    'style.css', // Keep style CSS
+    'app.vue', // Keep Vue.js file
+    'index.vue', // Keep Vue.js index file
+    // 🛠️ Miscellaneous Important Files
+    'README.txt', // Keep documentation in text format
+    'data.json', // Keep data JSON file
+    'data.yml', // Keep data YAML file
+    'env.json', // Keep environment JSON file
+    'env.yml', // Keep environment YAML file
+    '.env', // Keep environment variable files
+    '.env.local', // Keep local environment variables
+    '.env.production', // Keep production environment variables
+    '.env.development', // Keep development environment variables
+    // 🚧 Test-related files
+    'test.config.js', // Keep test config for testing frameworks
+    'test-utils.js', // Keep test utility files
+    'test.setup.js', // Keep setup files for tests
+    'jest.setup.js', // Keep Jest setup files
+    'mocha.setup.js', // Keep Mocha setup files
+    'karma.conf.js', // Keep Karma configuration for tests
+    'cypress.json', // Keep Cypress config for end-to-end testing
+    'karma.conf.js', // Keep Karma test runner config
+    'tests.js', // Keep test file
+    'tests.ts', // Keep TypeScript test file
+    'test.js', // Keep JavaScript test file
+    'test.ts', // Keep TypeScript test file
+];

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "scai",
-  "version": "0.1.24",
+  "version": "0.1.26",
   "type": "module",
   "bin": {
     "scai": "./dist/index.js"