apps/backend/src/scripts/ingest-drive.ts

import * as dotenv from 'dotenv';
import * as path from 'path';
import { ingestRepository } from '../services/GraphIngestor.js';
import { neo4jAdapter } from '../adapters/Neo4jAdapter.js';

dotenv.config({ path: path.resolve(process.cwd(), 'apps/backend/.env') });

const FORBIDDEN_PATHS = [
    /^c:[\/\\]?$/i, // C:, C:\, C:/
    /^c:[\/\\]windows/i,
    /^c:[\/\\]program files/i,
    /^c:[\/\\]program files \(x86\)/i,
    /^c:[\/\\]users[\/\\][^\\\/]+[\/\\]appdata/i // AppData
];

async function main() {
    const args = process.argv.slice(2);
    const targetPath = args[0];

    if (!targetPath) {
        console.error('❌ Usage: npm run ingest-drive -- <path>');
        console.error('Example: npm run ingest-drive -- "C:\\Users\\claus\\Documents"');
        process.exit(1);
    }

    // Safety checks
    const normalizedPath = path.normalize(targetPath);

    for (const forbidden of FORBIDDEN_PATHS) {
        if (forbidden.test(normalizedPath)) {
            console.error(`❌ DENIED: Ingesting '${normalizedPath}' is restricted for safety reasons.`);
            console.error('Please point to a specific subdirectory (e.g., Documents, Projects).');
            process.exit(1);
        }
    }

    console.log(`🚀 Starting ingestion of: ${normalizedPath}`);
    console.log('NOTE: This may take a while depending on drive size.');

    try {
        const result = await ingestRepository({
            rootPath: normalizedPath,
            maxDepth: 10,
            parseContent: true, // Read file content
            generateEmbeddings: true, // Generate vectors
            includePatterns: ['*'],
            excludePatterns: [
                'node_modules', '.git', 'dist', 'build', 'coverage',
                '$Recycle.Bin', 'System Volume Information', 'pagefile.sys', 'swapfile.sys',
                '.vscode', '.idea'
            ]
        });

        if (result.success) {
            console.log('✅ Ingestion completed successfully!');
            console.log('Stats:', result.stats);
        } else {
            console.error('❌ Ingestion finished with errors:', result.errors);
        }

    } catch (error) {
        console.error('💥 Fatal error:', error);
    } finally {
        await neo4jAdapter.close();
        process.exit(0);
    }
}

main();