broccolini-bot/scripts/bulk-lookup-users.js

#!/usr/bin/env node
/**
 * Bulk lookup Discord user information
 * 
 * Usage:
 *   node scripts/bulk-lookup-users.js <input_file> <output_file>
 * 
 * Input file: Text file with one user ID per line
 * Output file: JSON file with user lookup results
 */

const fs = require('fs');
const path = require('path');
const { Client, GatewayIntentBits } = require('discord.js');

// Load environment variables from repo root
const envPath = path.join(__dirname, '../../.env');
console.log(`Loading .env from: ${envPath}`);
const result = require('dotenv').config({ path: envPath });
if (result.error) {
    console.error(`Error loading .env: ${result.error.message}`);
    // Try broccolini-bot/.env as fallback
    require('dotenv').config({ path: path.join(__dirname, '../.env') });
}

const TOKEN = process.env.DISCORD_BOT_TOKEN || process.env.DISCORD_TOKEN;
const GUILD_ID = process.env.GUILD_ID || process.env.SERVER_ID;

if (!TOKEN) {
    console.error('Error: DISCORD_BOT_TOKEN must be set in .env');
    console.error('Available env vars:', Object.keys(process.env).filter(k => k.includes('DISCORD')));
    process.exit(1);
}

// Parse command line args
const args = process.argv.slice(2);
if (args.length < 2) {
    console.error('Usage: node scripts/bulk-lookup-users.js <input_file> <output_file>');
    process.exit(1);
}

const inputFile = args[0];
const outputFile = args[1];

// Read user IDs from input file
const userIds = fs.readFileSync(inputFile, 'utf-8')
    .split('\n')
    .map(line => line.trim())
    .filter(line => line.length > 0);

console.log(`Loaded ${userIds.length} user IDs from ${inputFile}`);

const client = new Client({
    intents: [
        GatewayIntentBits.Guilds,
        GatewayIntentBits.GuildMembers
    ]
});

const results = {};
let processed = 0;
let errors = 0;

async function lookupUser(userId) {
    try {
        // Add timeout to prevent hanging
        const timeoutPromise = new Promise((_, reject) => 
            setTimeout(() => reject(new Error('Lookup timeout')), 10000)
        );
        
        const fetchPromise = client.users.fetch(userId);
        const user = await Promise.race([fetchPromise, timeoutPromise]);
        
        return {
            success: true,
            id: user.id,
            username: user.username,
            globalName: user.globalName || user.username,
            tag: user.tag,
            bot: user.bot,
            avatar: user.displayAvatarURL()
        };
    } catch (error) {
        // Handle errors (not found, timeout, rate limit)
        if (error.message.includes('429')) {
            console.log(`    ⚠️  Rate limit hit for user ${userId}, will retry`);
        }
        return {
            success: false,
            id: userId,
            error: error.message,
            username: null,
            globalName: null,
            tag: null,
            bot: false
        };
    }
}

async function processUsers() {
    console.log('\nStarting bulk lookup...');
    console.log('This will take a few minutes for 2,428 users\n');
    
    const startTime = Date.now();
    
    // Process in batches to avoid rate limits
    const BATCH_SIZE = 3;  // Very small batches to avoid rate limits
    const DELAY_MS = 2000; // 2 seconds between batches
    
    for (let i = 0; i < userIds.length; i += BATCH_SIZE) {
        const batch = userIds.slice(i, i + BATCH_SIZE);
        
        // Lookup batch in parallel
        const promises = batch.map(userId => lookupUser(userId));
        const batchResults = await Promise.all(promises);
        
        // Store results
        batchResults.forEach(result => {
            if (!result.success) {
                errors++;
            }
            results[result.id] = result;
            processed++;
        });
        
        // Log every batch for debugging
        if (processed <= 50) {
            console.log(`  Batch complete: ${processed} users processed`);
        }
        
        // Progress update every 100 users
        if (processed % 100 === 0 || processed === userIds.length) {
            const elapsed = ((Date.now() - startTime) / 1000).toFixed(1);
            const rate = (processed / elapsed).toFixed(1);
            const remaining = ((userIds.length - processed) / rate).toFixed(0);
            console.log(`Progress: ${processed}/${userIds.length} (${errors} errors) - ${elapsed}s elapsed, ~${remaining}s remaining`);
        }
        
        // Wait before next batch to avoid rate limits
        if (i + BATCH_SIZE < userIds.length) {
            await new Promise(resolve => setTimeout(resolve, DELAY_MS));
        }
    }
    
    const totalTime = ((Date.now() - startTime) / 1000).toFixed(1);
    console.log(`\n✅ Completed in ${totalTime}s`);
    console.log(`   Successful: ${processed - errors}`);
    console.log(`   Errors: ${errors}`);
    
    // Save results
    const output = {
        timestamp: new Date().toISOString(),
        total_users: userIds.length,
        successful: processed - errors,
        errors: errors,
        users: results
    };
    
    fs.writeFileSync(outputFile, JSON.stringify(output, null, 2));
    console.log(`\n💾 Saved results to ${outputFile}`);
    
    process.exit(0);
}

client.once('ready', () => {
    console.log(`✅ Bot logged in as ${client.user.tag}`);
    processUsers();
});

client.on('error', (error) => {
    console.error('Discord client error:', error);
});

client.login(TOKEN);