fmode-voc-projects/scripts/tools/collect-tiktok-only.js

#!/usr/bin/env node
/**
 * 只跑 TikTok · 合并结果到已有 data/jiangzhong-*-voc.json
 * 用法：node scripts/tools/collect-tiktok-only.js [liver|probiotic|monkey|all]
 */
const fs = require('fs');
const path = require('path');
const { collectTikTokForProduct } = require('./collect-tiktok');

const PRODUCTS = [
    {
        id: 'liver',
        name: '江中肝纯片',
        keywords_en: ['milk thistle', 'liver support supplement', 'liver cleanse detox', 'silymarin'],
    },
    {
        id: 'probiotic',
        name: '儿童乳酸菌素片',
        keywords_en: ['kids probiotics', 'children probiotics gummy', 'probiotics gummies for kids', 'childrens digestive supplement'],
    },
    {
        id: 'monkey',
        name: '江中猴菇饮',
        keywords_en: ['lions mane mushroom supplement', 'lion mane mushroom', 'stomach gut health supplement', 'lions mane gummies'],
    },
];

async function main() {
    const target = process.argv[2] || 'all';
    const picks = target === 'all' ? PRODUCTS : PRODUCTS.filter((p) => p.id === target);
    if (!picks.length) {
        console.error(`❌ 未知 product: ${target}`);
        process.exit(1);
    }

    console.log('╔══════════════════════════════════════════════════════════╗');
    console.log('║   TikTok VOC 采集（Amazon 的内容端对标）                    ║');
    console.log('╚══════════════════════════════════════════════════════════╝');
    console.log(`🎯 目标：${picks.map((p) => p.id).join(', ')}`);

    for (const p of picks) {
        const jsonPath = path.resolve(__dirname, '..', '..', 'data', `jiangzhong-${p.id}-voc.json`);
        if (!fs.existsSync(jsonPath)) {
            console.error(`❌ ${jsonPath} 不存在，请先跑全量采集`);
            continue;
        }

        console.log(`\n━━━━ [${p.id}] ${p.name} ━━━━`);
        const t0 = Date.now();
        const tt = await collectTikTokForProduct(p);
        const elapsed = ((Date.now() - t0) / 1000).toFixed(1);

        const data = JSON.parse(fs.readFileSync(jsonPath, 'utf-8'));
        data.tiktok = tt;
        data._meta = data._meta || {};
        data._meta.tiktok_updated_at = new Date().toISOString();
        fs.writeFileSync(jsonPath, JSON.stringify(data, null, 2));

        const vcount = tt.merged_videos?.length || 0;
        const acount = tt.top_authors?.length || 0;
        const profiles = Object.keys(tt.author_profiles || {}).length;
        const postsAll = Object.values(tt.author_posts || {}).reduce((s, a) => s + (a?.length || 0), 0);
        const ccount = Object.values(tt.top_comments_by_video || {}).reduce((s, a) => s + (a?.length || 0), 0);
        const catUsers = Object.values(tt.category_users || {}).reduce((s, a) => s + (a?.length || 0), 0);

        console.log(`\n✅ [${p.id}] 完成 (${elapsed}s)`);
        console.log(`   · 视频 ${vcount} · 评论 ${ccount}`);
        console.log(`   · Top 作者 ${acount} · profiles ${profiles} · 作品 ${postsAll}`);
        console.log(`   · 品类用户 ${catUsers}`);
    }
}

main().catch((e) => { console.error('❌', e); process.exit(1); });