openclaw-voc-skill/scripts/tools/douyin-viral-script-analyzer.js

#!/usr/bin/env node
const fs = require('fs');
const path = require('path');

function parseArgs(argv) {
  const args = {};
  for (let i = 0; i < argv.length; i++) {
    const token = argv[i];
    if (!token.startsWith('--')) continue;
    const eq = token.indexOf('=');
    if (eq >= 0) {
      args[token.slice(2, eq)] = token.slice(eq + 1);
    } else {
      const key = token.slice(2);
      const next = argv[i + 1];
      if (next && !next.startsWith('--')) {
        args[key] = next;
        i++;
      } else {
        args[key] = true;
      }
    }
  }
  return args;
}

function usage() {
  return [
    'Usage:',
    '  node scripts/tools/douyin-viral-script-analyzer.js --input <douyin-json|dir> --output <out-dir>',
    '  node scripts/tools/douyin-viral-script-analyzer.js --text <transcript-or-desc> --output <out-dir>',
    '',
    'Outputs:',
    '  viral-script-analysis.json',
    '  viral-script-analysis.md'
  ].join('\n');
}

function ensureDir(dirPath) {
  fs.mkdirSync(dirPath, { recursive: true });
}

function readJson(filePath) {
  return JSON.parse(fs.readFileSync(filePath, 'utf8'));
}

function asArray(value) {
  if (!value) return [];
  return Array.isArray(value) ? value : [value];
}

function cleanText(value) {
  return String(value || '').replace(/\s+/g, ' ').trim();
}

function truncate(value, length = 120) {
  const text = cleanText(value);
  return text.length > length ? `${text.slice(0, length - 1)}…` : text;
}

function toNumber(value) {
  const number = Number(value || 0);
  return Number.isFinite(number) ? number : 0;
}

function splitList(value) {
  if (!value) return [];
  if (Array.isArray(value)) return value.map(String).map(item => item.trim()).filter(Boolean);
  const text = String(value).trim();
  if (!text) return [];
  if (text.startsWith('[')) {
    try {
      const parsed = JSON.parse(text);
      return Array.isArray(parsed) ? parsed.map(String).map(item => item.trim()).filter(Boolean) : [];
    } catch {
      return [];
    }
  }
  return text.split(/[,，;；|]/).map(item => item.trim()).filter(Boolean);
}

function loadInput(inputPath) {
  if (!inputPath) return {};
  const absolute = path.resolve(inputPath);
  const stat = fs.statSync(absolute);
  if (stat.isDirectory()) {
    const candidates = [
      'daily-report.json',
      'douyin-mini-voc.json',
      'raw.json',
      '_merged.json',
      'videos.json',
      'keyword-videos.json'
    ].map(name => path.join(absolute, name));
    const found = candidates.find(filePath => fs.existsSync(filePath));
    if (found) return readJson(found);
    const jsonFiles = fs.readdirSync(absolute)
      .filter(name => name.endsWith('.json'))
      .map(name => path.join(absolute, name));
    return jsonFiles.reduce((acc, filePath) => mergeData(acc, readJson(filePath)), {});
  }
  return readJson(absolute);
}

function mergeData(left, right) {
  return {
    metadata: { ...(left.metadata || {}), ...(right.metadata || {}) },
    videos: [
      ...asArray(left.videos),
      ...asArray(right.videos),
      ...asArray(right.accountVideos),
      ...asArray(right.items).filter(item => item.sourceType === 'video')
    ],
    comments: [...asArray(left.comments), ...asArray(right.comments)],
    replies: [...asArray(left.replies), ...asArray(right.replies)],
    transcripts: [...asArray(left.transcripts), ...asArray(right.transcripts)]
  };
}

function normalizeVideo(video, index) {
  const raw = video.raw || video;
  const statistics = video.statistics || raw.statistics || raw.stats || {};
  const author = video.author || raw.author || {};
  return {
    id: cleanText(video.aweme_id || video.awemeId || video.id || video.productId || raw.aweme_id || raw.awemeId || `video_${index}`),
    keyword: video.keyword || raw.keyword,
    title: cleanText(video.title || video.desc || video.content || raw.desc || raw.title),
    author: cleanText(typeof author === 'string' ? author : author.nickname || author.name || video.authorName),
    likeCount: toNumber(video.likeCount ?? video.digg_count ?? statistics.digg_count),
    commentCount: toNumber(video.commentCount ?? video.comment_count ?? statistics.comment_count),
    shareCount: toNumber(video.shareCount ?? video.share_count ?? statistics.share_count),
    playCount: toNumber(video.playCount ?? video.play_count ?? statistics.play_count),
    url: video.url || video.share_url || raw.share_url,
    raw
  };
}

function normalizeComment(comment, index) {
  const raw = comment.raw || comment;
  const user = comment.user || raw.user || {};
  return {
    id: cleanText(comment.cid || comment.commentId || comment.id || raw.cid || raw.id || `comment_${index}`),
    parentId: cleanText(comment.parentId || comment.aweme_id || raw.aweme_id || raw.item_id),
    text: cleanText(comment.text || comment.content || raw.text || raw.content),
    author: cleanText(comment.author || user.nickname || user.name),
    likeCount: toNumber(comment.likeCount ?? comment.digg_count ?? raw.digg_count),
    replyCount: toNumber(comment.replyCount ?? comment.reply_comment_total ?? raw.reply_comment_total),
    raw
  };
}

function normalizeTranscript(transcript) {
  if (!transcript) return undefined;
  if (typeof transcript === 'string') return { text: cleanText(transcript), segments: [] };
  return {
    awemeId: transcript.awemeId || transcript.aweme_id,
    text: cleanText(transcript.text),
    segments: asArray(transcript.segments)
  };
}

function resolveRecords(data, args) {
  const videos = asArray(data.videos)
    .concat(asArray(data.accountVideos))
    .concat(asArray(data.items).filter(item => item.sourceType === 'video'))
    .map(normalizeVideo);
  if (args.text) {
    videos.push({
      id: args.awemeId || args['aweme-id'] || 'manual_text',
      keyword: args.keyword,
      title: cleanText(args.text),
      author: args.author || '',
      likeCount: 0,
      commentCount: 0,
      shareCount: 0,
      playCount: 0,
      url: args.url || '',
      raw: {}
    });
  }
  const comments = asArray(data.comments).concat(asArray(data.replies)).map(normalizeComment);
  const transcripts = asArray(data.transcripts)
    .map(normalizeTranscript)
    .filter(Boolean);
  if (args.transcript) {
    const transcriptPath = path.resolve(args.transcript);
    const transcript = transcriptPath.endsWith('.json')
      ? readJson(transcriptPath)
      : { text: fs.readFileSync(transcriptPath, 'utf8') };
    transcripts.push(normalizeTranscript(transcript));
  }
  return { videos, comments, transcripts };
}

function firstSentence(text) {
  const cleaned = cleanText(text);
  const parts = cleaned.split(/[。！？!?；;\n]/).map(item => item.trim()).filter(Boolean);
  return parts[0] || cleaned.slice(0, 40);
}

function inferHookType(text) {
  const t = cleanText(text);
  if (/[?？]/.test(t) || /为什么|怎么|如何|到底|有没有/.test(t)) return '问题钩子';
  if (/别再|不要|千万|避坑|踩雷|错了|误区/.test(t)) return '反常识/避坑钩子';
  if (/我发现|亲测|真实|讲个|经历|以前|后来/.test(t)) return '个人经历钩子';
  if (/\d|一[个-龥]?招|三[个-龥]?点|5个|10个|清单/.test(t)) return '清单/数字钩子';
  if (/但是|其实|反而|不是.*而是|看起来/.test(t)) return '反转钩子';
  if (/爆|火|涨粉|成交|转化|收入|结果/.test(t)) return '结果承诺钩子';
  return '场景共鸣钩子';
}

function inferConflict(text) {
  const t = cleanText(text);
  const candidates = [
    ['避坑', '用户害怕踩坑，需要一个清晰判断标准'],
    ['焦虑', '用户有焦虑或不确定，需要降低决策风险'],
    ['贵', '价值感和价格接受度存在冲突'],
    ['没用', '用户担心结果不确定，需要证明和边界'],
    ['不会', '用户缺少操作路径，需要拆成步骤'],
    ['但是', '内容中存在反转或认知冲突，可作为口播中段转折'],
    ['为什么', '内容适合用问题驱动展开']
  ];
  const hit = candidates.find(([needle]) => t.includes(needle));
  return hit ? hit[1] : '围绕用户已有认知和真实结果制造轻冲突';
}

function evidenceFromStats(video) {
  const parts = [];
  if (video.playCount) parts.push(`播放 ${video.playCount}`);
  if (video.likeCount) parts.push(`点赞 ${video.likeCount}`);
  if (video.commentCount) parts.push(`评论 ${video.commentCount}`);
  if (video.shareCount) parts.push(`分享 ${video.shareCount}`);
  return parts.join(' / ') || '互动数据未提供';
}

function topQuestions(comments, limit = 3) {
  return comments
    .filter(comment => /[?？]|怎么|哪里|多少|能不能|有没有|适合|为什么/.test(comment.text))
    .sort((a, b) => (b.likeCount + b.replyCount * 2) - (a.likeCount + a.replyCount * 2))
    .slice(0, limit);
}

function commentsForVideo(comments, videoId, totalVideos) {
  const linked = comments.filter(comment => comment.parentId && comment.parentId === videoId);
  if (linked.length) return { scope: 'linked', comments: linked };
  if (totalVideos === 1 && comments.length) return { scope: 'single_video_unlinked', comments };
  return { scope: 'unlinked_ignored', comments: [] };
}

function inferEmotion(comments) {
  const text = comments.map(comment => comment.text).join(' ');
  if (/贵|便宜|值|性价比|价格/.test(text)) return '价值感犹豫';
  if (/怕|担心|焦虑|纠结|不敢|踩雷/.test(text)) return '风险规避';
  if (/哪里|怎么买|链接|求|想要|试试/.test(text)) return '行动意愿';
  if (/哈哈|笑|真实|太对|共鸣/.test(text)) return '共鸣互动';
  return '好奇追问';
}

function reusableFrame(video, hookType, conflict, questions) {
  const question = questions[0]?.text || '评论区最常见的问题';
  return [
    `开头：用「${truncate(firstSentence(video.title), 36)}」切入，先给出一个具体判断。`,
    `冲突：指出「${conflict}」，避免直接讲大道理。`,
    `论证：拆 2-3 个可验证原因，配合案例、数字或亲身观察。`,
    `互动：结尾抛出「${truncate(question, 34)}」这类问题，引导用户留言。`
  ];
}

function analyzeVideo(video, allComments, transcripts, totalVideos) {
  const transcript = transcripts.find(item => item.awemeId && item.awemeId === video.id) || transcripts[0];
  const sourceText = transcript?.text || video.title;
  const commentScope = commentsForVideo(allComments, video.id, totalVideos);
  const comments = commentScope.comments;
  const questions = topQuestions(comments);
  const hookType = inferHookType(sourceText);
  const conflict = inferConflict(sourceText);
  const openingPattern = firstSentence(sourceText);
  const frame = reusableFrame(video, hookType, conflict, questions);
  const warnings = [];
  if (!transcript?.text) warnings.push('未提供逐字稿，本条为基于视频描述、互动数据和评论的结构推断。');
  if (comments.length < 3) warnings.push('评论样本较少，评论触发点需谨慎使用。');
  if (commentScope.scope === 'unlinked_ignored') warnings.push('评论未匹配到该视频，未纳入本条拆解证据。');
  if (commentScope.scope === 'single_video_unlinked') warnings.push('评论缺少视频 ID，仅因样本只有一条视频而暂时归并。');

  return {
    videoId: video.id,
    keyword: video.keyword,
    title: video.title,
    author: video.author,
    url: video.url,
    hookType,
    openingPattern,
    conflict,
    proofPoint: evidenceFromStats(video),
    audienceEmotion: inferEmotion(comments),
    commentTrigger: questions.map(item => item.text),
    reusableFrame: frame,
    riskNotes: warnings,
    evidenceRefs: [
      { type: 'video', id: video.id, summary: evidenceFromStats(video), url: video.url },
      { type: 'comment_scope', scope: commentScope.scope, count: comments.length },
      ...questions.slice(0, 2).map(item => ({ type: 'comment', id: item.id, text: item.text }))
    ]
  };
}

function renderMarkdown(result) {
  const lines = [];
  lines.push('# 抖音爆款口播结构拆解');
  lines.push('');
  lines.push(`生成时间：${result.generatedAt}`);
  lines.push(`样本数：${result.summary.videoCount} 条视频 / ${result.summary.commentCount} 条评论 / ${result.summary.transcriptCount} 条逐字稿`);
  lines.push('');
  result.analyses.forEach((item, index) => {
    lines.push(`## ${index + 1}. ${item.title || item.videoId}`);
    lines.push('');
    lines.push(`- 作者：${item.author || '未标注'}`);
    lines.push(`- 关键词：${item.keyword || '未标注'}`);
    lines.push(`- 钩子类型：${item.hookType}`);
    lines.push(`- 开头模式：${item.openingPattern}`);
    lines.push(`- 核心冲突：${item.conflict}`);
    lines.push(`- 证明点：${item.proofPoint}`);
    lines.push(`- 用户情绪：${item.audienceEmotion}`);
    lines.push('');
    lines.push('### 可复用口播框架');
    item.reusableFrame.forEach(step => lines.push(`- ${step}`));
    if (item.commentTrigger.length) {
      lines.push('');
      lines.push('### 评论触发点');
      item.commentTrigger.forEach(text => lines.push(`- ${text}`));
    }
    if (item.riskNotes.length) {
      lines.push('');
      lines.push('### 风险边界');
      item.riskNotes.forEach(text => lines.push(`- ${text}`));
    }
    lines.push('');
  });
  return lines.join('\n');
}

function main() {
  const args = parseArgs(process.argv.slice(2));
  if (args.help) {
    console.log(usage());
    return;
  }
  const outputDir = path.resolve(args.output || path.join(process.cwd(), 'douyin-viral-script-analysis'));
  ensureDir(outputDir);
  const data = args.input ? loadInput(args.input) : {};
  const records = resolveRecords(data, args);
  const analyses = records.videos
    .slice(0, Number(args.limit || 10))
    .map(video => analyzeVideo(video, records.comments, records.transcripts, records.videos.length));
  const result = {
    status: analyses.length ? 'ok' : 'needs_data',
    generatedAt: new Date().toISOString(),
    summary: {
      videoCount: records.videos.length,
      commentCount: records.comments.length,
      transcriptCount: records.transcripts.length,
      analysisCount: analyses.length
    },
    analyses
  };
  const jsonPath = path.join(outputDir, 'viral-script-analysis.json');
  const mdPath = path.join(outputDir, 'viral-script-analysis.md');
  fs.writeFileSync(jsonPath, JSON.stringify(result, null, 2), 'utf8');
  fs.writeFileSync(mdPath, renderMarkdown(result), 'utf8');
  const final = { ...result.summary, status: result.status, outputDir, files: [jsonPath, mdPath] };
  console.log(JSON.stringify(final, null, 2));
}

try {
  main();
} catch (error) {
  console.error(error.message);
  process.exit(1);
}