tihao-ai/claude-code-tihao-sourcing/mcp/src/features/tihao-sourcing/brief-parser.js

const path = require('path');
const { readTextMaybe } = require('../../core/files');
const { extractPatternHits } = require('./brief-rule-patterns');

function parseBrief(input = {}) {
  const productContextText = normalizeTextInput(input.productIntro || input.productDescription || input.productContext || input.product);
  const chatContextText = normalizeTextInput(input.chatText || input.chatRecords || input.customerChat || input.conversation);
  const briefText = [
    input.briefText || input.text || '',
    input.brief || input.briefPath ? readTextMaybe(path.resolve(input.brief || input.briefPath)) : '',
    productContextText,
    chatContextText
  ].filter(Boolean).join('\n\n');

  const text = [
    briefText,
    input.brand,
    input.category,
    Array.isArray(input.keywords) ? input.keywords.join(' ') : input.keyword
  ].filter(Boolean).join('\n\n');

  const platforms = resolvePlatforms(input, text);
  const platformTargets = resolvePlatformTargets(text);
  const keywords = resolveKeywords(input, text);
  const targetCount = Number(input.targetCount || sumTargets(platformTargets) || resolveTargetCount(text) || 10);
  const budgetRange = resolveBudget(input, text);
  const fanRange = resolveFanRange(input, text);
  const exclusions = resolveList(input.excludeTags || input.exclusions, text, [
    '硬广', '泛美妆', '医美', '整形', '虚假宣传', '品牌纠纷', '夸大减肥', '低质搬运', '争议账号',
    '纯娱乐', '泛娱乐', '宠物', '装修', '婚礼', '纯自拍', '纯娱乐账号'
  ]);
  const blockedCreators = resolveList(input.blockedCreators, text, []);
  const referenceLinks = resolveReferenceLinks(input, text);
  const referenceAccounts = resolveReferenceAccounts(input, text);
  const referenceStyleAnchors = resolveReferenceStyleAnchors(text, referenceLinks, referenceAccounts);
  const referenceSignals = resolveReferenceSignals(text, referenceStyleAnchors);

  return {
    rawText: text,
    source: input.brief || input.briefPath || (briefText ? 'inline' : 'input'),
    contextSources: buildContextSources({ input, briefText, productContextText, chatContextText, referenceLinks, referenceAccounts }),
    brand: input.brand || resolveBrand(text) || resolveProduct(text) || '未命名品牌',
    category: input.category || inferCategory(text),
    platforms,
    platformTargets,
    keywords,
    targetCount,
    budgetRange,
    fanRange,
    region: input.region || inferRegion(text),
    gender: input.gender || (text.includes('女性') ? 'female' : ''),
    stylePreferences: resolveList(input.stylePreferences || input.preferredStyles, text, [
      '真实体验', '素人感', '成分拆解', '敏感肌', '通勤生活', '空瓶记录', '真实测评',
      '真实试吃', '配料表解释', '场景化种草', '生活化开箱', '真实吃法', '成分党',
      '真实使用', '前后对比', '步骤教程', '家居清洁', '收纳', '租房生活',
      '教程型', '案例型', '职场效率', '效率工具', '创业', '企业管理',
      '本地生活', '探店', '餐饮创业', '同城转化', '门店拍摄',
      'Vlog', '生活分享', '生活记录', '风景', '社交场景', '颜值', '自拍',
      '美食', '店铺', '门面', '美妆', '妆教', '妆容教程', '产品合集', '日常妆容',
      '时尚穿搭', '港风', '成熟风', '通勤穿搭', '女装', '男装', '精致生活'
    ]),
    referenceLinks,
    referenceAccounts,
    referenceStyleAnchors,
    referenceSignals,
    blockedCreators,
    exclusions,
    performanceRequirements: resolvePerformanceRequirements(text),
    briefRulePatterns: extractPatternHits(text),
    deliveryRequirements: ['账号名称', '平台', '粉丝数', '报价', '推荐理由', '风险提示', '分级']
  };
}

function resolvePerformanceRequirements(text) {
  const requirements = [];
  const cpe = text.match(/CPE\s*[≤<=]\s*(\d+(?:\.\d+)?)/i);
  if (cpe) requirements.push(`CPE≤${cpe[1]}`);
  const cp1 = text.match(/CP1\s*[≤<=]\s*(\d+(?:\.\d+)?)/i);
  if (cp1) requirements.push(`CP1≤${cp1[1]}`);
  const cpm = text.match(/CPM\s*[≤<=]\s*(\d+(?:\.\d+)?)/i);
  if (cpm) requirements.push(`CPM≤${cpm[1]}`);
  const cpc = text.match(/CPC\s*[≤<=<]\s*(\d+(?:\.\d+)?)/i);
  if (cpc) requirements.push(`CPC≤${cpc[1]}`);
  const cpv = text.match(/CPV\s*[≤<=<:：]\s*(\d+(?:\.\d+)?)/i);
  if (cpv) requirements.push(`CPV≤${cpv[1]}`);
  const recentPlays = text.match(/近\s*3\s*条[^\n]{0,10}(?:平均)?播放\s*[>＞≥]\s*(\d+(?:\.\d+)?)\s*万/);
  if (recentPlays) requirements.push(`近3条平均播放>${recentPlays[1]}万`);
  const readMedian = text.match(/阅读中位数\s*[≧≥>=]\s*(\d+(?:,\d{3})*)/);
  if (readMedian) requirements.push(`阅读中位数≥${readMedian[1]}`);
  const averageLikes = text.match(/(?:平均点赞|月均点赞|近\s*30\s*天平均点赞|近一个月平均点赞)[^\d]{0,12}(\d+(?:,\d{3})*)/);
  if (averageLikes) requirements.push(`近30天平均点赞≥${averageLikes[1]}`);
  if (/百赞以下|100赞以下|点赞低于\s*100|平均点赞低于\s*100/.test(text)) requirements.push('近30天平均点赞≥100');
  if (/近一个月(?:没有|没|未)发|停更一个月|一个月(?:没有|没|未)发布|近\s*30\s*天(?:没有|没|未)更新/.test(text)) requirements.push('近30天需有更新');
  if (/重复评论|同一个人评论|刷评|数据造假|数据真实性|数据波动/.test(text)) requirements.push('复核数据真实性/重复评论风险');
  const femaleRatio = text.match(/(?:女粉|女性粉丝|女性占比|女粉占比)[^\d]{0,8}(\d+)%/);
  if (femaleRatio) requirements.push(`女粉占比≥${femaleRatio[1]}%`);
  const maleRatio = text.match(/(?:男粉|男性粉丝|男性占比|男粉占比)[^\d]{0,8}(\d+)%/);
  if (maleRatio) requirements.push(`男粉占比≥${maleRatio[1]}%`);
  if (/报价.*(?:300|三百).*以上|300\s*元以上|高于\s*300|超过\s*300/.test(text)) requirements.push('报价300元以上需复核封面清晰度/背景统一/内容精致度');
  const readCost = text.match(/预估阅读单价\s*[≤<=]\s*(\d+(?:\.\d+)?)/);
  if (readCost) requirements.push(`预估阅读单价≤${readCost[1]}`);
  const homepageInteraction = text.match(/(?:主页互动数据|主页互动|互动数据|互动数)[^\d]{0,12}(\d+(?:,\d{3})*)\s*\+?/);
  if (homepageInteraction) requirements.push(`主页互动数据≥${homepageInteraction[1]}`);
  const audience = text.match(/(\d{2}-\d{2}岁占比\d+%以上)/);
  if (audience) requirements.push(audience[1]);
  const ageRatio = text.match(/(\d{2})-(\d{2})岁[^\n]{0,8}占比\s*(\d+)%\s*以上/);
  if (ageRatio) requirements.push(`${ageRatio[1]}-${ageRatio[2]}岁占比${ageRatio[3]}%以上`);
  const underAgeRatio = text.match(/(\d+)岁以下[^\n]{0,6}(?:的|占比)?[^\n]{0,6}(?:不能超过|不超过|≤|＜)\s*(\d+)%/);
  if (underAgeRatio) requirements.push(`${underAgeRatio[1]}岁以下占比≤${underAgeRatio[2]}%`);
  const interestTags = text.match(/兴趣[^\n]{0,6}(?:前三)?标签[^\n]{0,6}(?:含|包含)\s*(\d+)\s*个/);
  if (interestTags) requirements.push(`兴趣前三标签含${interestTags[1]}个`);
  const ipMustBe = text.match(/(?:达人)?\s*ip\s*(?:必须|要求|需|为|在)\s*([^\s，,。；\n]+)/i);
  if (ipMustBe) requirements.push(`达人IP必须${ipMustBe[1]}`);
  if (/(?:同|同一)个?省(?:份)?[^\n]{0,12}(?:不要|不|禁止)?\s*跨区域/.test(text)) requirements.push('同省KOL不要跨区域');
  return unique(requirements);
}

function normalizeTextInput(value) {
  if (!value) return '';
  if (Array.isArray(value)) return value.map(normalizeTextInput).filter(Boolean).join('\n');
  if (typeof value === 'object') return Object.values(value).map(normalizeTextInput).filter(Boolean).join('\n');
  return String(value || '').trim();
}

function buildContextSources({ input, briefText, productContextText, chatContextText, referenceLinks = [], referenceAccounts = [] }) {
  return {
    brief: Boolean(input.brief || input.briefPath || input.briefText || input.text || briefText),
    product: Boolean(productContextText),
    chat: Boolean(chatContextText),
    reference: Boolean((input.referenceLinks || []).length || (input.referenceAccounts || []).length || referenceLinks.length || referenceAccounts.length)
  };
}

function resolvePlatformTargets(text) {
  const targets = {};
  const tableRows = text.split(/\r?\n/)
    .filter(line => /^\|\s*(小红书|抖音|B站|bilibili|xiaohongshu|douyin)\s*\|/.test(line));
  for (const line of tableRows) {
    const cells = line.split('|').map(cell => cell.trim()).filter(Boolean);
    const platform = normalizePlatforms([cells[0]])[0];
    const count = Number((cells[1] || '').match(/(\d+)/)?.[1] || 0);
    if (platform && count) targets[platform] = count;
  }
  for (const match of text.matchAll(/(小红书|抖音|B站|bilibili|xiaohongshu|douyin)\s*[：:]?\s*(\d+)\s*(?:人|位|个|账号|达人)/g)) {
    const platform = normalizePlatforms([match[1]])[0];
    const count = Number(match[2] || 0);
    if (platform && count) targets[platform] = Math.max(Number(targets[platform] || 0), count);
  }
  return targets;
}

function sumTargets(platformTargets) {
  return Object.values(platformTargets || {}).reduce((sum, count) => sum + Number(count || 0), 0);
}

function resolveBrand(text) {
  const lines = text.split(/\r?\n/).map(cleanMarkdownLine).filter(Boolean);
  const explicit = lines.find(line => /^(客户名称|客户|品牌名称|品牌)[：:]/.test(line));
  if (explicit) return explicit.replace(/^(客户名称|客户|品牌名称|品牌)[：:]\s*/, '').trim();
  const productLine = lines.find(line => /本次推广产品|推广产品/.test(line));
  if (productLine) {
    const product = productLine.replace(/^(本次推广产品|推广产品)[：:]\s*/, '').trim();
    const match = product.match(/^(.{2,12}?)(?:修护|精华|面膜|口红|产品)/);
    if (match) return match[1];
  }
  return '';
}

function resolveProduct(text) {
  const lines = text.split(/\r?\n/).map(cleanMarkdownLine).filter(Boolean);
  const explicit = lines.find(line => /^(产品|推广产品|本次推广产品)[：:]/.test(line));
  return explicit ? explicit.replace(/^(产品|推广产品|本次推广产品)[：:]\s*/, '').trim() : '';
}

function cleanMarkdownLine(line) {
  return String(line || '')
    .replace(/^\s*[-*#]+\s*/, '')
    .replace(/\*\*/g, '')
    .replace(/\s+/g, ' ')
    .trim();
}

function resolveTargetCount(text) {
  const tableRows = text.split(/\r?\n/)
    .filter(line => /^\|\s*(小红书|抖音|B站|bilibili|xiaohongshu|douyin)\s*\|/.test(line));
  const counts = tableRows
    .map(line => {
      const cells = line.split('|').map(cell => cell.trim()).filter(Boolean);
      const countCell = cells[1] || '';
      const match = countCell.match(/(\d+)\s*(?:人|位|个|账号|达人)?/);
      return match ? Number(match[1]) : 0;
    })
    .filter(Boolean);
  if (counts.length) return counts.reduce((sum, count) => sum + count, 0);

  const direct = text.match(/(?:需求人数|推荐数量|目标数量)[^\d]{0,12}(\d+)\s*(?:人|位|个|账号|达人)?/);
  return direct ? Number(direct[1]) : 0;
}

function resolvePlatforms(input, text) {
  if (Array.isArray(input.platforms) && input.platforms.length) return normalizePlatforms(input.platforms);
  const platforms = [];
  if (/小红书|xiaohongshu|xhs/i.test(text)) platforms.push('xiaohongshu');
  if (/抖音|douyin/i.test(text)) platforms.push('douyin');
  if (/B站|bilibili|哔哩/i.test(text)) platforms.push('bilibili');
  return platforms.length ? platforms : ['xiaohongshu', 'douyin'];
}

function normalizePlatforms(platforms) {
  const map = {
    xhs: 'xiaohongshu',
    '小红书': 'xiaohongshu',
    xiaohongshu: 'xiaohongshu',
    '抖音': 'douyin',
    douyin: 'douyin',
    'B站': 'bilibili',
    bilibili: 'bilibili'
  };
  return [...new Set(platforms.map(item => map[String(item)] || String(item)))];
}

function resolveKeywords(input, text) {
  const provided = Array.isArray(input.keywords) ? input.keywords : String(input.keyword || '').split(/[,\s，、]+/);
  const categoryTermMap = {
    护肤美妆: ['敏感肌', '屏障', '修护', '护肤', '成分党', '成分拆解', '美妆', '测评', '空瓶', '胶原蛋白', '面膜', '精华', '防晒', '洁面', '妆教', '妆容教程', '步骤教程', '产品合集', '日常妆容', '口红', '彩妆'],
    母婴: ['DHA', '营养师', '育婴师', '妈妈', '母婴', '萌娃', '记录生活', '618合集', '干货科普', '儿童营养', '宝宝', '备孕'],
    食品餐饮: ['食品', '零食', '低糖', '健康', '控糖', '健身', '办公室零食', '试吃', '开箱', '真实吃法', '配料表', '成分党', '燕麦', '蛋白', '餐饮', '美食', '探店', '本地生活', '同城转化', '门店拍摄', '餐饮创业', '小面', '加盟', '店铺', '门面', '店面', '店招', '门头'],
    家居生活: ['家居', '家清', '清洁', '收纳', '租房', '家务', '厨房', '油污', '日化', '卫生间', '真实使用', '前后对比', '步骤教程', '妈妈'],
    B2B软件: ['SaaS', 'B2B', '项目管理', '效率工具', '创业', '企业管理', '职场效率', '教程型', '案例型', '协同办公', '软件', '工具'],
    服饰穿搭: ['时尚穿搭', '穿搭', '港风', '成熟风', '通勤穿搭', '女装', '男装', '服饰', '搭配', '精致生活', '职场穿搭', '轻熟', '高级感'],
    生活方式: ['种草', '国货', '通勤', '生活方式', '测评', '真实体验', '素人感', '开箱', '教程', '案例', 'Vlog', '生活分享', '生活记录', '风景', '社交场景', '颜值', '自拍', '精致生活']
  };
  const category = inferCategory(text);
  const commonTerms = [
    '种草', '测评', '开箱', '教程', '真实体验', '素人感', '生活方式',
    '女性', '妈妈', '职场', '同城', '本地生活', 'Vlog', '生活分享', '生活记录', '颜值'
  ];
  const domainTerms = [
    ...(categoryTermMap[category] || []),
    ...commonTerms,
    ...Object.values(categoryTermMap).flat()
  ];
  const explicitTerms = extractLabeledTerms(text);
  const hits = domainTerms.filter(term => text.includes(term));
  return unique([...provided, ...hits, ...explicitTerms]
    .map(cleanKeywordTerm)
    .filter(Boolean)).slice(0, 12);
}

function extractLabeledTerms(text) {
  const terms = [];
  for (const line of String(text || '').split(/\r?\n/).map(cleanMarkdownLine).filter(Boolean)) {
    const kv = parseKeyValueLine(line);
    if (!kv || !/关键词|关键字|风格|偏好|要求|重点|目标|任务|人群|内容/.test(kv.key)) continue;
    terms.push(...splitKeywordTerms(kv.value));
  }
  return terms;
}

function splitKeywordTerms(value) {
  return String(value || '')
    .replace(/找(小红书|抖音|B站|bilibili|xiaohongshu|douyin)?\s*\d+\s*(?:个|位|人|账号|达人)?/gi, ' ')
    .replace(/(小红书|抖音|B站|bilibili|xiaohongshu|douyin)\s*\d+\s*(?:个|位|人|账号|达人)?/gi, ' ')
    .replace(/\d+(?:\.\d+)?\s*(?:万|k|K)?\s*(?:-|到|至|~)\s*\d+(?:\.\d+)?\s*(?:万|k|K)?/g, ' ')
    .split(/[，,、；;\/|和与&\s]+/)
    .map(cleanKeywordTerm)
    .filter(Boolean);
}

function cleanKeywordTerm(value) {
  const cleaned = String(value || '')
    .replace(/^(找|寻找|需要|招募|重点看|要求|偏好|风格偏|不要|排除)/, '')
    .replace(/(?:博主|达人|账号|内容创作者|类内容创作者|类达人|类博主|流量号|经验|优先)$/g, '')
    .replace(/类$/g, '')
    .trim();
  if (cleaned.length < 2 || cleaned.length > 12) return '';
  if (/^(预算|粉丝|平台|人数|需求|目标|任务|产品|品牌|小红书|抖音|B站|bilibili|xiaohongshu|douyin)$/i.test(cleaned)) return '';
  return cleaned;
}

function resolveBudget(input, text) {
  if (input.budgetRange) return normalizeRange(input.budgetRange, { min: 0, max: 20000 });
  const upperBound = text.match(/(?:单个预算|预算|报价|单人预算|单篇预算)[^\d]{0,12}([\d,.]+)\s*(?:元|块|rmb|RMB)?\s*(?:内|以内|以下|封顶|上限)/i);
  if (upperBound) return { min: 0, max: toNumber(upperBound[1]) };
  const money = [...text.matchAll(/(?:¥|￥)?\s*([\d,.]+)\s*(?:-|到|至|~)\s*(?:¥|￥)?\s*([\d,.]+)/g)]
    .map(match => [toNumber(match[1]), toNumber(match[2])])
    .filter(([min, max]) => max > 0);
  const picked = money.find(([min, max]) => max >= 1000) || money[0];
  return picked ? { min: picked[0], max: picked[1] } : { min: 0, max: 20000 };
}

function resolveFanRange(input, text) {
  if (input.fanRange) return normalizeRange(input.fanRange, { min: 0, max: 500000 });
  const ranges = [];
  const lines = text.split(/\r?\n/);
  let inFanSection = false;
  for (const line of lines) {
    if (/粉丝量范围|粉丝范围|粉丝数范围/.test(line)) {
      inFanSection = true;
      continue;
    }
    if (inFanSection && /^#{1,4}\s+/.test(line) && !/粉丝/.test(line)) break;
    if (!inFanSection && !/粉丝|小红书|抖音|B站/.test(line)) continue;
    const match = line.match(/(\d+(?:\.\d+)?)\s*万?\s*(?:-|到|至|~)\s*(\d+(?:\.\d+)?)\s*万?\s*(?:粉|粉丝)?/);
    if (!match) continue;
    const hasWan = match[0].includes('万') || /粉丝/.test(line);
    const min = Number(match[1]) * (hasWan ? 10000 : 1);
    const max = Number(match[2]) * (hasWan ? 10000 : 1);
    if (max >= 10000) ranges.push({ min, max });
  }
  if (!ranges.length) return { min: 0, max: 500000 };
  return {
    min: Math.min(...ranges.map(range => range.min)),
    max: Math.max(...ranges.map(range => range.max))
  };
}

function normalizeRange(value, fallback) {
  return {
    min: Number(value.min ?? fallback.min),
    max: Number(value.max ?? fallback.max)
  };
}

function inferCategory(text) {
  if (/SaaS|B2B|项目管理|效率工具|企业管理|职场效率|协同办公|软件|CRM|企业级|数字化/.test(text)) return 'B2B软件';
  if (/家清|清洁|家居|收纳|洗衣|厨房|油污|卫生间|日化|租房|家务/.test(text)) return '家居生活';
  if (/食品|饮料|餐饮|美食|零食|低糖|控糖|燕麦|蛋白|试吃|配料表|探店|本地生活|同城|门店|小面|加盟|餐饮创业/.test(text)) return '食品餐饮';
  if (/母婴|宝宝|孕|育婴|备孕|DHA|儿童营养/.test(text)) return '母婴';
  if (/护肤|面膜|敏感肌|屏障|修护|精华|美妆|口红|胶原蛋白|洁面|防晒|泛红|刺痛/.test(text)) return '护肤美妆';
  if (/时尚穿搭|穿搭|港风|成熟风|通勤穿搭|女装|男装|服饰|搭配|轻熟|高级感/.test(text)) return '服饰穿搭';
  return '生活方式';
}

function inferRegion(text) {
  const cities = ['北京', '上海', '广州', '深圳', '杭州', '成都', '南京', '武汉', '重庆', '西安'];
  return cities.filter(city => text.includes(city));
}

function resolveList(inputList, text, vocabulary) {
  const provided = normalizeListInput(inputList);
  const hits = vocabulary.filter(item => text.includes(item));
  return [...new Set([...provided, ...hits].map(String).filter(Boolean))];
}

function normalizeListInput(inputList) {
  if (Array.isArray(inputList)) return inputList.flatMap(normalizeListInput);
  if (inputList == null || inputList === false) return [];
  if (typeof inputList === 'string') {
    return inputList.split(/[,，；;|]/).map(item => item.trim()).filter(Boolean);
  }
  return [inputList];
}

function toNumber(value) {
  return Number(String(value || '').replace(/,/g, '')) || 0;
}

function resolvePlatformTargets(text) {
  const targets = {};
  for (const line of text.split(/\r?\n/).map(cleanMarkdownLine).filter(Boolean)) {
    const cells = splitCells(line);
    if (cells.length >= 2) {
      const platform = normalizePlatforms([cells[0]])[0];
      const count = Number((cells[1] || '').match(/(\d+)/)?.[1] || 0);
      if (isKnownPlatform(platform) && count) targets[platform] = Math.max(Number(targets[platform] || 0), count);
    }
    for (const match of line.matchAll(/(小红书|抖音|B站|bilibili|xiaohongshu|douyin)\s*[：:]?\s*(\d+)\s*(?:人|位|个|账号|达人)/gi)) {
      const platform = normalizePlatforms([match[1]])[0];
      const count = Number(match[2] || 0);
      if (isKnownPlatform(platform) && count) targets[platform] = Math.max(Number(targets[platform] || 0), count);
    }
  }
  return targets;
}

function resolveBrand(text) {
  const lines = text.split(/\r?\n/).map(cleanMarkdownLine).filter(Boolean);
  for (const line of lines) {
    const kv = parseKeyValueLine(line);
    if (!kv) continue;
    if (/^(客户名称|客户|品牌名称|品牌)$/i.test(kv.key)) return kv.value;
  }
  for (const line of lines) {
    const kv = parseKeyValueLine(line);
    if (!kv || !/^(产品|推广产品|本次推广产品)$/i.test(kv.key)) continue;
    const match = kv.value.match(/^(.{2,18}?)(?:修护|精华|面膜|口红|产品|（|\()/);
    if (match) return match[1].trim();
    return kv.value;
  }
  return '';
}

function resolveProduct(text) {
  const lines = text.split(/\r?\n/).map(cleanMarkdownLine).filter(Boolean);
  for (const line of lines) {
    const kv = parseKeyValueLine(line);
    if (kv && /^(产品|推广产品|本次推广产品)$/i.test(kv.key)) return kv.value;
  }
  return '';
}

function resolveTargetCount(text) {
  const counts = Object.values(resolvePlatformTargets(text)).map(Number).filter(Boolean);
  if (counts.length) return counts.reduce((sum, count) => sum + count, 0);
  const direct = text.match(/(?:需求人数|推荐数量|目标数量)[^\d]{0,12}(\d+)\s*(?:人|位|个|账号|达人)?/);
  return direct ? Number(direct[1]) : 0;
}

function resolvePlatforms(input, text) {
  if (Array.isArray(input.platforms) && input.platforms.length) return normalizePlatforms(input.platforms);
  const platforms = [];
  if (/小红书|xiaohongshu|xhs/i.test(text)) platforms.push('xiaohongshu');
  if (/抖音|douyin/i.test(text)) platforms.push('douyin');
  if (/B站|bilibili|哔哩/i.test(text)) platforms.push('bilibili');
  return platforms.length ? [...new Set(platforms)] : ['xiaohongshu', 'douyin'];
}

function normalizePlatforms(platforms) {
  const map = {
    xhs: 'xiaohongshu',
    小红书: 'xiaohongshu',
    xiaohongshu: 'xiaohongshu',
    抖音: 'douyin',
    douyin: 'douyin',
    B站: 'bilibili',
    b站: 'bilibili',
    bilibili: 'bilibili',
    哔哩哔哩: 'bilibili'
  };
  return [...new Set(platforms.map(item => map[String(item).trim()] || String(item).trim()))];
}

function resolveFanRange(input, text) {
  if (input.fanRange) return normalizeRange(input.fanRange, { min: 0, max: 500000 });
  const ranges = [];
  for (const line of text.split(/\r?\n/).map(cleanMarkdownLine).filter(Boolean)) {
    if (!/粉丝|小红书|抖音|B站|bilibili|xiaohongshu|douyin/i.test(line)) continue;
    const cells = splitCells(line);
    const scan = isolateFanRangeText(cells.length >= 2 && /粉丝/.test(cells[0]) ? cells.slice(1).join(' ') : line);
    for (const match of scan.matchAll(/(\d+(?:\.\d+)?)\s*(万|w|W|k|K)?\s*(?:-|到|至|~)\s*(\d+(?:\.\d+)?)\s*(万|w|W|k|K)?/g)) {
      const min = normalizeFanCount(match[1], match[2], match[4]);
      const max = normalizeFanCount(match[3], match[4], match[2]);
      if (max >= 1000 && min > 0 && min <= max) ranges.push({ min, max });
    }
  }
  if (!ranges.length) return { min: 0, max: 500000 };
  return {
    min: Math.min(...ranges.map(range => range.min)),
    max: Math.max(...ranges.map(range => range.max))
  };
}

function isolateFanRangeText(line) {
  const text = String(line || '');
  const fanIndex = text.search(/粉丝|fans?/i);
  const platformOnly = fanIndex < 0;
  const start = platformOnly ? 0 : fanIndex;
  let scan = text.slice(start);
  const budgetIndex = scan.search(/预算|报价|单篇|费用|价格|投放|CPE|CPM|阅读单价|合作费|坑位费/i);
  if (budgetIndex >= 0) scan = scan.slice(0, budgetIndex);
  return scan;
}

function normalizeFanCount(value, unit, oppositeUnit) {
  const numeric = Number(value || 0);
  if (!numeric) return 0;
  const normalizedUnit = String(unit || '').toLowerCase();
  const normalizedOppositeUnit = String(oppositeUnit || '').toLowerCase();
  if (normalizedUnit === '万' || normalizedUnit === 'w') return Math.round(numeric * 10000);
  if (normalizedUnit === 'k') return Math.round(numeric * 1000);
  if (normalizedOppositeUnit === '万' || normalizedOppositeUnit === 'w') {
    return numeric >= 1000 ? Math.round(numeric) : Math.round(numeric * 10000);
  }
  if (normalizedOppositeUnit === 'k') return numeric >= 1000 ? Math.round(numeric) : Math.round(numeric * 1000);
  return Math.round(numeric);
}

function parseKeyValueLine(line) {
  const cells = splitCells(line);
  if (cells.length >= 2) return { key: cells[0], value: cells.slice(1).join(' | ').trim() };
  const match = line.match(/^(.{2,14}?)[：:]\s*(.+)$/);
  return match ? { key: match[1].trim(), value: match[2].trim() } : null;
}

function splitCells(line) {
  return String(line || '')
    .split('|')
    .map(cell => cell.trim())
    .filter(Boolean)
    .filter(cell => !/^[-:]{2,}$/.test(cell));
}

function isKnownPlatform(platform) {
  return platform === 'xiaohongshu' || platform === 'douyin' || platform === 'bilibili';
}

function resolveReferenceLinks(input, text) {
  const provided = Array.isArray(input.referenceLinks) ? input.referenceLinks.map(normalizeReferenceLink).filter(Boolean) : [];
  const urls = [...String(text || '').matchAll(/https?:\/\/[^\s|，,；;）)]+/gi)]
    .map(match => match[0].replace(/[。.!?]+$/, ''));
  const parsed = unique(urls).map(url => normalizeReferenceLink(url));
  return dedupeObjects([...provided, ...parsed], item => item.url || `${item.platform}:${item.noteId}`);
}

function normalizeReferenceLink(item) {
  if (!item) return null;
  if (typeof item === 'string') {
    return {
      url: item,
      platform: inferPlatformFromUrl(item),
      noteId: extractXhsNoteId(item),
      contentType: inferContentTypeFromText(item),
      needsEnrichment: true
    };
  }
  const url = String(item.url || item.link || '').trim();
  if (!url) return null;
  return {
    url,
    platform: item.platform || inferPlatformFromUrl(url),
    noteId: item.noteId || item.note_id || extractXhsNoteId(url),
    contentType: item.contentType || item.type || inferContentTypeFromText(`${item.sourceText || ''} ${url}`),
    sourceText: item.sourceText || '',
    needsEnrichment: item.needsEnrichment !== false
  };
}

function resolveReferenceAccounts(input, text) {
  const provided = Array.isArray(input.referenceAccounts) ? input.referenceAccounts : [];
  const accounts = provided.map(item => normalizeReferenceAccount(item)).filter(Boolean);
  const lines = String(text || '').split(/\r?\n/).map(cleanMarkdownLine).filter(Boolean);
  for (const line of lines) {
    if (!/参考|对标|标杆|账号|博主|达人|redId|小红书号|抖音号/i.test(line)) continue;
    if (/^(账号类型|.*参考账号|.*参考博主)[:：]?$/.test(line)) continue;
    const withoutUrls = line.replace(/https?:\/\/\S+/gi, '').trim();
    const kv = parseKeyValueLine(withoutUrls);
    if (kv && /^(账号类型|参考账号|参考博主|风格参考)$/i.test(kv.key)) continue;
    const value = kv ? kv.value : withoutUrls;
    for (const part of value.split(/[、，,；;\/]+/).map(item => item.trim()).filter(Boolean)) {
      if (/^(图文|视频|名单字段|关键词参考|风格参考|参考账号|参考博主|账号类型)$/.test(part)) continue;
      if (/参考账号|参考博主|账号类型/.test(part)) continue;
      if (part.length < 2 || part.length > 40) continue;
      accounts.push({
        name: part,
        platform: inferPlatformFromText(line),
        sourceText: line,
        needsEnrichment: true
      });
    }
  }
  return dedupeObjects(accounts, item => `${item.platform || ''}:${item.name || item.redId || item.userId || item.url || ''}`);
}

function normalizeReferenceAccount(item) {
  if (!item) return null;
  if (typeof item === 'string') {
    return { name: item, platform: inferPlatformFromText(item), sourceText: item, needsEnrichment: true };
  }
  return {
    name: item.name || item.nickname || item.displayName || '',
    redId: item.redId || item.red_id || '',
    userId: item.userId || item.user_id || '',
    platform: item.platform || inferPlatformFromText(`${item.url || ''} ${item.sourceText || ''}`),
    url: item.url || '',
    sourceText: item.sourceText || '',
    needsEnrichment: item.needsEnrichment !== false
  };
}

function resolveReferenceStyleAnchors(text, referenceLinks, referenceAccounts) {
  const anchors = [];
  let currentBucket = '';
  for (const rawLine of String(text || '').split(/\r?\n/)) {
    const line = cleanMarkdownLine(rawLine);
    if (!line) continue;
    if (/参考账号|参考博主|对标账号|标杆账号|账号类型/.test(line)) {
      currentBucket = line.replace(/[:：]\s*$/, '');
    }
    if (/https?:\/\//i.test(line)) {
      const links = (referenceLinks || []).filter(item => line.includes(item.url));
      anchors.push({
        bucket: currentBucket || inferReferenceBucket(line),
        contentType: /视频|video/i.test(line) ? 'video' : /图文|image|note/i.test(line) ? 'image_text' : (links[0]?.contentType || ''),
        platform: inferPlatformFromText(line) || links[0]?.platform || '',
        sourceText: line,
        links: links.map(item => item.url),
        needsSocialAnalysisEnrichment: true
      });
    }
  }
  for (const account of referenceAccounts || []) {
    anchors.push({
      bucket: inferReferenceBucket(account.sourceText || account.name || ''),
      platform: account.platform || '',
      sourceText: account.sourceText || account.name || '',
      accountHints: [account.name || account.redId || account.userId].filter(Boolean),
      needsSocialAnalysisEnrichment: true
    });
  }
  return dedupeObjects(anchors, item => `${item.bucket}:${item.sourceText}:${(item.links || []).join('|')}:${(item.accountHints || []).join('|')}`);
}

function resolveReferenceSignals(text, anchors) {
  const vocabulary = [
    '干货科普', '营养师', '育婴师', '妈妈', '母婴', '618合集', '合集', '图文', '视频',
    '素人感', '真实体验', '真实测评', '通勤生活', '成分党', '敏感肌', '开箱体验',
    '不硬广', '好物分享', '生活化', '专业背书', '选品建议'
  ];
  const source = [
    text,
    ...(anchors || []).map(item => `${item.bucket || ''} ${item.sourceText || ''}`)
  ].join('\n');
  return vocabulary.filter(item => source.includes(item));
}

function inferReferenceBucket(text) {
  if (/618|合集/.test(text)) return '618合集参考';
  if (/干货|科普|营养师|育婴师/.test(text)) return '干货科普参考';
  if (/风格/.test(text)) return '风格参考';
  return '参考账号';
}

function inferPlatformFromUrl(url) {
  if (/xhslink|xiaohongshu/i.test(url)) return 'xiaohongshu';
  if (/douyin/i.test(url)) return 'douyin';
  if (/bilibili|b23\.tv/i.test(url)) return 'bilibili';
  return '';
}

function inferPlatformFromText(text) {
  const value = String(text || '');
  if (/小红书|xhs|xiaohongshu|xhslink/i.test(value)) return 'xiaohongshu';
  if (/抖音|douyin/i.test(value)) return 'douyin';
  if (/B站|bilibili|b23\.tv/i.test(value)) return 'bilibili';
  return inferPlatformFromUrl(value);
}

function inferContentTypeFromText(text) {
  const value = String(text || '');
  if (/视频|video|douyin|bilibili|b23\.tv/i.test(value)) return 'video';
  if (/图文|image|note|xhslink|xiaohongshu/i.test(value)) return 'image_text';
  if (/主页|profile|account|home/i.test(value)) return 'account_home';
  return '';
}

function extractXhsNoteId(url) {
  const match = String(url || '').match(/\/(?:discovery\/item|explore)\/([0-9a-f]{16,32})/i);
  return match ? match[1] : '';
}

function unique(items) {
  return [...new Set(items.filter(Boolean).map(String))];
}

function dedupeObjects(items, keyFn) {
  const seen = new Set();
  const result = [];
  for (const item of items.filter(Boolean)) {
    const key = keyFn(item);
    if (seen.has(key)) continue;
    seen.add(key);
    result.push(item);
  }
  return result;
}

module.exports = {
  parseBrief
};