nuxt-app/server/service/agent-tool/executors/search/parse.ts

export interface SearchResultItem {
  title: string;
  url: string;
  snippet: string;
  displayUrl?: string;
}

export interface ParsedSearchResults {
  items: SearchResultItem[];
  totalResults?: string;
  engine: string;
}

export function parseBingResults(html: string, maxResults: number, snippetLength: number): ParsedSearchResults {
  const items: SearchResultItem[] = [];

  // Bing 结果项在 <li class="b_algo"> 中
  const itemRegex = /<li[^>]*class="[^"]*b_algo[^"]*"[^>]*>([\s\S]*?)<\/li>/gi;
  let match: RegExpExecArray | null;

  while ((match = itemRegex.exec(html)) !== null && items.length < maxResults) {
    const block = match[1];

    // 标题和链接在 <h2><a href="..." ...>标题</a></h2>
    const linkMatch = block.match(/<a[^>]*href="(https?:\/\/[^"]+)"[^>]*>([\s\S]*?)<\/a>/i);
    if (!linkMatch) continue;

    const url = linkMatch[1];
    const title = stripTags(linkMatch[2]).trim();
    if (!title || !url) continue;

    // 摘要在 <p class="b_lineclamp..."> 或 <div class="b_caption"><p>
    const snippetMatch =
      block.match(/<p[^>]*class="[^"]*b_lineclamp[^"]*"[^>]*>([\s\S]*?)<\/p>/i) ||
      block.match(/<div[^>]*class="[^"]*b_caption[^"]*"[^>]*>[\s\S]*?<p[^>]*>([\s\S]*?)<\/p>/i);
    const snippet = snippetMatch ? truncate(stripTags(snippetMatch[1]).trim(), snippetLength) : "";

    // 显示URL
    const displayMatch = block.match(/<cite[^>]*>([\s\S]*?)<\/cite>/i);
    const displayUrl = displayMatch ? stripTags(displayMatch[1]).trim() : undefined;

    items.push({ title, url, snippet, displayUrl });
  }

  return { items, engine: "bing" };
}

export function parseBaiduResults(html: string, maxResults: number, snippetLength: number): ParsedSearchResults {
  const items: SearchResultItem[] = [];

  // 百度结果项在 <div class="result ..."> 或 <div class="c-container ...">
  const itemRegex = /<div[^>]*class="[^"]*(?:result|c-container)[^"]*"[^>]*>([\s\S]*?)(?=<div[^>]*class="[^"]*(?:result|c-container)|$)/gi;
  let match: RegExpExecArray | null;

  while ((match = itemRegex.exec(html)) !== null && items.length < maxResults) {
    const block = match[1];

    const linkMatch = block.match(/<a[^>]*href="([^"]+)"[^>]*>([\s\S]*?)<\/a>/i);
    if (!linkMatch) continue;

    const url = linkMatch[1];
    const title = stripTags(linkMatch[2]).trim();
    if (!title || !url) continue;

    const snippetMatch = block.match(/<span[^>]*class="[^"]*content-right[^"]*"[^>]*>([\s\S]*?)<\/span>/i) ||
      block.match(/<div[^>]*class="[^"]*c-abstract[^"]*"[^>]*>([\s\S]*?)<\/div>/i) ||
      block.match(/<span[^>]*>([\s\S]*?)<\/span>/i);
    const snippet = snippetMatch ? truncate(stripTags(snippetMatch[1]).trim(), snippetLength) : "";

    items.push({ title, url, snippet });
  }

  return { items, engine: "baidu" };
}

export function parseGoogleResults(html: string, maxResults: number, snippetLength: number): ParsedSearchResults {
  const items: SearchResultItem[] = [];

  // Google 结果项在 <div class="g"> 中
  const itemRegex = /<div[^>]*class="[^"]*\bg\b[^"]*"[^>]*>([\s\S]*?)<\/div>\s*(?=<div[^>]*class="[^"]*\bg\b|<div[^>]*id="foot"|$)/gi;
  let match: RegExpExecArray | null;

  while ((match = itemRegex.exec(html)) !== null && items.length < maxResults) {
    const block = match[1];

    const linkMatch = block.match(/<a[^>]*href="\/url\?q=(https?:\/\/[^&"]+)&[^"]*"[^>]*>([\s\S]*?)<\/a>/i) ||
      block.match(/<a[^>]*href="(https?:\/\/[^"]+)"[^>]*>([\s\S]*?)<\/a>/i);
    if (!linkMatch) continue;

    const url = decodeURIComponent(linkMatch[1]);
    const title = stripTags(linkMatch[2]).trim();
    if (!title || !url) continue;

    const snippetMatch = block.match(/<span[^>]*>([\s\S]*?)<\/span>/i);
    const snippet = snippetMatch ? truncate(stripTags(snippetMatch[1]).trim(), snippetLength) : "";

    items.push({ title, url, snippet });
  }

  return { items, engine: "google" };
}

function stripTags(html: string): string {
  return html
    .replace(/<[^>]+>/g, "")
    .replace(/&amp;/g, "&")
    .replace(/&lt;/g, "<")
    .replace(/&gt;/g, ">")
    .replace(/&quot;/g, '"')
    .replace(/&#39;/g, "'")
    .replace(/&nbsp;/g, " ")
    .replace(/\s+/g, " ")
    .trim();
}

function truncate(str: string, maxLen: number): string {
  if (str.length <= maxLen) return str;
  return str.slice(0, maxLen) + "...";
}