memind/scripts/build-news002-real-design-sample.mjs

#!/usr/bin/env node
/**
 * 用当天已生成的真实新闻页制作 news002 视觉样图：
 * - 正文与来源仍来自当天真实页
 * - 配图只取对应来源文章的 og:image/twitter:image
 * - 无真实同源图片的卡片不进入样图
 * - 不写生产目录，不调用微信 API
 */
import fs from 'node:fs';
import path from 'node:path';
import process from 'node:process';
import { NEWS002_STYLE_BLOCK } from '../news-morning-templates.mjs';
import {
  auditNews002ImageCoverage,
  enrichNews002HtmlWithSourceImages,
  extractNews002StoryCards,
} from '../wechat-news-morning-images.mjs';

const DEFAULT_SOURCE =
  'https://m.tkmind.cn/MindSpace/a70ff537-8908-486e-9b6c-042e07cc25db/public/daily-news-0921.html';
const sourceUrl = process.env.NEWS002_SAMPLE_SOURCE_URL || DEFAULT_SOURCE;
const outputPath = path.resolve(
  process.argv[2] || path.join(process.cwd(), '.tmp', 'news002-real-design-sample.html'),
);

function cardHasImage(raw) {
  return /<div\b[^>]*class="[^"]*\bmedia\b[^"]*"[^>]*>[\s\S]*?<img\b/i.test(raw);
}

function stripRanges(source, ranges) {
  let output = source;
  for (const range of [...ranges].sort((a, b) => b.start - a.start)) {
    output = output.slice(0, range.start) + output.slice(range.end);
  }
  return output;
}

function keepOnlyStrictCards(html, maxStoryCards = 28) {
  const cards = extractNews002StoryCards(html).filter((card) => card.title);
  const keep = [];
  const remove = [];
  for (const card of cards) {
    if (!cardHasImage(card.raw) || keep.length >= maxStoryCards) {
      remove.push(card);
    } else {
      keep.push(card);
    }
  }
  return { html: stripRanges(html, remove), keptCount: keep.length, removedCount: remove.length };
}

function normalizeHeadlineCards(html) {
  let output = String(html).replace(
    /<a\b[^>]*class="[^"]*\blead-card\b[^"]*"[^>]*>[\s\S]*?<\/a>\s*(?=<ol class="brief")/i,
    '',
  );
  const candidates = extractNews002StoryCards(output)
    .filter((card) => card.title && cardHasImage(card.raw))
    .slice(0, 3);
  for (const card of [...candidates].sort((a, b) => b.start - a.start)) {
    const promoted = card.raw.replace(
      /^<div class="card"/i,
      '<div class="lead-card"',
    );
    output = output.slice(0, card.start) + promoted + output.slice(card.end);
  }
  return output;
}

function applyCanonicalNews002Design(html) {
  const withoutStyles = String(html).replace(/<style\b[^>]*>[\s\S]*?<\/style>/gi, '');
  const withStyle = withoutStyles.replace('</head>', `${NEWS002_STYLE_BLOCK}
<style>
  .wrap{max-width:720px;margin:0 auto;padding:0 14px 36px}
  .anchors{position:sticky;top:0;z-index:50;display:flex;gap:6px;overflow-x:auto;
    padding:9px 10px;margin:0 -14px 18px;background:rgba(250,247,242,.95);
    border-top:1px solid var(--line);border-bottom:1px solid var(--line);backdrop-filter:blur(12px)}
  .anchors a{flex-shrink:0;padding:5px 10px;border-radius:999px;color:var(--ink-2);
    font-size:13px;text-decoration:none}
  .wrap>.brief{margin:0 0 28px}
  footer{padding:22px 16px 34px;text-align:center;font-size:12px;color:var(--ink-3);
    line-height:1.9;border-top:1px solid var(--line)}
  .sample-note{max-width:720px;margin:12px auto 0;padding:10px 16px;border-radius:12px;
    background:#172F57;color:#fff;font-size:12px;line-height:1.7;text-align:center}
  .sample-note b{color:#FFE3A3}
</style>
</head>`);
  return withStyle
    .replace(/<title>([\s\S]*?)<\/title>/i, '<title>news002 真实新闻视觉样图 · $1</title>')
    .replace(
      /<body([^>]*)>/i,
      '<body$1><div class="sample-note"><b>news002 视觉样图</b> · 正文来自当天真实新闻，图片来自对应文章来源；尚未推送微信。</div>',
    );
}

async function main() {
  const response = await fetch(sourceUrl, {
    signal: AbortSignal.timeout(20_000),
    headers: { accept: 'text/html', 'user-agent': 'TKMind-News002-Sample/1.0' },
  });
  if (!response.ok) throw new Error(`读取真实新闻页失败：HTTP ${response.status}`);
  const source = await response.text();
  if (!/class="[^"]*\bmasthead\b/i.test(source)) {
    throw new Error('来源页不是 news002 masthead 结构，拒绝制作误导样图');
  }

  const enriched = await enrichNews002HtmlWithSourceImages(source, {
    timeoutMs: 8000,
    concurrency: 8,
  });
  const normalized = normalizeHeadlineCards(enriched.html);
  const strict = keepOnlyStrictCards(normalized, 28);
  const designed = applyCanonicalNews002Design(strict.html);
  const audit = auditNews002ImageCoverage(designed);
  if (!audit.ok) {
    throw new Error(
      `视觉样图未通过 news002 闸门：${audit.violations.join(', ')}`
      + `；尝试配图 ${enriched.attemptedCount}，解析成功 ${enriched.resolvedCount}，`
      + `保留卡片 ${strict.keptCount}，头条 ${audit.leadCount}`,
    );
  }

  fs.mkdirSync(path.dirname(outputPath), { recursive: true });
  fs.writeFileSync(outputPath, designed);
  console.log(JSON.stringify({
    outputPath,
    sourceUrl,
    attemptedImages: enriched.attemptedCount,
    resolvedImages: enriched.resolvedCount,
    removedCards: strict.removedCount,
    audit,
  }, null, 2));
}

main().catch((error) => {
  console.error(error instanceof Error ? error.message : error);
  process.exit(1);
});