@huyuan-ai/cli 1.0.0 → 1.0.1

package/vendor/opencli/dist/clis/1688/store.js

@@ -0,0 +1,226 @@
+import { CommandExecutionError, EmptyResultError } from '../../errors.js';
+import { cli, Strategy } from '../../registry.js';
+import { FACTORY_BADGE_PATTERNS, SERVICE_BADGE_PATTERNS, assertAuthenticatedState, buildDetailUrl, buildProvenance, canonicalizeSellerUrl, canonicalizeStoreUrl, cleanMultilineText, cleanText, extractAddress, extractBadges, extractMemberId, extractMetric, extractOfferId, extractShopId, extractYearsOnPlatform, gotoAndReadState, guessTopCategories, resolveStoreUrl, uniqueNonEmpty, } from './shared.js';
+function normalizeStorePayload(input) {
+    const storePayload = input.storePayload;
+    const contactPayload = input.contactPayload;
+    const seed = input.seed;
+    const contactText = cleanMultilineText(contactPayload?.bodyText);
+    const storeText = cleanMultilineText(storePayload?.bodyText);
+    const seedText = cleanMultilineText(seed?.bodyText);
+    const combinedText = [contactText, storeText, seedText].filter(Boolean).join('\n');
+    const sellerUrlRaw = cleanText(seed?.seller?.winportUrl
+        ?? seed?.seller?.sellerWinportUrlMap?.defaultUrl
+        ?? storePayload?.href
+        ?? input.resolvedUrl);
+    const storeUrl = safeCanonicalStoreUrl(sellerUrlRaw || input.resolvedUrl) ?? input.resolvedUrl;
+    const sellerUrl = canonicalizeSellerUrl(sellerUrlRaw) ?? storeUrl;
+    const companyUrl = pickCompanyUrl(contactPayload?.href, storeUrl);
+    const memberId = cleanText(seed?.seller?.memberId)
+        || input.explicitMemberId
+        || extractMemberId(input.resolvedUrl)
+        || extractMemberId(storePayload?.href ?? '')
+        || null;
+    const shopId = extractShopId(sellerUrl) ?? extractShopId(storeUrl);
+    const companyName = cleanText(seed?.seller?.companyName)
+        || firstNamedLine(contactText)
+        || firstNamedLine(storeText)
+        || null;
+    const serviceBadges = uniqueNonEmpty([
+        ...extractBadges(combinedText, SERVICE_BADGE_PATTERNS),
+        ...((seed?.services ?? []).map((service) => cleanText(service.serviceName))),
+    ]);
+    const factoryBadges = extractBadges(combinedText, FACTORY_BADGE_PATTERNS);
+    return {
+        member_id: memberId,
+        shop_id: shopId,
+        store_name: companyName,
+        store_url: storeUrl,
+        company_name: companyName,
+        company_url: companyUrl,
+        business_model_text: firstMetric(combinedText, ['经营模式', '生产加工', '主营产品']),
+        years_on_platform_text: extractYearsOnPlatform(combinedText),
+        location: extractAddress(contactText) ?? extractAddress(storeText),
+        staff_size_text: firstMetric(combinedText, ['员工人数', '员工总数']),
+        factory_badges: factoryBadges,
+        service_badges: serviceBadges,
+        response_rate_text: firstMetric(combinedText, ['响应率', '回复率', '响应速度']),
+        return_rate_text: extractReturnRate(combinedText),
+        top_categories: guessTopCategories(combinedText),
+        phone_text: extractMetric(contactText, '电话'),
+        mobile_text: extractMetric(contactText, '手机'),
+        ...buildProvenance(cleanText(contactPayload?.href) || cleanText(storePayload?.href) || input.resolvedUrl),
+    };
+}
+function safeCanonicalStoreUrl(url) {
+    try {
+        return canonicalizeStoreUrl(url);
+    }
+    catch {
+        return null;
+    }
+}
+function pickCompanyUrl(contactHref, storeUrl) {
+    const fromPage = cleanText(contactHref);
+    if (fromPage) {
+        const normalized = buildContactUrl(fromPage);
+        if (normalized)
+            return normalized;
+    }
+    return buildContactUrl(storeUrl);
+}
+function buildContactUrl(storeUrl) {
+    try {
+        const parsed = new URL(storeUrl);
+        if (!parsed.hostname.endsWith('.1688.com'))
+            return null;
+        return `${parsed.protocol}//${parsed.hostname}/page/contactinfo.html`;
+    }
+    catch {
+        return null;
+    }
+}
+function firstNamedLine(text) {
+    return text
+        .split('\n')
+        .map((line) => cleanText(line))
+        .find((line) => line.includes('有限公司') || line.includes('商行') || line.includes('工厂'))
+        ?? null;
+}
+function firstMetric(text, labels) {
+    for (const label of labels) {
+        const value = extractMetric(text, label);
+        if (value)
+            return value;
+    }
+    return null;
+}
+function extractReturnRate(text) {
+    const inline = text.match(/回头率\s*([0-9.]+%)/);
+    if (inline)
+        return cleanText(inline[0]);
+    const multiline = text.match(/回头率\s*\n\s*([0-9.]+%)/);
+    if (!multiline)
+        return null;
+    return `回头率${cleanText(multiline[1])}`;
+}
+function firstOfferId(links) {
+    for (const link of links) {
+        const offerId = extractOfferId(link);
+        if (offerId)
+            return offerId;
+    }
+    return null;
+}
+function firstContactUrl(links) {
+    for (const link of links) {
+        const url = buildContactUrl(link);
+        if (url)
+            return url;
+    }
+    return null;
+}
+async function readStorePayload(page, url, action) {
+    const state = await gotoAndReadState(page, url, 2500, action);
+    assertAuthenticatedState(state, action);
+    return await page.evaluate(`
+    (() => ({
+      href: window.location.href,
+      title: document.title || '',
+      bodyText: document.body ? document.body.innerText || '' : '',
+      offerLinks: Array.from(document.querySelectorAll('a[href*="detail.1688.com/offer/"], a[href*="offerId="]'))
+        .map((anchor) => anchor.href)
+        .filter(Boolean),
+      contactLinks: Array.from(document.querySelectorAll('a[href*="contactinfo"]'))
+        .map((anchor) => anchor.href)
+        .filter(Boolean),
+    }))()
+  `);
+}
+async function readItemSeed(page, offerId) {
+    const itemUrl = buildDetailUrl(offerId);
+    const state = await gotoAndReadState(page, itemUrl, 2500, 'store seed item');
+    assertAuthenticatedState(state, 'store seed item');
+    const seed = await page.evaluate(`
+    (() => {
+      const model = window.context?.result?.global?.globalData?.model ?? null;
+      const toJson = (value) => JSON.parse(JSON.stringify(value ?? null));
+      return {
+        href: window.location.href,
+        bodyText: document.body ? document.body.innerText || '' : '',
+        seller: toJson(model?.sellerModel),
+        services: toJson(model?.shippingServices?.fields?.buyerProtectionModel ?? []),
+      };
+    })()
+  `);
+    const hasSellerContext = !!cleanText(seed?.seller?.memberId) || !!cleanText(seed?.seller?.winportUrl);
+    if (!hasSellerContext) {
+        throw new CommandExecutionError('1688 store seed item did not expose seller context', '当前 tab 非商品详情上下文，请切到 detail.1688.com 商品页并重试');
+    }
+    return seed;
+}
+function hasAnyEvidence(storePayload, contactPayload, seed) {
+    return !!cleanText(storePayload?.bodyText)
+        || !!cleanText(contactPayload?.bodyText)
+        || !!cleanText(seed?.bodyText);
+}
+cli({
+    site: '1688',
+    name: 'store',
+    description: '1688 店铺/供应商公开信息（联系方式、主营、入驻年限、公开服务信号）',
+    domain: 'www.1688.com',
+    strategy: Strategy.COOKIE,
+    navigateBefore: false,
+    args: [
+        {
+            name: 'input',
+            required: true,
+            positional: true,
+            help: '1688 店铺 URL 或 member ID（如 b2b-22154705262941f196）',
+        },
+    ],
+    columns: ['store_name', 'years_on_platform_text', 'location', 'return_rate_text'],
+    func: async (page, kwargs) => {
+        const rawInput = String(kwargs.input ?? '');
+        const resolvedUrl = resolveStoreUrl(rawInput);
+        const explicitMemberId = extractMemberId(rawInput);
+        const storePayload = await readStorePayload(page, resolvedUrl, 'store');
+        const contactUrl = firstContactUrl(storePayload.contactLinks ?? []) || buildContactUrl(storePayload.href || resolvedUrl);
+        const contactPayload = contactUrl ? await readStorePayload(page, contactUrl, 'store contact') : null;
+        const offerId = extractOfferId(rawInput)
+            || firstOfferId(storePayload.offerLinks ?? [])
+            || firstOfferId(contactPayload?.offerLinks ?? []);
+        let seed = null;
+        if (offerId) {
+            try {
+                seed = await readItemSeed(page, offerId);
+            }
+            catch (error) {
+                if (!(error instanceof CommandExecutionError))
+                    throw error;
+            }
+        }
+        if (!hasAnyEvidence(storePayload, contactPayload, seed)) {
+            throw new EmptyResultError('1688 store', 'Store page is reachable but no visible fields were extracted. Open the store page in Chrome and retry.');
+        }
+        return [
+            normalizeStorePayload({
+                resolvedUrl,
+                storePayload,
+                contactPayload,
+                seed,
+                explicitMemberId,
+            }),
+        ];
+    },
+});
+export const __test__ = {
+    normalizeStorePayload,
+    safeCanonicalStoreUrl,
+    buildContactUrl,
+    firstNamedLine,
+    firstMetric,
+    extractReturnRate,
+    firstOfferId,
+    firstContactUrl,
+};

package/vendor/opencli/dist/clis/1688/store.test.d.ts

@@ -0,0 +1 @@
+export {};

package/vendor/opencli/dist/clis/1688/store.test.js

@@ -0,0 +1,62 @@
+import { describe, expect, it } from 'vitest';
+import { __test__ } from './store.js';
+describe('1688 store normalization', () => {
+    it('merges store contact text with seller seed data', () => {
+        const result = __test__.normalizeStorePayload({
+            resolvedUrl: 'https://yinuoweierfushi.1688.com/?offerId=887904326744',
+            explicitMemberId: null,
+            storePayload: {
+                href: 'https://yinuoweierfushi.1688.com/page/index.html',
+                bodyText: `
+          青岛沁澜衣品服装有限公司
+          联系方式
+          地址：山东省青岛市即墨区环秀街道办事处湘江二路97号甲
+        `,
+                offerLinks: ['https://detail.1688.com/offer/887904326744.html'],
+            },
+            contactPayload: {
+                href: 'https://yinuoweierfushi.1688.com/page/contactinfo.html',
+                bodyText: `
+          青岛沁澜衣品服装有限公司
+          电话：86 0532 86655366
+          手机：15963238678
+          地址：山东省青岛市即墨区环秀街道办事处湘江二路97号甲
+        `,
+            },
+            seed: {
+                bodyText: `
+          入驻13年
+          主营：大码女装
+          店铺回头率
+          87%
+          延期必赔
+          品质保障
+        `,
+                seller: {
+                    companyName: '青岛沁澜衣品服装有限公司',
+                    memberId: 'b2b-1641351767',
+                    winportUrl: 'https://yinuoweierfushi.1688.com/page/index.html?spm=abc',
+                },
+                services: [{ serviceName: '延期必赔' }, { serviceName: '品质保障' }],
+            },
+        });
+        expect(result.member_id).toBe('b2b-1641351767');
+        expect(result.store_url).toBe('https://yinuoweierfushi.1688.com');
+        expect(result.company_url).toBe('https://yinuoweierfushi.1688.com/page/contactinfo.html');
+        expect(result.years_on_platform_text).toBe('入驻13年');
+        expect(result.location).toBe('山东省青岛市即墨区环秀街道办事处湘江二路97号甲');
+        expect(result.return_rate_text).toContain('87%');
+        expect(result.top_categories).toEqual(['大码女装']);
+        expect(result.service_badges).toEqual(['延期必赔', '品质保障']);
+    });
+    it('builds contact urls and extracts offer ids', () => {
+        expect(__test__.safeCanonicalStoreUrl('https://yinuoweierfushi.1688.com/page/index.html?spm=foo')).toBe('https://yinuoweierfushi.1688.com');
+        expect(__test__.buildContactUrl('https://yinuoweierfushi.1688.com')).toBe('https://yinuoweierfushi.1688.com/page/contactinfo.html');
+        expect(__test__.firstOfferId([
+            'https://detail.1688.com/offer/887904326744.html',
+        ])).toBe('887904326744');
+        expect(__test__.firstContactUrl([
+            'https://yinuoweierfushi.1688.com/page/contactinfo.html?spm=1',
+        ])).toBe('https://yinuoweierfushi.1688.com/page/contactinfo.html');
+    });
+});

package/vendor/opencli/dist/clis/36kr/article.d.ts

@@ -0,0 +1 @@
+export {};

package/vendor/opencli/dist/clis/36kr/article.js

@@ -0,0 +1,62 @@
+/**
+ * 36kr article detail — INTERCEPT strategy.
+ *
+ * Fetches the full content of a 36kr article given its ID or URL.
+ */
+import { cli, Strategy } from '../../registry.js';
+import { CliError } from '../../errors.js';
+/** Extract article ID from a full URL or a bare numeric ID string */
+function parseArticleId(input) {
+    const m = input.match(/\/p\/(\d+)/);
+    return m ? m[1] : input.replace(/\D/g, '');
+}
+cli({
+    site: '36kr',
+    name: 'article',
+    description: '获取36氪文章正文内容',
+    domain: 'www.36kr.com',
+    strategy: Strategy.INTERCEPT,
+    args: [
+        { name: 'id', positional: true, required: true, help: 'Article ID or full 36kr article URL' },
+    ],
+    columns: ['field', 'value'],
+    func: async (page, args) => {
+        const articleId = parseArticleId(String(args.id ?? ''));
+        if (!articleId) {
+            throw new CliError('INVALID_ARGUMENT', 'Invalid article ID or URL');
+        }
+        await page.installInterceptor('36kr.com/api');
+        await page.goto(`https://www.36kr.com/p/${articleId}`);
+        await page.wait(5);
+        const data = await page.evaluate(`
+      (() => {
+        // Title: 36kr uses class "article-title" on h1
+        const title = document.querySelector('.article-title, h1')?.textContent?.trim() || '';
+        // Author: second .author-name (first is empty nav link, second has real name)
+        const authorEls = document.querySelectorAll('.author-name');
+        const author = Array.from(authorEls).map(el => el.textContent?.trim()).filter(Boolean)[0] || '';
+        // Date: 36kr uses class "title-icon-item item-time" for the publish date
+        const dateRaw = document.querySelector('.item-time')?.textContent?.trim() || '';
+        const date = dateRaw.replace(/^[·\s]+/, '').trim();
+        // Article body paragraphs
+        const bodyEls = document.querySelectorAll('[class*="article-content"] p, [class*="rich-text"] p, .article p');
+        const body = Array.from(bodyEls)
+          .map(el => el.textContent?.trim())
+          .filter(t => t && t.length > 10)
+          .join(' ')
+          .slice(0, 800);
+        return { title, author, date, body };
+      })()
+    `);
+        if (!data?.title) {
+            throw new CliError('NOT_FOUND', 'Article not found or failed to load', 'Check the article ID');
+        }
+        return [
+            { field: 'title', value: data.title },
+            { field: 'author', value: data.author || '-' },
+            { field: 'date', value: data.date || '-' },
+            { field: 'url', value: `https://36kr.com/p/${articleId}` },
+            { field: 'body', value: data.body || '-' },
+        ];
+    },
+});

package/vendor/opencli/dist/clis/36kr/hot.d.ts

@@ -0,0 +1,3 @@
+declare function getShanghaiDate(date?: Date): string;
+declare function buildHotListUrl(listType: string, date?: Date): string;
+export { buildHotListUrl, getShanghaiDate };

package/vendor/opencli/dist/clis/36kr/hot.js

@@ -0,0 +1,80 @@
+/**
+ * 36kr hot-list — INTERCEPT strategy.
+ *
+ * Navigates to the 36kr hot-list page and scrapes rendered article links.
+ * Supports category types: renqi (人气), zonghe (综合), shoucang (收藏), catalog (综合热门).
+ */
+import { cli, Strategy } from '../../registry.js';
+import { CliError } from '../../errors.js';
+const TYPE_MAP = {
+    renqi: '人气榜',
+    zonghe: '综合榜',
+    shoucang: '收藏榜',
+    catalog: '热门资讯',
+};
+function getShanghaiDate(date = new Date()) {
+    // Shanghai stays on UTC+8 year-round, so a fixed offset is sufficient here
+    // and avoids the slow Intl timezone path that timed out on Windows CI.
+    return new Date(date.getTime() + 8 * 60 * 60 * 1000).toISOString().slice(0, 10);
+}
+function buildHotListUrl(listType, date = new Date()) {
+    if (listType === 'catalog') {
+        return 'https://www.36kr.com/hot-list/catalog';
+    }
+    return `https://www.36kr.com/hot-list/${listType}/${getShanghaiDate(date)}/1`;
+}
+cli({
+    site: '36kr',
+    name: 'hot',
+    description: '36氪热榜 — trending articles (renqi/zonghe/shoucang/catalog)',
+    domain: 'www.36kr.com',
+    strategy: Strategy.INTERCEPT,
+    args: [
+        { name: 'limit', type: 'int', default: 20, help: 'Number of items (max 50)' },
+        {
+            name: 'type',
+            type: 'string',
+            default: 'catalog',
+            help: 'List type: renqi (人气), zonghe (综合), shoucang (收藏), catalog (热门资讯)',
+        },
+    ],
+    columns: ['rank', 'title', 'url'],
+    func: async (page, args) => {
+        const count = Math.min(Number(args.limit) || 20, 50);
+        const listType = String(args.type ?? 'catalog');
+        if (!TYPE_MAP[listType]) {
+            throw new CliError('INVALID_ARGUMENT', `Unknown type "${listType}". Valid types: ${Object.keys(TYPE_MAP).join(', ')}`);
+        }
+        const url = buildHotListUrl(listType);
+        await page.installInterceptor('36kr.com/api');
+        await page.goto(url);
+        await page.waitForCapture(6);
+        // Scrape rendered article links from DOM (deduplicated)
+        const domItems = await page.evaluate(`
+      (() => {
+        const seen = new Set();
+        const results = [];
+        const links = document.querySelectorAll('a[href*="/p/"]');
+        for (const el of links) {
+          const href = el.getAttribute('href') || '';
+          const title = el.textContent?.trim() || '';
+          if (!title || title.length < 5 || seen.has(href) || seen.has(title)) continue;
+          seen.add(href);
+          seen.add(title);
+          results.push({ title, url: href.startsWith('http') ? href : 'https://36kr.com' + href });
+        }
+        return results;
+      })()
+    `);
+        const items = Array.isArray(domItems) ? domItems : [];
+        if (items.length === 0) {
+            throw new CliError('NO_DATA', 'Could not retrieve 36kr hot list', '36kr may have changed its DOM structure');
+        }
+        return items.slice(0, count).map((item, i) => ({
+            rank: i + 1,
+            title: item.title,
+            url: item.url,
+        }));
+    },
+});
+export { buildHotListUrl, getShanghaiDate };

package/vendor/opencli/dist/clis/36kr/hot.test.d.ts

@@ -0,0 +1 @@
+export {};

package/vendor/opencli/dist/clis/36kr/hot.test.js

@@ -0,0 +1,15 @@
+import { describe, expect, it } from 'vitest';
+import { buildHotListUrl, getShanghaiDate } from './hot.js';
+describe('36kr/hot date routing', () => {
+    it('formats dates in Asia/Shanghai instead of UTC', () => {
+        const date = new Date('2026-03-25T18:30:00.000Z');
+        expect(getShanghaiDate(date)).toBe('2026-03-26');
+    });
+    it('builds dated hot-list routes with Shanghai-local date', () => {
+        const date = new Date('2026-03-25T18:30:00.000Z');
+        expect(buildHotListUrl('renqi', date)).toBe('https://www.36kr.com/hot-list/renqi/2026-03-26/1');
+    });
+    it('keeps catalog on the static route', () => {
+        expect(buildHotListUrl('catalog')).toBe('https://www.36kr.com/hot-list/catalog');
+    });
+});

package/vendor/opencli/dist/clis/36kr/news.d.ts

@@ -0,0 +1 @@
+export {};

package/vendor/opencli/dist/clis/36kr/news.js

@@ -0,0 +1,51 @@
+/**
+ * 36kr latest news — public RSS feed, no browser needed.
+ */
+import { cli, Strategy } from '../../registry.js';
+cli({
+    site: '36kr',
+    name: 'news',
+    description: 'Latest tech/startup news from 36kr (36氪)',
+    domain: 'www.36kr.com',
+    strategy: Strategy.PUBLIC,
+    args: [
+        { name: 'limit', type: 'int', default: 20, help: 'Number of articles (max 50)' },
+    ],
+    columns: ['rank', 'title', 'summary', 'date', 'url'],
+    func: async (_page, kwargs) => {
+        const count = Math.min(kwargs.limit || 20, 50);
+        const resp = await fetch('https://www.36kr.com/feed', {
+            headers: { 'User-Agent': 'Mozilla/5.0 (compatible; opencli/1.0)' },
+        });
+        if (!resp.ok)
+            return [];
+        const xml = await resp.text();
+        const items = [];
+        const itemRegex = /<item>([\s\S]*?)<\/item>/g;
+        let match;
+        while ((match = itemRegex.exec(xml)) && items.length < count) {
+            const block = match[1];
+            const title = block.match(/<title>([\s\S]*?)<\/title>/)?.[1]?.trim() ?? '';
+            const url = block.match(/<link><!\[CDATA\[(.*?)\]\]>/)?.[1] ??
+                block.match(/<link>(.*?)<\/link>/)?.[1] ??
+                '';
+            const pubDate = block.match(/<pubDate>(.*?)<\/pubDate>/)?.[1]?.trim() ?? '';
+            const date = pubDate.slice(0, 10);
+            // Extract plain-text summary from HTML description (first ~120 chars)
+            const rawDesc = block.match(/<description><!\[CDATA\[([\s\S]*?)\]\]>/)?.[1] ?? '';
+            const summary = rawDesc
+                .replace(/<[^>]+>/g, ' ')
+                .replace(/&nbsp;/g, ' ')
+                .replace(/&amp;/g, '&')
+                .replace(/&lt;/g, '<')
+                .replace(/&gt;/g, '>')
+                .replace(/\s+/g, ' ')
+                .trim()
+                .slice(0, 120);
+            if (title) {
+                items.push({ rank: items.length + 1, title, summary, date, url: url.trim() });
+            }
+        }
+        return items;
+    },
+});

package/vendor/opencli/dist/clis/36kr/news.test.d.ts

@@ -0,0 +1 @@
+export {};

package/vendor/opencli/dist/clis/36kr/news.test.js

@@ -0,0 +1,85 @@
+import { describe, it, expect, vi, afterEach } from 'vitest';
+const SAMPLE_RSS = `<?xml version="1.0" encoding="UTF-8"?>
+<rss version="2.0"><channel><title>36氪</title>
+<item>
+  <title>红杉中国领投AI公司「示例」，金额近2亿元</title>
+  <link><![CDATA[https://36kr.com/p/1111111111111111?f=rss]]></link>
+  <pubDate>2026-03-26 10:00:00  +0800</pubDate>
+</item>
+<item>
+  <title>马斯克旗下xAI估值突破1000亿美元</title>
+  <link><![CDATA[https://36kr.com/p/2222222222222222?f=rss]]></link>
+  <pubDate>2026-03-26 09:00:00  +0800</pubDate>
+</item>
+<item>
+  <title>OpenAI发布GPT-5，多模态能力大幅提升</title>
+  <link><![CDATA[https://36kr.com/p/3333333333333333?f=rss]]></link>
+  <pubDate>2026-03-25 20:00:00  +0800</pubDate>
+</item>
+</channel></rss>`;
+afterEach(() => {
+    vi.restoreAllMocks();
+});
+describe('36kr/news RSS parsing', () => {
+    it('parses RSS feed into ranked news items', async () => {
+        vi.spyOn(globalThis, 'fetch').mockResolvedValue({
+            ok: true,
+            text: async () => SAMPLE_RSS,
+        });
+        // Direct RSS parse test using the same regex logic as news.ts
+        const xml = SAMPLE_RSS;
+        const items = [];
+        const itemRegex = /<item>([\s\S]*?)<\/item>/g;
+        let match;
+        while ((match = itemRegex.exec(xml)) && items.length < 10) {
+            const block = match[1];
+            const title = block.match(/<title>([\s\S]*?)<\/title>/)?.[1]?.trim() ?? '';
+            const url = block.match(/<link><!\[CDATA\[(.*?)\]\]>/)?.[1] ??
+                block.match(/<link>(.*?)<\/link>/)?.[1] ??
+                '';
+            const pubDate = block.match(/<pubDate>(.*?)<\/pubDate>/)?.[1]?.trim() ?? '';
+            const date = pubDate.slice(0, 10);
+            if (title)
+                items.push({ rank: items.length + 1, title, date, url: url.trim() });
+        }
+        expect(items).toHaveLength(3);
+        expect(items[0].rank).toBe(1);
+        expect(items[0].title).toBe('红杉中国领投AI公司「示例」，金额近2亿元');
+        expect(items[0].date).toBe('2026-03-26');
+        expect(items[0].url).toBe('https://36kr.com/p/1111111111111111?f=rss');
+    });
+    it('respects limit — returns at most N items', async () => {
+        const xml = SAMPLE_RSS;
+        const limit = 2;
+        const items = [];
+        const itemRegex = /<item>([\s\S]*?)<\/item>/g;
+        let match;
+        while ((match = itemRegex.exec(xml)) && items.length < limit) {
+            const block = match[1];
+            const title = block.match(/<title>([\s\S]*?)<\/title>/)?.[1]?.trim() ?? '';
+            const url = block.match(/<link><!\[CDATA\[(.*?)\]\]>/)?.[1] ?? '';
+            const pubDate = block.match(/<pubDate>(.*?)<\/pubDate>/)?.[1]?.trim() ?? '';
+            const date = pubDate.slice(0, 10);
+            if (title)
+                items.push({ rank: items.length + 1, title, date, url: url.trim() });
+        }
+        expect(items).toHaveLength(2);
+    });
+    it('skips items with empty title', async () => {
+        const xml = `<rss><channel>
+      <item><title></title><link>https://36kr.com/p/0</link><pubDate>2026-01-01</pubDate></item>
+      <item><title>有标题的文章</title><link>https://36kr.com/p/1</link><pubDate>2026-01-01</pubDate></item>
+    </channel></rss>`;
+        const items = [];
+        const itemRegex = /<item>([\s\S]*?)<\/item>/g;
+        let match;
+        while ((match = itemRegex.exec(xml))) {
+            const block = match[1];
+            const title = block.match(/<title>([\s\S]*?)<\/title>/)?.[1]?.trim() ?? '';
+            if (title)
+                items.push({ title });
+        }
+        expect(items).toHaveLength(1);
+        expect(items[0].title).toBe('有标题的文章');
+    });
+});

package/vendor/opencli/dist/clis/36kr/search.d.ts

@@ -0,0 +1 @@
+export {};