fix(route): cw (#11613)

This commit is contained in:
Tony
2023-01-12 21:42:59 -05:00
committed by GitHub
parent a008fbeff8
commit 4e7c633dc2
5 changed files with 91 additions and 64 deletions

View File

@@ -1,26 +1,27 @@
const cheerio = require('cheerio');
const { parseDate } = require('@/utils/parse-date');
const { Cookie, CookieJar } = require('tough-cookie');
let cookieJar;
const config = require('@/config').value;
const { getCookies, setCookies } = require('@/utils/puppeteer-utils');
let cookie;
const baseUrl = 'https://www.cw.com.tw';
const got = require('@/utils/got').extend({
headers: {
'User-Agent': config.trueUA,
},
});
const getCookie = async () => {
const response = await got(`${baseUrl}/user/get/cookie-bar`);
const cookies = response.headers['set-cookie'];
if (Array.isArray(cookies)) {
cookieJar = cookies.map(Cookie.parse);
} else {
cookieJar = [Cookie.parse(cookieJar)];
const getCookie = async (browser, tryGet) => {
if (!cookie) {
cookie = await tryGet('cw:cookie', async () => {
const page = await browser.newPage();
await page.setRequestInterception(true);
page.on('request', (request) => {
request.resourceType() === 'document' || request.resourceType() === 'script' ? request.continue() : request.abort();
});
await page.goto(`${baseUrl}/user/get/cookie-bar`, {
waitUntil: 'domcontentloaded',
});
cookie = await getCookies(page);
await page.close();
return cookie;
});
}
cookieJar = CookieJar.fromJSON({ cookies: cookieJar });
return cookie;
};
const parseList = ($, limit) =>
@@ -36,15 +37,24 @@ const parseList = ($, limit) =>
})
.slice(0, limit);
const parseItems = (list, tryGet) =>
const parseItems = (list, browser, tryGet) =>
Promise.all(
list.map((item) =>
tryGet(item.link, async () => {
const { data: response } = await got(item.link, {
cookieJar,
const page = await browser.newPage();
await page.setRequestInterception(true);
page.on('request', (request) => {
request.resourceType() === 'document' || request.resourceType() === 'script' ? request.continue() : request.abort();
});
await setCookies(page, cookie, 'cw.com.tw');
await page.goto(item.link, {
waitUntil: 'domcontentloaded',
});
const response = await page.evaluate(() => document.documentElement.innerHTML);
await page.close();
const $ = cheerio.load(response);
const meta = JSON.parse($('head script[type="application/ld+json"]').eq(0).text());
$('.article__head .breadcrumb, .article__head h1, .article__provideViews, .ad').remove();
$('img.lazyload').each((_, img) => {
@@ -67,9 +77,8 @@ const parseItems = (list, tryGet) =>
module.exports = {
baseUrl,
cookieJar,
got,
getCookie,
setCookies,
parseList,
parseItems,
};