* feat(route/cna): declare headlines as supported https://www.cna.com.tw/list/headlines.aspx works the same as other categories (aall, aipl, ...) - the same HTML structure and API * fix(route/cna): ignore youtube links during full-text extraction Sometimes a youtube link instead of a news page is used directly as a news item (ex: 晚間影音新聞 during 6~7pm UTC+8 on weekdays). That may result in HTTP 429. Full-text extraction returns nothing, anyway. * fix(route/cna): note about where the API is from
29 lines
772 B
TypeScript
29 lines
772 B
TypeScript
import { load } from 'cheerio';
|
|
|
|
import got from '@/utils/got';
|
|
|
|
export async function getFullText(item) {
|
|
if (item.link.startsWith('https://www.youtube.com/')) {
|
|
return item;
|
|
}
|
|
|
|
const detailResponse = await got({
|
|
method: 'get',
|
|
url: item.link,
|
|
});
|
|
const content = load(detailResponse.data);
|
|
content('div.SubscriptionInner').remove();
|
|
content('.gmailNews').remove();
|
|
const topImage = content('.fullPic').html();
|
|
|
|
item.description = (topImage === null ? '' : topImage) + content('.paragraph').eq(0).html();
|
|
item.category = [
|
|
...content("meta[property='article:tag']")
|
|
.toArray()
|
|
.map((e) => e.attribs.content),
|
|
content('.active > a').text(),
|
|
];
|
|
|
|
return item;
|
|
}
|