fix(route): 国家药品监督管理局医疗器械技术审评中心 获取全文 (#10357)

* Update index.js

* Update index.js

* Update index.js

* Update index.js

* Update cdrh.js

* Revert "Update cdrh.js"

This reverts commit 18b5174f109f8e7a26c974d47357b6309ec64781.

* Update lib/v2/cmde/index.js


* Update index.js
This commit is contained in:
terrytw 2022-08-02 21:32:04 +08:00 committed by GitHub
parent 681150af88
commit 85afee25fd
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 60 additions and 29 deletions

View File

@ -1,44 +1,75 @@
const cheerio = require('cheerio');
const timezone = require('@/utils/timezone');
const { parseDate } = require('@/utils/parse-date');
const rootURL = 'https://www.cmde.org.cn';
module.exports = async (ctx) => {
const cate = ctx.params.cate ?? 'xwdt/zxyw';
const link = `${rootURL}/${cate}/`;
const url = `${rootURL}/${cate}/`;
const browser = await require('@/utils/puppeteer')({ stealth: true });
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(link);
await page.waitForSelector('.child-list .list ul');
const html = await page.evaluate(() => document.querySelector('.child-list .list ul').innerHTML);
const title = await page.evaluate(() => document.title);
browser.close();
const data = await ctx.cache.tryGet(
url,
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(url, {
waitUntil: 'domcontentloaded',
});
await page.waitForSelector('.list');
const html = await page.evaluate(() => document.documentElement.innerHTML);
await page.close();
const $ = cheerio.load(html);
const list = $('li');
const $ = cheerio.load(html);
const data = {
title,
link,
item: list
.map((_, item) => ({
title: $(item).find('a').attr('title'),
link: new URL($(item).find('a').attr('href'), link).href,
pubDate: parseDate($(item).find('span:last-child').text().replace(/\(|\)/g, '')),
}))
.get(),
};
return {
title: $('head title').text(),
description: $('meta[name=ColumnDescription]').attr('content'),
items: $('.list ul li')
.toArray()
.map((item) => {
item = $(item);
return {
title: $(item).find('a').attr('title'),
link: new URL($(item).find('a').attr('href'), url).href,
};
}),
};
},
);
ctx.state.data = data;
const items = await Promise.all(
data.items.map((item) =>
ctx.cache.tryGet(item.link, 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(item.link, {
waitUntil: 'domcontentloaded',
});
await page.waitForSelector('.text');
ctx.state.json = {
const html = await page.evaluate(() => document.documentElement.innerHTML);
await page.close();
const $ = cheerio.load(html);
item.description = $('.text').html();
item.pubDate = timezone(parseDate($('meta[name="PubDate"]').attr('content')), +8);
return item;
})
)
);
await browser.close();
ctx.state.data = {
title: data.title,
link: data.link,
item: data.item.map((item) => item.title),
description: data.description,
link: url,
item: items,
};
};