김대선

cheerio, sanitize-html 설치 및 네이버 채용정보 데이터를 받아와서, 항목별로 나누기 까지 구현

const express = require('express')
const router = express.Router()
const puppeteer = require('puppeteer')
const cheerio = require('cheerio')
const sanitizeHtml =require('sanitize-html')
const splitData = (string)=>{
return string.replace(/(<([^>]+)>)*(\\t)?/gi, "").split("\t").filter((ele)=> ele != '')
}
const makeObject = (array)=>{
const result = [];
for(let i of array){
result.push(splitData(i))
}
return result
}
router.get('/', (req, res, error)=>{
res.send("naver home")
})
router.get('/recruitment', async (req, res, error)=>{
const browser = await puppeteer.launch();
const page = await browser.newPage();
await page.goto('https://recruit.navercorp.com/naver/job/list/developer')
let content = await page.content()
let temp = null;
while(true){
if (temp == content) break;
else{
temp = await page.content()
await page.click('#moreDiv > button').catch((error)=>{
})
await page.waitForTimeout(300)
content = await page.content()
}
}
let $ = cheerio.load(content, {decodeEntities: true})
const selector = '#jobListDiv > ul > li'
let result = sanitizeHtml($(selector), {
parser : {
decodeEntities: true
}
})
resArr = result.split('</li><li>')
result = makeObject(resArr)
console.log(result)
res.send(result)
})
module.exports = router
\ No newline at end of file