ershisan99/scrape-icp
1
0
Fork 0
mirror of https://github.com/ershisan99/scrape-icp.git synced 2025-12-18 05:09:32 +00:00
Code Issues Packages Projects Releases Wiki Activity

try another url

Browse Source
This commit is contained in:
andres
2023-03-14 01:52:45 +01:00
parent d9fdbe3c59
commit fa96716e8a
1 changed files with 11 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files

16
src/index.ts
View File

@@ -1,5 +1,5 @@
import { chromium, type Page } from 'playwright' import { chromium, type Page } from 'playwright'
import { type Browser, expect } from '@playwright/test' import { type BrowserContext, expect } from '@playwright/test'
import { Markup, Telegraf } from 'telegraf' import { Markup, Telegraf } from 'telegraf'
import dotenv from 'dotenv' import dotenv from 'dotenv'
import { type InputFile } from 'telegraf/types' import { type InputFile } from 'telegraf/types'
@@ -7,8 +7,8 @@ import { message } from 'telegraf/filters'
import cron from 'node-cron' import cron from 'node-cron'
dotenv.config() dotenv.config()
const url = 'https://google.com' // const url = 'https://google.com'
// const url = 'https://icp.administracionelectronica.gob.es/icpplustiem/citar?p=28&locale=es&appkey=null' const url = 'https://icp.administracionelectronica.gob.es/icpplustiem/citar?p=28&locale=es&appkey=null'
const { NIE, NAME, COUNTRY_CODE } = process.env const { NIE, NAME, COUNTRY_CODE } = process.env
const NO_APPOINTMENT = 'no hay citas disponibles' const NO_APPOINTMENT = 'no hay citas disponibles'
const delay = async (ms: number) => await new Promise(resolve => setTimeout(resolve, ms)) const delay = async (ms: number) => await new Promise(resolve => setTimeout(resolve, ms))
@@ -78,7 +78,7 @@ const step4 = async (page: Page) => {
await sendScreenshot(bot, page, 'step 4') await sendScreenshot(bot, page, 'step 4')
} }
const step5 = async (page: Page, browser: Browser) => { const step5 = async (page: Page, browser: BrowserContext) => {
try { try {
if (await page.getByText(NO_APPOINTMENT).count() > 0) { if (await page.getByText(NO_APPOINTMENT).count() > 0) {
await sendScreenshot(bot, page, 'step 5 failed') await sendScreenshot(bot, page, 'step 5 failed')
@@ -103,7 +103,7 @@ const step6 = async (page: Page) => {
} }
const scrape = async () => { const scrape = async () => {
const browser = await chromium.launch({ const chr = await chromium.launch({
timeout: 200000, timeout: 200000,
logger: { logger: {
isEnabled: (name, severity) => name === 'browser', isEnabled: (name, severity) => name === 'browser',
@@ -111,6 +111,12 @@ const scrape = async () => {
log: (name, severity, message, args) => { console.log(`${name} ${message}`) } log: (name, severity, message, args) => { console.log(`${name} ${message}`) }
} }
}) })
const browser = await chr.newContext({
geolocation: { longitude: -3.703790, latitude: 40.416775 },
userAgent: 'Chrome/91.0.4472.77',
timezoneId: 'Europe/Madrid',
locale: 'es-ES'
})
try { try {
const page = await browser.newPage() const page = await browser.newPage()
console.log('scraping...') console.log('scraping...')