Puppeteer rewrite (#119)

* Moving to puppeteer | removing scrapingAnt
This commit is contained in:
Christian Kellner
2024-12-17 12:38:28 +01:00
committed by GitHub
parent 58965a6f1b
commit 214e714c03
30 changed files with 1367 additions and 1190 deletions

View File

@@ -11,8 +11,9 @@ function normalize(o) {
const price = o.price || 'N/A €';
const title = o.title || 'No title available';
const address = o.address || 'No address available';
const link = shortenLink(o.link);
const id = buildHash(parseId(shortenLink(o.link)), o.price);
const shortLink = shortenLink(o.link);
const link = `https://www.immobilien.de/${shortLink}`;
const id = buildHash(parseId(shortLink), o.price);
return Object.assign(o, { id, price, size, title, address, link });
}
function applyBlacklist(o) {
@@ -22,9 +23,11 @@ function applyBlacklist(o) {
}
const config = {
url: null,
crawlContainer: '.estates_list .list_immo a._ref',
crawlContainer: '._ref',
sortByDateParam: 'sort_col=*created_ts&sort_dir=desc',
waitForSelector: 'body',
crawlFields: {
id: '@href', //will be transformed later
price: '.list_entry .immo_preis .label_info',
size: '.list_entry .flaeche .label_info | removeNewline | trim',
title: '.list_entry .part_text h3 span',
@@ -32,7 +35,6 @@ const config = {
link: '@href',
address: '.list_entry .place',
},
paginate: '.list_immo .blocknav .blocknav_list li.next a@href',
normalize: normalize,
filter: applyBlacklist,
};