External/fredy
1
0
Fork 0
mirror of https://github.com/orangecoding/fredy.git synced 2026-06-16 12:31:07 +00:00
Code Issues Packages Projects Releases Wiki Activity

Puppeteer rewrite (#119)

Browse Source 
* 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
Download Patch File Download Diff File Expand all files Collapse all files

14
lib/provider/einsAImmobilien.js
View File

@@ -2,14 +2,10 @@ import utils, { buildHash } from '../utils.js';
let appliedBlackList = [];
function normalize(o) {
let size = `${o.size.replace(' Wohnfläche ', '').trim()}`;
if (o.rooms != null) {
size += ` / / ${o.rooms.trim()}`;
}
const link = `https://www.1a-immobilienmarkt.de/expose/${o.id}.html`;
const price = normalizePrice(o.price);
const id = buildHash(o.id, price);
return Object.assign(o, { id, price, size, link });
return Object.assign(o, { id, price, link });
}
/**
@@ -39,12 +35,12 @@ const config = {
url: null,
crawlContainer: '.tabelle',
sortByDateParam: 'sort_type=newest',
waitForSelector: 'body',
crawlFields: {
id: '.inner_object_data input[name="marker_objekt_id"]@value | int',
price: '.tabelle .inner_object_data .single_data_price | removeNewline | trim',
size: '.tabelle .inner_object_data .data_boxes div:nth-child(1)',
rooms: '.tabelle .inner_object_data .data_boxes div:nth-child(2)',
title: '.tabelle .inner_object_data .tabelle_inhalt_titel_black | removeNewline | trim',
price: '.inner_object_data .single_data_price | removeNewline | trim',
size: '.tabelle .tabelle_inhalt_infos .single_data_box | removeNewline | trim',
title: '.inner_object_data .tabelle_inhalt_titel_black | removeNewline | trim',
},
normalize: normalize,
filter: applyBlacklist,