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

fixing immoswp

Browse Source
This commit is contained in:
Christian Kellner
2024-06-12 14:15:21 +02:00
parent b075e09ac2
commit 06c4ebb975
2 changed files with 35 additions and 34 deletions
Download Patch File Download Diff File Expand all files Collapse all files

67
lib/provider/immoswp.js
View File

@@ -1,44 +1,47 @@
import utils from '../utils.js'; import utils from '../utils.js';
let appliedBlackList = []; let appliedBlackList = [];
function normalize(o) { function normalize(o) {
const id = o.id.substring(o.id.indexOf('-') + 1, o.id.length); const id = o.id.substring(o.id.indexOf('-') + 1, o.id.length);
const size = o.size || 'N/A m²'; const size = o.size || 'N/A m²';
const price = (o.price || '--- €').replace('Preis auf Anfrage', '--- €'); const price = (o.price || '--- €').replace('Preis auf Anfrage', '--- €');
const address = o.address || 'No address available'; const title = o.title || 'No title available';
const title = o.title || 'No title available'; const link = `https://immo.swp.de/immobilien/${id}`;
const link = `https://immo.swp.de/immobilien/${id}`; const description = o.description;
const description = o.description; return Object.assign(o, {id, price, size, title, link, description});
return Object.assign(o, { id, address, price, size, title, link, description });
} }
function applyBlacklist(o) { function applyBlacklist(o) {
const titleNotBlacklisted = !utils.isOneOf(o.title, appliedBlackList); const titleNotBlacklisted = !utils.isOneOf(o.title, appliedBlackList);
const descNotBlacklisted = !utils.isOneOf(o.description, appliedBlackList); const descNotBlacklisted = !utils.isOneOf(o.description, appliedBlackList);
return titleNotBlacklisted && descNotBlacklisted; return titleNotBlacklisted && descNotBlacklisted;
} }
const config = { const config = {
url: null, url: null,
crawlContainer: '.js-serp-item', crawlContainer: '.js-serp-item',
sortByDateParam: 's=most_recently_updated_first', sortByDateParam: 's=most_recently_updated_first',
crawlFields: { crawlFields: {
id: '@id', id: '.js-bookmark-btn@data-id',
price: 'div.item__spec.item-spec-price | trim', price: 'div.align-items-start div:first-child | trim',
size: 'div.item__spec.item-spec-area | trim', size: 'div.align-items-start div:nth-child(3) | trim',
title: 'a.js-item-title-link@title', title: '.card-title h2 | trim',
address: 'div.item__locality | removeNewline | trim', link: '.ci-search-result__link@href',
description: 'div.item__main-info-points.clearfix p small | removeNewline | trim', description: '.js-show-more-item-sm | removeNewline | trim',
}, },
paginate: 'li.page-item.pagination__item a.page-link@href', paginate: 'li.page-item.pagination__item a.page-link@href',
normalize: normalize, normalize: normalize,
filter: applyBlacklist, filter: applyBlacklist,
}; };
export const init = (sourceConfig, blacklist) => { export const init = (sourceConfig, blacklist) => {
config.enabled = sourceConfig.enabled; config.enabled = sourceConfig.enabled;
config.url = sourceConfig.url; config.url = sourceConfig.url;
appliedBlackList = blacklist || []; appliedBlackList = blacklist || [];
}; };
export const metaInformation = { export const metaInformation = {
name: 'Immo Südwest Presse', name: 'Immo Südwest Presse',
baseUrl: 'https://immo.swp.de/', baseUrl: 'https://immo.swp.de/',
id: 'immoswp', id: 'immoswp',
}; };
export { config }; export {config};

2
test/provider/immoswp.test.js
View File

@@ -25,12 +25,10 @@ describe('#immoswp testsuite()', () => {
expect(notify.size).to.be.a('string'); expect(notify.size).to.be.a('string');
expect(notify.title).to.be.a('string'); expect(notify.title).to.be.a('string');
expect(notify.link).to.be.a('string'); expect(notify.link).to.be.a('string');
expect(notify.address).to.be.a('string');
/** check the values if possible **/ /** check the values if possible **/
expect(notify.price).that.does.include('€'); expect(notify.price).that.does.include('€');
expect(notify.title).to.be.not.empty; expect(notify.title).to.be.not.empty;
expect(notify.link).that.does.include('https://immo.swp.de'); expect(notify.link).that.does.include('https://immo.swp.de');
expect(notify.address).to.be.not.empty;
}); });
resolve(); resolve();
}); });