mirror of
https://github.com/orangecoding/fredy.git
synced 2026-06-16 12:31:07 +00:00
770d30af95
Fredy 2.0.0 introduces the concept of search jobs. You can now configure multiple of these jobs running in the same instance of Fredy. Also a new API has been created 🎉
49 lines
1.4 KiB
JavaScript
Executable File
49 lines
1.4 KiB
JavaScript
Executable File
const utils = require('../utils');
|
|
|
|
let appliedBlackList = [];
|
|
|
|
function normalize(o) {
|
|
const id = parseInt(o.id.substring(o.id.indexOf('_') + 1, o.id.length));
|
|
const size = o.size != null ? o.size.replace('Wohnfläche ', '') : 'N/A m²';
|
|
const price = o.price.replace('Kaufpreis ', '');
|
|
const address = o.address.split(' • ')[1];
|
|
const title = o.title || 'No title available';
|
|
|
|
return Object.assign(o, { id, address, price, size, title });
|
|
}
|
|
|
|
function applyBlacklist(o) {
|
|
const titleNotBlacklisted = !utils.isOneOf(o.title, appliedBlackList);
|
|
const descNotBlacklisted = !utils.isOneOf(o.description, appliedBlackList);
|
|
|
|
return titleNotBlacklisted && descNotBlacklisted;
|
|
}
|
|
|
|
const config = {
|
|
enabled: null,
|
|
url: null,
|
|
crawlContainer: '#result-list-stage .item',
|
|
crawlFields: {
|
|
id: '@id',
|
|
price: 'div[id*="selPrice_"] | trim',
|
|
size: 'div[id*="selArea_"] | trim',
|
|
title: '.item a img@title',
|
|
link: 'a[id*="lnkImgToDetails_"]@href',
|
|
address: '.item .box-25 .ellipsis .text-100 | removeNewline | trim'
|
|
},
|
|
paginate: '#idResultList .margin-bottom-6.margin-bottom-sm-12 .panel a.pull-right@href',
|
|
normalize: normalize,
|
|
filter: applyBlacklist
|
|
};
|
|
|
|
exports.init = (sourceConfig, blacklist) => {
|
|
config.enabled = sourceConfig.enabled;
|
|
config.url = sourceConfig.url;
|
|
appliedBlackList = blacklist;
|
|
};
|
|
|
|
//must match the id of the source given in the config!
|
|
exports.id = () => 'immonet';
|
|
|
|
exports.config = config;
|