upgrading dependencies, fixing immowelt, using hash of price and id as unique identifier for listings

This commit is contained in:
Christian Kellner
2024-09-05 13:34:14 +02:00
parent 1bf012f13e
commit 35feb772d7
15 changed files with 680 additions and 446 deletions

View File

@@ -1,44 +1,49 @@
import utils from '../utils.js';
import utils, {buildHash} from '../utils.js';
let appliedBlackList = [];
let appliedBlacklistedDistricts = [];
function normalize(o) {
const size = o.size || '--- m²';
return Object.assign(o, { size });
const size = o.size || '--- m²';
const id = buildHash(o.id, o.price);
return Object.assign(o, {id, size});
}
function applyBlacklist(o) {
const titleNotBlacklisted = !utils.isOneOf(o.title, appliedBlackList);
const descNotBlacklisted = !utils.isOneOf(o.description, appliedBlackList);
const isBlacklistedDistrict =
appliedBlacklistedDistricts.length === 0 ? false : utils.isOneOf(o.description, appliedBlacklistedDistricts);
return !isBlacklistedDistrict && titleNotBlacklisted && descNotBlacklisted;
const titleNotBlacklisted = !utils.isOneOf(o.title, appliedBlackList);
const descNotBlacklisted = !utils.isOneOf(o.description, appliedBlackList);
const isBlacklistedDistrict =
appliedBlacklistedDistricts.length === 0 ? false : utils.isOneOf(o.description, appliedBlacklistedDistricts);
return !isBlacklistedDistrict && titleNotBlacklisted && descNotBlacklisted;
}
const config = {
url: null,
crawlContainer: '#srchrslt-adtable .ad-listitem ',
//sort by date is standard oO
sortByDateParam: null,
crawlFields: {
id: '.aditem@data-adid | int',
price: '.aditem-main--middle--price-shipping--price | removeNewline | trim',
size: '.aditem-main .text-module-end span:nth-child(2) | removeNewline | trim',
title: '.aditem-main .text-module-begin a | removeNewline | trim',
link: '.aditem-main .text-module-begin a@href | removeNewline | trim',
description: '.aditem-main p:not(.text-module-end) | removeNewline | trim',
address: '.aditem-main--top--left | trim | removeNewline',
},
paginate: '#srchrslt-pagination .pagination-next@href',
normalize: normalize,
filter: applyBlacklist,
url: null,
crawlContainer: '#srchrslt-adtable .ad-listitem ',
//sort by date is standard oO
sortByDateParam: null,
crawlFields: {
id: '.aditem@data-adid | int',
price: '.aditem-main--middle--price-shipping--price | removeNewline | trim',
size: '.aditem-main .text-module-end span:nth-child(2) | removeNewline | trim',
title: '.aditem-main .text-module-begin a | removeNewline | trim',
link: '.aditem-main .text-module-begin a@href | removeNewline | trim',
description: '.aditem-main p:not(.text-module-end) | removeNewline | trim',
address: '.aditem-main--top--left | trim | removeNewline',
},
paginate: '#srchrslt-pagination .pagination-next@href',
normalize: normalize,
filter: applyBlacklist,
};
export const metaInformation = {
name: 'Ebay Kleinanzeigen',
baseUrl: 'https://www.kleinanzeigen.de/',
id: 'kleinanzeigen',
name: 'Ebay Kleinanzeigen',
baseUrl: 'https://www.kleinanzeigen.de/',
id: 'kleinanzeigen',
};
export const init = (sourceConfig, blacklist, blacklistedDistricts) => {
config.enabled = sourceConfig.enabled;
config.url = sourceConfig.url;
appliedBlacklistedDistricts = blacklistedDistricts || [];
appliedBlackList = blacklist || [];
config.enabled = sourceConfig.enabled;
config.url = sourceConfig.url;
appliedBlacklistedDistricts = blacklistedDistricts || [];
appliedBlackList = blacklist || [];
};
export { config };
export {config};