ebkScraper.js

// Generated by CoffeeScript 1.6.3
(function() {
  var EbkScraper, baseUrl, cheerio, request,
    __bind = function(fn, me){ return function(){ return fn.apply(me, arguments); }; },
    __hasProp = {}.hasOwnProperty,
    __extends = function(child, parent) { for (var key in parent) { if (__hasProp.call(parent, key)) child[key] = parent[key]; } function ctor() { this.constructor = child; } ctor.prototype = parent.prototype; child.prototype = new ctor(); child.__super__ = parent.prototype; return child; },
    __indexOf = [].indexOf || function(item) { for (var i = 0, l = this.length; i < l; i++) { if (i in this && this[i] === item) return i; } return -1; };

  cheerio = require('cheerio');

  request = require('request');

  baseUrl = "http://kleinanzeigen.ebay.de";

  EbkScraper = (function(_super) {
    __extends(EbkScraper, _super);

    function EbkScraper(opts) {
      this.opts = opts;
      this._loadOffer = __bind(this._loadOffer, this);
      this.scrape = __bind(this.scrape, this);
    }

    EbkScraper.prototype.scrape = function() {
      var self;
      self = this;
      return request.get(baseUrl + this.opts.targetUrl, self._loadOffer);
    };

    EbkScraper.prototype._loadOffer = function(err, header, body) {
      var $, self;
      self = this;
      if (err) {
        console.error(err);
      }
      $ = cheerio.load(body);
      return $('#srchrslt-adtable > li').each(function() {
        return self._extOffer({
          titel: $(this).find('.ad-title').html(),
          insertTime: $(this).find('.ad-listitem-addon').html().toString().replace(/\s/g, ""),
          district: $(this).find('.c-h-adtble-lctn').html().slice(9),
          postcode: $(this).find('.c-h-adtble-lctn').html().slice(0, 5),
          netRent: $(this).find('.ad-listitem-details > strong').html().replace(/\./g, "").match(/\d*/)[0],
          href: baseUrl + $(this).find('.ad-title')['0'].attribs.href,
          subtitle: $(this).find('p').html()
        }, function(offer) {
          return self._checkOffer(offer);
        });
      });
    };

    EbkScraper.prototype._extOffer = function(offer, cb) {
      return request.get(offer.href, function(err, header, body) {
        var $;
        if (err) {
          console.error(err);
        }
        $ = cheerio.load(body);
        offer.street = $('#street-address').text();
        offer.insertDate = $($('.c-attrlist > dd')[1]).text();
        offer.id = $($('.c-attrlist > dd')[2]).text();
        offer.rooms = $($('.c-attrlist > dd')[3]).find('span').text().replace(/\s/g, "");
        offer.size = $($('.c-attrlist > dd')[4]).find('span').text().replace(/\s/g, "");
        offer.description = $('p[itemprop="description"]').text();
        offer.phone = $($('.viewad-contact-phone')[0]).text();
        return cb(offer);
      });
    };

    EbkScraper.prototype._checkOffer = function(offer) {
      var _ref, _ref1;
      if (!((_ref = offer.district, __indexOf.call(this.opts.district, _ref) >= 0) || (_ref1 = offer.postcode, __indexOf.call(this.opts.postcode, _ref1) >= 0))) {
        return;
      }
      if (offer.netRent < this.opts.netRent.min || offer.netRent > this.opts.netRent.max) {
        return;
      }
      if (offer.rooms < this.opts.rooms.min || offer.rooms > this.opts.rooms.max) {
        return;
      }
      if (offer.size < this.opts.size.min || offer.size > this.opts.size.max) {
        return;
      }
      if (this.opts.regex && !offer.description.match(this.opts.regex)) {
        return;
      }
      return this.emit('match', offer);
    };

    return EbkScraper;

  })(require('events').EventEmitter);

  module.exports = EbkScraper;

}).call(this);