Add ssl to sequelize

Improve olx scrapper

See merge request saburly/marketalarm/web!113

app/config/appConfig.js

@@ -45,6 +45,11 @@ const USER_AGENT =
   process.env.USER_AGENT ||
   "Mozilla/5.0 (Windows NT 6.1; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/65.0.3325.181 Safari/537.36";
 
+const USE_SCRAPER_API = process.env.USE_SCRAPER_API === undefined ? 1 : parseInt(process.env.USE_SCRAPER_API);
+const SCRAPER_API_KEY = process.env.SCRAPER_API_KEY || "";
+const SCRAPER_API_BASE_URL = process.env.SCRAPER_API_BASE_URL || "";
+const NODE_FETCH_TIMEOUT_MS = parseInt(process.env.NODE_FETCH_TIMEOUT_MS) || 60000
+
 module.exports = {
   APP_PORT,
   APP_URL,
@@ -59,5 +64,9 @@ module.exports = {
   STAGING,
   CHECK_UP_DAYS,
   PROSTOR_LOGIN,
-  USER_AGENT
+  USER_AGENT,
+  USE_SCRAPER_API,
+  SCRAPER_API_KEY,
+  SCRAPER_API_BASE_URL,
+  NODE_FETCH_TIMEOUT_MS
 };

app/crawler/crawl.js

@@ -10,6 +10,7 @@ const RentalCrawler = require("./specificCrawlers/rental");
 const ProstorCrawler = require("./specificCrawlers/prostor");
 const AktidoCrawler = require("./specificCrawlers/aktido");
 const SaljicCrawler = require("./specificCrawlers/saljic");
+const { logDebug } = require("../helpers/log");
 
 const {
   OLX_CONFIG,
@@ -75,7 +76,9 @@ async function crawlAll() {
 
   for (const crawler of crawlers) {
     try {
+      logDebug('Starting crawler: ', crawler);
       const newRealEstatesFromSingleCrawler = await crawler.crawl();
+      logDebug('Crawler done: ', crawler);
       if (Array.isArray(newRealEstatesFromSingleCrawler)) {
         newRealEstates.push(...newRealEstatesFromSingleCrawler);
       }

app/crawler/specificCrawlers/aktido.js

@@ -159,7 +159,7 @@ class AktidoCrawler {
     }
 
     try {
-      const res = await fetch(url);
+      const res = await fetch(url, {}, false);
       const body = await res.text();
       const $ = cheerio.load(body);
       let hrefs = [];
@@ -202,6 +202,10 @@ class AktidoCrawler {
       const body = await adPageSource.text();
       const $ = cheerio.load(body);
 
+      if (body.indexOf('<html') === -1) {
+        throw { message: 'Failed to fetch page !' }
+      }
+
       const mapElementParent = $(".box-map").parent();
       const scriptElement = $("script", mapElementParent);
       if (

app/crawler/specificCrawlers/olx.js

@@ -1,6 +1,7 @@
 "use strict";
 
 const fetch = require("../../helpers/fetchWrapper");
+const { logDebug } = require("../../helpers/log");
 const cheerio = require("cheerio");
 const Promise = require("bluebird");
 const moment = require("moment-timezone");
@@ -44,6 +45,16 @@ const OLX_ENUMS = {
 
 const { OLX_FORCE_CRAWL } = require("../specificConfigs/olx");
 
+const chunk = (array, size = 10) => {
+  let i, j ,temparray;
+  const result = []
+  for (i=0,j=array.length; i<j; i+=size) {
+    temparray = array.slice(i,i+size);
+    result.push(temparray);
+  }
+  return result;
+}
+
 class OlxCrawler {
   constructor(
     savers = [],
@@ -52,7 +63,7 @@ class OlxCrawler {
     maxPages = 1000,
     maxResultsPerPage = 100,
     ignoredUsernames = [],
-    delayBetweenPages = 1000
+    delayBetweenPages = 500
   ) {
     this.savers = savers;
     this.baseUrl = "https://www.olx.ba/pretraga?sort_order=desc&sort_po=datum";
@@ -65,6 +76,7 @@ class OlxCrawler {
   }
 
   async crawl() {
+    logDebug("Starting OLX crawl");
     const crawlAdCategories = this.crawlerAdCategories;
 
     const newRealEstates = [];
@@ -88,14 +100,32 @@ class OlxCrawler {
         const entries = singlePageResults.entries();
 
         for (const [index, { value: singlePageResult }] of entries) {
+          if (PRINT_CRAWLER_DEBUG) {
+            console.log("================================");
+            console.log("Category Indexer index : ", index);
+          }
+
           if (singlePageResult) {
+            console.log("\tTotal entries          : ", singlePageResult.length)
             const saveResults = await this.saveCrawledResults(singlePageResult);
             const { newRecords, existingRecords } = saveResults;
 
+            if (PRINT_CRAWLER_DEBUG) {
+              console.log("--------------------------");
+              console.log("\tNew record URLs [", newRecords.length, "] :");
+
+              for(const newRecord of newRecords) {
+                console.log("\t\t",newRecord.url);
+              }
+
+              console.log("\t-------------------------");
+              console.log("\tExisting record URLs [", existingRecords.length, "] :");
+            }
+
             newRealEstates.push(...newRecords);
 
             for (const existingRecord of existingRecords) {
-              const { publishedDate, renewedDate } = existingRecord;
+              const { publishedDate, renewedDate, url } = existingRecord;
 
               const publishedDateMoment = moment.utc(publishedDate);
               const renewedDateMoment = moment.utc(renewedDate);
@@ -105,13 +135,25 @@ class OlxCrawler {
                 "minute"
               );
 
+              if (PRINT_CRAWLER_DEBUG) {
+                console.log("\t\t", url);
+                console.log("\t\t\tPublished date         : ", publishedDate);
+                console.log("\t\t\tRenewed date           : ", renewedDate);
+                console.log("\t\t\tIs same (up to minute) : ", stopCrawlingThisCategory);
+              }
+
               if (stopCrawlingThisCategory && !OLX_FORCE_CRAWL) {
                 generatorsToRemove[index] = true;
-                // console.log("\tGenerator ", index + 1, "has no more new ads");
+                if (PRINT_CRAWLER_DEBUG) {
+                  console.log("\t\t\tStopping this category indexer");
+                }
                 break;
               }
             }
           } else {
+            if (PRINT_CRAWLER_DEBUG) {
+              console.log("\tNo more entries in this category, stopping!");
+            }
             //Generator returned undefined, remove this generator from array
             generatorsToRemove[index] = true;
             // console.log("Generator ", index + 1, "has no more pages");
@@ -136,31 +178,36 @@ class OlxCrawler {
   }
 
   async *categoryIndexer(adCategory) {
-    let pageToIndex = 1;
+    try {
+      let pageToIndex = 1;
 
-    const urlAdTypePart = OLX_ENUMS.OLX_AD_TYPE[this.crawlerAdTypes];
-    const urlCategoryPart = OLX_ENUMS.OLX_AD_CATEGORY[adCategory];
-    if (urlAdTypePart !== undefined && urlCategoryPart !== undefined) {
-      while (true) {
-        const urlPageToCrawl = `${this.baseUrl}${urlAdTypePart}${urlCategoryPart}&stranica=${pageToIndex}`;
-        const singlePageResults = await this.indexSinglePage(
-          urlPageToCrawl,
-          this.maxResultsPerPage
-        );
+      const urlAdTypePart = OLX_ENUMS.OLX_AD_TYPE[this.crawlerAdTypes];
+      const urlCategoryPart = OLX_ENUMS.OLX_AD_CATEGORY[adCategory];
+      if (urlAdTypePart !== undefined && urlCategoryPart !== undefined) {
+        while (true) {
+          const urlPageToCrawl = `${this.baseUrl}${urlAdTypePart}${urlCategoryPart}&stranica=${pageToIndex}`;
+          const singlePageResults = await this.indexSinglePage(
+            urlPageToCrawl,
+            this.maxResultsPerPage
+          );
 
-        if (Array.isArray(singlePageResults) && singlePageResults.length > 0) {
-          yield singlePageResults;
-        } else {
-          return undefined;
-        }
-
-        ++pageToIndex;
-        if (pageToIndex === this.maxPages) {
-          return undefined;
+          await this.sleep(this.delayBetweenPages);
+          if (Array.isArray(singlePageResults) && singlePageResults.length > 0) {
+            yield singlePageResults;
+          } else {
+            return undefined;
+          }
+
+          ++pageToIndex;
+          if (pageToIndex === this.maxPages) {
+            return undefined;
+          }
         }
+      } else {
+        return undefined;
       }
-    } else {
-      return undefined;
+    } catch (e) {
+      console.log('Error inside generator: ', e);
     }
   }
 
@@ -170,8 +217,10 @@ class OlxCrawler {
     }
 
     try {
-      const res = await fetch(url);
+      const res = await fetch(url, {}, false);
+      logDebug("Got category results for: ", url);
       const body = await res.text();
+      logDebug("Got category results text for: ", url);
       const $ = cheerio.load(body);
       let hrefs = [];
 
@@ -192,26 +241,44 @@ class OlxCrawler {
 
       const asyncScraping = [];
       for (let i = 0; i < actualNoOfResults; i++) {
-        asyncScraping.push(this.scrapeAd(hrefs[i]));
+        asyncScraping.push(hrefs[i]);
       }
 
-      const scrapedData = await Promise.all(asyncScraping);
-      const filteredScrapedData = scrapedData.filter(adData => !!adData);
+      const allChunks = chunk(asyncScraping, 2);
+      const dataResults = []
+      for (let i = 0; i < allChunks.length; i++) {
+        const singleChunk = allChunks[i];
+        const promises = singleChunk.map(c => this.scrapeAd(c))
+        const chunkResults = await Promise.all(promises);
+        await this.sleep(this.delayBetweenPages);
+        dataResults.push(...chunkResults);
+        logDebug("Chunk results len:", chunkResults.length);
+      }
+
+
+      const filteredScrapedData = dataResults.filter(adData => !!adData);
+      logDebug("Filtered scraped data length: ", filteredScrapedData.length);
+
       return filteredScrapedData;
     } catch (e) {
-      console.error("Exception caught:" + e);
+      console.error("Exception caught, index single page: " + e);
       return [];
     }
   }
 
   async scrapeAd(url) {
-    // console.log("Scraping : ", url);
+    logDebug("Scraping : ", url);
     try {
       const adPageSource = await fetch(url);
       const body = await adPageSource.text();
       const $ = cheerio.load(body);
       let status = AD_STATUS.STATUS_NORMAL;
 
+      if (body.indexOf('<html') === -1) {
+        console.error("This is the body: ", body);
+        throw { message: 'Failed to fetch page !' }
+      }
+
       const propertySelectors = {
         username:
           "#lg > div.desno2.profil > div:nth-child(2) > div.vrsta1.vrsta_desno > a > div.username > span",
@@ -238,37 +305,25 @@ class OlxCrawler {
 
       //====== PRICE DETECTION AND EXTRACTION =====
       let price = null;
-      const normalPriceValue = $("#pc > p:nth-child(2)").text();
-      const urgentPriceValue = $(
-        "#artikal_glavni_div > div.artikal_lijevo > div:nth-child(5) > p"
-      )
-        .text()
-        .trim();
 
-      if (normalPriceValue && normalPriceValue.length > 0) {
-        price = normalPriceValue;
-        if (
-          $("#pc > p.n")
-            .text()
-            .indexOf("Hitna") !== -1
-        ) {
-          status = AD_STATUS.STATUS_URGENT;
-        } else {
-          status = AD_STATUS.STATUS_NORMAL;
-        }
-      } else if (urgentPriceValue && urgentPriceValue.length > 0) {
-        const priceValues = urgentPriceValue.split("KM");
-        //priceValues will contain values like ["100000", "90000", ...], second element is urgent price
-        if (priceValues.length > 1) {
-          price = priceValues[1].trim();
-          status = AD_STATUS.STATUS_DISCOUNTED;
-        } else {
-          throw { message: "Can't find urgent price" };
-        }
-      } else {
-        throw {
-          message: "Can't find price (it is not normal nor urgent price ?)"
-        };
+      const priceHeader = $("#pc > p.n").text().trim();
+      const priceValue = $("#pc > p:nth-child(2)").text().trim();
+      price = priceValue;
+
+      if (priceHeader.indexOf('Hitn') !== -1) {
+        // Urgent price
+        status = AD_STATUS.STATUS_URGENT;
+      }
+
+      const discountPriceTag = $("#artikal_glavni_div > div.artikal_lijevo > p:nth-child(4)").text().trim();
+      if (discountPriceTag.indexOf('Akcij') !== -1) {
+        status = AD_STATUS.STATUS_DISCOUNTED;
+        const discountPriceValues = $("#artikal_glavni_div > div.artikal_lijevo > div:nth-child(5) > p").text().trim();
+        // discountPriceValues contain string like "10.000 KM 7.500 KM"
+        // First price is regular, second is currently active (discounted) price
+        const bothPrices = discountPriceValues.split('KM');
+        // Now, currently active price is second element of bothPrices array
+        price = bothPrices[1] ? bothPrices[1].trim() : null;
       }
 
       //====== OTHER AD INFORMATION ===============
@@ -278,7 +333,7 @@ class OlxCrawler {
 
       let otherInformationDivId;
       //We need to locate DIV ID where other information are stored
-      for (let possibleId = 10; possibleId <= 20; possibleId++) {
+      for (let possibleId = 1; possibleId <= 30; possibleId++) {
         const adTypeFieldTitle = $(
           `#artikal_glavni_div > div.artikal_lijevo > div:nth-child(${possibleId}) > div:nth-child(2) > div.df1`
         )
@@ -650,10 +705,12 @@ class OlxCrawler {
         distanceToRiver,
         numberOfViewsAgency
       };
+      //
+      //console.log("Scraped data:", data);
 
       return data;
     } catch (e) {
-      console.error("Exception caught: " + e.message, "\r\nURL:", url);
+      console.error("Exception caught scrapeAd : " + e.message, "\r\nURL:", url);
     }
     return null;
   }
@@ -768,6 +825,9 @@ class OlxCrawler {
     if (!priceText) {
       return NaN;
     }
+    if (priceText === "Po dogovoru") {
+      return null;
+    }
     const formattedPriceText = priceText.replace(".", "").replace(",", ".");
     return parseFloat(formattedPriceText);
   }

app/crawler/specificCrawlers/prostor.js

@@ -4,6 +4,7 @@ const fetch = require("../../helpers/fetchWrapper");
 const cheerio = require("cheerio");
 const moment = require("moment-timezone");
 const FormData = require("form-data");
+const nodeFetch = require("node-fetch");
 
 const {
   AD_TYPE,
@@ -62,13 +63,19 @@ class ProstorCrawler {
 
   async crawl() {
     const crawlAdCategories = this.crawlerAdCategories;
+    const crawlAdTypes = this.crawlerAdTypes;
+    if (!crawlAdCategories || !crawlAdTypes) {
+      return []
+    }
+
+    const newRealEstates = [];
     //We need session cookie to use login privileges
     const prostorCookie = await this.getCookies();
-    //New tag to check if crawler loged in
+    //New tag to check if crawler logged in
     const login = await this.loginForScraping(PROSTOR_LOGIN, prostorCookie);
-    const newRealEstates = [];
+
     //Crawl only if login was successful
-    if (crawlAdCategories && login) {
+    if (login) {
       const indexGenerators = [];
       for (const adCategory of crawlAdCategories) {
         indexGenerators.push(this.categoryIndexer(adCategory, prostorCookie));
@@ -134,6 +141,11 @@ class ProstorCrawler {
         prostorCookie
       );
 
+      if (!Array.isArray(listOfAllRealEstates)){
+        console.log('[PROSTOR] Could not find real estate JSON data, check selector !');
+        return undefined;
+      }
+
       let elementToStartIndexFrom = 0;
       while (true) {
         const realEstatesForSinglePage = listOfAllRealEstates.slice(
@@ -197,12 +209,16 @@ class ProstorCrawler {
 
     // console.log("[PROSTOR] Scraping : ", url);
     try {
-      const adPageSource = await fetch(url, {
+      const adPageSource = await nodeFetch(url, {
         headers: { Cookie: prostorCookie }
       });
       const body = await adPageSource.text();
       const $ = cheerio.load(body);
 
+      if (body.indexOf('<html') === -1) {
+        throw { message: 'Failed to fetch page !' }
+      }
+
       // link contains part of the URL in the format of : /prodaja/stan/stup/9556
       // general form is : /actionType/realEstateType/location/realEstateID
       // linkParts contains : ['', 'actionType', 'realEstateType', 'location', 'realEstateID']
@@ -427,14 +443,14 @@ class ProstorCrawler {
     }
 
     try {
-      const res = await fetch(url, {
+      const res = await nodeFetch(url, {
         headers: { Cookie: prostorCookie }
       });
       const body = await res.text();
       const $ = cheerio.load(body);
 
       const scriptElement = $(
-        "body > div > div.container-fluid > script:nth-child(7)"
+          "body > div.content > div.container-fluid > script:nth-child(6)"
       );
 
       if (
@@ -591,7 +607,7 @@ class ProstorCrawler {
     formData.append("email", PROSTOR_LOGIN.EMAIL);
     formData.append("password", PROSTOR_LOGIN.PASSWORD);
 
-    return fetch("https://prostor.ba/moj-prostor/prijava", {
+    return nodeFetch("https://prostor.ba/moj-prostor/prijava", {
       method: "POST",
       body: formData,
       headers: { Cookie: prostorCookie }
@@ -618,9 +634,12 @@ class ProstorCrawler {
       });
   }
   async getCookies() {
-    const getResponse = await fetch("https://prostor.ba/moj-prostor/prijava", {
-      headers: { Cookie: "" }
-    });
+    const getResponse = await nodeFetch(
+      "https://prostor.ba/moj-prostor/prijava",
+      {
+        headers: { Cookie: "" }
+      }
+    );
     const raw = getResponse.headers.raw()["set-cookie"];
     const cookie = raw
       .map(datastring => {

app/crawler/specificCrawlers/rental.js

@@ -159,7 +159,7 @@ class RentalCrawler {
     }
 
     try {
-      const res = await fetch(url);
+      const res = await fetch(url, {} , false);
       const body = await res.text();
       const $ = cheerio.load(body);
       let hrefs = [];
@@ -202,6 +202,10 @@ class RentalCrawler {
       const body = await adPageSource.text();
       const $ = cheerio.load(body);
 
+      if (body.indexOf('<html') === -1) {
+        throw { message: 'Failed to fetch page !' }
+      }
+
       const mapElementParent = $(".box-map").parent();
       const scriptElement = $("script", mapElementParent);
       if (

app/crawler/specificCrawlers/saljic.js

@@ -1,8 +1,10 @@
 "use strict";
 
 const fetch = require("../../helpers/fetchWrapper");
+const { getUrlParams } = require("../../helpers/url");
 const cheerio = require("cheerio");
 const moment = require("moment-timezone");
+const PromisePool = require('@supercharge/promise-pool');
 
 const {
   AD_TYPE,
@@ -46,12 +48,13 @@ class SaljicCrawler {
     maxPages = 5000,
     maxResultsPerPage = 5000,
     ignoredUsernames = [],
-    delayBetweenPages = 1000
+    delayBetweenPages = 500
   ) {
     this.savers = savers;
     this.baseUrl = "https://www.saljicnekretnine.ba/v2/nekretnine_search";
     this.crawlerAdTypes = crawlerAdTypes;
     this.crawlerAdCategories = crawlerAdCategories;
+    this.maxPages = maxPages
     this.maxResultsPerPage = maxResultsPerPage;
     this.delayBetweenPages = delayBetweenPages;
   }
@@ -157,7 +160,7 @@ class SaljicCrawler {
     }
 
     try {
-      const res = await fetch(url);
+      const res = await fetch(url, {}, false);
       const body = await res.text();
       const $ = cheerio.load(body);
       let hrefs = [];
@@ -205,11 +208,23 @@ class SaljicCrawler {
 
       const asyncScraping = [];
       for (let i = 0; i < actualNoOfResults; i++) {
-        asyncScraping.push(this.scrapeAd(hrefsAbs[i], adTypes[i]));
+        asyncScraping.push([hrefsAbs[i], adTypes[i]]);
       }
 
-      const scrapedData = await Promise.all(asyncScraping);
-      const filteredScrapedData = scrapedData.filter(adData => !!adData);
+      
+
+      const dataResults = []
+      const { scrapedData, errors } = await PromisePool
+        .withConcurrency(2)
+        .for(asyncScraping)
+        .process(async data => {
+          const result = await this.scrapeAd(...data)
+          await this.sleep(this.delayBetweenPages);
+          dataResults.push(result)
+          return result; //TODO: this does not work, scrapedData is null, dataResults works
+        })
+
+      const filteredScrapedData = dataResults.filter(adData => !!adData);
       return filteredScrapedData;
     } catch (e) {
       console.error("[SALJIC] Exception caught:" + e);
@@ -224,10 +239,20 @@ class SaljicCrawler {
       const body = await adPageSource.text();
       const $ = cheerio.load(body);
 
+      if (body.indexOf('<html') === -1) {
+        throw { message: 'Failed to fetch page !' }
+      }
+
       // No information for status ex. PRODAN
       const status = AD_STATUS.STATUS_NORMAL;
       //Extracting agency ID from url
-      const agencyObjectId = parseInt(url.substring(46, url.length));
+      const agencyObjectId = url
+        ? parseInt(url.substring(46, url.length))
+        : null;
+
+      if (!agencyObjectId) {
+        throw { message : 'No agency object ID - URL changed?'}
+      }
 
       //Extracting main properties
       const propertySelectors = {
@@ -237,11 +262,10 @@ class SaljicCrawler {
           "div.content-wrap > div.container > div.col-md-8.nobottommargin > div.single-post > div.entry > div.topmargin-sm.single-product > div.product > div.product-price > ins",
         streetName:
           "div.content-wrap > div.container > div.col-md-8.nobottommargin > div.single-post > div.entry > div.entry-content.topmargin > p",
-
         descriptions:
           "div.content-wrap > div.container > div.col-md-8.nobottommargin > div.single-post > div.entry > div.entry-content.topmargin > div.toggle.toggle-bg > div.togglec >p:nth-child(1)",
         latAndLong:
-          "div.content-wrap > div.container > div.col-md-8.nobottommargin > div.single-post > div.entry > div.entry-content.topmargin > div.gmap.bottommargin > iframe"
+          "iframe"
       };
       const title = $(propertySelectors.title)
         .text()
@@ -272,14 +296,26 @@ class SaljicCrawler {
         .trim();
 
       const latAndLongSrc = $(propertySelectors.latAndLong).attr("src");
-      const latText = latAndLongSrc.substring(
-        latAndLongSrc.indexOf("marker=") + 7,
-        latAndLongSrc.indexOf("%2C", latAndLongSrc.indexOf("marker="))
-      );
-      const longText = latAndLongSrc.substring(
-        latAndLongSrc.indexOf("%2C", latAndLongSrc.indexOf("marker=")) + 3,
-        latAndLongSrc.length
-      );
+      let latText;
+      let longText;
+      if (latAndLongSrc){
+        const mapParams = getUrlParams(latAndLongSrc);
+        if (mapParams) {
+          if (mapParams['marker']){
+            const marker = mapParams['marker'].split(',');
+            latText = marker[0] ? marker[0] : undefined;
+            longText = marker[1] ? marker[1] : undefined;
+          }else{
+            if (mapParams['mlat']) {
+              latText = mapParams['mlat'];
+            }
+            if (mapParams['mlon']) {
+              longText = mapParams['mlon'];
+            }
+          }
+        }
+      }
+
       const locationLat = parseFloat(latText) || null;
       const locationLong = parseFloat(longText) || null;
 
@@ -328,11 +364,10 @@ class SaljicCrawler {
       let numberOfViewsKivi = null;
       let streetNumber = 0;
       let adStatus = status;
-      let shortDescription = descriptions.substring(
-        0,
-        descriptions.indexOf(".")
-      );
-      let longDescription = descriptions;
+      let shortDescription = descriptions
+        ? descriptions.substring(0, descriptions.indexOf("."))
+        : "";
+      let longDescription = descriptions || "";
       //Extracting data - Glavne karakteristike
       let mainFieldIndex = 1;
       do {
@@ -343,10 +378,14 @@ class SaljicCrawler {
           .replace(/[\n\r\t]/gm, "")
           .trim();
 
-        const mainFieldTitle = mainField.substring(0, mainField.indexOf("  "));
+        const mainFieldTitle = mainField
+          ? mainField.substring(0, mainField.indexOf("  "))
+          : "";
         const mainFieldValue = mainField
-          .substring(mainField.indexOf("  "), mainField.length)
-          .trim();
+          ? mainField
+              .substring(mainField.indexOf("  "), mainField.length)
+              .trim()
+          : "";
 
         switch (mainFieldTitle) {
           case "Površina":
@@ -363,7 +402,7 @@ class SaljicCrawler {
             numberOfRooms = parseInt(mainFieldValue);
             break;
           case "Broj spratova":
-            numberOfFloors = parseInt(mainFieldValue);
+            numberOfFloors = this.parseNumberOfFloors(mainFieldValue);
             break;
           case "Sprat":
             floor = parseInt(mainFieldValue);
@@ -409,6 +448,9 @@ class SaljicCrawler {
             )
             .trim();
           realEstateType = this.getAdCategoryId(categoryTmp);
+          if (!realEstateType) {
+            throw { message: 'No real estate type - page body not loaded correctly or page changed?' }
+          }
         } else {
           switch (additionalField) {
             case "Internet":
@@ -565,7 +607,7 @@ class SaljicCrawler {
 
       return data;
     } catch (e) {
-      console.error("Exception caught: " + e.message, "\r\nURL:", url);
+      console.error("[SALJIC] Exception caught: " + e.message, "\r\nURL:", url);
     }
     return null;
   }
@@ -610,6 +652,21 @@ class SaljicCrawler {
     }
   }
 
+  parseNumberOfFloors(numberOfFloorsText) {
+    const tryNumericalValue = parseInt(numberOfFloorsText);
+    if (!isNaN(tryNumericalValue)){
+      return tryNumericalValue;
+    }
+
+    // Guess number of floors based on number of + sign concatenations
+    // e.g. P+S+Pt -> 3 floors
+    if (typeof numberOfFloorsText === 'string' && numberOfFloorsText.indexOf('+') > 0) {
+      return numberOfFloorsText.split('+').length + 1
+    }
+
+    return null
+  }
+
   async sleep(ms) {
     return new Promise(resolve => setTimeout(resolve, ms));
   }

app/helpers/db/searchRequest.js

@@ -3,6 +3,7 @@ const db = require("../../models/index");
 const sequelize = require("sequelize");
 const Op = sequelize.Op;
 const { AD_CATEGORY } = require("../../common/enums");
+const { CHECK_UP_DAYS } = require("../../config/appConfig");
 
 const getSearchRequest = async searchRequestId => {
   try {
@@ -16,6 +17,22 @@ const getSearchRequest = async searchRequestId => {
 const createSearchRequest = async (searchRequestFields = {}) => {
   return await db.SearchRequest.create(searchRequestFields);
 };
+const findAllRequestsForCheckUp = async () => {
+  const checkUpOffset = 24 * 60 * 60 * 1000 * CHECK_UP_DAYS; //in miliseconds
+  const checkupDate = new Date();
+  checkupDate.setTime(checkupDate.getTime() - checkUpOffset);
+
+  const dateQuery = {
+    notifiedAt: {
+      [Op.lte]: checkupDate
+    }
+  };
+  const allRequestsForCheckUp = await db.SearchRequest.findAll({
+    where: dateQuery
+  });
+
+  return allRequestsForCheckUp;
+};
 
 const findSearchRequestsForRealEstate = async realEstate => {
   const {
@@ -459,5 +476,6 @@ const findSearchRequestsForRealEstate = async realEstate => {
 module.exports = {
   getSearchRequest,
   createSearchRequest,
-  findSearchRequestsForRealEstate
+  findSearchRequestsForRealEstate,
+  findAllRequestsForCheckUp
 };

app/helpers/db/searchRequestMatch.js

@@ -2,7 +2,6 @@
 const db = require("../../models/index");
 const sequelize = require("sequelize");
 const Op = sequelize.Op;
-const { CHECK_UP_DAYS } = require("../../config/appConfig");
 
 const findRealEstatesForSearchRequest = async searchRequestId => {
   const query = {
@@ -43,42 +42,6 @@ const findNotNotifiedMatches = async () => {
 
   return matchingRecords;
 };
-const findAllRequestsForCheckUp = async () => {
-  //First we find IDs of search request that don't need to be emailed for check up - to EXCLUDE
-  //The ones that received notification for real estate CHECK_UP_DAYS days from now
-  const date = new Date();
-  const checkUpDate = date.getDate() - CHECK_UP_DAYS;
-  date.setDate(checkUpDate);
-  const dateQuery = {
-    createdAt: {
-      [Op.gte]: date
-    }
-  };
-
-  const excludedMatches = await db.SearchRequestMatch.findAll({
-    attributes: ["searchRequestId"],
-    where: dateQuery,
-    order: [["searchRequestId", "ASC"]]
-  });
-
-  const excludedRequestsAll = excludedMatches.map(match => {
-    return match.dataValues.searchRequestId;
-  });
-  //Removing duplicate search request id-s for optimization
-  const excludedRequests = [...new Set(excludedRequestsAll)];
-
-  const query = {
-    subscribed: true,
-    id: {
-      [Op.notIn]: excludedRequests
-    }
-  };
-  const allRequestsForCheckUp = await db.SearchRequest.findAll({
-    where: query
-  });
-
-  return allRequestsForCheckUp;
-};
 
 const addMatches = async matchingRecords => {
   return await db.SearchRequestMatch.bulkCreate(matchingRecords, {
@@ -89,6 +52,5 @@ const addMatches = async matchingRecords => {
 module.exports = {
   findRealEstatesForSearchRequest,
   addMatches,
-  findNotNotifiedMatches,
-  findAllRequestsForCheckUp
+  findNotNotifiedMatches
 };

app/helpers/emailContentGenerator.js

@@ -9,6 +9,7 @@ const { AD_CATEGORY, AD_TYPE, EMAIL_FREQUENCY } = require("../common/enums");
 
 //Tag to recognize staging from development
 const stagingTag = STAGING ? "[STAGING] " : "";
+const wordOfMouthRequest = `Molimo vas <strong>recite svojim prijateljima</strong> za Kivi - što više korisnika budemo imali, moći ćemo više agencija uključiti i više nekretnina imati u bazi. Hvala!`
 
 const generateEmailFooter = (searchRequestId, emailFrequencyTitle) => {
   return `  <div>Trenutno ste prijavljeni da obavještenja o novim nekretninama primate <strong>${emailFrequencyTitle.toLowerCase()} </strong>.</div>
@@ -69,6 +70,9 @@ const generateNotificationEmail = (
   ${moreRealEstates}
   </div>
   <br/>
+  ${wordOfMouthRequest}
+  <br/>
+  <br/>
   ${emailFooter}`;
 };
 
@@ -132,6 +136,10 @@ const generateNewSearchRequestEmail = (searchRequest, matchingRealEstates) => {
   </div>
   ${matchingRealEstates.length > 0 ? instantRealEstatesText : ""}
   <br/>
+  <br/>
+  ${wordOfMouthRequest}
+  <br/>
+  <br/>
   ${emailFooter}`;
 };
 

app/helpers/fetchWrapper.js

@@ -1,13 +1,58 @@
 const nodeFetch = require("node-fetch");
-const { USER_AGENT } = require("../config/appConfig");
+const AbortController = require('abort-controller');
+const FetchCache = require('@sozialhelden/fetch-cache').default;
+
+console.log("Fc ", FetchCache)
+
+const {
+  USER_AGENT,
+  USE_SCRAPER_API,
+  SCRAPER_API_KEY,
+  SCRAPER_API_BASE_URL,
+  NODE_FETCH_TIMEOUT_MS
+} = require("../config/appConfig");
+
+const timeout = (ms) => {
+    return new Promise(resolve => setTimeout(resolve, ms));
+}
+
+const fetchCache = new FetchCache({
+  fetch: nodeFetch,
+  cacheOptions: {
+    // Don't save more than 100 responses in the cache. Allows infinite responses by default
+    maximalItemCount: 10000,
+    // When should the cache evict responses when its full?
+    evictExceedingItemsBy: 'age', // Valid values: 'lru' or 'age'
+    defaultTTL: 6 * 60 * 60 * 1000  // 6 hours
+    // ...see https://github.com/sozialhelden/hamster-cache for all possible options
+  },
+});
+
+
+const fetch = async (url, options = {}, useCache = true) => {
+  const controller = new AbortController();
 
-const fetch = async (url, options = {}) => {
   const newOptions = Object.assign({}, options);
   if (!newOptions["headers"]) {
     newOptions["headers"] = {};
   }
-  newOptions["headers"]["User-Agent"] = USER_AGENT;
-  return nodeFetch(url, newOptions);
+
+  newOptions.signal = controller.signal;
+
+ // newOptions["headers"]["User-Agent"] = USER_AGENT;
+
+  let urlToFetchThroughAPI = Buffer.from(url).toString('base64'); 
+  if (SCRAPER_API_BASE_URL.includes('scraperapi')) { 
+    urlToFetchThroughAPI = url; 
+  }
+
+  const urlAdaptedForScraping = USE_SCRAPER_API
+    ? `${SCRAPER_API_BASE_URL}?api_key=${SCRAPER_API_KEY}&url=${urlToFetchThroughAPI}`
+    : url;
+  const result = useCache ? fetchCache.fetch(urlAdaptedForScraping, newOptions) : nodeFetch(urlAdaptedForScraping, newOptions);
+
+  const timeoutId = setTimeout(() => controller.abort(), NODE_FETCH_TIMEOUT_MS);
+  return result;
 };
 
 module.exports = fetch;

app/helpers/log.js

@@ -0,0 +1,13 @@
+const {
+  PRINT_CRAWLER_DEBUG
+} = require("../config/appConfig");
+
+const logDebug = (...args) => {
+  if (PRINT_CRAWLER_DEBUG) {
+    console.log(...args);
+  }
+}
+
+module.exports = {
+  logDebug
+};

app/helpers/url.js

@@ -7,6 +7,26 @@ const currentSearchRequest = async req => {
 
   return await getSearchRequest(searchRequestId);
 };
-module.exports = {
-  currentSearchRequest
+
+const getUrlParams = function (url) {
+  if (typeof url === 'string' && url.length > 0){
+    const params = {};
+    const questionMarkIndex = url.indexOf('?');
+    if (questionMarkIndex === -1) {
+      return undefined;
+    }
+    const query = url.substring(questionMarkIndex+1);
+    const vars = query.split('&');
+    for (let i = 0; i < vars.length; i++) {
+      const pair = vars[i].split('=');
+      params[pair[0]] = decodeURIComponent(pair[1]);
+    }
+    return params;
+  }
+  return undefined;
+};
+
+module.exports = {
+  currentSearchRequest,
+  getUrlParams
 };

app/migrations/20200215002846-add-column-notified-to-searchrequests.js

@@ -0,0 +1,14 @@
+"use strict";
+
+module.exports = {
+  up: (queryInterface, Sequelize) => {
+    return queryInterface.addColumn("SearchRequests", "notifiedAt", {
+      type: Sequelize.DATE,
+      defaultValue: new Date()
+    });
+  },
+
+  down: (queryInterface, Sequelize) => {
+    return queryInterface.removeColumn("SearchRequests", "notifiedAt");
+  }
+};

app/models/index.js

@@ -16,7 +16,7 @@ config.logging = parseInt(process.env.SEQUELIZE_LOGGING) ? console.log : false;
 
 let sequelize;
 if (config.use_env_variable) {
-  sequelize = new Sequelize(process.env[config.use_env_variable], config);
+  sequelize = new Sequelize(process.env[config.use_env_variable] + "?ssl=true", config);
 } else {
   sequelize = new Sequelize(
     config.database,

app/models/searchRequest.js

@@ -82,7 +82,11 @@ module.exports = (sequelize, DataTypes) => {
     floorMin: DataTypes.INTEGER,
     floorMax: DataTypes.INTEGER,
     accessRoadType: DataTypes.TEXT,
-    heatingType: DataTypes.TEXT
+    heatingType: DataTypes.TEXT,
+    notifiedAt: {
+      type: DataTypes.DATE,
+      defaultValue: new Date()
+    }
   });
 
   return SearchRequest;

app/services/notificationService.js

@@ -15,9 +15,10 @@ const {
 } = require("../helpers/emailContentGenerator");
 const {
   findNotNotifiedMatches,
-  findAllRequestsForCheckUp,
   findRealEstatesForSearchRequest
 } = require("../helpers/db/searchRequestMatch");
+const { findAllRequestsForCheckUp } = require("../helpers/db/searchRequest");
+
 const { sendEmail } = require("../services/emailService");
 
 const notifyForNewRealEstates = async newRealEstates => {
@@ -35,7 +36,7 @@ const notifyForNewSearchRequest = async searchRequest => {
     matchingRealEstates
   );
   const { email } = searchRequest;
-
+  //In case of the new search req, notifiedAt column is populated with default value - now (moment of creation)
   await sendEmail(
     email,
     `${stagingTag} Kivi - novi zahtjev za pretragu`,
@@ -76,6 +77,10 @@ const notifyMatches = async (matches, dailyNotification = false) => {
         sendEmailPromise.catch(err =>
           console.log("[Email Sending Failed]", err)
         );
+
+        //Change time of notified At for searchReq
+        searchRequest.notifiedAt = new Date();
+        searchRequest.save();
       }
     }
   }
@@ -143,8 +148,12 @@ const checkUpNotify = async () => {
     const sendEmailPromise = sendEmail(email, emailSubject, emailContent);
     asyncSendEmailActions.push(sendEmailPromise);
     sendEmailPromise.catch(err => console.log("[Email Sending Failed]", err));
+
+    //Change time of notified At for searchReq
+    searchRequest.notifiedAt = new Date();
+    searchRequest.save();
   }
-  await Promise.all(asyncSendEmailActions); */
+  await Promise.all(asyncSendEmailActions);*/
 };
 
 module.exports = {

development.env

@@ -22,6 +22,11 @@ GA_ID=Google Analytics ID
 #=============== GOOGLE MAPS =============#
 API_MAP_KEY=(your-key-here)
 
+#=============== SCRAPER API SUPORT =============#
+USE_SCRAPER_API= To turn it on (1) or off (0)
+SCRAPER_API_KEY= Key for Scraper api
+SCRAPER_API_BASE_URL= Base url without question mark (example: http://sabur.kivi.ba:1337)
+
 #=============== AWS SDK EMAIL SETTINGS =======#
 AWS_KEY_ID=(your-key-here)
 AWS_SECRET_ACCESS_KEY=(your-key-here)
@@ -67,6 +72,7 @@ AKTIDO_IGNORED_USERNAMES=!!! This is not used for aktido crawler !!!
 AKTIDO_DELAY_BETWEEN_PAGES=time in miliseconds to wait before indexing next page
 AKTIDO_FORCE_CRAWL=Non-zero value will force crawler to crawl all pages without stopping when known real estate is found
 #==SALJIC NEKRETNINE==
+SALJIC_MAX_PAGES=Restrict crawler to this number of pages
 SALJIC_MAX_RESULTS_PER_PAGE=For Saljic crawler, this represents how many ads are crawled at once
 SALJIC_CRAWLER_AD_TYPE=enum name of what type of ads should be crawled, check common/enums.js file for valid values
 SALJIC_CRAWLER_AD_CATEGORIES=comma separated list of enum names of categories to be included, check common/enums.js file for valid values

index.js

@@ -4,6 +4,7 @@ const bodyParser = require("body-parser");
 const layout = require("express-layout");
 const compression = require("compression");
 const forceSSL = require("./app/helpers/forceSSL");
+const { logDebug } = require("./app/helpers/log");
 
 const {
   APP_PORT,
@@ -38,11 +39,17 @@ app.listen(APP_PORT, () =>
 
 let crawlerRunning = STOP_CRAWLER;
 const crawl = () => {
+  logDebug("Crawl start. crawlerRunning: ", crawlerRunning);
   if (!crawlerRunning) {
     crawlerRunning = true;
     crawlAll().then(newRealEstates => {
-      crawlerRunning = false;
+      logDebug("crawlAll done, new real estate len: ", newRealEstates.length);
       notifyForNewRealEstates(newRealEstates);
+    }).catch(e => {
+      console.error('Error happened: ', e);
+    }).finally(()=> {
+      crawlerRunning = false;
+      logDebug('Finally done!');
     });
   }
 };

package-lock.json

@@ -40,6 +40,32 @@
         "@sendgrid/helpers": "^6.3.0"
       }
     },
+    "@sozialhelden/fetch-cache": {
+      "version": "2.0.1",
+      "resolved": "https://registry.npmjs.org/@sozialhelden/fetch-cache/-/fetch-cache-2.0.1.tgz",
+      "integrity": "sha512-vMlsdT5JQCGjx1fcFxmMNh7ZKppjjsfUAeZEhhNwhEL7GaqbZXsD1OXEyx2IcRa25ZuZtvJSV6Q3rE77VRdLvg==",
+      "requires": {
+        "@sozialhelden/hamster-cache": "^1.0.0"
+      }
+    },
+    "@sozialhelden/hamster-cache": {
+      "version": "1.0.0",
+      "resolved": "https://registry.npmjs.org/@sozialhelden/hamster-cache/-/hamster-cache-1.0.0.tgz",
+      "integrity": "sha512-/TEGA8mdMawZp4Yq/GrkL+72YL5EGuSeVXC3pKW12YY1t3C+zCN/HZ0HRp4zWF/e67svXcxuz/B0AEQxEdvi7A=="
+    },
+    "@supercharge/goodies": {
+      "version": "1.4.0",
+      "resolved": "https://registry.npmjs.org/@supercharge/goodies/-/goodies-1.4.0.tgz",
+      "integrity": "sha512-Np6u2qjRwiA3wTgzz4n2yduydIjSXqtJWP5cOnNqjdlCR/EUAK86LAOhEcU+YW211D1ksugns3GqpARJDoXQ7g=="
+    },
+    "@supercharge/promise-pool": {
+      "version": "1.3.0",
+      "resolved": "https://registry.npmjs.org/@supercharge/promise-pool/-/promise-pool-1.3.0.tgz",
+      "integrity": "sha512-9/EVrJevSPEqI4i/gRH8Dt7C+FQT65wRRYuu0MDaGmSLZ2aTel0jOGu8Ae84fPiQ+Ah0B80RPFUxk+K+Cz48DA==",
+      "requires": {
+        "@supercharge/goodies": "~1.4.0"
+      }
+    },
     "@types/caseless": {
       "version": "0.12.2",
       "resolved": "https://registry.npmjs.org/@types/caseless/-/caseless-0.12.2.tgz",
@@ -79,6 +105,14 @@
       "resolved": "https://registry.npmjs.org/abbrev/-/abbrev-1.1.1.tgz",
       "integrity": "sha512-nne9/IiQ/hzIhY6pdDnbBtz7DjPTKrY00P/zvPSm5pOFkl6xuGrGnXn/VtTNNfNtAfZ9/1RtehkszU9qcTii0Q=="
     },
+    "abort-controller": {
+      "version": "3.0.0",
+      "resolved": "https://registry.npmjs.org/abort-controller/-/abort-controller-3.0.0.tgz",
+      "integrity": "sha512-h8lQ8tacZYnR3vNQTgibj+tODHI5/+l06Au2Pcriv/Gmet0eaj4TwWH41sO9wnHDiQsEj19q0drzdWdeAHtweg==",
+      "requires": {
+        "event-target-shim": "^5.0.0"
+      }
+    },
     "accepts": {
       "version": "1.3.5",
       "resolved": "https://registry.npmjs.org/accepts/-/accepts-1.3.5.tgz",
@@ -1074,6 +1108,11 @@
         "es5-ext": "~0.10.14"
       }
     },
+    "event-target-shim": {
+      "version": "5.0.1",
+      "resolved": "https://registry.npmjs.org/event-target-shim/-/event-target-shim-5.0.1.tgz",
+      "integrity": "sha512-i/2XbnSz/uxRCU6+NdVJgKWDTM427+MqYbkQzD321DuCQJUqOuJKIA0IM2+W2xtYHdKOmZ4dR6fExsd4SXL+WQ=="
+    },
     "events": {
       "version": "1.1.1",
       "resolved": "https://registry.npmjs.org/events/-/events-1.1.1.tgz",

package.json

@@ -17,7 +17,8 @@
     "checkup-notify": "cd app/npmScripts && node npmCheckUpNotify.js",
     "test-search": "cd test && node searchTest.js",
     "test-olx-scraper": "cd test && node olxScrapeTest.js",
-    "test-rental-scraper": "cd test && node rentalScrapeTest.js"
+    "test-rental-scraper": "cd test && node rentalScrapeTest.js",
+    "test-saljic-scraper": "cd test && node saljicScrapeTest.js"
   },
   "repository": {
     "type": "git",
@@ -31,6 +32,9 @@
   "dependencies": {
     "2checkout-node": "0.0.1",
     "@sendgrid/mail": "^6.3.1",
+    "@sozialhelden/fetch-cache": "^2.0.1",
+    "@supercharge/promise-pool": "^1.3.0",
+    "abort-controller": "^3.0.0",
     "aws-sdk": "^2.422.0",
     "bluebird": "^3.5.5",
     "cheerio": "^1.0.0-rc.2",

test/olxScrapeTest.js

@@ -9,7 +9,7 @@ if (urlToScrape) {
 
   (async () => {
     const data = await crawler.scrapeAd(urlToScrape);
-    console.log(data);
+    console.log("Scraped data:", data);
   })();
 } else {
   console.log("No URL to scrape. Use like this : ");

test/saljicScrapeTest.js

@@ -0,0 +1,17 @@
+"use strict";
+
+const saljicCrawler = require("../app/crawler/specificCrawlers/saljic");
+
+const urlToScrape = process.argv[2] || undefined;
+
+if (urlToScrape) {
+  const crawler = new saljicCrawler();
+
+  (async () => {
+    const data = await crawler.scrapeAd(urlToScrape);
+    console.log("Scraped data:", data);
+  })();
+} else {
+  console.log("No URL to scrape. Use like this : ");
+  console.log("npm run test-saljic-scraper -- URL_TO_SCRAPE");
+}