adding similarity check (#29)

* adding similarity check

* adding paging

* fixing tests

* docu

* better error handling

* fixing tests

* adjusting page limit

* fixing login screen

* cleanup

* upgrade browser list

* prevent spamming the log

* fixing tests

* removing job listings when removing a job or the user

test/provider/einsAImmobilien.test.js

@@ -1,3 +1,4 @@
+const similarityCache = require('../../lib/services/similarity-check/similarityCache');
 const mockNotification = require('../mocks/mockNotification');
 const providerConfig = require('./testProvider.json');
 const mockStore = require('../mocks/mockStore');
@@ -6,6 +7,10 @@ const expect = require('chai').expect;
 const provider = require('../../lib/provider/einsAImmobilien');
 
 describe('#einsAImmobilien testsuite()', () => {
+  after(() => {
+    similarityCache.stopCacheCleanup();
+  });
+
   provider.init(providerConfig.einsAImmobilien, [], []);
 
   const Fredy = proxyquire('../../lib/FredyRuntime', {
@@ -17,7 +22,7 @@ describe('#einsAImmobilien testsuite()', () => {
 
   it('should test einsAImmobilien provider', async () => {
     return await new Promise((resolve) => {
-      const fredy = new Fredy(provider.config, null, provider.metaInformation.id, 'test1');
+      const fredy = new Fredy(provider.config, null, provider.metaInformation.id, 'test1', similarityCache);
       fredy.execute().then((listings) => {
         expect(listings).to.be.a('array');
 

test/provider/immonet.test.js

@@ -1,3 +1,4 @@
+const similarityCache = require('../../lib/services/similarity-check/similarityCache');
 const mockNotification = require('../mocks/mockNotification');
 const providerConfig = require('./testProvider.json');
 const mockStore = require('../mocks/mockStore');
@@ -6,6 +7,10 @@ const expect = require('chai').expect;
 const provider = require('../../lib/provider/immonet');
 
 describe('#immonet testsuite()', () => {
+  after(() => {
+    similarityCache.stopCacheCleanup();
+  });
+
   provider.init(providerConfig.immonet, [], []);
   const Fredy = proxyquire('../../lib/FredyRuntime', {
     './services/storage/listingsStorage': {
@@ -16,7 +21,7 @@ describe('#immonet testsuite()', () => {
 
   it('should test immonet provider', async () => {
     return await new Promise((resolve) => {
-      const fredy = new Fredy(provider.config, null, provider.metaInformation.id, 'test1');
+      const fredy = new Fredy(provider.config, null, provider.metaInformation.id, 'test1', similarityCache);
       fredy.execute().then((listing) => {
         expect(listing).to.be.a('array');
 

test/provider/immoscout.test.js

@@ -1,3 +1,4 @@
+const similarityCache = require('../../lib/services/similarity-check/similarityCache');
 const mockNotification = require('../mocks/mockNotification');
 const providerConfig = require('./testProvider.json');
 const mockStore = require('../mocks/mockStore');
@@ -7,6 +8,9 @@ const provider = require('../../lib/provider/immoscout');
 const scrapingAnt = require('../../lib/services/scrapingAnt');
 
 describe('#immoscout testsuite()', () => {
+  after(() => {
+    similarityCache.stopCacheCleanup();
+  });
   provider.init(providerConfig.immoscout, [], []);
   const Fredy = proxyquire('../../lib/FredyRuntime', {
     './services/storage/listingsStorage': {
@@ -25,7 +29,7 @@ describe('#immoscout testsuite()', () => {
         return;
       }
 
-      const fredy = new Fredy(provider.config, null, provider.metaInformation.id, 'test1');
+      const fredy = new Fredy(provider.config, null, provider.metaInformation.id, 'test1', similarityCache);
       fredy.execute().then((listing) => {
         expect(listing).to.be.a('array');
 

test/provider/immowelt.test.js

@@ -1,3 +1,4 @@
+const similarityCache = require('../../lib/services/similarity-check/similarityCache');
 const mockNotification = require('../mocks/mockNotification');
 const providerConfig = require('./testProvider.json');
 const mockStore = require('../mocks/mockStore');
@@ -6,6 +7,9 @@ const expect = require('chai').expect;
 const provider = require('../../lib/provider/immowelt');
 
 describe('#immowelt testsuite()', () => {
+  after(() => {
+    similarityCache.stopCacheCleanup();
+  });
   it('should test immowelt provider', async () => {
     provider.init(providerConfig.immowelt, [], []);
     const Fredy = proxyquire('../../lib/FredyRuntime', {
@@ -16,7 +20,7 @@ describe('#immowelt testsuite()', () => {
     });
 
     return await new Promise((resolve) => {
-      const fredy = new Fredy(provider.config, null, provider.metaInformation.id, 'test1');
+      const fredy = new Fredy(provider.config, null, provider.metaInformation.id, 'test1', similarityCache);
       fredy.execute().then((listing) => {
         expect(listing).to.be.a('array');
 

test/provider/kleinanzeigen.test.js

@@ -1,3 +1,4 @@
+const similarityCache = require('../../lib/services/similarity-check/similarityCache');
 const mockNotification = require('../mocks/mockNotification');
 const providerConfig = require('./testProvider.json');
 const mockStore = require('../mocks/mockStore');
@@ -6,6 +7,9 @@ const expect = require('chai').expect;
 const provider = require('../../lib/provider/kleinanzeigen');
 
 describe('#kleinanzeigen testsuite()', () => {
+  after(() => {
+    similarityCache.stopCacheCleanup();
+  });
   it('should test kleinanzeigen provider', async () => {
     provider.init(providerConfig.kleinanzeigen, [], []);
     const Fredy = proxyquire('../../lib/FredyRuntime', {
@@ -16,7 +20,7 @@ describe('#kleinanzeigen testsuite()', () => {
     });
 
     return await new Promise((resolve) => {
-      const fredy = new Fredy(provider.config, null, provider.metaInformation.id, 'test1');
+      const fredy = new Fredy(provider.config, null, provider.metaInformation.id, 'test1', similarityCache);
       fredy.execute().then((listing) => {
         expect(listing).to.be.a('array');
 

test/provider/neubauKompass.test.js

@@ -1,3 +1,4 @@
+const similarityCache = require('../../lib/services/similarity-check/similarityCache');
 const mockNotification = require('../mocks/mockNotification');
 const providerConfig = require('./testProvider.json');
 const mockStore = require('../mocks/mockStore');
@@ -6,6 +7,9 @@ const expect = require('chai').expect;
 const provider = require('../../lib/provider/neubauKompass');
 
 describe('#neubauKompass testsuite()', () => {
+  after(() => {
+    similarityCache.stopCacheCleanup();
+  });
   provider.init(providerConfig.neubauKompass, [], []);
   const Fredy = proxyquire('../../lib/FredyRuntime', {
     './services/storage/listingsStorage': {
@@ -16,7 +20,7 @@ describe('#neubauKompass testsuite()', () => {
 
   it('should test neubauKompass provider', async () => {
     return await new Promise((resolve) => {
-      const fredy = new Fredy(provider.config, null, provider.metaInformation.id, 'test1');
+      const fredy = new Fredy(provider.config, null, provider.metaInformation.id, 'test1', similarityCache);
       fredy.execute().then((listing) => {
         expect(listing).to.be.a('array');
 

test/provider/testProvider.json

@@ -21,7 +21,7 @@
     "enabled": true
   },
   "kleinanzeigen": {
-    "url": "https://www.ebay-kleinanzeigen.de/s-wohnung-kaufen/duesseldorf/anzeige:angebote/preis::420000/wohnung/k0c196l2068r5+wohnung_kaufen.qm_d:90,+wohnung_kaufen.zimmer_d:3.5,",
+    "url": "https://www.ebay-kleinanzeigen.de/s-immobilien/duesseldorf/anzeige:angebote/wohnung/k0c195l2068r5",
     "enabled": true
   },
   "neubauKompass": {

test/provider/wgGesucht.test.js

@@ -1,3 +1,4 @@
+const similarityCache = require('../../lib/services/similarity-check/similarityCache');
 const mockNotification = require('../mocks/mockNotification');
 const providerConfig = require('./testProvider.json');
 const mockStore = require('../mocks/mockStore');
@@ -6,6 +7,9 @@ const expect = require('chai').expect;
 const provider = require('../../lib/provider/wgGesucht');
 
 describe('#wgGesucht testsuite()', () => {
+  after(() => {
+    similarityCache.stopCacheCleanup();
+  });
   provider.init(providerConfig.wgGesucht, [], []);
   const Fredy = proxyquire('../../lib/FredyRuntime', {
     './services/storage/listingsStorage': {
@@ -16,7 +20,7 @@ describe('#wgGesucht testsuite()', () => {
 
   it('should test wgGesucht provider', async () => {
     return await new Promise((resolve) => {
-      const fredy = new Fredy(provider.config, null, provider.metaInformation.id, 'test1');
+      const fredy = new Fredy(provider.config, null, provider.metaInformation.id, 'test1', similarityCache);
       fredy.execute().then((listing) => {
         expect(listing).to.be.a('array');
         const notificationObj = mockNotification.get();

test/similarity/similarity.test.js

@@ -0,0 +1,39 @@
+const SimilarityCacheEntry = require('../../lib/services/similarity-check/SimilarityCacheEntry');
+const expect = require('chai').expect;
+
+describe('similarityCheck', () => {
+  describe('#similarityCheck()', () => {
+    it('should be false', () => {
+      const check = new SimilarityCacheEntry(0);
+      check.setCacheEntry('Hallo');
+      expect(check.hasSimilarEntries('Welt')).to.be.false;
+    });
+    it('should be true', () => {
+      const check = new SimilarityCacheEntry(0);
+      check.setCacheEntry('Hallo');
+      expect(check.hasSimilarEntries('hallo')).to.be.true;
+    });
+    it('should be true', () => {
+      const check = new SimilarityCacheEntry(0);
+      check.setCacheEntry('Selling an incredible house in san francisco');
+      expect(check.hasSimilarEntries('incredible house in san francisco for sale')).to.be.true;
+    });
+    it('should be true', () => {
+      const check = new SimilarityCacheEntry(0);
+      check.setCacheEntry('a');
+      check.setCacheEntry('b');
+      check.setCacheEntry('c');
+      check.setCacheEntry('d');
+      expect(check.hasSimilarEntries('b')).to.be.true;
+    });
+    it('should be false', () => {
+      const check = new SimilarityCacheEntry(0);
+      check.setCacheEntry(
+        'The index is known by several other names, especially Sørensen–Dice index,[3] Sørensen index and Dice\'s coefficient. Other variations include the "similarity coefficient" or "index", such as Dice similarity coefficient (DSC). Common alternate spellings for Sørensen are Sorenson, Soerenson and Sörenson, and all three can also be seen with the –sen ending.'
+      );
+      check.setCacheEntry(
+        'where |X| and |Y| are the cardinalities of the two sets (i.e. the number of elements in each set). The Sørensen index equals twice the number of elements common to both sets divided by the sum of the number of elements in each set.'
+      );
+    });
+  });
+});