test suite

test/test-collector.js

@@ -0,0 +1,131 @@
+var _ = require('underscore');
+var r = require('redis').createClient();
+var collector = require('../collector');
+
+var doc1 = {id:1, text:"I like pie."};
+var doc2 = {id:2, text:"I like potatoes."};
+var doc3 = {id:3, text:"I have an irrational love, yes I do, of flan."};
+
+var testCase = require('nodeunit').testCase;
+var test_db = '_test_tfidf';
+
+module.exports = testCase({
+  setUp: function (callback) {
+      r.select(test_db, function(err, ok) {
+        if (err) callback(err);
+        r.flushdb(function (err, ok) {
+          // index our three documents
+          collector.indexDocument(doc1.id, doc1.text, function() {
+            collector.indexDocument(doc2.id, doc2.text, function() {
+              collector.indexDocument(doc3.id, doc3.text, callback);
+            });
+          });
+        });
+      });
+  },
+  tearDown: function (callback) {
+    // clean up
+    r.flushdb(callback);
+  },
+
+  testN: function (test) {
+    // there should be three documents now
+    test.expect(1);
+    r.get('N', function(err, n) {
+      test.equal(n, 3);
+    });
+    test.done();
+  },
+
+  testTF: function(test) {
+    // check the frequency of some terms
+    test.expect(4);
+
+    // Occurs once in doc1
+    r.zscore('tf:i', 1, function(err, score) {
+      test.equal(score, 1);
+    });
+
+    // Occurs twice in doc3
+    r.zscore('tf:i', 3, function(err, score) {
+      test.equal(score, 2);
+    });
+
+    // A stemmed word
+    r.zscore('tf:irrat', 3, function(err, score) {
+      test.equal(score, 1);
+    });
+
+    // An unstemmed word that should not be in there
+    r.zscore('tf:irrational', 3, function(err, score) {
+      test.equal(score, null);
+    });
+
+    test.done();
+  },
+
+  testDF: function(test) {
+    // check document (overall) frequency for terms
+    test.expect(3);
+
+    // 'i' is in three documents
+    r.smembers('df:i', function(err, members) {
+      test.equal(members.length, 3);
+    });
+
+    // 'pie' is in one document
+    r.smembers('df:pie', function(err, members) {
+      test.equal(members.length, 1);
+    });
+
+    // 'glug' is in no documents
+    r.smembers('df:glug', function(err, members) {
+      test.equal(members.length, 0);
+    });
+
+    test.done();
+
+  },
+
+  testLEN: function(test) { 
+    // we know how many terms there are for each document
+    test.expect(3);
+
+    r.zscore('len', 1, function(err, score) {
+      test.equal(score, 3);
+    });
+
+    r.zscore('len', 2, function(err, score) {
+      test.equal(score, 3);
+    });
+
+    r.zscore('len', 3, function(err, score) {
+      test.equal(score, 10);
+    });
+
+    test.done();
+  },
+
+  testTERMS: function(test) {
+    // we should have a correct list of terms
+    test.expect(1);
+
+    r.smembers('terms', function(err, terms) {
+      // Terms should be:
+      // [ 'an', 'do', 'flan', 'have',
+      //   'i', 'irrat', 'like', 'love',
+      //  'of', 'pie', 'potato', 'ye' ]
+      test.equal(terms.length, 12);
+    });
+
+    test.done();
+  },
+
+  testWEIGHT: function(test) {
+    test.equal('You must write me', 'dude');
+    test.done();
+  }
+
+
+});
+