Multiple New Functionalities

AFINN-111.json

@@ -506,7 +506,7 @@
   "courageous": 2,
   "courteous": 2,
   "courtesy": 2,
-  "cover-up": -3,
+  "cover up": -3,
   "coward": -2,
   "cowardly": -2,
   "coziness": 2,
@@ -761,8 +761,9 @@
   "dodging": -2,
   "dodgy": -2,
   "does not work": -3,
+  "doesn't work": -3,
   "dolorous": -2,
-  "dont like": -2,
+  "don't like": -2,
   "doom": -2,
   "doomed": -2,
   "doubt": -1,
@@ -1487,7 +1488,7 @@
   "lackadaisical": -2,
   "lagged": -2,
   "lack": -2,
-  "made-up": -1,
+  "made up": -1,
   "madly": -3,
   "madness": -3,
   "mandatory": -1,
@@ -1587,7 +1588,7 @@
   "merry": 3,
   "naive": -2,
   "natural": 1,
-  "na�ve": -2,
+  "naïve": -2,
   "needy": -2,
   "negative": -2,
   "negativity": -2,
@@ -1634,7 +1635,7 @@
   "offline": -1,
   "oks": 2,
   "ominous": 3,
-  "once-in-a-lifetime": 3,
+  "once in a lifetime": 3,
   "opportunities": 2,
   "opportunity": 2,
   "oppressed": -2,
@@ -1987,8 +1988,8 @@
   "sedition": -2,
   "seditious": -2,
   "seduced": -1,
-  "self-confident": 2,
-  "self-deluded": -2,
+  "self confident": 2,
+  "self deluded": -2,
   "selfish": -3,
   "selfishness": -3,
   "sentence": -2,
@@ -2014,8 +2015,8 @@
   "shocking": -2,
   "shocks": -2,
   "shoot": -1,
-  "short-sighted": -2,
-  "short-sightedness": -2,
+  "short sighted": -2,
+  "short sightedness": -2,
   "shortage": -2,
   "shortages": -2,
   "shrew": -4,
@@ -2074,7 +2075,7 @@
   "solving": 1,
   "somber": -2,
   "some kind": 0,
-  "son-of-a-bitch": -5,
+  "son of a bitch": -5,
   "soothe": 3,
   "soothed": 3,
   "soothing": 3,
@@ -2477,4 +2478,4 @@
   "zealot": -2,
   "zealots": -2,
   "zealous": 2
-}
+}

README.md

@@ -10,18 +10,43 @@
 
 ## Usage
 ```javascript
-var sentimentAnalysis = require('sentiment-analysis');
-
-sentimentAnalysis('Dinosaurs are awesome!'); // +0.4
-sentimentAnalysis('Everything is stupid');  // -0.2
-sentimentAnalysis('Windows is very unstable');  // -0.2
-sentimentAnalysis('London is gloomy today because of all the smog');  // -0.4
-sentimentAnalysis('I am so grateful for all the presents, thank you!');  // +0.5
-sentimentAnalysis('Really enjoying the warm weather');  // +0.3
-sentimentAnalysis('It was a catastrophic disaster');  // -0.6
+var sentimentAnalysisLib = require('sentiment-analysis');
+var sentimentAnalysis = new sentimentAnalysisLib();
+
+sentimentAnalysis.analyseSentence('Dinosaurs are awesome!'); // +0.4
+sentimentAnalysis.analyseSentence('Everything is stupid');  // -0.2
+sentimentAnalysis.analyseSentence('Windows is very unstable');  // -0.2
+sentimentAnalysis.analyseSentence('London is gloomy today because of all the smog');  // -0.4
+sentimentAnalysis.analyseSentence('I am so grateful for all the presents, thank you!');  // +0.5
+sentimentAnalysis.analyseSentence('Really enjoying the warm weather');  // +0.3
+sentimentAnalysis.analyseSentence('It was a catastrophic disaster');  // -0.6
 ```
 sentiment-analysis will return a score between -1 and +1, where negative numbers represent a negative overall sentiment.
 
+### Options
+
+*Custom file*
+```javascript
+var sentimentAnalysisLib = require('sentiment-analysis');
+var sentimentAnalysis = new sentimentAnalysisLib({
+	customWordsFile : 'myCustomAfinn.json'
+});
+
+sentimentAnalysis.analyseSentence('Dinosaurs are awesome!');
+```
+
+*Custom words*
+```javascript
+var sentimentAnalysisLib = require('sentiment-analysis');
+var sentimentAnalysis = new sentimentAnalysisLib({
+	customWords : {
+		dinosaurs : -4
+	}
+});
+
+sentimentAnalysis.analyseSentence('Dinosaurs are awesome!'); // 0
+```
+
 
 ## Testing
 `npm test`
@@ -32,7 +57,7 @@ See unit test, integration testing results on [Travis CI]
 See the `gulpfile.js` for documentation of build process.
 
 ## License
-MIT � [Alicia Sykes](http://aliciasykes.com)
+MIT � [Alicia Sykes](http://aliciasykes.com), [Maurizio Carboni](https://www.linkedin.com/in/mauriziocarboni)
 
 [AFINN-111]: <http://www2.imm.dtu.dk/pubdb/views/publication_details.php?id=6010>
-[Travis CI]: <https://travis-ci.org/Lissy93/sentiment-analysis>
+[Travis CI]: <https://travis-ci.org/Lissy93/sentiment-analysis>

gulpfile.js

@@ -18,7 +18,8 @@ var watch   = require('gulp-watch');
 require('coffee-script/register');
 
 var footerTxt = "\/* (C) Alicia Sykes <[email protected]> 2015           " +
-    "*\\\r\n\\* MIT License. Read full license at: https:\/\/goo.gl\/IL4lQJ *\/";
+    "*\\\r\n|* (C) Maurizio Carboni <[email protected]> 2016          " +
+    "*|\r\n\\* MIT License. Read full license at: https:\/\/goo.gl\/IL4lQJ *\/";
 
 /* Delete the files currently in finished directory */
 gulp.task('clean', function () {
@@ -37,7 +38,7 @@ gulp.task('build', ['clean'],  function(){
 });
 
 /* Run unit tests and generate coverage report */
-gulp.task('test', function (cb) {
+gulp.task('test', ['build'], function (cb) {
     gulp.src(['./index.js'])
         .pipe(istanbul())
         .pipe(istanbul.hookRequire())
@@ -62,4 +63,4 @@ gulp.task('test-after-build',['build'],function(){
 });
 
 /* Defualt gulp task, deletes old files, compiles source files and runs tests */
-gulp.task('default', ['test-after-build', 'watch']);
+gulp.task('default', ['test-after-build', 'watch']);

index.coffee

@@ -1,58 +1,86 @@
+sentimentAnalysisPath = __dirname
 
-afinnWordList = require __dirname + '/AFINN-111.json' # Get the AFINN-111 list
-
-# Returns a boolean true if given word is found in word list
-doesWordExist = (word)->
-  if word of afinnWordList then true else false
-
-# Returns an integer value + or - sentiment score for given word
-getScoreOfWord = (word)->
-  if afinnWordList[word] then afinnWordList[word] else 0
-
-# Formats sentence and returns a lowercase a-z array of words
-getWordsInSentence = (sentence)->
-  sentence = if sentence? then sentence else '' # Double check is defined
-  sentence = if typeof sentence == 'string' then sentence else ''
-  sentence = sentence.toLowerCase()
-  sentence = sentence.replace(/(?:https?|ftp):\/\/[\n\S]+/g, '') # Remove URLs
-  sentence = sentence.replace(/[^\w\s]/gi, '')  # Remove special characters
-  sentence = sentence.split(' ')                # Split into an array
-  sentence = sentence.filter((n) -> n != '')    # Remove blanks
-  sentence = removeDuplicates(sentence)
-
-# Remove Duplicates
-removeDuplicates = (arr) ->
-  if arr.length == 0
-    return []
-  res = {}
-  res[arr[key]] = arr[key] for key in [0..arr.length-1]
-  value for key, value of res
-
-# Ensure score is in a valid range between -1 to +1
-scaleScore = (score)->
-  score = if score > 10 then 10 else score
-  score = if score < -10 then -10 else score
-  score/10
-
-# Returns an overall sentiment score for sentence
-analyseSentence = (sentence) ->
-  score = 0
-  wordsArr = getWordsInSentence(sentence)
-  for word in wordsArr
-    if doesWordExist(word)
-      score += getScoreOfWord(word)
-  scaleScore(score)
-
-module.exports = analyseSentence # Export main method as module
-
-
-# If we're developing/ testing then export the private methods too
-if process.env.NODE_ENV == 'test'
-  module.exports =
-    main: analyseSentence
-    _private:
-      scaleScore: scaleScore
-      doesWordExist: doesWordExist
-      getScoreOfWord: getScoreOfWord
-      removeDuplicates: removeDuplicates
-      getWordsInSentence: getWordsInSentence
+module.exports = class Analizer
+  constructor: ( opts ) ->
+    # Include inside the class the wordlist
+    @afinnWordList = require(sentimentAnalysisPath + '/AFINN-111.json')
+    if opts
+      # If using a custom file, load it
+      if opts.customWordsFile
+        opts.customWords = require(opts.customWordsFile)
+      # if using custom words
+      # ( in case of custom file, this words are loaded from the file )
+      if opts.customWords
+        # Overwrite existing words
+        for i,word of opts.customWords
+          @afinnWordList[i] = word
+    @afinnPhrases = []
+    @afinnPhrasesCamel = []
+    for word,value of @afinnWordList
+      wordA = @constructor.transformPlainApostrophe(word)
+      if ( wordA != word )
+        @afinnWordList[wordA] = value
+        delete @afinnWordList[word]
+        word = wordA
+      if @constructor.isPhrase(word)
+        compressedWord = @constructor.compressPhrase(word)
+        @afinnPhrases.push word
+        @afinnPhrasesCamel.push compressedWord
+        @afinnWordList[compressedWord] = value
+
+  # Returns an overall sentiment score for sentence
+  analyseSentence: ( sentence ) ->
+    score = 0
+    wordsArr = @getWordsInSentence(sentence)
+    for word in wordsArr
+      if @doesWordExist(word)
+        score += @getScoreOfWord(word)
+    @constructor.scaleScore(score)
+
+  # Returns a boolean true if given word is found in word list
+  doesWordExist: ( word ) ->
+    word of @afinnWordList
+
+  # Returns an integer value + or - sentiment score for given word
+  getScoreOfWord: ( word ) ->
+    @afinnWordList[word] || 0
+
+
+  # Formats sentence and returns a lowercase a-z array of words
+  getWordsInSentence: ( sentence ) ->
+    sentence = sentence || '' # Double check is defined
+    sentence = if typeof sentence == 'string'then sentence.toLowerCase() else ''
+    sentence = sentence.replace(/(?:https?|ftp):\/\/[\n\S]+/g, '') # Remove URLs
+    sentence = sentence.replace(/[\n\r\t]/gi, ' ')  # Transform \n and \t
+    sentence = @constructor.transformPlainApostrophe(sentence) # Transform '
+    sentence = sentence.replace(/[^\w\s']/gi, '')  # Remove special characters
+    # Replace phrases with the camelized version
+    for word,i in @afinnPhrases
+      sentence = sentence.replace( new RegExp(word,'g'), @afinnPhrasesCamel[i])
+    sentence = sentence.split(' ')                # Split into an array
+    sentence = sentence.filter((n) -> n != '')    # Remove blanks
+    @constructor.removeDuplicates(sentence)
+
+  # Remove Duplicates
+  @removeDuplicates: ( arr ) ->
+    if arr.length == 0
+      return []
+    res = {}
+    res[arr[key]] = arr[key] for key in [0..arr.length-1]
+    value for key, value of res
+
+  # Ensure score is in a valid range between -1 to +1
+  @scaleScore: ( score ) ->
+    score = if score > 10 then 10 else score
+    score = if score < -10 then -10 else score
+    score/10
+
+  @isPhrase: ( word ) ->
+    word.indexOf(' ') != -1
+
+  @compressPhrase: ( phrase ) ->
+    (w[0].toUpperCase() + w[1..-1].toLowerCase() for w in phrase.split /\s+/)
+    .join ''
+
+  @transformPlainApostrophe: ( phrase ) ->
+    phrase.replace(/[`"]/gi, '\'')