Merge pull request #61 from fergiemcdowall/hifiaz-add-id

Adding test and updating README.md

README.md

@@ -49,7 +49,7 @@ const newString = sw.removeStopwords(oldString, [ 'even', 'a', 'custom', 'stopwo
 
 ### <language code>
 
-Arrays of stopwords for the following 31 languages are supplied: 
+Arrays of stopwords for the following 32 languages are supplied: 
 
 * `af` - Afrikaans
 * `ar` - Modern Standard Arabic
@@ -65,6 +65,7 @@ Arrays of stopwords for the following 31 languages are supplied:
 * `ha` - Hausa
 * `he` - Hebrew
 * `hi` - Hindi
+* `id` - Indonesian
 * `it` - Italian
 * `ja` - Japanese
 * `lgg` - Lugbara (without diacritics)

package.json

@@ -1,7 +1,7 @@
 {
   "name": "stopword",
-  "version": "0.2.3",
-  "description": "A module for node.js that takes in text and returns text that is stripped of stopwords. Has pre-defined stopword lists for 31 languages and also takes lists with custom stopwords as input.",
+  "version": "0.2.4",
+  "description": "A module for node.js that takes in text and returns text that is stripped of stopwords. Has pre-defined stopword lists for 32 languages and also takes lists with custom stopwords as input.",
   "main": "lib/stopword.js",
   "scripts": {
     "test": "mocha"

test/test.js

@@ -212,6 +212,12 @@ describe('general stopwordiness:', function () {
     newString.should.eql(['alhoewel', 'oop', 'see', 'sowat', 'uur', 'per', 'motor', 'Bremen', 'af', 'lê', 'getye', 'Weserrivier', 'twee', 'keer', 'per', 'dag', 'duidelik', 'waarneembaar'])
   })
 
+  it('should remove indonesian stopwords', function () {
+    const oldString = 'seorang anak perempuan di Bantul Yogyakarta sedang memasak nasi dengan bantuan tungku'.split(' ')
+    const newString = sw.removeStopwords(oldString, sw.id)
+    newString.should.eql(['anak', 'perempuan', 'Bantul', 'Yogyakarta', 'memasak', 'nasi', 'bantuan', 'tungku'])
+  })
+
 
   // Right to Left languages
   it('should remove arabic stopwords', function () {