added maxLength to double metaphone

ajtulloch · Feb 20, 2012 · 95ff89f · 95ff89f
2 parents 456d80f + d6ec04b
commit 95ff89f
Show file tree

Hide file tree

Showing 4 changed files with 48 additions and 16 deletions.
diff --git a/README.md b/README.md
@@ -170,7 +170,8 @@ the raw phonetics are obtained with phonetics()
 
     console.log('phonetics'.phonetics());   
 
-DoubleMetaphone deals with two encodings returned in an array
+DoubleMetaphone deals with two encodings returned in an array. This
+feature is experimental and subject to change.
 
     var natural = require('natural'),
     var dm = natural.DoubleMetaphone;

diff --git a/lib/natural/phonetics/double_metaphone.js b/lib/natural/phonetics/double_metaphone.js
@@ -29,11 +29,18 @@ function isVowel(c) {
 	return c.match(/[aeiouy]/i);
 }
 
-function process(token, spyCallback) {
+function truncate(string, length) {
+    if(string.length >= length)
+        string = string.substring(0, length);
+
+    return string;
+}
+
+function process(token, maxLength) {
 	token = token.toUpperCase();
 	var primary = '', secondary = '';	
     var pos = 0;
-    var spy = {};
+    maxLength == maxLength || 32;
 
     function subMatch(startOffset, stopOffset, terms) {
         return subMatchAbsolute(pos + startOffset, pos + stopOffset, terms);
@@ -325,7 +332,7 @@ function process(token, spyCallback) {
 
             pos += 2;            
         } else if(pos == token.length - 1
-                && subMatch(-2, pos, ['AI', 'OI'])) {
+                && subMatch(-2, 0, ['AI', 'OI'])) {
             addSecondary('', 'S');            
         } else if(token[pos + 1] != 'L' && (
                 token[pos - 1] != 'A' && token[pos - 1] != 'I')) {
@@ -402,11 +409,9 @@ function process(token, spyCallback) {
     var san = (token.substring(0, 3) == 'SAN');
     var startsWithVowel = isVowel(token[0]);
     var slavoGermanic = token.match(/(W|K|CZ|WITZ)/);
-    spy.slavoGermanic = slavoGermanic;
 
     if(subMatch(0, 2, ['GN', 'KN', 'PN', 'WR', 'PS'])) {
     	pos++;
-    	spy.initialSilentConsonantSkipped = true;
     }
 
     while(pos < token.length) {
@@ -478,13 +483,14 @@ function process(token, spyCallback) {
 	        	break;
     	}
 
-    	pos++;
-    }
+        if(primary.length >= maxLength && secondary.length >= maxLength) {
+            break;
+        }
 
-    if(spyCallback)
-    	spyCallback(spy);
+    	pos++;
+    }    
 
-    return [primary, secondary];
+    return [truncate(primary, maxLength), truncate(secondary, maxLength)];
 }
 
 function compare(stringA, stringB) {

diff --git a/package.json b/package.json
@@ -1,7 +1,7 @@
 {
   "name": "natural",
   "description": "General natural language (tokenizing, stemming, classification, inflection, phonetics, tfidf, WordNet) facilities for node.",
-  "version": "0.0.67",
+  "version": "0.0.68",
   "homepage": "https://github.com/NaturalNode/natural",
   "engines": {
     "node": ">=0.4.9"
@@ -17,7 +17,11 @@
   "author": "Chris Umbel <chris@chrisumbel.com>",
   "keywords": ["natural", "language", "porter", "lancaster", "stemmer", "bayes",
     "classifier", "phonetic", "metaphone", "inflector", "wordnet", "tf-idf", 
+<<<<<<< HEAD
     "logistic", "regression", "doublemetaphone"],
+=======
+    "logistic", "regression", "doublemetaphone", "double"],
+>>>>>>> d6ec04b5db5a52da9d3b9fd47b7d3e851cb9f769
   "main": "./lib/natural/index.js",
   "maintainers": [{
       "name": "Chris Umbel",

diff --git a/spec/double_metaphone_spec.js b/spec/double_metaphone_spec.js
@@ -700,12 +700,33 @@ describe('double metaphone', function() {
 	    	expect(encodings[1]).toMatch(/APRPRT/);
 
 			encodings = doubleMetaphone.process('intervention');
-	    	expect(encodings[0]).toMatch(/ANTRFNXN/);
-	    	expect(encodings[1]).toMatch(/ANTRFNXN/);		    	    		
 
-			encodings = doubleMetaphone.process('français');
-			console.log(encodings);
+	    	expect(encodings[0]).toBe('ANTRFNXN');
+	    	expect(encodings[1]).toBe('ANTRFNXN');		    	    		
 
+			encodings = doubleMetaphone.process('Français');
+	    	expect(encodings[0]).toBe('FRNS');
+	    	expect(encodings[1]).toBe('FRNSS');
+    	});
+
+    	it('should truncate codes if specified', function() {
+			var encodings = doubleMetaphone.process('Matrix', 4);
+	    	expect(encodings[0]).toBe('MTRK');
+	    	expect(encodings[1]).toBe('MTRK');
+
+			encodings = doubleMetaphone.process('Français', 4);
+	    	expect(encodings[0]).toBe('FRNS');
+	    	expect(encodings[1]).toBe('FRNS');
+    	});
+
+    	it('should not truncate code is shorter than specification', function() {
+			var encodings = doubleMetaphone.process('Matrix', 32);
+	    	expect(encodings[0]).toBe('MTRKS');
+	    	expect(encodings[1]).toBe('MTRKS');
+
+			encodings = doubleMetaphone.process('Français', 5);
+	    	expect(encodings[0]).toBe('FRNS');
+	    	expect(encodings[1]).toBe('FRNSS');
     	});
 
     	it('should compare', function() {