Skip to content

Commit

Permalink
correctly tokenize strings
Browse files Browse the repository at this point in the history
  • Loading branch information
kbrsh committed May 23, 2017
1 parent 8662611 commit 58d9087
Show file tree
Hide file tree
Showing 3 changed files with 35 additions and 7 deletions.
20 changes: 17 additions & 3 deletions dist/wade.js
Expand Up @@ -77,6 +77,20 @@
}
}

var getWords = function(str) {
var lastIndex = str.length - 1;

if(str[0] === " ") {
str = str.substring(1);
}

if(str[lastIndex] === " ") {
str = str.substring(0, lastIndex);
}

return str.split(" ");
}

var lowercase = function(str) {
return str.toLowerCase();
}
Expand All @@ -86,7 +100,7 @@
}

var removeStopWords = function(str) {
var words = str.split(" ");
var words = getWords(str);

for(var i = 0; i < words.length; i++) {
if(stopWords.indexOf(words[i]) !== -1) {
Expand All @@ -100,7 +114,7 @@
var Wade = function(data) {
var search = function(item) {
var index = search.index;
var keywords = Wade.process(item).split(" ");
var keywords = getWords(Wade.process(item));
var keywordsLength = keywords.length;
var fullwordsLength = keywordsLength - 1;
var scoreIncrement = 1 / keywordsLength;
Expand Down Expand Up @@ -141,7 +155,7 @@
Wade.index = function(data) {
var index = {};
for(var i = 0; i < data.length; i++) {
var str = data[i].split(" ");
var str = getWords(data[i]);
for(var j = 0; j < str.length; j++) {
var item = str[j];
var itemLength = item.length - 1;
Expand Down
2 changes: 1 addition & 1 deletion dist/wade.min.js

Some generated files are not rendered by default. Learn more about how customized files appear on GitHub.

20 changes: 17 additions & 3 deletions src/index.js
Expand Up @@ -67,6 +67,20 @@ var containsPrefix = function(pattern, index, results, resultsLocations, scoreIn
}
}

var getWords = function(str) {
var lastIndex = str.length - 1;

if(str[0] === " ") {
str = str.substring(1);
}

if(str[lastIndex] === " ") {
str = str.substring(0, lastIndex);
}

return str.split(" ");
}

var lowercase = function(str) {
return str.toLowerCase();
}
Expand All @@ -76,7 +90,7 @@ var removePunctuation = function(str) {
}

var removeStopWords = function(str) {
var words = str.split(" ");
var words = getWords(str);

for(var i = 0; i < words.length; i++) {
if(stopWords.indexOf(words[i]) !== -1) {
Expand All @@ -90,7 +104,7 @@ var removeStopWords = function(str) {
var Wade = function(data) {
var search = function(item) {
var index = search.index;
var keywords = Wade.process(item).split(" ");
var keywords = getWords(Wade.process(item));
var keywordsLength = keywords.length;
var fullwordsLength = keywordsLength - 1;
var scoreIncrement = 1 / keywordsLength;
Expand Down Expand Up @@ -131,7 +145,7 @@ Wade.process = function(item) {
Wade.index = function(data) {
var index = {};
for(var i = 0; i < data.length; i++) {
var str = data[i].split(" ");
var str = getWords(data[i]);
for(var j = 0; j < str.length; j++) {
var item = str[j];
var itemLength = item.length - 1;
Expand Down

0 comments on commit 58d9087

Please sign in to comment.