javascript split string multiple ways

Question

I am looking for a way to split a sting like "StringBuilder SB = new StringBuilder();" into an array with seperate words, symbols and spaces

var string = "StringBuilder SB = new StringBuilder();";
var array = string.split(...);

output should be like:

array = ["StringBuilder", " ", "SB", " ", "=", " ", "new", " ", "StringBuilder", "(", ")", ";"];

• Mike Cluck: Cool. What have you tried?
• Jake Smith: msdn.microsoft.com/en-us/library/…
• Blue Boy: stackoverflow.com/questions/650022/…
• SteamDev: developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/…
• vp_arth: Just write your own tokenizer. char by char..

Answer 1

Not sure this applies all your needs:

"StringBuilder SB = new StringBuilder();".match(/(\s+|\w+|.)/g);
["StringBuilder", " ", "SB", " ", "=", " ", "new", " ", "StringBuilder", "(", ")", ";"]

• Drew: So this works simply by copy pasting your snippet, but I am curious how "/(\s|\w+|.)/g" this does what it does?
• vp_arth: It just matches tokens by alternation. One of them (| separated) should be matched. and if nothing matched last one . means "every single character". So, we split input by spaces(\s+), words(\w+) and other characters(.).

Answer 2

easy-customizable solution:

function mySplit(str) {
 function isChar(ch) {
 return ch.match(/[a-z]/i);
 }
 var i;
 var result = [];
 var buffer = "";
 var current;
 var onWord = false;
 for (i = 0; i < str.length; ++i) {
 current = str[i];
 if (isChar(current)) {
 buffer += current;
 onWord = true;
 } else {
 if (onWord) {
 result.push(buffer);
 }
 result.push(current);
 onWord = false;
 }
 }
 if(onWord) {
 result.push(buffer);
 }
 return result;
}