number splitter

README.md

Splits numbers with commas to make them more readable. Thanks to @maettig for being awesome

annotated.js

function(n){
return                                //return
(""+n)                                //the number converted to a sting
.replace(/\B(?=(...)+$)/g,",")        //with a comma inserted every 3 characters
}

index.js

//simple version:
function(n){return(""+n).replace(/\B(?=(...)+$)/g,",")}
//Does not fuck decimals up version:
function f(a,b,c){return b||(c?a+',':a+'').replace(/(\.\d*)|\d(?=(\d{3})+\b)/g,f)}

package.json

{
  "name": "numSplit",

  "description": "Splits numbers with commas to make them more readable.",

  "keywords": [
    "number",
    "comma",
    "math",
    "format"
  ]
}

test.html

<script>
numSplit = function(n){return(""+n).replace(/\B(?=(...)+$)/g,",")}
    
document.write(numSplit(3255325235523632))​​​​​
</script>

It's hard to explain but it works. Edit: Not enough test cases. "12345" becomes "1234,5" because rule number 3 matches first.

1. Find a dot followed by at least four digits (\.\d{3}\B). Insert a comma after the third digit. Note: \B is a "non-boundary" where a word-character (\w) and another word-character met.
2. Or find a digit that is followed by groups of 3 digits (\d(\d{3})+\b).
3. Continue from either 1. or 2. and insert a comma every three digits. Because of the leading \B this only happens if a digit was matched before.

function(a){return(''+a).replace(/(\.|\B)\d{3}\B|\d(?=(\d{3})+\b)/g,'$&,')}

@maettig
Thats pretty cool :D
But imo it's kinda pointless.
Feel free to fork tho.

Making numSplit(1234.5678) work is pointless?

@maettig
Well, it doesn't really make it work, it adds commas to decimal places... I don't think thats how english works. Then again, I might be wrong about that.

try numSplit(125252.2421415) in Firefox => "1252,52.242,141,5"

Correct would be 125,252.2421415

You are right. I missed that in my tests. What about this?

function(a){return(''+a).replace(/\.\d*|\B(?=(\d{3})+\b)/g,function(a){return a||','})}

Merging the two functions is possible but makes the code very ugly. Edit: OK, here it is. This is smaller and more reliable because it does not insert a comma in "U100".

function f(a,b,c){return b?b:c?a+',':(''+a).replace(/(\.\d*)|\d(?=(\d{3})+\b)/g,f)}

@maettig
Sweet!
1 byte smaller:

function f(a,b,c){return b||(c?a+',':a+'').replace(/(\.\d*)|\d(?=(\d{3})+\b)/g,f)}

Yep, that's cool. This will start recursive calls of replace but it will stop immediately because a contains single digits only. When you do a benchmark you will see it's slower but this can be ignored because of the fast regular expression engines nowadays.

I like it!

What about saving 2 bytes with split?

function(n){return''+(''+n).split(/\B(?=(?:...)+$)/)}

123456789..toLocaleString() seems doing the trick.