RegEx needed to split javascript string on "|" but not "\|"

Question

We would like to split a string on instances of the pipe character |, but not if that character is preceded by an escape character, e.g. \|.

ex we would like to see the following string split into the following components

1|2|3\|4|5

1
2
3\|4
5

I'm expecting to be able to use the following javascript function, split, which takes a regular expression. What regex would I pass to split? We are cross platform and would like to support current and previous versions (1 version back) of IE, FF, and Chrome if possible.

The MDN docs suggest javascript REs support lookahead assertions but not lookbehind assertions. Anyone know if this is accurate? Because a lookbehind assertion is the trivial solution here. — Lily Ballard, Oct 05 '12 at 21:34
@KevinBallard unfortunately MDN is right. See e.g. http://blog.stevenlevithan.com/archives/mimic-lookbehind-javascript — Kijewski, Oct 06 '12 at 02:52

Bart Kiers · Accepted Answer · 2012-10-08T20:16:56.410

9

Instead of a split, do a global match (the same way a lexical analyzer would):

match anything other than \\ or |
or match any escaped char

Something like this:

var str = "1|2|3\\|4|5";
var matches = str.match(/([^\\|]|\\.)+/g);

A quick explanation: ([^\\|]|\\.) matches either any character except '\' and '|' (pattern: [^\\|]) or (pattern: |) it matches any escaped character (pattern: \\.). The + after it tells it to match the previous once or more: the pattern ([^\\|]|\\.) will therefor be matches once or more. The g at the end of the regex literal tells the JavaScript regex engine to match the pattern globally instead of matching it just once.

edited Oct 08 '12 at 20:16

answered Oct 05 '12 at 21:52

Bart Kiers

166,582
36
299
288

I guess that is a good way to do it. The other thing you can do is to find the matches for the pattern `[^\\]([|])` and save everything that is not group(1). The advantage is that the regular expression is a bit easier on the eyes. – Maarten Bodewes Oct 06 '12 at 00:44
could you explain the expression that you've provided above? I'm quite new to regex and related technologies. – MedicineMan Oct 08 '12 at 19:52

score 1 · Answer 2 · answered Oct 05 '12 at 21:44

1

What you're looking for is a "negative look-behind matching regular expression".

This isn't pretty, but it should split the list for you:

var output = input.replace(/(\\)?|/g, function($0,$1){ return $1?$1:$0+'\n';});

This will take your input string and replace all of the '|' characters NOT immediately preceded by a '\' character and replace them with '\n' characters.

answered Oct 05 '12 at 21:44

Andrew Coonce

1,557
11
19

Even shorter: `return $1||$0+'\n'`. But does this solution work? – elclanrs Oct 05 '12 at 21:46
@elclanrs `$1` will be `""` if the pipe was not preceded by a backslash, otherwise it will contain that very character. An empty string evaluates to false. – Kijewski Oct 06 '12 at 02:50

Mamsaac · Answer 3 · 2012-10-05T22:19:10.513

A regex solution was posted as I was looking into this. So I just went ahead and wrote one without it. I did some simple benchmarks and it is -slightly- faster (I expected it to be slower...).

Without using Regex, if I understood what you desire, this should do the job:

function doSplit(input) {
    var output = [];
    var currPos = 0,
        prevPos = -1;
    while ((currPos = input.indexOf('|', currPos + 1)) != -1) {
        if (input[currPos-1] == "\\") continue;
        var recollect = input.substr(prevPos + 1, currPos - prevPos - 1);
        prevPos = currPos;
        output.push(recollect);
    }
    var recollect = input.substr(prevPos + 1);
    output.push(recollect);
    return output;
}
doSplit('1|2|3\\|4|5'); //returns [ '1', '2', '3\\|4', '5' ]

RegEx needed to split javascript string on "|" but not "\|"

3 Answers3

Linked