Sanitize string of regex characters before RegExp build?

Question

I want to use a string to perform a global regex, but it might have regex characters in it. What's the best way to escape all regex characters in a string before building a regex with it?

Basically I might have something like this;

var test = 'test.';
var regex = new RegExp(test, 'ig');

I need 'test.' to become 'test\.' so it doesn't behave in unexpected ways.

Similar to a question I asked a few months ago! I never really got an answer though so maybe you will... stackoverflow.com/questions/3614440/… — Tyler
– Tyler, Commented Jun 9, 2011 at 22:55

Nickolay · Accepted Answer · 2023-12-11 21:44:30Z

23

new RegExp(test.replace(/[#-.]|[[-^]|[?|{}]/g, '\\$&'));

This escapes the following:

The characters between # and ., so: #$%&'()*+,-.
The four characters [, \, ], and ^
These characters ?, |, {, }

...by prepending a backslash before each occurrence (as $& Inserts the matched substring, and the g flag means to replace all instances.)

Alternatively, if you're looking for a shorter version, try:

new RegExp(test.replace(/[#-}]/g, '\\$&'));

This will end up escaping a lot more than it should, but it won't harm anything.

edited Dec 11, 2023 at 21:44

Nickolay

32.5k13 gold badges111 silver badges195 bronze badges

answered Jun 9, 2011 at 22:59

user578895

Sign up to request clarification or add additional context in comments.

3 Comments

youurayy Over a year ago

you forgot the /g global flag, to replace all the occurrences (I tried to edit your post, but it tells me that the change must be at least 6 characters)

Christian Bankester Over a year ago

I had some issues with the second one (the regex would not match some things it should for some reason), but the first one works great!

ratijas Over a year ago

Apparently, the short form doesn't work on Qt 5.15, so it might be a compatibility issue.

mVChr · Accepted Answer · 2011-06-09 22:55:18Z

2

This will replace all special RegExp characters with their escaped version:

test = test.replace(/[\\\.\+\*\?\^\$\[\]\(\)\{\}\/\'\#\:\!\=\|]/ig, "\\$&");

answered Jun 9, 2011 at 22:55

mVChr

50.3k11 gold badges111 silver badges105 bronze badges

6 Comments

Tyler Over a year ago

But some of those characters are only special in certain contexts, aren't they? Does that cause a problem at all?

mVChr Over a year ago

In other contexts escaping the character won't harm the RegExp match string, it will just end up being unnecessary.

user578895 Over a year ago

no need to escape #, /, ' or :. You also need - and ,, and you don't need to escape most of those characters in the context you're in. It can simply become /[-\\.,_*+?^$[\](){}!=|]/

Aaron Gillion Over a year ago

Also no need for case insensitive /ig in answer just /g. @MarkKahn why comma and underscore?

user578895 Over a year ago

, is for repetition ranges: {1,3}. Not sure why I added an underscore, it was 6 years ago :D Seems unnecessary

|

Collectives™ on Stack Overflow

Sanitize string of regex characters before RegExp build?

2 Answers 2

3 Comments

6 Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

3 Comments

6 Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related