160

I need a regular expression that will match any character that is not a letter or a number. Once found I want to replace it with a blank space.

Aryan Beezadhur
  • 4,503
  • 4
  • 21
  • 42
James Jeffery
  • 3,365
  • 4
  • 19
  • 13

10 Answers10

208

To match anything other than letter or number you could try this:

[^a-zA-Z0-9]

And to replace:

var str = 'dfj,dsf7lfsd .sdklfj';
str = str.replace(/[^A-Za-z0-9]/g, ' ');
Darin Dimitrov
  • 1,023,142
  • 271
  • 3,287
  • 2,928
  • 25
    `\w` is for Word characters and is exactly the same as `[a-zA-Z0-9_]` (notice that underscore is considered a word character.) ...so the shorthand would be `str.replace(/[^\w]/g, ' ')` – Joel Mellon Aug 30 '13 at 16:41
  • but it will include many unicode letters, too! is there any way to exclude unicode letters? – Clite Tailor May 04 '17 at 12:06
  • 2
    To include unicode characters, you can use [^\p{L}0-9] – Dave Aug 09 '17 at 21:29
  • 1
    @Dave: As of 2018 you can't without a polyfill, apparently... https://stackoverflow.com/questions/280712/javascript-unicode-regexes – Nickolay Jun 07 '18 at 16:39
  • is there a way, instead of repalce with a space... replace with "\" followed by the character that was been identified? Like this: make this dfj,dsf7lfsd .sdklfj into this dfj\,dsf7lfsd \.sdklfj? – CrazySpy Nov 26 '19 at 13:52
61

This regular expression matches anything that isn't a letter, digit, or an underscore (_) character.

\W

For example in JavaScript:

"(,,@,£,() asdf 345345".replace(/\W/g, ' '); // Output: "          asdf 345345"
Boris Verkhovskiy
  • 14,854
  • 11
  • 100
  • 103
sbmaxx
  • 856
  • 5
  • 6
  • I believe he is looking for /(_|\W)/g, to match anything not a digit or letter (english language) – kennebec Jun 07 '10 at 18:44
  • @sbmaxx I want to replace all except &, (, ) these characters. how could i add this condition in the current regex. – Kunal Pal Feb 21 '19 at 11:24
17

You are looking for:

var yourVar = '1324567890abc§$)%';
yourVar = yourVar.replace(/[^a-zA-Z0-9]/g, ' ');

This replaces all non-alphanumeric characters with a space.

The "g" on the end replaces all occurrences.

Instead of specifying a-z (lowercase) and A-Z (uppercase) you can also use the in-case-sensitive option: /[^a-z0-9]/gi.

Sean Kendle
  • 3,538
  • 1
  • 27
  • 34
favo
  • 5,426
  • 9
  • 42
  • 61
  • Will it match spaces aswell? I need spaces to be kept. Thanks. – James Jeffery Jun 07 '10 at 18:17
  • Space characters would match, but then would be replaced by space characters, so in effect it would leave them alone (a space will stay a space). – jimbo Jun 07 '10 at 18:42
14

This is way way too late, but since there is no accepted answer I'd like to provide what I think is the simplest one: \D - matches all non digit characters.

var x = "123 235-25%";
x.replace(/\D/g, '');

Results in x: "12323525"

See https://developer.mozilla.org/en-US/docs/Web/JavaScript/Guide/Regular_Expressions

tobi-or-not
  • 566
  • 4
  • 13
13
  • Match letters only /[A-Z]/ig
  • Match anything not letters /[^A-Z]/ig
  • Match number only /[0-9]/g or /\d+/g
  • Match anything not number /[^0-9]/g or /\D+/g
  • Match anything not number or letter /[^A-Z0-9]/ig

There are other possible patterns

Youssef AbouEgla
  • 1,445
  • 10
  • 21
5

Just for others to see:

someString.replaceAll("([^\\p{L}\\p{N}])", " ");

will remove any non-letter and non-number unicode characters.

Source

Richie
  • 53
  • 2
  • 4
  • I'm not sure where the two upvotes came from, but in JavaScript (which this question is about) there's no `replaceAll`, and the `\p{..}` property escapes are not widely implemented. – Nickolay Jun 07 '18 at 16:43
  • 1
    @Nickolay There is now! :) – leo May 15 '23 at 09:43
5

try doing str.replace(/[^\w]/); It will replace all the non-alphabets and numbers from your string!

Edit 1: str.replace(/[^\w]/g, ' ')

aayushi
  • 359
  • 4
  • 11
  • 2
    A working answer would be `str.replace(/[^\w]/g, ' ')`. If you don't include `/g` flag it will only replace the first occurence. And if you don't define a replacement string, here a blank space `' '`, it will replace by `undefined` all over the place. Finally, underscores will not be replaced because they match `\w`. This answer is not a perfect fit. – Julien Lirochon Feb 18 '17 at 17:25
4

To match anything other than letter or number or letter with diacritics like é you could try this:

[^\wÀ-úÀ-ÿ]

And to replace:

var str = 'dfj,dsf7é@lfsd .sdklfàj1';
str = str.replace(/[^\wÀ-úÀ-ÿ]/g, '_');

Inspired by the top post with support for diacritics

source

aloisdg
  • 22,270
  • 6
  • 85
  • 105
2

Have you tried str = str.replace(/\W|_/g,''); it will return a string without any character and you can specify if any especial character after the pipe bar | to catch them as well.

var str = "1324567890abc§$)% John Doe #$@'.replace(/\W|_/g, ''); it will return str = 1324567890abcJohnDoe

or look for digits and letters and replace them for empty string (""):

var str = "1324567890abc§$)% John Doe #$@".replace(/\w|_/g, ''); it will return str = '§$)% #$@';

Juan
  • 603
  • 7
  • 15
-1

Working with unicode, best for me:

text.replace(/[^\p{L}\p{N}]+/gu, ' ');
Alexander Goncharov
  • 1,572
  • 17
  • 20