linux find regex

Question

I'm having trouble using the regex of the find command. Probably something I don't understand about escaping on the command line.

Why are these not the same?

find -regex '.*[1234567890]'
find -regex '.*[[:digit:]]'

Bash, Ubuntu

What output do you have to indicate that they are not? – shellter Apr 12 '11 at 13:16 — shellter, Apr 12 '11 at 13:16

score 72 · Answer 1 · answered Apr 12 '11 at 13:15

You should have a look on the -regextype argument of find, see manpage:

      -regextype type
          Changes the regular expression syntax understood by -regex and -iregex 
          tests which occur later on the command line.  Currently-implemented  
          types  are  emacs (this is the default), posix-awk, posix-basic, 
          posix-egrep and posix-extended.

I guess the emacs type doesn't support the [[:digit:]] construct. I tried it with posix-extended and it worked as expected:

find -regextype posix-extended -regex '.*[1234567890]'
find -regextype posix-extended -regex '.*[[:digit:]]'

I thought it might be that, but I checked the emacs regex definitions and they appear to support [:digit:]. http://www.gnu.org/software/emacs/manual/html_node/elisp/Char-Classes.html#Char-Classes — Dijkstra, Apr 12 '11 at 13:19

score 57 · Accepted Answer · answered Apr 12 '11 at 13:21

57

Regular expressions with character classes (e.g. [[:digit:]]) are not supported in the default regular expression syntax used by find. You need to specify a different regex type such as posix-extended in order to use them.

Take a look at GNU Find's Regular Expression documentation which shows you all the regex types and what they support.

answered Apr 12 '11 at 13:21

dogbane

266,786
75
396
414

1

Thanks. That document is the one I should have been looking for. All I could find was http://www.gnu.org/software/emacs/manual/html_node/elisp/Char-Classes.html#Char-Classes - which seemed to suggest that character classes *were* supported in "emacs regex mode". – Dijkstra Apr 12 '11 at 13:24
1

Also check [Regular_expression#Character_classes](http://en.wikipedia.org/wiki/Regular_expression#Character_classes) - very helpful resource – asoundmove Jul 23 '13 at 15:48

score 30 · Answer 3 · edited Nov 01 '15 at 20:23

30

Note that -regex depends on whole path.

 -regex pattern
              File name matches regular expression pattern.  
              This is a match on the whole path, not a search.

You don't actually have to use -regex for what you are doing.

find . -iname "*[0-9]"

edited Nov 01 '15 at 20:23

yegor256

102,010
123
446
597

answered Apr 12 '11 at 13:18

kurumi

25,121
5
44
52

2

The excplicit dot is unneeded here, too ;) – Al.G. Nov 22 '15 at 17:33

score 1 · Answer 4 · answered Apr 12 '11 at 13:16

1

Well, you may try this '.*[0-9]'

answered Apr 12 '11 at 13:16

StKiller

7,631
10
43
56

linux find regex

4 Answers4

Linked