How to match case insensitive patterns with ls?

Question

I would like to list all files matching a certain pattern while ignoring the case.

For example, I run the following commands:

ls *abc*

I want to see all the files that have "abc" as a part of the file name, ignoring the case, like

-rw-r--r-- 1 mtk mtk 0 Sep 21 08:12 file1abc.txt
-rw-r--r-- 1 mtk mtk 0 Sep 21 08:12 file2ABC.txt

Note

I have searched the man page for case, but couldn't find anything.

Answer 1

This is actually done by your shell, not by ls.

In bash, you'd use:

shopt -s nocaseglob

and then run your command.

Or in zsh:

unsetopt CASE_GLOB

Or in yash:

set +o case-glob

and then your command.

You might want to put that into .bashrc, .zshrc or .yashrc, respectively.

Alternatively, with zsh:

setopt extendedglob
ls -d -- (#i)*abc*

(that is turn case insensitive globbing on a per-wildcard basis)

With ksh93:

ls -d -- ~(i:*abc*)

You want globbing to work different, not ls, as those are all files passed to ls by the shell.

Answer 2

As explained by polemon, it is the shell (not ls) that extends *abc* to a list of files. This is called Pattern Matching.

Aside from changing the whole Pattern Matching behavior to ignore case, you could use another form of pattern matching than the *. The following would do what you want in bash:

ls *[aA][bB][cC]*

From bash man:

[...] Matches any one of the enclosed characters.

This allows more fine grain matching where you could use *[aA][bB]c* to match abc or ABc but not abC or ABC. Or an example in French, where I could want to match all instances of the e character:

ls *[eéèêëEÉÈÊË]*

Answer 3

You can also add -i (--ignore-case) option to grep to get and the below output.

[root@localhost ~]# ls -l | grep -i abc
-rw-r--r--  1 root root    0 Feb 25 20:41 fileabc.txt
-rw-r--r--  1 root root    0 Feb 25 20:41 fileABC.txt

Answer 4

I don't think this is possible with ls, but you could use find instead:

find . -maxdepth 1 -iname '*abc*'

Answer 5

use the following code:

LC_COLLATE=en_US ; export LC_COLLATE
ls *abc*
LC_COLLATE=C ; export LC_COLLATE

Answer 6

You can GREP the file using -i which adds the insensitive option.

For example to find both abc and ABC you can run this command

    [root@mtk bin]# ls -l | grep -i abc
    640K -rw-r--r-- 1 mtk mtk 639K Sep 21 08:12 file1abc.txt
    676K -rw-r--r-- 1 mtk mtk 674K Sep 21 08:12 file2ABC.txt

Answer 7

In tcsh, set an alias in .aliases

alias lsnc "ls BS!* | sort -f "

("BS" = one "backslash"; I had to use this to get it to show up in my browser.)

and use lsnc instead of ls

In bash, try an alias in .bash_aliases like

lsnc() { ls $1 | sort -f ; }

With options to ls, e.g., ls -l, here is a simple fix:

tcsh:

alias lsncl "ls -l BS!* | sort -f -k 9"

bash:

lsncl() { ls -l $1 | sort -f -k 9 ; }