Can I use shell wildcards to select filenames ranging across double-digit numbers (e.g., from foo_1.jpg to foo_54.jpg)?

Question

I have a directory with image files foo_0.jpg to foo_99.jpg. I would like to copy files foo_0.jpg through foo_54.jpg.

Is this possible just using bash wildcards?

I am thinking something like cp foo_[0-54].jpg but I know this selects 0-5 and 4 (right?)

Also, if it is not possible (or efficient) with just wildcards what would be a better way to do this?

Thank you.

Rabah

Answer 1

I assume you want to copy these files to another directory:

cp -t target_directory foo_{0..54}.jpg

Conventional

+1: Like the use of the {0..54}. That's much cleaner than the globbing – Duelist

You are amazing! Would you happen to know where in the bash manual that is described? I can't find it but I now remember seeing it somewhere. Nevermind. Found it under Brace Expansion. – Tipi

Apparently this brace expansion sequence expression is used for file name matching but not for regex. – Tipi

Answer 2

I like glenn jackman answer, but if you really want to use globbing, following might also work for you:

$ shopt -s extglob
$ cp foo_+([0-9]).jpg $targetDir

In extended globbing +() matches one or more instances of whatever expression is in the parentheses.

Now, this will copy ALL files that are named foo_ followed by any number, followed by .jpg. This will include foo_55.jpg, foo_139.jpg, and foo_1223218213123981237987.jpg.

On second thought, glenn jackman has the better answer. But, it did give me a chance to talk about extended globbing.

Duelist

Answer 3

ls foo_[0-9].jpg foo_[1-4][0-9].jpg foo_5[0-4].jpg

Try it with ls and if that looks good to you then do the copy.

Tipi

@DigitalRoss: That's good except it will match files like foo_a.jpg if they exist. Perhaps replacing your ?'s with [0-9]'s is the best solution. – Tipi

Answer 4

for i in `seq 0 54`; do cp foo_$i.jpg <target>; done

Stercoraceous

Answer 5

An extension answer of @grok12's answer above

$ ls foo_([0-9]|[0-4][0-9]|5[0-4]).jpg

Basically the regex above will match below

• anything with a single digit OR
• two digits and that first digit must be between 0-4 and second digit between 0-9 OR
• two digits and that first digit is 5 and second digit between 0-9

Alternatively you can achieve similar result with regex below

$ ls file{[0-9],[0-4][0-9],5[0-4]}.txt

Tog