So I am supposed to find all the files in the given directory with surffix of .C or cc, and change their name to .C.blah or .cc.blah
Is there a command to do that?
Or should I implement this with the command find?
34 Answers
There are a few approaches, depending on what is available on your *nix.
There two different rename commands around, one provided through perl, and one as part of standard utils.
Perl version:
rename 's/(\.cc$|\.C$)/$ *.cc *.C utils version - you might be able to do this with one line:
rename .C .C.blah *.C rename .cc .cc.blah *.cc for loop:
for i in *.C *.cc ; do mv $i $i.blah ; done use a simple for-loop:
for file in *.txt; do echo $file $file.blah; done This command echoes the old file name and the new file to the terminal. If you're getting expected results, go for the real command:
for file in *.txt; do mv $file $file.blah; done Anyway, this won't rename *.txt files in subdirectories. Use find for it.
I know you already found your own acceptable solution, but I want to post what I would normally use at work. This syntax works for GNU findutils, which is part of any modern linux distro.
## xargs is much faster than -exec, xargs can run parallel on # cores with -P # find ./ -type f -name '*.C' -o -name '*.cc' | xargs -I '{}' mv '{}' '{}'.BAK ## If I'm specifying more than two extensions, I would generally use a regex find ./ -type f -regex ".*\.\(C\|cpp\)$" | xargs -I '{}' mv '{}' '{}'.BAK Especially with the now common quad-core SMP servers, you can usually half running time with -p 4. Even with only one core, xargs avoids the extraneous forking find ...-exec would initiate. With a desktop, and unlimited time, it's not really a problem. On the other hand using server resources, and working with many thousands of files, avoiding unnecessary overhead is a priority.
There are two different rename commands.
On Ubuntu/Debian/... distros:
rename -v 's/$/.blah/' *.C *.cc On RedHat/CentOS/... distros:
rename -v $ .blah *.C *.cc 
