Linux – How to Copy Range of Files in Command Line (ZSH/BASH)

bashcopylinuxzsh

I have a list of files with consecutive numbers as suffixes. I would like to copy only a range of these files. How can I specify a range as part of my cp command.

$ls
P1080272.JPG* P1080273.JPG* P1080274.JPG* P1080275.JPG* P1080276.JPG* P1080277.JPG*
P1080278.JPG* P1080279.JPG* P1080280.JPG* P1080281.JPG* P1080282.JPG* P1080283.JPG*

I would like to copy files from P1080275.JPG to P1080283.JPG with something similar to:

$cp P10802[75-83].JPG ~/Images/.

Is there a way to do this?

Best Answer

You were very close. Your question was almost the correct syntax:

cp P10802{75..83}.JPG ~/Images

Related Solutions

Bash – Unique Features of bash compared to zsh

zsh is for vulcans. ;-)

Seriously: bash 4.0 has some features previously only found in zsh, like ** globbing:

% ls /usr/src/**/Makefile

is equivalent to:

% find /usr/src -name "Makefile"

but obviously more powerful.

In my experience bash's programmable completion performs a bit better than zsh's, at least for some cases (completing debian packages for aptitude for example).

bash has Alt + . to insert !$

zsh has expansion of all variables, so you can use e.g.



% rm !$<Tab>


for this. zsh can also expand a command in backtics, so

% cat `echo blubb | sed 's/u/a/'`<Tab>


yields

% cat blabb


I find it very useful to expand rm *, as you can see what would be removed and can maybe remove one or two files from the commmand to prevent them from being deleted.

Also nice: using the output from commands for other commands that do not read from stdin but expect a filename:

% diff <(sort foo) <(sort bar)


From what I read bash-completion also supports completing remote filenames over ssh if you use ssh-agent, which used to be a good reason to switch to zsh.

Aliases in zsh can be defined to work on the whole line instead of just at the beginning:

% alias -g ...="../.."
% cd ...

Bash – Create ‘Virtual File’ from Command Output

This is the cleanest way to do what you want:

mutt admin@example.org -a <(mysqldump mysqldumpoptions)

The <() operator is what you were asking for; it creates a FIFO (or /dev/fd) and forks a process and connects stdout to the FIFO. >() does the same, except connects stdin to the FIFO instead. In other words, it does all the mknod stuff for you behind the scenes; or on a modern OS, does it in an even better way.

Except, of course, that doesn't work with mutt, it says:

/dev/fd/63: unable to attach file.

I suspect the problem is that mutt is trying to seek in the file, which you can't do on a pipe of any sort. The seeking is probably something like scanning the file to figure out what MIME type it is and what encodings might work (ie, whether the file is 7bit or 8bit), and then seeking to the beginning of the file to actually encode it into the message.

If what you want to send is plain text, you could always do something like this to make it the main contents of the email instead (not ideal, but it actually works):

mysqldump mysqldumpoptions | mutt -s "Here's that mysqldump" admin@example.org

Best Answer

Related Solutions

Bash – Unique Features of bash compared to zsh

Bash – Create ‘Virtual File’ from Command Output

Related Topic