Run a script over multiple files in unix

问题:

Started to learn some shell scripting. I have a perl script that runs on one file. How would I write a shell script to run the perl script a bunch of times on all files with keyword "filename" in it?

So in English,

for /filename/ in filenames  
    perl myscript.pl completefilename

Thanks.

回答1:

find . -name "filename*" -exec perl myscript.pl '{}' \; 

回答2:

for i in $(\ls -d filenames)
do
    perl myscript.pl $i
done

The backslash in front of the 'ls' command is to temporarily disable any aliases.

HTH

回答3:

find . -path "\*filename\*" -exec perl myscript.pl {} \;

edit: escaped stars, didn't want the markup here

回答4:

And if you have spaces in your filenames, use the old standby

find . -maxdepth 1 -print0 | xargs -0 perl myscript.pl

回答5:

In bash:

files=`ls -1 *`
for $file in $files;
do
    perl myscript.pl $file;
done

回答6:

One liner:

$ for file in filename1 filename2 filename3; do perl myscript $file; done

Instead of the space separated list of filenames you can also use wildcards, for instance:

$ for file in *.txt *.csv; do perl myscript $file; done

回答7:

FILES="keyword"

for f in "$FILES" do perl myscript.pl $f done

From http://www.cyberciti.biz/faq/bash-loop-over-file/

回答8:

I personally use the zsh shell, which gives you a very nice way to run a command recursively on a set of subdirectories. It also allows you to change the suffix of the file, which is handly when using lame to create MP3 files of .wav files:

for i in **/*.wav; lame $i $i:r.mp3

You can also pipe the output of one command to another, which I something I use frequently when I'm downloading a number of bittorrent files and want to see the percentage that each download has completed:

for i in **/*.txt; grep -H percent $i | tail -1