5 ms·

Linux Find Command Examples

27 points by mlLK 17y ago

philh 17y ago
A mention of xargs might be worth slipping into this. A lot of the time it doesn't really matter whether you use that or -exec, but it's faster on large lists of files. And some commands (grep) act differently if there are multiple files specified as input.
- noss 17y ago
  xargs was my first thought as well. I think I type in this line almost daily (why have i not made an alias already?): find . -type f | xargs grep needle gnu find and xargs have the -print0 and -0 arguments for piping nul terminated filenames, so special chars in file names wont bite you.
  - kree10 17y ago
    I used to use the find | xargs grep idiom a lot too until I installed ack ( http://betterthangrep.com/ http://betterthangrep.com/ ).
  - Harkins 17y ago
    That example is the equivalent of 'grep -r needle .' if you have GNU grep.
antirez 17y ago
I always use it in the form: find . -iname '*.someextension' -exec ls -l {} \; {} gets expanded to the matching file names. Of course ls -l is just an example of command to run against the file name.
jcsalterego 17y ago
Reading this article I realized I always put `find .`, i.e. explicitly specifying the current directory. I guess this is like saying `ls .` all the time, but oh well. Sometimes I find that I need to chain a few actions for my search results that aren't possible with multiple -execs, so I employ the shell script loop: for js in `find . -name '*.js'`; do dir=`dirname $js`; cd $dir && make; java -jar jsmin.jar "$i"; done; Unfortunately if you've got spaces in your filenames, this will throw off /bin/sh and you have to specify delimiters or work around that. Lastly, I've found with find that if you have a n-sized conditional where n > 1 and an -exec, you'll want to use escaped parentheses or otherwise the conditionals won't evaluate properly. Edits: addressing formatting woes.
piramida 17y ago
man find expanded into article? nice.
williebee 17y ago
sudo find / -size +1G -exec ls -lah '{}' \; | awk '{print $8" : "$5}' | tee files_over_1G.txt ### This I use alot.

27 points by mlLK 17y ago

philh 17y ago

A mention of xargs might be worth slipping into this. A lot of the time it doesn't really matter whether you use that or -exec, but it's faster on large lists of files. And some commands (grep) act differently if there are multiple files specified as input.

noss 17y ago
xargs was my first thought as well. I think I type in this line almost daily (why have i not made an alias already?): find . -type f | xargs grep needle gnu find and xargs have the -print0 and -0 arguments for piping nul terminated filenames, so special chars in file names wont bite you.
- kree10 17y ago
  I used to use the find | xargs grep idiom a lot too until I installed ack ( http://betterthangrep.com/ http://betterthangrep.com/ ).
- Harkins 17y ago
  That example is the equivalent of 'grep -r needle .' if you have GNU grep.

antirez 17y ago

I always use it in the form: find . -iname '*.someextension' -exec ls -l {} \; {} gets expanded to the matching file names. Of course ls -l is just an example of command to run against the file name.

jcsalterego 17y ago

Reading this article I realized I always put `find .`, i.e. explicitly specifying the current directory. I guess this is like saying `ls .` all the time, but oh well. Sometimes I find that I need to chain a few actions for my search results that aren't possible with multiple -execs, so I employ the shell script loop: for js in `find . -name '*.js'`; do dir=`dirname $js`; cd $dir && make; java -jar jsmin.jar "$i"; done; Unfortunately if you've got spaces in your filenames, this will throw off /bin/sh and you have to specify delimiters or work around that. Lastly, I've found with find that if you have a n-sized conditional where n > 1 and an -exec, you'll want to use escaped parentheses or otherwise the conditionals won't evaluate properly. Edits: addressing formatting woes.

piramida 17y ago

man find expanded into article? nice.

williebee 17y ago

sudo find / -size +1G -exec ls -lah '{}' \; | awk '{print $8" : "$5}' | tee files_over_1G.txt ### This I use alot.