I want to get the size of all directories within a specific directory. I was thinking something like
find . -type d -exec du -sh {} \;
But that returns all directories recursively. How can I limit the depth?
asked Nov 29, 2012 at 6:59
Add a comment
|
6 Answers
Why use find at all and not simply glob for directories?
du -shc */
answered Nov 29, 2012 at 7:37
-
Well, the find option is more cross-platform, with the same command working on GNU/Linux and BSD/MacOSX systems (provided that
-maxdepth 1is specified before the-type dargument). The OSX version ofduwill give a syntax error withdu -shc */. Nov 29, 2012 at 7:49 -
1Fair point, but do note that the poster used linux as a tag, so I was assuming linux in my answer. Nov 29, 2012 at 7:51
-
This one is fine if you don't have hidden directories to include. Nov 30, 2012 at 9:08
-
Easily adapted to hidden directories:
du -shc */ .??*/(The ?? is to exclude . and ..) Nov 30, 2012 at 10:17 -
This actually IS best on Linux, because it allows you to ignore the current working directory, only listing the subdirectories (what the OP specified). Find includes the
.directory. It also gives you the choice to see the sum total or not. Dec 3, 2012 at 19:38
Add -maxdepth 1 to your find parameters.
This one should do the job efficiently :
du -hc --max-depth=1
One big difference I think of is that, when encountering hardlinked files, they will be counted only once. In a find loop, they will be counted once per base directory. [Is it correct english?]
answered Nov 29, 2012 at 7:40
You can use the -maxdepth option.
I'm using this one,
ls | xargs du -sh
basically there are many ways to skin a cat :)
You can also do this:
du -ah .
What it does is to take the subdirectories of the current directory, find its individual's size and in the end, it prints the total size.