script to find largest files on system

[jwriter]

1) How can I get a list of say the 20 largest files on my system (hidden, system, whatever)?

2) Same for 20 largest folders?

thank you

[acme.mail.order]

for files: (replace the obvious with the size limit you want)

find / -type f -size +20M

for directories:

du | sort -n | tail -30

[jwriter]

Quote: Originally Posted by acme.mail.order for directories: du | sort -n | tail -30

This is very cool, largest folders listed by size.

Quote: Originally Posted by acme.mail.order for files: (replace the obvious with the size limit you want) find / -type f -size +20M

I am getting this error...

blah:~ G4$ find / -type f -size +200M
find: -size: +200M: illegal trailing character

Forgive my lack of Unix knowledge - am I entering the file size correctly?

[hayne]

Quote: Originally Posted by jwriter blah:~ G4$ find / -type f -size +200M find: -size: +200M: illegal trailing character

It seems to complaining about the "M" - but that should be fine.
Two ideas:

1) Maybe there is some strange invisible character that you inadvertently typed. Maybe just try typing it in again.
If you can reproduce the problem, or if you still have the command you previously entered on your screen, then copy & paste it and then edit the command-line to insert "echo " (without the quotes) at the beginning and add the following at the end (copy & paste it from here into your Terminal):

| hexdump -C

(I.e. the command should look like:

echo find / -type f -size +200M | hexdump -C

Finally, press Return to execute that command, then copy & paste the command and its results back here so we can see.

2) Maybe you have some other (non-Apple-supplied) version of 'find' that doesn't take such arguments.
See what you get from:

command -V find

[jwriter]

Quote: Originally Posted by hayne (I.e. the command should look like: echo find / -type f -size +200M | hexdump -C

What I got was...

blah:~ G4$ echo find / -type f -size +200M  | hexdump -C
00000000  66 69 6e 64 20 2f 20 2d  74 79 70 65 20 66 20 2d  |find / -type f -|
00000010  73 69 7a 65 20 2b 32 30  30 4d 0a                 |size +200M.|
0000001b
blah:~ G4$

I'm wondering about the period after the M.

[jwriter]

Quote: Originally Posted by hayne 2) Maybe you have some other (non-Apple-supplied) version of 'find' that doesn't take such arguments.

It seems to run like this...

blah:~ G4$ find / -type f -size +200000000

except I get a lot of permission denied, such as...

find: /private/var/db/shadow: Permission denied
find: /private/var/db/Spotlight-V100: Permission denied
find: /private/var/db/TokenCache: Permission denied

Should I use sudo?

[hayne]

Quote: Originally Posted by jwriter blah:~ G4$ echo find / -type f -size +200M | hexdump -C 00000000 66 69 6e 64 20 2f 20 2d 74 79 70 65 20 66 20 2d |find / -type f -| 00000010 73 69 7a 65 20 2b 32 30 30 4d 0a |size +200M.| 0000001b blah:~ G4$ I'm wondering about the period after the M.

That period is what 'hexdump' outputs for non-displayable characters.
The character after the M is a newline character. (hexadecimal 0A is \n)
I'm not sure how you got that into the command. Normally a newline character (entered via the Return key) shows up as a new line.

Have you tried just retyping the command-line?

[acme.mail.order]

Quote: Originally Posted by jwriter blah:~ G4$ find / -type f -size +200M find: -size: +200M: illegal trailing character Forgive my lack of Unix knowledge - am I entering the file size correctly?

The command is correct as-is. You've inadvertently copied something else.

There are various GUI utilities to do things like this. If you are looking to clear some space then searching for the biggest single files won't help much (unless you've got tons of media). Removing additional languages clears up nearly a gigabyte and unneeded printers is about the same - both resources are distributed across the filesystem.

[baf]

find differs between tiger and leopard. jwriter if you are still running tiger then +20M is not supported. you must use:
-size n[c]
True if the file's size, rounded up, in 512-byte blocks is n. If n is followed by a c, then the primary is true if the file's size is n bytes (characters).

[tw]

Quote: Originally Posted by hayne The character after the M is a newline character. (hexadecimal 0A is \n) I'm not sure how you got that into the command. Normally a newline character (entered via the Return key) shows up as a new line.

that's a copy/paste error - one of the annoying things about copying out of Safari is that it tends to include any html newlines when you copy, unless you're really careful with your selection. happens to me all the time.

[jwriter]

Quote: Originally Posted by acme.mail.order If you are looking to clear some space then searching for the biggest single files won't help much (unless you've got tons of media). Removing additional languages clears up nearly a gigabyte and unneeded printers is about the same - both resources are distributed across the filesystem.

Thanks to all who helped with this. I understand this is not a good way to find likely candidates for clearing space - I will look at the printers and localization. But this was helpful for the future in case I need to copy commands out of a browser. Thanks again.

[yellow]

You might try OmniDiskSweeper instead..

[ghostdog74]

ls -laR /home | awk '
/total/{next}
/^\.$/{next}
/^\.\.$/{next}
/^\// && f{f=0}
/^\//{        
    f=1
    gsub(/:$/,"")
    old=$0          
    next
}
f && $1 ~/^-/{
   # for files
   $1=$2=$3=$4=$6=$7=$8=""     
   m=split($0,s , " ")
   size=s[1]
   fold = old "/" s[2]
   print size,fold 
}' | sort -rn|head -20

[smninos]

So, why is this available as a search option via Finder if it doesn't actually work? The minute you (I) enter the size of the folder I am searching for all the results that were showing up, disappear? It does work for files, just not folders. Am I missing something?

[Kirok]

Not sure if this is still alive and it seems to answered already but what I use is

BLOCKSIZE=1048576; du -x | sort -nr | head -10