if statement for file size

[Ratzie]

is there a simple if statement for a file size?

if [ file size -gt 60 ]
than
blah
fi

Doesnot work.

[A. Clay Stephenson · Accepted solution]

No directly: You need to leverage ls -l and awk (as but one method):

FNAME=myfile
typeset -i10 SZ=$(ls -l ${FNAME} | awk '{print $5}')
if [[ ${SZ} -gt 60 ]]
then
blah
blah
fi

The only built-in test is
if [[ -s ${FNAME} ]]
which tests id ${FNAME exists and has a size > 0.

[Hai Nguyen_1]

You got spelling problem. Should have read:

if [ file size -gt 60 ]
then
blah
fi

Hai

[Kent Ostby]

ll | awk '{$5>60)'

Would get you all the files > then 60 bytes.

Best regards,

Kent Ostby

[Dave La Mar]

find file -size +61c >/dev/null
if [ $? -lt 1 ]
then
blah
fi

Best regards,

dl

[RAC_1]

ll|awk '{if($5>=60) print}'

Will print files of 60 bytes and more.

[Tom Smith_9]

Try this

LOGSIZE=$((`du -k $LOG | awk '{print $1}'`))

Assuming $LOG is the name of the file.

Hope it helps.

[Tom Smith_9]

Sorry my last post was missing this,

if [ $(( $LOGSIZE > $LOGSIZELIMIT )) != 0 ] ; then

[Hein van den Heuvel]

In perl there is: -s

Usage example:

perl -e 'while (<*>) { print -s," $_\n" if (-s > 6000)}'

Hein.

[Elmar P. Kolkman]

One thing to bear in mind: most shell commands have problems calculating and perhaps comparing with values larger the 2G.
If your file might be in that range, you can alter Clay's shell solution to divide the size by 1024, for instance, to get the size in kb in your awk part (awk '/file$/{printf "%d\n",$5/1024}).

You could also use things like 'wc -b ', though I wouldn't recommend this for large files (it will literally count bytes in the file!) or 'du -k | cut -f1' to get the size in kb.
ls -s is also a solution, but will get you the size of the file in filesystemblocksize, which might not be what you want.