how to get file size 2

[ediforu]

how can i get the filesize of a file from inside the
awk program?

Thanks in advance
edi

 

[vgersh99]

hint: any OS commands are accessible from within awk with "system" call. vlad
+---------------------------+
|#include<disclaimer.h> |
+---------------------------+

 

[ediforu]

yes, i know
but how do i get that value into a variable back in the
awk program?

 

[vgersh99]

&quot;date&quot; | getline curDate; close(&quot;date&quot; vlad
+---------------------------+
|#include<disclaimer.h> |
+---------------------------+

 

[marsd]

requires gawk 3.1+

function dirlist(dir,pat, array1,x) {
if (!pat || !dir) {
perror(&quot;No pattern or directory for file matching.&quot;, 1)
}
command = &quot;ls -l &quot; dir pat &quot; | sort&quot;
while ((command |& getline) > 0) {
array1[$9] = $5 / 1024
}
parray(array1,x)
tlist = array_to_list(array1)
return tlist
}

function array_to_list(arrname, tlist,m) {
for (m in arrname) {
tlist = length(tlist) < 1 ? arrname[m] : tlist&quot; &quot;arrname[m]
}
return tlist
}

function perror(msg,flg) {
printf &quot;%s\n&quot;, msg

if (flg) {
exit
}
return
}


function parray(arrname,len, x) {
while (x <= len) {
print arrname[x++]
}
return
}

BEGIN {
x = 0;
xlist = dirlist(&quot;/home/me/&quot;,&quot;*.txt&quot
if (xlist) {
split(xlist,arr)
for (x in arr) {
print x , arr[x]
}
}
}
{
more stuff
}


It's a lot easier in the shell.

 

[marsd]

Disregard the parray line, artifact of an earlier script
using these functions, sorry.

 

[ediforu]

thanks vlad, that answers part of my question
could be so kind to take a look at this:
I am parsing a log file

Oct 2 05:50:07 /mniprod/xmit/o.coneg.1002020346
Oct 2 05:50:15 /mniprod/xmit/o.coneg.1002020347
Oct 2 05:50:27 /mniprod/xmit/o.coneg.1002020356
Oct 2 05:51:08 /tsiprod/xmit/o.tetra.1002020422

I need to get the file size of the file in the log.

thanks in advance

 

[ediforu]

Got it !
thanks vlad marsd
you are cool

 

[vgersh99]

something like that to start you off - for Solaris:
[no error checking]

nawk -f getSize.awk getSize.txt

#--------------- getSize.txt
Oct 2 05:50:07 /bin/ls
Oct 2 05:50:15 /bin/nawk
Oct 2 05:50:27 /bin/tar
Oct 2 05:51:08 /bin/cut

#--------------- getSize.awk
BEGIN {
lsSIZE=5
filePath=4
CMD=&quot;ls -l&quot;
}

{
cmd=CMD &quot; &quot; $filePath
cmd | getline fileSpec; close(cmd);
split(fileSpec, fileSpecARR, &quot; &quot;
filesARR[$filePath]=fileSpecARR[lsSIZE];

}

END {
for (i in filesARR)
printf(&quot;%s -> [%s]\n&quot;, i, filesARR);
}
vlad
+---------------------------+
|#include<disclaimer.h> |
+---------------------------+

 

[marsd]

Nice Vlad.
Have a good weekend.