AWK write file to different directory

[stevenawk]

Hi,

program:

echo -n "Enter the file to process: "
read inputFile
date
nawk '
/^\[.*\]/ {
id=$3
f=id".txt" # assign f
sub(/{/,"",f) # replace { with nothing, in variable f
sub(/}/,"",f) # replace } with nothing, in variable f
time=$1" "$2 # timestamp is column 1 and 2
print>>f # use append command for writing line
close(f) # close the file
next
}
{
print time" "id" "$0>>f
close(f) # close file
}
' $inputFile
date # display time to see how long it took


Why oh why can i not get it to make the files in another directory (results for e.g.). I have tried every combo of quotes.

Please help,

S.

 

[feherke]

Hi

Why oh why can i not get it to make the files in another directory (results for e.g.).

While you posted none of the tried combination I can only guess what you done.

[gray]# wrong[/gray]

print>>"results/"f

[gray]# correct[/gray]

print>>[red]([/red]"results/"f[red])[/red]

Feherke.

http://rootshell.be/~feherke/

 

[stevenawk]

Hi

thanks again Feherke, i did try your wrong combo, but i have never even heard of brackets around it!

cheers

Steve

 

[Annihilannic]

I don't think it's documented anywhere either, but the >> operator seems to have tighter binding than the implicit string concatenation in awk. I was equally confused when I encountered this one...

Another workaround would be:

file="results/"f
print>>file

Annihilannic.