[SOLVED] Specifying file for awk to parse.

tombro · 01-17-2018, 10:32 AM

Good Folks,
I have a gawk script that parses a file created in the BEGIN section.
The script works perfectly, but I have to include the dynamically created file on the command line.
An example begin section:

Code:

BEGIN {
  system("ls -1 > /tmp/foo.tmp")
}

If the script name is foo.awk, the command line must be:

Quote:

foo.awk /tmp/foo.tmp

I know I can write a shell script that fires off the above command line, but is there a way to somehow specify the input file within the gawk script itself?

Stéphane Ascoët · 01-18-2018, 02:08 AM

I don't understand what you want to do exactly. The obvious for me would be to get the $1 coming from the command line, but since you're asking I guess I probably miss the whole problem...

fatmac · 01-18-2018, 06:31 AM

I agree, better to use a replacable parameter, as then you can use any input file just by naming it when you run awk.

pan64 · 01-18-2018, 06:45 AM

you do not need a file at all, use:

Code:

awk '
.....
cmd = "ls -l"
while ( ( cmd | getline ) > 0 )
 ....'

but if you want to specify anything, use:

Code:

awk -v var=filename '
......
system("ls -l > " filename)
....
'

smallpond · 01-19-2018, 01:45 PM

What's wrong with not having a file?

Code:

ls -1 |awk ...

firstfire · 01-21-2018, 11:54 PM

Hi.
First of all, I agree with all opinions above and suggest you to consider them carefully.
But, it is indeed possible to specify a file to process from the awk script itself. Here's an example:

Code:

$ awk 'BEGIN{ARGV[1]="/proc/cpuinfo"; ARGC=2} {print}'
processor       : 0
vendor_id       : GenuineIntel
cpu family      : 6
model           : 58
model name      : Intel(R) Core(TM) i7-3517U CPU @ 1.90GHz
stepping        : 9
microcode       : 0x15
cpu MHz         : 1291.218
cache size      : 4096 KB
physical id     : 0
siblings        : 4
core id         : 0
.....

And, of course, one can also read a file manually using getline.