most of the solutions out there call awk from python. But i want to do it the other way round. I have a python script that extracts information from a file. However said filename is referenced in a column of the awk script.
How do i pass python the argument "%s20s", filename and get the input from the standard output? i want to add the output as several columns more.
thanks for your examples
Cheers
The awk variable FILENAME gives the name of the current file being processed (or '-' if stdin). However, this is not available in the BEGIN block, but you can use ARGV[1] instead (assuming you are only passing one filename):
#!/bin/awk -f
BEGIN {
cmd = "./myscript.py '\"%s20s\"' " ARGV[1]
print cmd
cmd | getline var
print var
}
The python script (py3) I used for testing was:
#!/usr/bin/python
import sys
print(sys.argv)
So I get the following output:
/home/user1> runit.awk afile
./myscript.py "%s20s" afile
['./myscript.py', '"%s20s"', 'afile']
You can call external commands using the system function. Does that solve your problem?
$ awk 'BEGIN { system("echo something") }'
something
But that only provides the return code. If you need to capture stdin you can do this:
$ awk 'BEGIN { "echo something" | getline; print "output: "$0 }'
output: something
Getline works line-by-line so if you want multiple lines:
$ awk 'BEGIN { while ("cat multi_line_file" | getline) { print "output: "$0 } }'
output: line 1
output: line 2
output: line 3
