Tcl variables inside awk (using awk inside Tcl)

Question

I need to use awk command inside a Tcl script. I have a file like the following. Let's call it giri.txt.

1  what
2  Why
3  When
4  who

I wrote the following code, thinking it will print like this:

what did
why did
when did
who did

set a did
exec cat giri.txt | awk {{ print $2 " " $a }}

But it prints like this:

what what
why why
when when
who who

Can anyone explain why it's printing like this? Can any one suggest a way to use a Tcl variable inside awk?

Diego Torres Milano · Accepted Answer · 2013-07-24 05:30:19Z

1

You can probably solve everything tcl without invoking awk, but if you really want to do it:

set a did
exec awk "{print \$2, a}" a=$a < giri.txt

If you have doubts about how the command will be expanded replace exec by echo and check the result

echo awk "{print \$2, a}" a=$a

answered Jul 24, 2013 at 5:30

Diego Torres Milano

69.9k9 gold badges116 silver badges145 bronze badges

Sign up to request clarification or add additional context in comments.

3 Comments

GIRI MURALI Over a year ago

If we use awk inside tcl ,is the same shell awk code is working ? Or is it a tcl version of awk ?

Diego Torres Milano Over a year ago

A new process is created by exec

GIRI MURALI Over a year ago

if i want to print $a and $2 without space how can i do it ?

potrzebie · Accepted Answer · 2013-07-25 00:58:28Z

0

Do it without awk:

set a "did"
set stream [open "giri.txt"]
while {[gets $stream line] >= 0} {
  puts "$line $a"
}
close $stream

Even better with a reusable procedure:

proc forEachLine {file args} {
  set stream [open $file]
  while {[gets $stream line] >= 0} {
    {*}$args $line
  }
  close $stream
}

set a "did"
forEachLine "giri.txt" apply {{line} {
  global a
  puts "$line $a"
}}

answered Jul 25, 2013 at 0:58

potrzebie

1,8381 gold badge13 silver badges28 bronze badges

2 Comments

GIRI MURALI Over a year ago

The first code, you given is just reading the file line by line and printing it along with $a. It is not selecting the column . To use operations similar to awk, we need somthing like set fields [regexp -all -inline {\S+} $line] . And with awk commad we can do this whole thing bu two lines right ? does it have some performace advantage if we right in Tcl its self

potrzebie Over a year ago

@GIRIMURALI Less lines of code doesn't neccessary mean less execution time, you'll have to test performance difference. If performance is very important, consider using C. I can think of lots of other disadvantages with calling awk though: Your script will use two processes instead of one, wich is not good on a system where the number of processes are limited. You'll have a harder time debugging two processes. Anyone that maintains the code need to know awk. There may be different versions of awk on different systems, and sometimes no awk at all.

Collectives™ on Stack Overflow

Tcl variables inside awk (using awk inside Tcl)

2 Answers 2

3 Comments

2 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

3 Comments

2 Comments

Your Answer

Sign up or log in

Post as a guest

Related