FS = "\t" doesn't work with nawk

[Guest_imported]

A script I try to use reads fields separated with \t from a file. The script uses nawk and in the script the field separator is set to FS="\t". However when I run it, this doesn't work, the fields are not separated. I've changed the field separator in the script and the file to * instead and then it works. What could be wrong?

Best regards
-Palle Zetterstrand
Palle.Zetterstrand@emw.ericsson.se

 

[flogrr]

It could be that the output field separator ( OFS )
defaults to a space, so that even though you set the
FS to a tab the output is separated by a space.

Try this: FS = OFS = "\t"

This should retain the tabs in your output the same
as in your input.

Hope this helps you.

 

[Guest_imported]

It was even simpler. The echo command that created the file put \t between the fields instead of tabs. I didn't see this at first. Maybe /usr/ucb/echo was used instead of /usr/bin/echo, when I changed to /usr/bin/echo it worked. As a UNIX newbie I learn stuff everyday... thanks for the help

-Palle Zetterstrand

 

[flogrr]

Yeah, the ucb stuff can sometimes mess things up.

The best way I have found to handle this is to
put this line into your .cshrc as the last line
in the file:

set path = ($path /usr/ucb)

Of course, take any reference to ucb stuff out of
the direct path so this can be appended as the very
last directory searched. This should find the
/usr/bin stuff first without doing without the ucb
stuff should you need any progs included in ucb.

Glad you worked it out.