help about delimiters

[talun]

Good morning to all.
I'm a AWK beginner and I would like to know if it is possible to define the FS variable in a such way that I'm able to read fields delimited by different delimiters.
For example I need to extract the numeric values from a record set where the values are delimited by spaces, brackets and commas (see below)

Trace Length [m] 1.34E+00 +- 4.85E-01 (5.38E-01, 2.29E+00)

I'm using gawk with Windows XP.

Can anyone help me?

Thanks!

Sergio

 

[feherke]

Hi

[tt]gawk[/tt] ( and as far as I heard [tt]mawk[/tt] too ) accept regular expressions as field delimiter. Theoretically this should work :

[blue]c:\>[/blue] type script.awk
BEGIN {
  FS="[ (),]"
}
{
  for (i=0;i<=NF;i++) print i,$i
}
[blue]c:\>[/blue] echo "Trace Length [m]  1.34E+00 +- 4.85E-01  (5.38E-01, 2.29E+00)" | awk -f script.awk
0 Trace Length [m]  1.34E+00 +- 4.85E-01  (5.38E-01, 2.29E+00)
1 Trace
2 Length
3 [m]
4 
5 1.34E+00
6 +-
7 4.85E-01
8 
9 
10 5.38E-01
11 
12 2.29E+00
13

Feherke.

http://rootshell.be/~feherke/

 

[talun]

Hi feherke, many thanks for your reply.

bye
Sergio

 

[futurelet]

All modern awks accept regular expressions as field delimiter. Mawk and gawk are alone in accepting regexps as record-separators. Not even Perl or Ruby will do that. (Probably because regexps are more powerful in those languages.)