topic Re: regular expression in awk in Operating System - HP-UX

regular expression in awk

David Gartner — Fri, 18 May 2001 00:18:15 GMT

Hi

I'm using awk to search for a particular pattern in a text file. The script is very basic as I'm just trying some things currently.

/^2..abc...ANI$/ {print "some text" }

I save the above in a file called awktest and run it using;

awk -f awktest inputfile

Now the problem I am having occurs while I try and anchor the regex with the $ It doesn't produce what I'm expecting

a sample input line in my file would be

2lmabcXYZANI 12345 890 NO

I just want the field "2lmabcXYZANI" but I get no output.

Re: regular expression in awk

A. Clay Stephenson — Fri, 18 May 2001 01:00:28 GMT

Hi David,
You are operating under a misconception about how awk works. In your example, the input record (line) would have to begin with '2' and the line would have to end with 'ANI'. 'ANI' in the middle of the input record does not match. What you need to do is remove the '$' and thus the input record passes the match.
You then would enter a block to process that line and find where your patern begins and ends and print that section. I would use the the built-in functions index or match to find the positions and substr to extract the desired positions followed by a print or printf. You just need a bit more practice with it.

This should get you started, Clay

Re: regular expression in awk

federico_3 — Fri, 18 May 2001 05:55:57 GMT

try like this:

/^2/ && /ANI$/ {print "some text" }

federico

Re: regular expression in awk

Vincent Stedema — Fri, 18 May 2001 06:26:35 GMT

If you're using a space as the FS:

$1 ~/^2.*ANI$/ {print}
or
$1 ~/^2..abc...ANI$/ {print}

Vincent

Re: regular expression in awk

Stefan Schulz — Fri, 18 May 2001 07:22:45 GMT

If you like to get the first field i would use the following:

awk '/^2..abc...ANI/ {print $1} ' inputfile

this should give you the desired output.

Hope the helps. Stefan

Re: regular expression in awk

Carlos Fernandez Riera — Fri, 18 May 2001 07:45:27 GMT

Use substr function of awk:

awk '

substr ( $1, 1,2 ) == "xxxx" && substr ($1,13,4 ) { print .....}'

See man awk for substr parameters .