topic Re: script help in Operating System - HP-UX

script help

Thi Vu — Wed, 16 Jan 2002 15:55:03 GMT

Hello everyone,

I have a report that have 10 different kinds of data (i.e: data: CMTP
info about the data ....
....
....

data: MDCA
info about the data....
....
....)
and each data has various info that some can be 1 page or up to 3 pages long. So instead of have over 100 pages of report to see only 2 of the data info that is require - I'm writing a script to parse the info for the 2 types of data that are needed and send it to another file so that the user can print it if he/she wants. I had tried grep (but it only show the line where it matches not the info after that) I'd read that sed can do the trick. Can someone help me with this. Thank you.

Thi

Re: script help

Justo Exposito — Wed, 16 Jan 2002 16:24:38 GMT

Re: script help

Rodney Hills — Wed, 16 Jan 2002 16:45:33 GMT

It depends a lot on what defines the start of the info and the end. If it is like you show, then the following will fetch that block of info.

sed -n -e '/^data: /,/.*)$/p' newfile

-- Rod Hills

Re: script help

David Lodge — Wed, 16 Jan 2002 17:01:54 GMT

you can also use awk to rip between two patterns:

awk '/data: DATA/,/end/ { print $0 }' file.dat

Re: script help

harry d brown jr — Wed, 16 Jan 2002 17:39:53 GMT

#!/usr/bin/perl -d
#
$ReportFile = "/tmp/4";
#
open(REPORTLINES, "<".$ReportFile) || die "can not open $ReportFile";
#
LINE: while(){
$lineofdata = $_;
if (substr($lineofdata,0,5) eq "Data:") {
$reportname = substr($lineofdata,6,length($lineofdata)-6);
$ReportFileName=$ReportFile . $reportname;
close(NEWREPORT);
open(NEWREPORT, ">".$ReportFileName) || die "can not open $ReportFile$repo
rtname";
next LINE;
}
#
printf NEWREPORT $lineofdata;
}
close(REPORTFIXED);

Re: script help

Thi Vu — Wed, 16 Jan 2002 20:34:39 GMT

Thank you for all your help. I can only use one suggestion and need more help, this is what my script look like:

list="CTMT GRPP MDDA"

for file in $list
do
awk '/DATA: $file/,/Report/ {print $0}' >>
done

This will allow me to go through loop and print 3 types of data and append it to a newfile. However, when I ran this sript the variable $file did not get pass to the awk command line at all. The error report from awk is:

awk: There is a regular expression error.
Invalid pattern.
The input line number is 1.
The source line number is 1.

Is there a way to pass a variable to the awk command in a script ? Again thank you for your help.

Thi

Re: script help

Rodney Hills — Wed, 16 Jan 2002 20:49:41 GMT

Passing args to an awk script requires using the -F option, Change your-
awk '/DATA: $file/,/Report/ {print $0}' >>

to

awk -Ffile="DATA: $file" 'file,/Report/ {print $0}' ...

The -F option sets awk variable "file" to "DATA: $file" and then you can use that awk variable in your awk script.

-- Rod Hills

Re: script help

Rodney Hills — Wed, 16 Jan 2002 20:50:36 GMT

Sorry, that is the -V option (not -F).

-- Rod Hills

Re: script help

James R. Ferguson — Wed, 16 Jan 2002 21:08:54 GMT

Hi:

Try this:

# file=/tmp/myfile
# echo awk '$0~DATA: file,/Report/ {print $0}' file=$file >>

Regards!

...JRF...