topic Re: Question about script in Operating System - HP-UX

Question about script

Anthony_69 — Thu, 26 Jun 2003 12:38:38 GMT

Hi....wondering if you could help....
i have a file with block of approx 20 lines similar to the below, repeated 100's of times througout the file:

type: 114
id: 3
version: 3
date: 2003-06-25 10:30:55
node: 0
application: 0
length: 75345
class: 234
service: 0
content: 0
report: 0
timer: 0
count time: 0
input field: 0

The problem i that I need to output the line listed "class" and the line listed "date" together to a seperate file. I need this to be repeated from each block of 20 lines throughout the file. Any ideas?
Thanks

Re: Question about script

Kenneth_19 — Thu, 26 Jun 2003 12:41:39 GMT

Try:

grep -E 'class|date' input_file > output_file

Kenneth

Re: Question about script

Bill Douglass — Thu, 26 Jun 2003 12:42:19 GMT

If you're putting this pair of lines from each block into the same file, then

grep -e ^date: -e ^class: > /some/file

should do the trick.

Re: Question about script

Bill Douglass — Thu, 26 Jun 2003 12:43:41 GMT

Sorry, typo in my replyt:

grep -e ^date: -e ^class: /path/to/file > /some/file

Re: Question about script

Steve Steel — Thu, 26 Jun 2003 12:43:44 GMT

Hi

grep -e class -e date file > file.out

Steve Steel

Re: Question about script

Dario_1 — Thu, 26 Jun 2003 12:56:39 GMT

Hi!

Use the grep command to do what you want. See from the man page:

grep, egrep, fgrep - search a file for a pattern

SYNOPSIS
Plain call with pattern
grep [-E|-F] [-c|-l|-q] [-bhinsvwx] pattern [file ...]

Call with (multiple) -e pattern
grep [-E|-F] [-c|-l|-q] [-bhinsvwx] -e pattern... [-e pattern] ...
[file ...]

Look at the line with the word multiple for details and re-direct to a new file like every other person mentioned (grep -e class -e date file > newfile) Check the man pages for more information.

Regards,

Dario

Re: Question about script

Anthony_69 — Thu, 26 Jun 2003 12:58:21 GMT

Thanks guys.
one final thing....
the resultant output is something like this..
.
.
.
.
class: 1234
date: 2003-6-26@08:01am
class: 0000
date: 2003-6-26@08:02am
class: 9999
date: 2003-6-26@08:03am
class: 0000
date: 2003-6-26@08:04am
class: 1234
date: 2003-6-26@08:05am
class: 0000
date: 2003-6-26@08:06am
.
.

From this file, I need to be able to list only the lines with class 0000 and their corresponding date

Re: Question about script

Jean-Louis Phelix — Thu, 26 Jun 2003 13:02:28 GMT

hi,

awk '/class: 0000/{print ; getline ; print}' file

Regards.

Re: Question about script

Jean-Louis Phelix — Thu, 26 Jun 2003 13:05:14 GMT

Or ...

sed -n '/class: 0000/ {
N
p
}' file

Regards.

Re: Question about script

Hai Nguyen_1 — Thu, 26 Jun 2003 13:06:34 GMT

Anthony,

Assume that you have the output in a file named output.txt. Below is a script to do what you want:

#!/bin/sh

while read CLASS
do
read DATE
if [ "$CLASS" = "class: 0000" ]
then
echo $CLASS
echo $DATE
fi
done < output.txt

Hai

Re: Question about script

James R. Ferguson — Thu, 26 Jun 2003 13:08:02 GMT

Hi:

With regard to your last question, using your file as presented:

# awk '/^class: 0000/ {print $0;getline;print $0}' inputfile > outputfile

Regards!

...JRF...

Re: Question about script

Ionut Grigorescu_2 — Fri, 27 Jun 2003 09:15:45 GMT

Hi,

awk 'BEGIN{var=0}{if ($2=="0000"){print;var=NR+1} if (NR==var) print}' your.file

have fun, ionut

Re: Question about script

Caesar_3 — Fri, 27 Jun 2003 16:19:27 GMT

Hello!

Use with awk:
awk '/class: 0000/ {print $0; getline; print $0}' orig_file > new_file

Caesar