UNIX for Dummies Questions & Answers

Hi,

Is it possible to grep only on a certain part of a file? Say I have file in.txt which contains below:

11/16 13:07:19.5436 --- ERROR 123 detected.
11/16 13:08:19.5436 --- Generating a <reading> event
11/16 13:08:19.7784 ---- Sending a <writing> event
11/16 14:08:37.4516 --- Generating a <reading> event
11/16 14:08:37.9199 ---- Sending a <writing> event
11/16 14:09:19.5436 --- ERROR 123 detected.
11/16 14:09:49.8452 --- Generating a <reading> event
11/16 14:09:50.0868 ---- Sending a <writing> event
11/16 16:10:51.6707 --- Generating a <reading> event
11/16 16:10:52.3232 ---- Sending a <writing> event
11/16 16:08:19.5436 --- Generating a <reading> event
11/16 16:08:19.7784 ---- Sending a <writing> event
11/16 16:08:19.9436 --- ERROR 123 detected.
11/16 16:08:37.4516 --- Generating a <reading> event
11/16 16:08:37.9199 ---- Sending a <writing> event
11/16 16:09:49.8452 --- Generating a <reading> event
11/16 16:09:50.0868 ---- Sending a <writing> event
11/16 16:10:51.6707 --- Generating a <reading> event
11/16 16:10:52.3232 ---- Sending a <writing> event


Now, I only want to get a certain part on the file.. say the part

11/16 14:09:49.8452 --- Generating a <reading> event
11/16 14:09:50.0868 ---- Sending a <writing> event
11/16 16:10:51.6707 --- Generating a <reading> event
11/16 16:10:52.3232 ---- Sending a <writing> event
11/16 16:08:19.5436 --- Generating a <reading> event
11/16 16:08:19.7784 ---- Sending a <writing> event

which is after the 2nd "ERROR 123 detected." line. Then put it in a file.

Also, is it possible to cut this file into chunk. The delimiter would be the "ERROR 123 detected."
Each chunk of file will be put into a new file. Im just new to unix and still learning and I have many ideas in mind that I could not execute due to insufficient knowledge in the language. Thanks.

Something like this:

Code:

awk '/ERROR [0-9]* detected/{close(f);f=($4"_"$5"_"++c".out")}f{print>f}' filename

Use nawk on Solaris.

Code:

awk '/ERROR [0-9]* detected/{++c};c==2' filename>second_error.out

Use nawk or /usr/xpg4/bin/awk on Solaris.

You could also use the split command to cut the file into chunks.

Code:

split -p "ERROR 123 detected"  filename outfileprefix

The 2nd outfile (ending in ab) contain the lines for the 2nd "ERROR 123 detected"