pattern match in each line and capture it question

UNIX for Advanced & Expert Users
1

I need a clarification on one of the scripts that i have written, I new to file handling and i need help:

I am trying to find a particular pattern in a file "****** PBX" in set of 5 files named:
xy.cc3
xv.cc3
xx.cc3
xr.cc3
xd.cc3

in a directory. If i find the files starting with these line i need to capture the full line and put it in a log file.

Can anyone help me please

Thanks,
Sandeep

2

try

grep '^...... PBX' xy.cc3 xv.cc3 xx.cc3 xr.cc3 xd.cc3 >logfile

This will prefix each "extracted-line" with the corresponding file-name

(The '^' character represents the start-of-the-line and each dot ('.') represents one-arbitrary-character, in regular-expresions used in line-identification in commands like grep, egrep, awk, sed, etc.)

3

thanks it works

4

I need one more info , I have written the script below with your inputs
---------------------------------------------------------------------
##!/usr/bin/sh -x
##Initialising variables

FROM_DIR=/usr/local/Sandeep_test/santest; export FROM_DIR

PBX_TXT="****** PBX"; export PBX_TXT

cd $FROM_DIR

for file in $FROM_DIR/*.cc3

do

grep '^...... PBX' $file > file_log

cut -c8-40 file_log

done
exit 0
--------------------------------------------------------

The output of the the above script is
Output:

PBX TYPE:san PBX-id: 555
PBX TYPE:san2 PBX-id: 556
PBX TYPE:san3 PBX-id: 557

now i need one more info:How do i catch the above result so that san and 555 are put it in a file ( example: file 1) like the below format:

san,555
san2,556
san3,557

and then comapare this line by line with an another file(example:file 2) that is as below:
san,555
san2,556
san3,558

now after the compare of the file 1 and file 2 it should say that san3,557 is not correct as this is not in file 2

Please help will be very gratefull

Thanks,
Sandeep

5

Add the following to the cut -c8-40 file-log:

| sed 's/ PBX-id: /,/' >file-1

and then use either diff or comm (w. file-1 file-2 as its two arguments) in order to see the difference.

6

cut -c8-40 | sed 's/ PBX-id: /,/' >file-log

added the above and it did not give any result. please help

7

sorry but this how the output looks from my script. i mistakingly sent the wrong yesterday:

PBX TYPE:IDC4 PBX-id: L14
PBX TYPE:CC3 PBX-id: LC4
PBX TYPE:CC3 PBX-id: LC8

aloso can u please explain the code, when u send it ........will be grateful.

Thanks,
sandeep

8

can someone help me out with this please?...

9

First, you could read "man cut" to find out how the cut command is working and how to change it according to you data.

Second, just change the 8-40 to 10-40!!

10

Third!! your original command was:

cut -c8-40 file_log

so you should have added the sed at the end as:

cut -c8-40 file_log | sed ....

Not after 40!?!?!

11

THnaks for the reply.

The output of the script is

PBX TYPE:CC3 ,NSH

note the space between CC3 and ,NSH

I would want this to be :

CC3,NSH

My script is like this. Please reveiw.

##!/usr/bin/sh -x
##Initialising variables

FROM_DIR=/usr/local/Sandeep_test/santest; export FROM_DIR

PBX_TXT="****** PBX"; export PBX_TXT

echo "----------------------------LIST OF INCORRECT PBX/SWTICH TYPE FOUND--------------------------"

echo "----------------------------LIST The CORRECT PBX/SWTICH TYPE---------------------------------"
cd $FROM_DIR

for file in $FROM_DIR/LEE*.cc3

do

grep '^...... PBX' $file > file_log

cut -c10-40| sed 's/ PBX-id: /,/' file_log

done
exit 0

12

This is NOT the script that you're running!!!

The "cut" has STILL no File-Parameter and is piped to the "sed" which has the File-Parameter.

You should have missed ONE OF THE BLANKS in "/ PBX-id: /,/" !!!

13

##!/usr/bin/sh -x
##Initialising variables

FROM_DIR=/usr/local/Sandeep_test/santest; export FROM_DIR

PBX_TXT="****** PBX"; export PBX_TXT

cd $FROM_DIR

for file in $FROM_DIR/LEE*.cc3

do

grep '^...... PBX' LEECOUNTY.cc3 > file_log

cut -c8-40 > file_log1

sed 's/ PBX-id: /,/' file_log1 > file_log2

done
exit 0

This is the script and when i cat file_log2 the output is:

PBX TYPE:IDC4 ,L14
PBX TYPE:IDC3 ,L12
PBX TYPE:IDC5 ,L13

Note there is lot of space between IDC4 and ,L14

Desired output would be:
IDC4,L14
IDC3,L12
IDC5,L13

14

OK apparantly you have more than ONE BLANK around the text!!

Please use:

sed 's/ *PBX-id: */,/' file_log1 > file_log2

Note the '*' after each blank.

15

Thanks so much.i really appreciate it. was help up for so many days.

Just one last thing , now the result is

PBX TYPE:IDC4,L14
PBX TYPE:IDC3,L12
PBX TYPE:IDC5,L13

What can i do to remove PBX TYPE: so that i have only

IDC4,L14
IDC3,L12
IDC5,L13
i can do a cut of this result again but can't we specify the elimination of the result in sed itself?

Thnx again:)

Regards,
Sandeep

16

Change "-c8-40" to "-c17-40" !!!

Don't you ever want to know how UNIX-Commands work?

17

Thanks again, I will do the change. I am so soryy to have troubled you so much. I am trying to and i have written some scripts as well. Those were using very easy commands and they look very amatuer. I am doing some googling and studying but since i have to spend most of my time with office work i am doing this outside my office timmings on interest. Also there is no one guiding as well. Thanks and with help of people like you one day i hope i can give solutions to others like you are doing

Thanks again

Regards,
Sandeep