UNIX for Dummies Questions & Answers

Hi Guys,

I am new to the forum and to scripting so bear with me.

Thanks, Gary.

I have 3 files - file1, file2, file3

I am trying to come up with a script that will check the output of these files and if the 1st nine fields are matched in all 3 files, echo "The following string had been matched in all 3 files - "string"

i.e.

file1

Code:

WARNING: The Job Test for Scheduler abc is late. A run should have been expected by
WARNING: The Tester for Scheduler def is late. A run was expected by

file2

Code:

WARNING: The Tester for Scheduler ghi is late. A run has not completed
WARNING: The Tester for Scheduler def is late. A run should have been expected by

file3

Code:

WARNING: The Tester for Scheduler def is late. A run was expected by

So, the output in the above instance would be

The following string had been matched in all 3 files - "WARNING: The Tester for Scheduler def is late. A"

Please note that abc, def and ghi can be any number of different combinations, so long as the 1st 9 fields are matched in all 3 files I want to report it.

Code:

perl -e '$hash{`head -c 9 $_`} = 1  foreach (`ls file*`);
> ((keys %hash) ==1) ? print  "all three files are same\n" : print "all three files are not same\n";'

HTH,
PL

Hi, Thanks for your reply. I ran the following command and got an error -

Code:

perl -e '$hash{`head -c 9 $_`} = 1  foreach (`ls test*`); ((keys %hash) ==1) ? print  "all three files are same\n" : print "all three files are not same\n";'
 
head: illegal option -- c
head: illegal option -- c
head: illegal option -- c

all three files are same

I tried with -n instead and it compares the 3 files fine but I am not wanting to check that the test* files are all the same. I am trying to check to see if the first 9 columns are matched in any line in all 3 files. The files themselves could differ each time. I hope that makes sense.

I can run head -c on my Linux server but not this server which is Solaris 10.

Code:

 perl -e '$hash{`head -n 9 $_`} = 1  foreach (`ls test*`); ((keys %hash) ==1) ? print  "all three files are same\n" : print "all three files are not same\n";'

all three files are same

---------- Post updated at 11:08 PM ---------- Previous update was at 09:47 AM ----------

I have found a nawk command which is near to getting what I want but still needs some tweaking. e.g.

nawk 'NR==FNR{a[$1$2$3$4$5$6$7$8$9]}{if ($1$2$3$4$5$6$7$8$9 in a) print $0}' warning1.log warning2.log warning3.log > newtest

It is checking the 1st 9 fields in warning1.log and matching any of those to the same fields in warning2.log and warning3.log, then outputting to newtest.

The problem I have is that is is outputting the complete warning1.log and appending the matched lines to the bottom. Any ideas how I would get it to only output to newtest if the 9 fields are matched in all 3 seperate warning.log files and not output the entire warning1.log file?

Thanks again.
Gary.

---------- Post updated at 11:28 PM ---------- Previous update was at 11:08 PM ----------

Sorry. I also tried the following, which outputs all matches but not the one from warning1.log in this instance using the following -

nawk 'NR==FNR{a[$1$2$3$4$5$6$7$8$9]=$1$2$3$4$5$6$7$8$9;next}a[$1$2$3$4$5$6$7$8$9]' warning1.log warning2.log warning3.log > newtest

I am also only wanting it to output one match in each of the 3 warning files so if it is matched twice in one file I only need it output the once to newtest from that file.

I am essentially going to be checking that the same 1st 9 fields were reported in all 3 warning files.