Move Values In Perl Array - perl

Question

I am having difficulty placing array columns in a format that is consistent. I have the following output:
Mon,Jun,25,14:39:29,2012,971,29,0,25,0,0,0,4,Mon,Jun,25,14:39:29,2012,25,mod_was_ap22_http.c
Mon,Jun,25,14:40:29,2012,972,28,0,25,0,0,0,3,Mon,Jun,25,14:40:29,2012,3,mod_sm22.cpp,22,mod_was_ap22_http.c
Mon,Jun,25,14:41:29,2012,973,27,0,24,0,0,0,3,Mon,Jun,25,14:41:29,2012,24,mod_was_ap22_http.c
Mon,Jun,25,14:42:29,2012,974,26,0,20,0,0,0,6,Mon,Jun,25,14:42:29,2012,1,mod_sm22.cpp,19,mod_was_ap22_http.c
Mon,Jun,25,14:43:29,2012,971,29,0,26,0,0,0,3,Mon,Jun,25,14:43:29,2012,2,mod_sm22.cpp,24,mod_was_ap22_http.c
Mon,Jun,25,14:44:30,2012,957,43,0,41,0,0,0,2,Mon,Jun,25,14:44:30,2012,1,mod_sm22.cpp,40,mod_was_ap22_http.c
Mon,Jun,25,14:45:30,2012,963,37,0,35,0,0,0,2,Mon,Jun,25,14:45:30,2012,2,mod_sm22.cpp,32,mod_was_ap22_http.c
Mon,Jun,25,14:46:30,2012,972,28,0,24,1,1,0,2,Mon,Jun,25,14:46:30,2012,24,mod_was_ap22_http.c,1,ApacheModule.cpp
Mon,Jun,25,14:47:30,2012,961,39,1,37,0,0,0,1,Mon,Jun,25,14:47:30,2012,37,mod_was_ap22_http.c,1,ApacheModule.cpp
Mon,Jun,25,14:48:30,2012,968,32,0,30,0,0,0,2,Mon,Jun,25,14:48:30,2012,30,mod_was_ap22_http.c
Mon,Jun,25,14:49:30,2012,972,28,0,25,0,0,0,3,Mon,Jun,25,14:49:30,2012,1,mod_sm22.cpp,24,mod_was_ap22_http.c
I would like the columns to display:
DayOfWeek,Month,Day,Time,Year,Rdy,Bsy,Rd,Wr,Ka,Log,Dns,Cls,AP22,SM22,ApacheModule
Currently the columns in bold are not in that order (the rest are correct). Each line isn't consistent with that format. The line sometimes has ap22 first, sometimes has sm22 first, and sometimes has none or all three modules. The number before the module relates to the module. How can I move the data into a consistent format?
Note that the 2nd date, mod_was_http.c, mod_sm22.cpp, and ApacheModule.cpp in each line will be removed in the final array.
Here is my code so far:
# This program parses a error log for necessary information and outputs in CSV format.
# chunks of your input to ignore, see below...
my %ignorables = map { $_ => 1 } qw([notice mpmstats: rdy bsy rd wr ka log dns cls bsy: in);
# 3-arg open is safer than 2, lexical my $fh better than a global FH glob
open my $error_fh, '<', 'iset_error_log';
sub findLines {
my($item,#result)=("");
# Iterates over the lines in the file, putting each into $_
while (<$error_fh>) {
# Select only those fields that have the word 'notice'
if (/\[notice/) {
# Place those lines with the word 'rdy' on the next line
if (/\brdy\b/){
push #result,"$item\n";
$item="";
}
else {
$item.=",";
}
# Split the line into fields, separated by spaces, skip the %ignorables
my #line = grep { not defined $ignorables{$_} } split /\s+/;
# More cleanup
s/|^\[|notice|[]]//g for #line; # remove unnecessary elements from the array
# Output the line.
#line = join(",", #line);
s/,,/,/g for #line;
map $item.=$_, #line;
}
}
#result
}
my #array = &findLines;
foreach $line (#array){
print $line; #This is where I would like to organize the lines if possible.
}
My input file looks like this:
[Mon Jun 25 07:51:17 2012] [notice] mpmstats: rdy 990 bsy 10 rd 0 wr 7 ka 0 log 0 dns 0 cls 3
[Mon Jun 25 07:51:17 2012] [notice] mpmstats: bsy: 2 in mod_sm22.cpp, 5 in mod_was_ap22_http.c
[Mon Jun 25 08:08:17 2012] [notice] mpmstats: rdy 974 bsy 26 rd 1 wr 24 ka 0 log 0 dns 0 cls 1
[Mon Jun 25 08:08:17 2012] [notice] mpmstats: bsy: 1 in mod_sm22.cpp, 23 in mod_was_ap22_http.c, 1 in ApacheModule.cpp Mon,Jun,25,14:38:29,2012,962,38,0,36,0,0,0,2,Mon,Jun,25,14:38:29,2012,3,mod_sm22.cpp,33,mod_was_ap22_http.c
[Mon Jun 25 21:54:41 2012] [notice] mpmstats: rdy 999 bsy 1 rd 0 wr 0 ka 0 log 0 dns 0 cls 1
[Mon Jun 25 21:55:41 2012] [notice] mpmstats: rdy 999 bsy 1 rd 0 wr 0 ka 0 log 0 dns 0 cls 1
[Mon Jun 25 21:59:41 2012] [notice] mpmstats: rdy 999 bsy 1 rd 0 wr 1 ka 0 log 0 dns 0 cls 0
[Mon Jun 25 21:59:41 2012] [notice] mpmstats: bsy: 1 in mod_was_ap22_http.c
[Mon Jun 25 22:00:41 2012] [notice] mpmstats: rdy 999 bsy 1 rd 0 wr 1 ka 0 log 0 dns 0 cls 0
[Mon Jun 25 22:00:41 2012] [notice] mpmstats: bsy: 1 in mod_was_ap22_http.c
[Mon Jun 25 22:03:41 2012] [notice] mpmstats: rdy 998 bsy 2 rd 0 wr 2 ka 0 log 0 dns 0 cls 0
[Mon Jun 25 22:03:41 2012] [notice] mpmstats: bsy: 2 in mod_was_ap22_http.c
[Mon Jun 25 22:08:42 2012] [notice] mpmstats: rdy 998 bsy 2 rd 0 wr 2 ka 0 log 0 dns 0 cls 0
[Mon Jun 25 22:08:42 2012] [notice] mpmstats: bsy: 2 in mod_was_ap22_http.c
[Mon Jun 25 22:21:42 2012] [notice] mpmstats: rdy 999 bsy 1 rd 0 wr 1 ka 0 log 0 dns 0 cls 0
[Mon Jun 25 22:21:42 2012] [notice] mpmstats: bsy: 1 in mod_was_ap22_http.c
[Mon Jun 25 22:22:42 2012] [notice] mpmstats: rdy 999 bsy 1 rd 0 wr 1 ka 0 log 0 dns 0 cls 0
[Mon Jun 25 22:22:42 2012] [notice] mpmstats: bsy: 1 in mod_was_ap22_http.c
[Mon Jun 25 22:31:42 2012] [notice] mpmstats: rdy 999 bsy 1 rd 0 wr 0 ka 0 log 0 dns 0 cls 1
[Mon Jun 25 22:32:42 2012] [notice] mpmstats: rdy 999 bsy 1 rd 0 wr 1 ka 0 log 0 dns 0 cls 0
[Mon Jun 25 22:32:42 2012] [notice] mpmstats: bsy: 1 in mod_was_ap22_http.c
[Mon Jun 25 23:06:43 2012] [notice] mpmstats: rdy 999 bsy 1 rd 0 wr 1 ka 0 log 0 dns 0 cls 0
[Mon Jun 25 23:06:43 2012] [notice] mpmstats: bsy: 1 in mod_was_ap22_http.c

Answer 1

You probably want to reorder the columns while you still have the split, before you use join to turn them back into a line of text.
You just need to do a swap.
# 0 ,1 ,2 ,3 ,4 ,5 ,6 ,7 ,8 ,9 ,10 ,11 ,12 ,13 ,14 ,15
# DayOfWeek,Month,Day,Time,Year,Rdy,Bsy,Rd,Wr,Ka,Log,Dns,Cls,AP22,SM22,ApacheModule
#
# Sometimes the last 2 fields are missing and 13 comes before 14 and 15 in the
# input, so fix that.
if (#line < 16) {
push #line, '', ''; # or whatever you want for blanks
}
#line = #line[0..12,14,15,13]; # rearrange the array
Also, your regular expression s/,,/,/g is going to break this if you use blank strings ('') as your empty fields. The short lines missing the last fields will go back to being short missing the correct 13 and 14 fields.
Based upon the kinds of questions being asked here and previously, I highly recommend you get a copy of Modern Perl (available for download or purchase) or Learning Perl to get a better grasp of the language as a whole. I have recently read much of the former and enjoyed it and gained most of my initial Perl knowledge from an earlier edition of the latter.