asked on

parse a text file

Each alarm starts with UNIT6, I'm trying to get all the rows of each alarm into one row and from there I can sort and do whatever I need. I've had some success using while loops but the problem is some alarms have 4 rows, some have 5,6, or 7. Once it gets beyond 4 rows I get duplicates or I only get alarms with 5 rows depending on how I set up the while loop. I attached the code I have been trying so far it is somewhat messy from all the retries.
Here are is a partial alarm log:
UNIT6 LOT-0086 BLOCK-0086 ENVIR 2007-11-14 00:09:19.37
* ALARM D142B
(10239) 7312 RATIO THRESHOLD
N

UNIT6 LOT-0086 ENVIR 2007-11-14 00:09:20.05
* ALARM
(10241) 7000 CABINET OPEN
Cabinet open.
02 01 00

UNIT6 LOT-0086 COMM 2007-11-14 00:09:20.46
** ALARM
(10242) 8999 BLOCKED FROM USE
PCBLOCKHub PCBLOCKHub CC bank
02 00 1d

#!/usr/bin/perl -w
use strict;
 
 
open DDDLOG,  "expx.txt" or die "Cannot open file $!";
open DDDRPT1, ">all.txt" or die "Cannot open file $!";
 
 
 
my @FFF; my $fff;
my @DDD; my $ddd;
my @NUM; my $num;
my @GGG; my $ggg;
my @DESC; my $desc;
my @DESC2; my $desc2;
my @DESC3; my $desc3;
 
 
########  PARSE RAW FILE ################
 
 
MAIN: while (<DDDLOG>){
            if (/\s+UNIT6\s+LOT-\d\d\d\d/){
            $fff = $_;
            @FFF = split /\s+/, $fff;
            #print "@FFF\n";
            #}
     
 
         LOOP2: while (<DDDLOG>){
                     redo MAIN if (/\s+UNIT6/);
                     if (/\*|\*\*|\*\*\*/){
                     $ddd = $_;
                     @DDD = split /\s+/, $ddd;
                     #print "@FFF @DDD\n";
                     #}
 
             LOOP3: while (<DDDLOG>){
                         redo MAIN if (/\s+UNIT6/);
                         if (/7\d\d\d|8\d\d\d/){
                         $num = $_;
                         @NUM = split /\s+/, $num;
                         #print DDDRPT1 "@FFF @DDD @NUM\n";
                         #}
 
                   LOOP4: while (<DDDLOG>){
                               redo MAIN if (/\s+UNIT6/);
                               if (/[a-zA-Z0-9]/){
                              $desc = $_;
                               #chomp($desc);
                               @DESC = split /\s+/, $desc;
                
                              #print DDDRPT1 "@FFF @DDD @NUM @DESC\n";
                               # print DDDRPT1 "$DESC[1] $DESC[2] $DESC[3]\n";
                 
                 
 
 
     LOOP5: while (<DDDLOG>){
                 redo MAIN if (/\s+UNIT6/);
                 if (/[a-zA-Z0-9]/){
                 $desc2 = $_;
                 #chomp($desc2);
                 @DESC2 = split /\s+/, $desc2;
                
               print DDDRPT1 "@FFF @DDD @NUM @DESC @DESC2\n";
                  # print DDDRPT1 "$DESC[1] $DESC[2] $DESC[3]\n";
                  #}  
                 #elsif(/ /){
                 #print DDDRPT1 "@FFF @DDD @NUM @DESC\n";
                 #last;
                 
            
     LOOP6: while (<DDDLOG>){
                 redo MAIN if (/\s+UNIT6/);
                 if (/[a-zA-Z0-9]/){
                 $desc3 = $_;
                 #chomp($desc3);
                 @DESC3 = split /\s+/, $desc3;
                
               #print DDDRPT1 "@FFF @DDD @NUM @DESC @DESC2 @DESC3\n";
                  # print DDDRPT1 "$DESC[1] $DESC[2] $DESC[3]\n";
                  }  
                 elsif(/ /){
                 #print DDDRPT1 "@FFF @DDD @NUM @DESC\n";
                 last;    
            
}}}}
   }
}
            }
            } 
            }
     }}
}
close DDDRPT1;
close DDDLOG;

Open in new window

ASKER CERTIFIED SOLUTION

ozo

membership

This solution is only available to members.

To access this solution, you must be a member of Experts Exchange.

Start Free Trial

omcr

ASKER

Ozo,
Amazing. I'm not going say how much time (days) I've spent on this.
If you can take a second to explain it that would help. About the only thing I think I understand is $e adds a "\n" to the end everytime it sees UNIT6. Are you chomping every line in the file, I think so, and then newline in $e signals where to end. But I don't understand how you get each line into $e and make it one line without the use of arrays?. Anyway there is one problem, the actual log file contains several lines of trash at the beginning and end. I'm going to try to enclose your code in a while loop and see what happens