Extracting From A Text File

[dexter48]

Hi I'm searching for a string occurance in a text file. I find the string ok and write the results to a log file. But on the line above is also some information I need. How can i get that. The string occurs a number of times, but not an exact match for the string above. Can you help please: see code

#!C:\\Perl\\bin 
 
use File::Copy;
use Archive::Zip qw( :ERROR_CODES :CONSTANTS );
 
###############################################################
#Get the date,# Months are 0-11!,# Years are 1900 out!
###############################################################
($sec,$min,$hour,$mday,$mon,$year,$wday,$yday) = gmtime(time);
$mon = $mon+1;
$year = $year+1900;
 
###############################################################
#set the variables
###############################################################
$drive = "Y:";
$dir = "backup";
$filename = "Alllogfiles_";
$date = sprintf"%s_%s_%s",$year,$mon,$mday;
$comp_name = "Misys-FromLab"; 
#$comp_name = "Misys-ToLab"; 
$searchstring = "EXCEPTION";
$errorfile = "MisysFromLabErrors$date.txt";
###############################################################
#Add them together for the archive location
###############################################################
$fopen = "$drive\\$dir\\$filename$date.zip";
 
###############################################################
#working directory
###############################################################
$perltemp = "c:\\perltemp";
 
###########################
#Get rid of any old files
###########################
system("del $perltemp\\*.log");
system("del $perltemp\\All*.Zip");
 
###############################################
#get the archive file
###############################################
copy ($fopen, $perltemp);
 
###################################################
#now open the archive
###################################################
my $opt = x;
my $ofile = "$perltemp\\$filename$date.zip";
 
# extract a zip file
if ($opt=~/x/i) {
    print "Extracting $ofile...\n";
    $zip = Archive::Zip->new();
    die 'Error reading zip file.' if $zip->read( $ofile ) != AZ_OK;
    my @members = $zip->members();
    foreach $element(@members) {
        print "$element\n";
        $zip->extractMember($element);
    }
    print "Done!\n";
}
 
#############################################################
#Search for Exceptions in FromLab
############################################################
opendir(DIR, "$perltemp") || die "Could not open $dir $!";
 
while($name = readdir(DIR)) {    
    ##Put the names in an array, latest files 1st
    if ($name =~ /$comp_name/) {
        print "Beginning search for Order Exceptions\n";
        print "puting name = $name in the array\n";
        unshift(@files, $name);
    }
}
 
closedir(DIR);
 
open (FHNDL, ">$errorfile") || die "ERROR Could not open $file: $!n";
 
foreach (@files) {
    my $logfile = $_;
    #print "searching  = $logile\n";
 
    open(SEARCHFILE, "<$perltemp/$logfile") || die "Could not open file $matchesboth : $!";
 
    while (<SEARCHFILE>) {
        #print "chomping file $dir, line = $line"; 
        if ($_ =~/$searchstring/){
 
            #print FHNDL "matched string = $1 - file = $logfile - $string = $_\n";
            print FHNDL "matched string in $logfile - $string = $_\n";
            print FHNDL "$###############################################################################################################################\n";
        }
    } 
}
close (FHNDL);
exit;
 

this gives me

matched string in Misys-FromLab-29 Aug 2007 15.log - = 29 Aug 2007 15:00:31.058848 EXCEPTION: System.OverflowException: SqlDbType.SmallDateTime overflow. Value '01/01/1890 00:00:00' is out of range. Must be between 1/1/1900 12:00:00 AM and 6/6/2079 11:59:59 PM.

################################################## ################
just above is anothe line i need

[KevinADC]

You can cache the lines into an array and use the array to get the previous line, this is totally untested but may work as-is, if not it will work after fixing any errors I made:

my @cached_lines = ();
while (<SEARCHFILE>) {
        #print "chomping file $dir, line = $line"; 
      push @cached_lines,$_; #cache the current line      
        if ($_ =~/$searchstring/){
         if (scalar @cached_lines < 2) {
            print "Not enough cached lines yet to check previous lines\n";
         }
         else {
            my $last_line = $cached_lines[-2];
            #print FHNDL "matched string = $1 - file = $logfile - $string = $_\n";
               print FHNDL "matched string in $logfile - $string = $_\n";
               print FHNDL "previous line in $logfile - $last_line = $_\n";
               print FHNDL "$###############################################################################################################################\n";
           }
       } 
   }
}             

there may be other concerns too like emptying some of the array if it grows too large during processing. If all you ever need is the previous line you could cache the line to a scalar instead of an array and not worry about the array growing too large.

[KevinADC]

You could also use Tie::File to access the search file instead of open() and a "while" loop. This way the file is treated like a perl array for the most part and you can use a for/foreach loop to loop through the file and access any other line in the file based on the current lines index number.

[dexter48]

Hi Kevin, this looks as though it works, thank you very much, I'll look into tie when I get a chance

[Kelicula]

The Tie::File is wonderful! I tried looping through a filehandle vs. using Tie::File and tie file was MUCH quicker. Also very easy to work with, most all methods (excuse my terminology, javascript throwback) of the array object are supported with tie file.

Good luck!