harry34 has asked for the wisdom of the Perl Monks concerning the following question:

I have the follwing code which extracts data from a file containing 30 data sets having the following pattern (see code). The problem is that it outputs the data in a very long list, I have no idea where the first data set finishes and the second starts.
I need to modify this code so that each time the code sees the string ' Final graph set matrix' it extracts the data (pattern as given in the code) up until it sees the string ' PLUTO4 finished'.
It should then break out the loop and start over separating each block of data by a new line or title e.g. data set1 ..... data setn.
If interested I could send the file from which I am working from.

cheers harry 
#!/usr/local/bin/perl 

$in_filename = "graph_set.out";
open (IN,"$in_filename") or die "Can't open $in_filename:$!\n";
local $/ = undef;       # undef record seperator
my $string = <IN>;         # read whole file into string    
close (IN);

my @gset_match = ($string =~ /([A-Z]\s\d+,\s\d+\(\s*\d+\))/g);

open (TEXT, ">graph_set.txt") or die "Can't create graph_set.txt: $!\n
+";
foreach $_(@gset_match) { print TEXT "$_\n"; }
close TEXT;

[download]

Replies are listed 'Best First'.
Re: pattern matching between two specific strings
by Zaxo (Archbishop) on Jul 08, 2003 at 10:23 UTC

    You can break up the input file into records by setting the record seperator, 

    #!/usr/local/bin/perl 

my ($in_filename, @records) = 'graph_set.out';
{
    local $/ = ' PLUTO4 finished';
    my $start = ' Final graph set matrix';
    open IN, '<' $in_filename or die "Can't open $in_filename: ", $!;
    @records = map {
        substr $_, index( $_, $start) + length( $start}
    } <IN>;
    chomp @records;
    close (IN);
}

    [download]
    Without seeing your data, I can't say much about your regex or how it parses.

    After Compline,
    Zaxo

[reply]
[d/l]
      I'm geting some errors with this code ?
      Is it possible for me to have your email and I can send you the file I'm working from
[reply]
Re: pattern matching between two specific strings
by zby (Vicar) on Jul 08, 2003 at 10:42 UTC
    I would do it on line basis (using the .. flip flop operator): 
    .
.
.
my $firsttime = 1;
while(<IN>){
  if(/Final graph set matrix/ .. /PLUTO4 finished/){
    if($firstime--){
      print "\nnew data set\n";
    }
    my @gset_match = ($string =~ ([A-Z]\s\d+,\s\d+\(\s*\d+\))/g);
    foreach $_(@gset_match) { print TEXT "$_\n"; }
  }
}
.
.
.

    [download]
    I feel there should be some more elegant way to print the delimiters, but can't figure it out.
[reply]
[d/l]
[select]
      I could not get the code you provided to work
      I've listed a portion of the file I'm working from which contains numerous datasets having the following format, see below: I need to extract the data between Final graph set matrix and PLUTO4 finished each time it is present in the file and seperated by a title of some sort.

      hope you can help, cheers 
      pattern:<br>
/([A-Z]\s\d+,\s\d+\(\s*\d+\))/g

      [download]
      FILE CONTAINS DATA IN THE FOLLOWING FORMAT

      Final graph set matrix
      ----------
      C 1, 1( 9)
      ----------
      C 2, 2(11) C 1, 1( 4)
      ----------
      C 2, 2(18) C 1, 2(11) C 1, 1( 9)
      R 2, 2( 8)
      ----------
      C 1, 2(11) C 2, 2(18) C 2, 2(11) C 1, 1( 4)
      R 2, 2(18)
      PLUTO4 finished


      Final graph set matrix
      ----------
      C 1, 1( 4)
      ----------
      C 2, 2(18) C 1, 1( 4)
      R 2, 2(18)
      PLUTO4 finished


      Final graph set matrix
      ----------
      C 1, 1( 4)
      ----------
      C 2, 2(18) C 1, 1( 4)
      R 2, 2(18)
      ----------
      C 2, 2(11) C 2, 2(11) C 1, 1( 9)
      ----------
      C 2, 2(11) C 2, 2(11) C 2, 2(18) C 1, 1( 9)
      R 2, 2( 4)
      PLUTO4 finished
[reply]
[d/l]
        Please post the code you derived from mine - I'll see what is the problem with it.
[reply]

Back to Seekers of Perl Wisdom