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

I have this code in my program and it's parsing a large log file. When it reaches at this point the regular expression seems to hang badly, there is a faster way to process the code to improve speed? Here is the code: 
my $p1= "iapw_p1";
my $p2= "iapw_p2";
my $p3= "iapw_p3";
my $b1= "iapw_b1";
my $p0= "iapw_p0";
my $c0= "iapw_c0";
my $c1= "iapw_c1";
my $c3= "iapw_c3";
my $h1= "iapw_h1";
my $p1b="<font color=\"navy\" size=\"2\"><b>P1</b></font><font color=\
+"#000000\" size=\"2\"><b><i> = (file inquiry - launch page)</i></b></
+font>";
my $p2b="<font color=\"navy\" size=\"2\"><b>P2</b><font color=\"#00000
+0\" size=\"2\"><b><i> = (file search - replace - operators)</i></b></
+font>";
my $p3b="<font color=\"navy\" size=\"2\"><b>P3</b><font color=\"#00000
+0\" size=\"2\"><b><i> = (file notepad)</i></b></font>";
my $b1b="<font color=\"navy\" size=\"2\"><b>B1</b><font color=\"#00000
+0\" size=\"2\"><b><i> = (file inquiry: old - owners and preowners)</i
+></b></font>";
my $p0b="<font color=\"navy\" size=\"2\"><b>P0</b><font color=\"#00000
+0\" size=\"2\"><b><i> = (file search)</i></b></font>";
my $c0b="<font color=\"navy\" size=\"2\"><b>C0</b><font color=\"#00000
+0\"e size=\"2\"><b><i> = (file search)</i></b></font>";
my $c1b="<font color=\"navy\" size=\"2\"><b>C1</b><font color=\"#00000
+0\" size=\"2\"><b><i> = (file inquiry)</i></b></font>";
my $c3b="<font color=\"navy\" size=\"2\"><b>C3</b><font color=\"#00000
+0\" size=\"2\"><b><i> = (file notepad)</i></b></font>";
my $h1b="<font color=\"navy\" size=\"2\"><b>H1</b><font color=\"#00000
+0\" size=\"2\"><b><i> = (file and arch inquiry and history)</i></b></
+font>";

for (@a_data) { 
    s/$p1/$p1b/;
    s/$p2/$p2b/;
    s/$p3/$p3b/;
    s/$b1/$b1b/;
    s/$p0/$p0b/;
    s/$c0/$c0b/;
    s/$c1/$c1b/;
    s/$c3/$c3b/;
    s/$h1/$h1b/;

}

[download]

edited: Wed Aug 14 15:08:05 2002 by jeffa - title change (was: Regular Expression)

Replies are listed 'Best First'.
Re: Regular Expression
by Abigail-II (Bishop) on Aug 13, 2002 at 16:23 UTC
    Untested: 
    s!iapw_(..)!qq {<font color = "navy" size = "2"><b>\U$1\E}     .
            qq {</b><font color = "#000000" size = "2"><b><i>} .
            qq { = (file inquiry: old - owners and preowners)} .
            qq {</i></b></font>}!eg for @a_data;

    [download]
    The only reason for the /e and the qq{} is readability.

    Abigail

[reply]
[d/l]
[select]
      Seems the text in the parens changes slightly for each of his sub strings. I changed Abigail-II's example to take this into acccount.
      my %text = ( p1 => '(file inquiry - launch page)',
             p2 => '(file search - replace - operators)',
             p3 => '(file notepad)',
             b1 => '(file inquiry: old - owners and preowners)',
             p0 => '(file search)',
             c0 => '(file search)',
             c1 => '(file inquiry)',
             c3 => '(file notepad)',
             h1 => '(file and arch inquiry and history)');
s!iapw_(\w\d)!qq {<font color = "navy" size = "2"><b>\U$1\E}     .
              qq {</b><font color = "#000000" size = "2"><b><i>} .
              qq { = $text{$1}</i></b></font>}!eg for @a_data;

      [download]

      (Oh, I also tested it -- it works!)
      Cheers,
      Shendal
[reply]
[d/l]
        It did work just like the first code I have. If I am using a small file it goes fast, but if I am parsing a text log file with 100.000 lines it hangs again!
[reply]

Back to Seekers of Perl Wisdom