Selective searching?

I'm not totally sure I understand problem in this case... is it that each HTML page has a common section that you don't want displayed as "description"; you'd rather skip ahead to where the actual content begins?

If so, this solution might work well for you:

Code Sample:

open FILE, "some.html" or
  die "can't open some.html: $!";
{ # this slurps the ENTIRE file into
  # a scalar variable (quickly, too!)
  local $/;
  $file = <FILE>;
}
close FILE;

$endcommon = << 'END';
This should be a string holding the END of
the "common sidebar" thing, or whatever.  As
soon as this string is found, the VERY NEXT
character in the file will be a candidate for
the description.
END

if (($p = index($file,$endcommon)) != -1) {
  while (length($descrip) < 200) {
    $descrip = substr($file,$p,200);
    $p += 200;
    # get next 200 chars
    # NOTE: THIS IS A VERY POOR HTML TAG
    # STRIPPING ROUTINE
    # it "breaks" for:
    # <img src="foo.gif" alt=" ---> ">
    $descrip =~ s/<[^>]*>//g;
  }
  $descrip .= "...";
}

$descrip | |= "No description";

------------------
--
MIDN 4/C PINYAN, NROTCURPI, US Naval Reserve

[This message has been edited by japhy (edited 16 December 1999).]

To see a much better example (I fixed the code up considerably), please go to this URL: http://www.pobox.com/~japhy/perl/forum_examples

The directory is the one that says "maverick". In there are a couple mock HTML files, and a program that displays their titles, descriptions, and a link to each.

I'll be working on it -- email me (or post in this forum) any questions, comments, concerns you have.

------------------
--
MIDN 4/C PINYAN, NROTCURPI, US Naval Reserve

Thanks very much.

Nice coding Japhy,

may I invite you to participate in TWF's BB project @ www.Boardzilla.org

Later,

Malte

------------------
Malte Ubl - www.Boardzilla.org
Communication: public<->programmers
of the Boardzilla BB