PERL Question

Ok.... Here's what I want to do...

I think this may have to do with matching (=~) but I'm not sure. I want to be able to--uh--extract text that is between certain HTML commands in a file. e.g, I want to get the titles for a group of files, so, I'll want to find the and get "This is the title" from <title>This is the title</title> without getting the whole document.

Any ideas?

Thanks,
Nik

[Nodrog]

You could use an XML parser. Alternatively, something like this...

#!/usr/bin/perl
my(@lines, $open, $close);
$open = "<".$ARGV[1]; 
$close = "</".$ARGV[1].">";
open(FILE, "$ARGV[0]") || die "couldnt open file"; 
@lines = <FILE>;
close(FILE);
foreach(@lines) {
	if(/$open[^>]*>(.*)$close/) {print $1."\n";}
}

will search a file for a tag, and print out the text between it.

For example...

"perl parser.pl what.html title"

will output the text between the <title> and </title> tag in the what.html file

"perl parser.pl blah.html b"

will output all the bold text in the file (ie all the text between <b> and </b>)



Obviously thats pretty basic, but you could play around with it till it does what you need.

edit - stupid mistake

[MT]

<b>owned</b> unbolded <b>is you</b>
heh

[Atamur]

($text) = $html =~ m#<blah>(.*?)</blah>#is;

[Nodrog]

Originally posted by MT

<b>owned</b> unbolded <b>is you</b>
heh

youre obviously gay

[MT]

#!/usr/bin/perl -w
use strict;
require HTML::TokeParser;
my ($tag,$doc) = @ARGV;
my $parser = new HTML::TokeParser->new($doc);
while (my $tok = $parser->get_tag("$tag")) {
  print $parser->get_trimmed_text("/$tag") . "\n";
}

Dont oppress me homobasher!

[W]

#!/usr/bin/perl
$starttag=shift(@ARGV);
$untag='</'.$starttag.'>';$startag='<'.$starttag;
$endtag='>';
$state=0;
while(<>)while(length){
$state=1,$_=$1 if $state=0 && m/$starttag(.*)/;
$state=2,$_=$1 if $state=1 && m/$endtag(.*)/;
$state=0,$_=$2,print $1 if $state=2 && m/(.*?)$untag(.*)/;
}

> ./myperlcode title index.html other.html

> ./myperlcode a links.html


You can notice that I like online algorithms.

[W]

Originally posted by Nodrog
edit - stupid mistake

Well, and your program will only find one tag per line. If someone is bright enough to remove all the newlines from their html, your program will ever only find the first tag.

[Atamur]

[quote]Originally posted by W
[b]

...............

why not simply gobble up the entire file into memory? it would be faster and simpler.

[W]

[quote]Originally posted by Atamur
[b]

Originally posted by W

...............

why not simply gobble up the entire file into memory? it would be faster and simpler.

As I said, I like online algorithms. Say it's a collection of eighty thousand 10kb html files?

[Atamur]

Originally posted by W
As I said, I like online algorithms. Say it's a collection of eighty thousand 10kb html files?

loading one file at a time in memory will be much faster then a line at a time.

[W]

Originally posted by Atamur
loading one file at a time in memory will be much faster then a line at a time.

No.