Tek-Tips is the largest IT community on the Internet today!

Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!

Register Log in
  • Congratulations TouchToneTommy on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!

Reading metatags

Status
Not open for further replies.
redstarfcs1

redstarfcs1

Programmer
Sep 21, 2005
7
RO
How to read metatags without any perl module?
 
Sort by date Sort by votes
this might help?

Code: 
[b]#!/usr/bin/perl[/b]

undef $/;
$_ = <DATA>;
$/ = "\n";

@metatags = m/(<meta[^>]+>)/ig;

foreach (@metatags) {
  ($name, $content) = m/name="?([^"]+)"? content="?([^"]+)"?/;
  print "$name\t$content\n";
}

[blue]__DATA__
<HEAD>
<TITLE>Stamp Collecting World</TITLE>
<META name="description" content="Everything you wanted to know about stamps, from prices to history.">
<META name="keywords" content="stamps, stamp collecting, stamp history, prices, stamps for sale">
</HEAD>[/blue]


Kind Regards
Duncan
 
Upvote 0 Downvote
I would maybe do it just a little bit different:

Code: 
#!perl
use strict;

$_ = do {undef $/; <DATA>};
my $len = index(lc($_),'</head>');
$_ = substr $_,0,$len;

my @metatags = m/(<meta[^>]+>)/ig;

foreach (@metatags) {
  my ($name, $content) = m/name="?([^"]+)"? content="?([^"]+)"?/;
  print "$name\t$content\n";
}

__DATA__
<HEAD>
<TITLE>Stamp Collecting World</TITLE>
<META name="description" content="Everything you wanted to know about stamps, from prices to history.">
<META name="keywords" content="stamps, stamp collecting, stamp history, prices, stamps for sale">
</HEAD>
<BODY>
bunch of stuff down here we shouldn't need to parse
</BODY>
</HTML>

no need to really check for meta tags after the closing </head> tag. Of course I would also just use ruby, could probably do it with a single character. [bigcheeks]
 
Upvote 0 Downvote
Hi

Just as a sad conclusion, Duncan, I sucked a lot with such things. Some people prefer to use single quotes in HTML tag attributes, other ones use different cases in attribute names.
[tt]
<META Name="description" Content='Everything you wanted to know about stamps, from prices to history.'>
<META NAME='keywords' CONTENT="stamps, stamp collecting, stamp history, prices, stamps for sale">
[/tt]
So usually this catch abit better :
Code: 
($sep, $name, $sep, $content) = m/name=(["']?)(.+)\1 content=(["']?)(.+)\3/i;
But stil useless if [tt]name[/tt] and [tt]content[/tt] are reversed :
[tt]
<META content='Everything you wanted to know about stamps, from prices to history.' name="description">
[/tt]
And there are cases, when a [tt]meta[/tt] tag is wrapped to multiple lines.
It is horrible what people can do to poor HTML...

Feherke.
 
Upvote 0 Downvote
Hi

A question. Should be better to use the regular expression somehow like this, to exclude de currently used quotation mark from the value :
Code: 
($sep, $name, $sep, $content) = m/name=(["']?)([red][^\1][/red]+)\1 content=(["']?)([red][^\3][/red]+)\3/i;
It is theory. Can someone correct it to work properly ?

Feherke.
 
Upvote 0 Downvote
This is really ugly, but it works with the data you provided and it doesn't matter which order the name and content fields are provided.

Code: 
local undef $/;
my ($text) = <DATA>;
my ($headers) = $text =~ m[<head>(.*)</head>]is;

while ($headers =~ m!<META\s+(\w+)\s*=\s*['"]([^'"]+)['"]\s+(\w+)\s*=\s*['"]([^'"]+)['"]!isg) {
    print "$2 - $4", "\n";
}
 
Upvote 0 Downvote
Hi

Yes, rharsh, I was thinking to something similar. But only the actually used quote has to be excluded from the value :
[tt]
<META name="description" content="It's my site">
[/tt]
And the [tt]meta[/tt] tags could have more than two attributes :
[tt]
<META lang="en" name="description" content="It's my site">
<META lang="it" name="description" content="E il mio sito">
[/tt]
Sorry, if seems to go offtopic, but I'm really interested in a bulletproof solution ( without character-by-character parsing ).

Feherke.
 
Upvote 0 Downvote
The most reliable way to do this would be to use a proper tag-aware parser such as HTML::TokeParser::Simple. Here's a quick example to extract the attributes from <meta> tags:
Code: 
#!/usr/bin/perl -w
use strict;
use HTML::TokeParser::Simple;

# create a new parser
my $p = new HTML::TokeParser::Simple( 'test.html' );

# just want to keep track of how many meta tags we've seen
my $tagno = 1;

# loop through all the tokens in the page
while ( my $t = $p->get_token ) {
   # if we've found a meta tag
   if ( $t->is_start_tag( 'meta' ) ) {

      # print some information
      print 'Tag #'.$tagno++."\n";

      # get the attributes from the tag
      my $attrs = $t->get_attr;

      # print the attributes
      print "$_ - $attrs->{ $_ }\n" for ( keys %$attrs );

      # print a line to distinguish it from the next tag
      print '-' x 30, "\n";
   }
}
 
Upvote 0 Downvote
Sorry feherke

I don't understand what you mean by:-

Just as a sad conclusion, Duncan, I sucked a lot with such things. Some people prefer to use single quotes in HTML tag attributes, other ones use different cases in attribute names.
Click to expand...


Kind Regards
Duncan
 
Upvote 0 Downvote
Hi

Duncan, your code is a good starting point, something like that was my first try too. But works well only in aseptic laboratory conditions and may fail parsing a HTML document written by others. My final conclusion was that is safer to use two regular expressions.

Feherke.
 
Upvote 0 Downvote
Kev - check the usernames - the OP who asked for a non-module solution has disappeared, hence the HTML::TokeParser::Simple recommendation - Gotcha!! :)
 
Upvote 0 Downvote
redstarfcs1

I have spent a bit more time looking into this for you and have come up with this:-

Code: 
[b]#!/usr/bin/perl[/b]

undef $/;
$_ = <DATA>;
$/ = "\n";

@metatags = m/(<meta[^>]+>)/ig;

foreach (@metatags) {
  if (s/(name|content)=['"]?([^'"]+)['"]? (name|content)=['"]?([^'"]+)['"]?/if(lc($1)eq"name"){$hash{name}=$2;$hash{content}=$4}else{$hash{name}=$4;$hash{content}=$2}/ie) {
    print $hash{name} . "\t" . $hash{content} . "\n"; 
  }
}

[blue]__DATA__
<META Name="name1" Content='Everything you wanted to know about stamps, from prices to history.'>
blah blah blah
<META Content="Everything you wanted to know about stamps, from prices to history." Name='name2'>
blah blah
<META NAME='name3' CONTENT="stamps, stamp collecting, stamp history, prices, stamps for sale">
blah blah blah blah
<META CONTENT='stamps, stamp collecting, stamp history, prices, stamps for sale' NAME="name4">[/blue]


Kind Regards
Duncan
 
Upvote 0 Downvote
the solution above is far from perfect - but at least it does not matter in which order the name/content are

this could be adapted to be more flexible


Kind Regards
Duncan
 
Upvote 0 Downvote
this is a bit better:-

Code: 
[b]#!/usr/bin/perl[/b]

undef $/;
$_ = <DATA>;
$/ = "\n";

@metatags = m/(<meta[^>]+>)/ig;

foreach (@metatags) {
  if (s/(name|content)=['"]?([^'"]+)['"]? (name|content)=['"]?([^'"]+)['"]?/$hash{lc($1)}=$2;$hash{lc($3)}=$4/ie) {
    print $hash{name} . "\t" . $hash{content} . "\n"; 
  }
}

[blue]__DATA__
<META Name="name1" Content='Everything you wanted to know about stamps, from prices to history.'>
blah blah blah
<META Content="Everything you wanted to know about stamps, from prices to history." Name='name2'>
blah blah
<META NAME='name3' CONTENT="stamps, stamp collecting, stamp history, prices, stamps for sale">
blah blah blah blah
<META CONTENT='stamps, stamp collecting, stamp history, prices, stamps for sale' NAME="name4">[/blue]


Kind Regards
Duncan
 
Upvote 0 Downvote
Hi

Much better Duncan, I like the idea. But there are... details to solve ( I hope this was not rude ) :
[ul]
[li][tt][highlight #eeeeff]<meta name="n" [red]x="x"[/red] content="c">[/highlight][/tt] - anything else between the two captured attribute[/li]
[li][tt][highlight #eeeeff]<meta name="n" content="c[red]'[/red]mon">[/highlight][/tt] - single quote in the value[/li]
[li][tt][highlight #eeeeff]<meta name="n"
content="c">[/highlight][/tt] - wrapped to multiple lines[/li]
[/ul]


Feherke.
 
Upvote 0 Downvote
Status
Not open for further replies.

Similar threads

ljsmith91
  • Locked
  • Question
Reading a Data Hash created via XML::Simple 2
Replies
6
Views
447
rharsh
rharsh
PinkeyNBrain
  • Locked
  • Question
Reading Event Logs in Win2008 2
Replies
1
Views
155
PinkeyNBrain
PinkeyNBrain
martinjburgess
  • Locked
  • Question
Reading a file
Replies
2
Views
155
martinjburgess
martinjburgess
spewn
  • Locked
  • Question
reading an xml file with GGI::Simple, with nested name/value pairs...
Replies
1
Views
180
Zhris
Zhris
steady12345
  • Locked
  • Question
Reading array location
Replies
3
Views
152
steady12345
steady12345

Part and Inventory Search

Sponsor

Back
Top