WWW::Mechanize reading HTML

[vit]

Dear Monks,
I tried to find how to feed WWW::Mechanize with HTML instead of URL without changing the module. One way is to read it from file like

my $url  = "file:///D:/webpage.html";

but then I need to indicate a root which I do not want because it is changing and also I need to write HTML to this file which is also not good.

Basically what I need is to parse HTML and get all links from there with some conditional filtering.

[afoken]

Basically what I need is to parse HTML and get all links from there with some conditional filtering.

Use a HTML parser, like WWW::Mechanize does. WWW::Mechanize uses HTML::Form and HTML::TokeParser. Involving WWW::Mechanize in analysing a HTML file is just nonsense.

Alexander

[vit]

Thanks a lot! I used HTML::TokeParser which inputs HTML. Works perfect.

[ikegami]

Relative to current work dir:

use URI::file qw( );

my $url = URI::file->new_abs("webpage.html");

Relative to script dir (assuming you didn't change work directory):

use Cwd       qw( realpath );
use URI::file qw( );

my $url = URI::file->new("webpage.html")->abs(realpath($0));

[Corion]

Have you looked at the WWW::Mechanize documentation? Especially the ->update_html method sounds like what you need.