0

I'm starting with Perl and I'm trying to do a script that logs me out of a page (I already did the login one). The idea is, I'm trying to use find_link function from WWW::Mechanize (I also tried WWW::Mechanize::Firefox) to find the logout link, but it's not working. So, my question is:

is the event onmouseover avoiding that link to be found? I think WWW::Mechanize::Firefox supports javascript but maybe that's not correct.

<a class="hb2" href="/goform/Logout?logout=Logout" onmouseover="(window.status='Logout');
return true">Logout</a>

Is there a way to do what I want? Any help or suggestions will be highly appreciated.

Serge Belov
  • 5,633
  • 1
  • 31
  • 40
Malincy Montoya
  • 87
  • 1
  • 13
  • `WWW::Mechanize::Firefox` supports Javascript. What happens if you use it? – choroba Dec 05 '12 at 23:48
  • The javascript `onmouseover` should not interfere with finding the link using plain `WWW::Mechanize`. A line like `my $logout_link = $mech->find_link( text => 'Logout' );` should work fine. Perhaps post your code so we can examine it more closely? – sjs Dec 06 '12 at 00:19

3 Answers3

1

Apparently my MozRepl (needed to use WWW::Mechanize::Firefox) add-on was not well installed. @gangabass, @andy I checked the HTML and it seemed to be in the source. So @choroba I checked why WWW::Mechanize::Firefox was acting the same as WWW::Mechanize and I reinstalled the add-on. Then everything worked perfect. Thanks a lot to all of you for taking time to help me.

Malincy Montoya
  • 87
  • 1
  • 13
  • When you say "in the source", at what point did you look at the source? From the browser after the page was loaded? If so, then that page could have been generated by JavaScript. What you really want to look at is the raw HTML content before any browser touches it. – Andy Lester Dec 06 '12 at 16:09
  • Well yes, I took it from the browser because I had no access to the raw HTML. Now I do, and I found out it's ASP generated. The line is `AspFormTag("/goform/Logout")` – Malincy Montoya Dec 06 '12 at 20:40
0

Are you sure you have this HTML fragment in your source HTML? I mean your HTML code fragment may be created dynamically via Javascript or may be you have frames in your source HTML.

So please show output of $mech->content() and your code to process this link (as stated above $mech->find_link( text => 'Logout' ); should work).

gangabass
  • 10,607
  • 2
  • 23
  • 35
0

Have you dumped the page with the mech-dump command line utility? Try that. If your link shows up in the output of mech-dump then you're calling find_all_links() wrong.

If mech-dump does not show your link, then you probably either have malformed HTML, or the HTML is generated in the browser with JavaScript.

Andy Lester
  • 91,102
  • 13
  • 100
  • 152
  • No, I haven't. My page is https and needs authentication, and I'm not sure how to call *mech-dump* from a script. I'll look for it – Malincy Montoya Dec 06 '12 at 15:20
  • If "from a script" you mean from inside your Perl program, then just call the `->dump_links()` method on your `$mech` object. And if you haven't heard of the `->dump_links()` method, then I suggest you read the entire documentation page for WWW::Mechanize because there are many little handy tricks (`dump_forms`, `save_content`, etc) like that in there. – Andy Lester Dec 06 '12 at 15:22
  • 1
    To compensate for malformed HTML I have occasionally found the need to do something like: $mech->update_html( html_fix_function( $mech->content() ) ); – ddoxey Dec 06 '12 at 16:05
  • 1
    @ddoxey: And that's why `->update_html()` exists. At least, that's the reason that made me create `->update_html()` in the first place. There are other uses for it especially in automated testing. – Andy Lester Dec 06 '12 at 16:11