From: Ramiro Diaz Trepat <ramirodt@...>
Date: 2007-12-15T14:48:01+09:00
Subject: Re: [noob] Parsing problems using https and redirects

------=_Part_4662_26078902.1197697683893
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable
Content-Disposition: inline

Thank you very much Konrad, it seems that I am on my way now.
The only weird thing that happened now with Mechanize is that it all works
perfectly on my Linux but it doesn=B4t on my Mac/Leopard.
Both have Ruby 1.8.6

On the mac I get the following error while trying to execute the first
Mechanize example:

./mechanize.rb:4: uninitialized constant WWW (NameError)
    from /Library/Ruby/Site/1.8/rubygems/custom_require.rb:27:in
`gem_original_require'
    from /Library/Ruby/Site/1.8/rubygems/custom_require.rb:27:in `require'
    from goog.rb:2


and the code is the first example of machanize:

require 'rubygems'
require 'mechanize'

agent =3D WWW::Mechanize.new
agent.user_agent_alias =3D 'Mac Safari'
page =3D agent.get("http://www.google.com/")
search_form =3D page.forms.with.name("f").first
search_form.q =3D "Hello"
search_results =3D agent.submit(search_form)
puts search_results.body


I really don't know why this constant is uninitialized and how could I
initialize it.  Besides it worries my that on Linux, after installing the
mechanize gem, everything worked out of the box.

Thanks again




On Dec 14, 2007 10:14 PM, Konrad Meyer <konrad@tylerc.org> wrote:

> Quoth Ramiro Diaz Trepat:
> > Hello list,
> >    I have to develop a simple script to parse some parts of a web site
> and I
> > thought it could be a good opportunity to start trying Ruby.
> >    I found that there are two network libraries that I could supposedly
> use
> > to retrieve the contents of the web site: open-uri and net-http.
> >
> > *First problem*
> >    This web site is accessed only with https and has a self issued
> > certificate.  This has made it impossible so far for me to access the
> > contents of the web site.
> >    Simple examples from the Hpricot html parsing library like this one:
> >
> > require 'hpricot'
> > require 'open-uri'
> > doc =3D Hpricot(open("https://xxxxxx"))
> >
> >    will not work because the open will fail because of problems due to
> > https.
> >
> > *Second problem*
> >    I need to know also how to handle redirection and cookies.  But to b=
e
> > fair, I still can do some further reading myself on these issues.
> >
> >    Thank you very much.
>
> 2) Look at mechanize.
>
> 1) Look at http-access2 (or whatever it's been renamed to).
>
> Regards,
> --
> Konrad Meyer <konrad@tylerc.org> http://konrad.sobertillnoon.com/
>

------=_Part_4662_26078902.1197697683893--