Rails core - Jun 2006 - RawPostDataTest#test_put_with_empty_content_type_ignores_body failure

Hi !

I get a failure for
RawPostDataTest#test_put_with_empty_content_type_ignores_body at
r4498:

  2) Error:
test_put_with_empty_content_type_ignores_body(RawPostDataTest):
NoMethodError: undefined method `downcase' for nil:NilClass
   
./test/controller/../../lib/action_controller/cgi_ext/raw_post_data_fix.rb:39:in
`initialize_query'
    C:/ruby/lib/ruby/1.8/cgi.rb:2270:in `initialize'
    ./test/controller/raw_post_test.rb:58:in `cgi_params'
    ./test/controller/raw_post_test.rb:43:in
`test_put_with_empty_content_type_ignores_body'

If we look here, we see the Content-Type header is corrected for POST requests:

http://dev.rubyonrails.org/browser/trunk/actionpack/lib/action_controller/cgi_ext/raw_post_data_fix.rb#L17

Should we correct PUT requests too, or is it better to not assume the
value and correctly implement the test ?

I'm on Win32, if that makes any difference.

Thanks !
-- 
François Beausoleil
http://blog.teksol.info/

_______________________________________________
Rails-core mailing list
Rails-core@lists.rubyonrails.org
http://lists.rubyonrails.org/mailman/listinfo/rails-core

This has been like this for quite some time now on Win32. The problem is
that, when you save an empty string into an ENV variable, it is later
returned as nil. Since the whole core team is on OS X they don''t
experience
this and therefore it was never fixed.

4 weeks ago I submitted a patch for this in ticket #5276 <
http://dev.rubyonrails.org/ticket/5276>

The patch is simple, it just adds checking for nil in a certain condition
that previously assumed a value is a string.

On 6/28/06, Francois Beausoleil <francois.beausoleil@gmail.com>
wrote:
>
> Hi !
>
> I get a failure for
> RawPostDataTest#test_put_with_empty_content_type_ignores_body at

_______________________________________________
Rails-core mailing list
Rails-core@lists.rubyonrails.org
http://lists.rubyonrails.org/mailman/listinfo/rails-core

On Wed, Jun 28, 2006 at 03:37:38PM +0200, Mislav Marohni??
wrote:
>    This has been like this for quite some time now on Win32. The problem is
>    that, when you save an empty string into an ENV variable, it is later
>    returned as nil. Since the whole core team is on OS X they
don''t
>    experience this and therefore it was never fixed.
> 
>    4 weeks ago I submitted a patch for this in ticket #5276
>    <http://dev.rubyonrails.org/ticket/5276>
> 
>    The patch is simple, it just adds checking for nil in a certain
condition
>    that previously assumed a value is a string.
Applied. Thanks for the fix.

marcel
-- 
Marcel Molina Jr. <marcel@vernix.org>