Skip Menu |

Preferred bug tracker

Please visit the preferred bug tracker to report your issue.

This queue is for tickets about the WWW-Mechanize CPAN distribution.

Report information
The Basics
Id: 8092
Status: resolved
Priority: 0/
Queue: WWW-Mechanize

People
Owner: MARKSTOS [...] cpan.org
Requestors: rob.casey [...] bluebottle.com
Cc:
AdminCc:

Bug Information
Severity: Wishlist
Broken in: (no value)
Fixed in: (no value)



Date: Thu, 21 Oct 2004 21:25:41 -0500
From: Mark Stosberg <mark [...] summersault.com>
To: bug-WWW-Mechanize [...] rt.cpan.org
CC: open-source [...] bluebottle.com
Subject: wish: improved support for attributes in ::Link
This patch is stalled while this issue gets worked out: Rob, In general this looks good. It turns out that yesterday I also worked on Link.pm, adding support for an 'alt' attribute. This patch will have a merge conflict with my work, but that's not a big deal. I created an alternative interface for Link.pm, because the 6 element (now 7 with attrs) was unwieldly. So you can also call like this now: my $link = WWW::Mechanize::Link->new({                 url => $url,                 text => $text,                 name => $name,                 tag => $tag,                 base => $self->base,                 alt => $alt,                                 attrs => $attrs, # not yet handled             }); ############### Now considering your patch as well, I think we simplify this even further if we want. We can get all the data we want with parameters: my $link = WWW::Mechanize::Link->new({                                 parser_obj    => $p,                                 parser_token  => $token,                                          base                  => $self->base,                         }); To me, this makes more sense because it moves a maximal amount of the link munging logic into ::Link. There would still be the option to pass an explicit 7 element array if you want to.   What do others think about this design alternative?         Mark ----- Forwarded message from Rob Casey <open-source@bluebottle.com> ----- User-Agent: Microsoft-Entourage/10.1.4.030702.0 Date: Thu, 30 Sep 2004 22:31:04 +1000 Subject: Re: [mech-dev]Anchor onclick handling From: Rob Casey <open-source@bluebottle.com> To: Andy Lester <andy@petdance.com> CC: Mark Stosberg <mark@summersault.com>, www-mechanize-development@lists.sourceforge.net Show quoted text
----- End forwarded message ----- -- http://mark.stosberg.com/
On 29/9/04 11:40 PM, "Andy Lester" <andy@petdance.com> wrote: Show quoted text
> The idea of "consensus" is one of the biggest problems with open source. > We stall on doing something, awaiting approval from "everyone", whoever > that may be. We spend more time talking about the action to be taken > instead of JFDIing it. > > Yes, write up the patch, and if it makes sense to me to put in, then > I'll put it in. I'm the only "everyone" that matters anyway. :-)
Attached is a very simple patch which implements the attr behaviour as suggested by Mark. Comments and feedback are welcomed. Regards, Rob -- Rob Casey Technical Manager Bluebottle Solutions Pty. Ltd. http://www.bluebottle.com/

Message body is not shown because sender requested not to inline it.

I have now added a patch for an attrs method, based on this exchange. Because I already implemented 'alt'... I didn't directly use your patch, but the spirit of it is there. :) I still gave you credit in the Changes file for helping. :) This is an CVS now and should appear in the next release. Mark