[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message]

Re: hurry GenX...

• To: Mike Fitzgerald <mikefz@w...>
• Subject: Re: hurry GenX...
• From: Tim Bray <tbray@t...>
• Date: Thu, 18 Mar 2004 10:55:07 -0800
• Cc: "xml-dev@l..." <xml-dev@l...>
• In-reply-to: <4059DBD9.7020601@w...>
• References: <4059DBD9.7020601@w...>

On Mar 18, 2004, at 9:26 AM, Mike Fitzgerald wrote:

> Hope Tim finishes up his latest changes to GenX by Friday. He's going 
> to get real busy after that. %^}

I've been at Sun since Monday.  I was hacking on Genx last night.  
Since you brought it up, here's the state of play.

David Tolpin & I did some work on running the whole Canonical XML test 
suite through it, which would be a damn good test IMHO.  Unfortunately 
Genx couldn't handle it, because I'd taken a puritan view of namespaces 
(only one prefix-ns binding allowed, no defaults) but the c14n test 
suite is full of things like

<a xmlns:foo="http://a.b">
<b xmlns:bar="http://a.b">
</b>
</a>

Upon consideration, I decided that the testing and confidence level 
we'll get from passing the C14n suite is of sufficient value that 
making Genx a little more complex is is worth the effort.  I'm about 
halfway there, I'll post an interim snapshot today that has all the 
methods implemented but without the full namespace declaration nesting 
as above.

I decided that Elliotte was right and nuked the addAllAttributes() call 
and the addQName() call, there's a genxGetNamespacePrefix() and if you 
wanna put QNames in content, build 'em yourself.

BTW, the copyright going forward will say

Copyright (c) Tim Bray and Sun Microsystems

but the license won't change -Tim

smime.p7s

• References:
  • hurry GenX...
    • From: Mike Fitzgerald <mikefz@w...>

• Prev by Date: RE: hurry GenX...
• Next by Date: Questions on XQuery operators
• Previous by thread: hurry GenX...
• Next by thread: RSS beyond the Blog: 1992 or 1999? - was Re: hurry GenX...
• Index(es):
  • Date
  • Thread