[microformats-discuss] funness -> validator

brian suda brian.suda at gmail.com
Tue Aug 16 15:17:59 PDT 2005


Lucas Gonze wrote:

>On 8/16/05, Tantek Çelik <tantek at cs.stanford.edu> wrote:
>  
>
>>On 8/16/05 11:59 AM, "Dr. Ernie Prabhakar" <drernie at opendarwin.org> wrote:
>>
>>    
>>
>>>On Aug 16, 2005, at 11:50 AM, Lucas Gonze wrote:
>>>      
>>>
>>>>One thing I'm finding as I go along is that a validator would be
>>>>intensely useful, because you can (and have to) get pretty creative,
>>>>and it's hard to be sure that the results are still parseable.
>>>>        
>>>>
>>>That's something I wondered about.  Is there a rigorous grammar for
>>>what is and is not a valid microformat, such that someone -could-
>>>write a validator?
>>>      
>>>
>>The short answer is no, we are still figuring it out.
>>    
>>
There was a discussion about some of the pros and cons previously on
this list
http://microformats.org/discuss/mail/microformats-discuss/2005-July/000328.html

The previous list posts go into more detail, but there are several
things that XMDP cannot do (yet). There is always and on going
discussion about this topic with new ideas poping-up every once and
awhile. I won't bore anyone unless they want an indepth description of
the current state of the XMDP validator and all the
issue that have slowed development.

A universal validator would be a golden egg. Once written it could look
parse any XMDP and validate against it without any prior knowledge of
the format. Currently, that is impossible, each microformat has some
sort of data that is unique in some fashion (which is NOT a bad thing,
but it makes validation more difficult).

This is the link to a BETA validator, the input/output is pretty ugly,
but the idea is that you enter a URL and it will be fetched, the <head
profile=""> parsed. Those URLs are fetched, the XMDPs are passed through
an XSLT and another dynamic-XSLT is created. Then the new Dynamic-XSLT
(which encompasses all the profiles you are using) is processed against
your page and the validation output is displayed.
http://suda.co.uk/projects/XMDP/
Currently, it only uses XFN and no-follow (partly because those
microformats are very flat, no nested elements)

Feedback is eggargly awaited.
-brian



More information about the microformats-discuss mailing list