Re: Validation of new document based on cnxml

Le 6 mai 2005 à 01:40, David Malcolm a écrit :

> On Wed, 2004-08-04 at 16:56 +0200, Michèle Garoche wrote:
>
>> Le 4 août 2004, à 15:29, Geert Stappers a écrit :
>>
>>> On Tue, Aug 03, 2004 at 03:48:42PM +0200, Mich?le Garoche wrote:
>>>
>>>> Here's a proposal for the contents, which validates:
>>>>
>>>> <?xml version="1.0"?>
>>>> <!DOCTYPE document SYSTEM
>>>> "http://cnx.rice.edu/cnxml/0.5/DTD/cnxml_plain.dtd" [
>>>> ]>
>>>> <document id="m0000" xmlns="http://cnx.rice.edu/cnxml"><name  
>>>> >Unknown
>>>> name</name><content><para id="unknown">This is an unknown
>>>> paragrah.</para></content></document>
>>>>
>>>> 1 - New doctype (type and uri)
>>>> 2 - document tag required with the specific id
>>>> 3 - name tag required
>>>> 4 - content tag required
>>>> 5 - para with id required
>>>>
>>>>
>>>
>>> And how do you see this related to the conglomerate.org project?
>>>
>> Well, it all depends on whether you drop the random part or not. I've
>> assumed (maybe incorrectly) that you want this random part to work.
>>
>> When it was put in conglomerate, cnxml was at version 0.4. Now, it is
>> at version 0.5, the uri has changed, the root level element too, as
>> well as some tags.
>>
>> 1 - So, now, all cnxml documents should have this doctype:
>>
>> <!DOCTYPE document SYSTEM
>> "http://cnx.rice.edu/cnxml/0.5/DTD/cnxml_plain.dtd" []>
>>
>> 2 - All cnxml documents should have a root level element. This root
>> level element is document with a required fixed id. This required  
>> id is
>> the one given by rice.edu, so I've assumed here too, that this was
>> given to conglomerate at the beginning, since it works (I may also be
>> wrong here). Here it is:
>>
>> document id="m0000"
>>
>> The xmlns should be now:
>>
>> http://cns.rice.edu/cnxml
>>
>> So that the whole document tag is:
>>
>> <document id="m0000" xmlns="http://cns.rice.edu/cnxml">
>>
>> 3 - All cnxml documents should have a name tag, which is just plain
>> text.
>>
>> <name>blah blah blah</name>
>>
>> 4 - All cnxml documents should have content tags. The real text is  
>> put
>> inside the beginning and end content tag.
>>
>> 5 - For all the tags used inside content, each tag should have a  
>> unique
>> id.
>>
>> With all those requirements, a minimalist cnxml document is the one I
>> gave.
>>
>> Tomorrow I'll create a more complex cnxml document which could  
>> serve as
>> template, because this one I gave is really too minimalist to give a
>> feeling of what cnxml can do.
>>
>
> (Coming back months later)
> Did any of this make it into CVS (or bugzilla)?
No you rejected the patch as you wanted that someone from the cnxml  
team confirmed it.

Cheers,
Michèle
<http://micmacfr.homeunix.org>