Re: [xml2rfc] No <t> in a <dd> !

[Carsten Bormann <[e-mail]>]

> On 2021-01-03, at 16:54, Roger Price <[e-mail]> wrote:
> 
> On Sun, 3 Jan 2021, Carsten Bormann wrote:
> 
>> On 2021-01-03, at 14:56, Roger Price <[e-mail]> wrote:
>>> 
>>> draft-rprice-UPS-management-protocol-00.rfc(81): Error: Element dd has extra
>>>  content: t, at /rfc/middle/section[1]/section[1]/dl/dd[1]/t
>>> draft-rprice-UPS-management-protocol-00.rfc(97): Error: Element dd has extra
>>>  content: dl, at /rfc/middle/section[1]/section[1]/dl/dd[3]/dl
>> 
>> The full story in v3.rnc is:
> 
> Thanks for the pointer to v3.rnc.  Can I take this as the true grammar for xml2rfc?

AFAIK, that is the grammar xml2rfc actually uses for checking the input for v3.

> Wow! XML makes hard work of it.  I guess its because the XML mixed content declaration restricts the use of #PCDATA.  Why does the v3.rnc definition use "text" rather than #PCDATA ?

What was called #PCDATA (“parsed character data”) in SGML, and thus in XML DTDs, is called “text” in Relax-NG.

>> So, once you have used a span-level item (plain text, spans like em and strong, etc.), you can’t move up to block-level items (t, dl).
>> Can you post an excerpt of what you have been trying to do?
> 
> My problem was that I wrote
> 
> <dl>
>  <dt>Pizza</dt>
>  <dd>Lots of flavours. Look at these
>      <dl>
>        <dt>Banana...</dt>
>        <dd>...</dd>
>      </dl>
>  </dd>
> </dl>
> 
> I fixed this by adding a <t> to turn the text into a block:
> 
> <dl>
>  <dt>Pizza</dt>
>  <dd><t>Lots of flavours. Look at these   <- I added a <t>
>      <dl>
>        <dt>Banana...</dt>
>        <dd>...</dd>
>      </dl>
>  </dd>
> </dl>

Exactly; you need to avoid what SGML called “mixed content”.

> xml2rfc accepts this but I found that nested <dl> are all presented at the same depth in HTML output, so I will change the way I mark up.

I can’t reproduce that — for me, they nest just fine, nicely progressing to the right.

Grüße, Carsten