Re: design question? or puzzle?

> On 11,Apr2021, at 7:19 PM, Liam R. E. Quin <liam@fromoldbooks.org> wrote:
> 
> On Sun, 2021-04-11 at 18:25 -0600, C. M. Sperberg-McQueen wrote:
>> 
>> Is there a way to write the ixml grammar so as to achieve the goal of
>> serializing a nonterminal as an element if and only if it has two
>> element children, and not to serialize it when it has only one
>> element child?
> 
> Although i don't have an answer to that, isn't it a virtue of ixml that
> we can use XSLT to achieve such a clearly-specifid transformation?

yes, I think so.

but if we push that argument a little harder, it seems to become an argument
for doing away with marks (serialization annotations) entirely and only ever
producing a raw parse tree.  I’d rather not do that, because I think it’s better if
the ixml parser produces output in the form that any downstream process 
would like to consume, or failing that some form that’s closer to what’s desired
than the raw parse tree will be.  

in principle, any notation makes some things easier and not other things, and
sometimes there are things a notation just cannot specify.  i’d like to understand
better the boundaries of the expressive power of the marks in ixml. 

perhaps something to discuss on tuesday.

Michael


********************************************
C. M. Sperberg-McQueen
Black Mesa Technologies LLC
cmsmcq@blackmesatech.com
http://www.blackmesatech.com
********************************************

Received on Monday, 12 April 2021 02:03:33 UTC