Re: Table Issue

[Mike Sangrey <mike sojurn lns pa us>]

>> Ali Abdin said:
>>> The problem is because there are no characters in between the <entry>
>>> tag the charactersSAXFunc never gets called so I can not print out a
>>> &nbsp;
>> 
>>> What you will need to do instead is: <entry>&nbsp;</entry>
  
> Mike Sangrey said:
>> Couldn't we have a call back which would be called upon an empty
>> tag. Seems to me this would be generally useful in the SAX parser for
>> error handling.

Daniel.Veillard@w3.org said:
>   As sent previously to Ali, I don't suggest to do that. Keeping the
> original form of the documentation as clean as possible is quite more
> important than "nice rendering" for one of the generated formats.
> Hacking the generator to output &nbsp; is a hack too but located in
> better place. 

Apparently I misunderstood.  I thought Ali's suggestion was to
force everyone to code their XML with <entry>&nbsp;</entry> tags.
My suggestion was to give the generator a chance to solve the problem
by adding a (*EmptyTagCallBack)() to libxml.  The idea was for the SAX
parser to check whether there was a callback installed, and if there was,
to call it and let the generator solve the problem.  In other words,
hacking the html generator was what I was suggesting, too.

I also thought having that callback would allow for other tools which
use libxml some control over empty tags when using SAX.  It just seemed
like a general solution with broad utility.  Designing a DTD so that
empty tags can't happen would not be generally useful (come to think
of it, I'm not even sure that's possible); tools will need the smarts.
Without that callback, the code which uses the SAX parser and needs to
insist on instances of non-empty tags gets ugly.

Sorry, Ali, that I misunderstood.