Re: New work-queue item: Conditional inclusion

Earlier I wrote:

> [concerning conditional marked sections]
>
> [...] if there are N possible conditions, it's
> necessary to validate source documents 2**N times, once for
> each possible combination of INCLUDE/IGNORE.  This becomes
> impractical at about N=3.

Arjun Ray pointed out that well-formedness _can_ in fact
be checked with a single pass in the presence of conditional
marked sections.  If element boundaries cannot cross
marked section boundaries, then changing an INCLUDE MS
to an IGNORE MS cannot make a WF document non-WF.

A further simple restriction on the content and context of
conditional MSs can also make one-pass validation tractable.


--Joe English

  jenglish@crl.com

Received on Tuesday, 10 June 1997 15:50:24 UTC