Replied inline...

On Friday, 13 January 2017 12:44:03 UTC+1, Michi Amsler wrote:
>
> given the case, that a word has no child at all, I would also like the 
> context to be true;
>
> - when I use the NOT keyword, the TESTFLAG is not added
> - when I use the NEGATE keyword, the TESTFLAG is set
> - also if NONE is used, the TESTFLAG is set
>
> it seems to me that if there is no child, the "NOT c MEMBEROFX" contextual 
> test seems to not yield true while NEGATED / NONE does? Why is that the 
> case?
>


I think this was a quirk that got blessed into design. We wanted to support 
the case of asking "there exists a child which does not match X" (NOT c X) 
separately from "any potential child must not match X" (NONE c X). I don't 
recall the exact debate.

Analogously, the C and ALL tests mean "there exists a child in which all 
readings match X" (cC X) versus "all children must match X" (ALL c X) 
versus "all readings in all children must match X" (ALL cC X).

The problem is more that NOT also handles the missing case for topological 
scanning tests, e.g. (NOT -1* X) succeeds if there are no cohorts to the 
left, so the behavior for dependency is slightly different. I wish I could 
turn NOT into only affecting set matching (basically make it (*) - X), but 
there's too much history to do that.

-- Tino Didriksen

-- 
You received this message because you are subscribed to the Google Groups 
"Constraint Grammar" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To post to this group, send email to [email protected].
Visit this group at https://groups.google.com/group/constraint-grammar.
For more options, visit https://groups.google.com/d/optout.

Reply via email to