Bug in REMCOHORT

[Paul Meurer]

Hi,

I think I discovered a bug in the REMCOHORT command.

A minimal example:

---
# cat remcohort-bug.vrt | vislcg3 --grammar remcohort-bug.cg3 --trace > remcohort-bug.dis

STRICT-TAGS += Noun Adj >AMOD ;

# this rule is triggered on "<2>", but not on "<4>".
WITH NOPARENT Adj
IF (-1 Noun) {
MAP >AMOD (*) ;
SETPARENT (*) TO (jC1 (*)) ;
REMCOHORT IGNORED WITHCHILD (*) (*) ;
} ;
---

Input:

---
"<1>"
"1" Noun
"<2>"
"2" Adj
"<3>"
"3" Noun
"<4>"
"4" Adj
---

Output:


---
"<1>"
"1" Noun #1->1
"<2>"
"2" Adj >AMOD #2->1 WITH:6 MAP:8 SETPARENT:9 REMCOHORT:10
"<3>"
"3" Noun #3->3
"<4>"
"4" Adj #4->4
---

As you can see, the rule is triggered on "<2>", but not on "<4>". When I remove the REMCOHORT line, it triggers on both cohorts, but of course I do need REMCOHORT.

Btw, an unrelated minor bug: NOPARENT does not seem to have an effect. I just replace it by (NOT p (*)).

- Paul Meurer

[Tino Didriksen]

RemCohort Ignored fixed in v1.5.5 (2e4b2b8a).

-- Tino Didriksen