Open ceedubs opened 3 months ago
dependents
returns numbered arguments as Name
, and delete.term
(and all the delete.*
commands) expects HQSplit'
. The conversion should be (and appears to be) lossless.
Unfortunately, dependents foo; delete 1
always fails prior to the structured arguments work because of #4898, so it’s hard to tell if this particular problem existed before.
The issue seems to be that the Name
s returned by dependents
aren’t qualified enough for HQSplit'
-consuming commands to find them (understandable that we don‘t want delete
to delete some random term with the same name that happens to be found).
I don’t think this is related to the numbered args changes, because I don’t see anything relevant around where dependents
’ Name
s are produced.
I think this is another face of #1298 – if we had absolute (or hash-qualified) names produced, there would be no confusion/ambiguity when they’re presented to delete.term
.
Describe and demonstrate the bug
If you run
dependents
and then try todelete.term
with a numbered result it doesn't work (or at least not always).Below you can see that it works if I manually select the fully-qualified name.
Screenshots
Environment (please complete the following information):
ucm --version
335512e331339cd496db97b6d3bead1efc090f42Additional context
It's possible that this is a regression that was introduced at the same time that #5055 was, but I haven't verified that. cc @sellout