Closed MichaelRawson closed 2 months ago
I like the initiative, but can't this lead to a name clash?
It's "namespaced" by putting _split
on the end. So unless we have other names of the form spN_split
, no.
Is there are chance these names could mix with any other term material? I am thinking of the evil user coming with a TPTP problem with a predicate sp5_split
(perhaps by coincidence) in the signature.
You are quite right - I thought it was just "already broken" and this wouldn't fix it (but would be considerably shorter) but I think addFreshFunction
does some linear search, which surprised me.
Let's close this and I'll keep this corner in mind when I get to look at some of my other gripes with Signature
.
There's a static
splPrefix
floating around inSplitter
, which is mirrored assplitPrefix
inInferenceStore
. It doesn't seem to do much and does not conform with the convention.Just compute this thing on-demand and conform to the convention.