Entity in utterance is not extracted if begins in single dot

I have a case where an entity within an utterance is not being extracted. The utterance looks like this: “Lookup example history .abc”, and the entity value is “.abc”.

The language training utterances are there. For example, I have all of these:
Lookup example history .a
Lookup example history .abc .def .ghi
Lookup example history .123abc
Lookup example history .abc123

And they’re all mapped to the correct entity.

However, if I try “Lookup example history .12”, or “Lookup example history …abc”, then it works.

So it seems that if it starts with a single dot and a letter it fails.

Does anyone know why?

Hey Chad, I want to test your scenario to see if I can replicate. Can you tell me the entity type you’re using (string?) and the channel (voice or chat)?