can't find tokenizer rule that keeps "A.B.C." together but splits "a.b.c." into "a.b.c" and "."

[jefhil]

I want to treat upper and lower case "a.b.c." the same but can't figure out where the rule that splits/keeps is located
TIA

[jefhil]

from spacy import blank
nlp = blank("en")
ABC = nlp.tokenizer.explain("A.B.C.")
abc = nlp.tokenizer.explain("a.b.c.")
print(f"{ABC=} {abc=}")

--> ABC=[('TOKEN', 'A.B.C.')] abc=[('TOKEN', 'a.b.c'), ('SUFFIX', '.')]