spacy tokenizer: is there a way to use regex as a key in custom exceptions for update_exc - spacy

Question

It is possible to add custom exceptions to spacy tokenizer. And these exceptions work fine.
However, as far as I know, it's possible to use only strings as keys to match for those exceptions. It's done this way:
import spacy
from spacy.tokens import Doc, Span, Token
from spacy.util import update_exc
from spacy.lang.tokenizer_exceptions import BASE_EXCEPTIONS
from spacy.symbols import ORTH, NORM, LEMMA, POS, TAG
CUSTOM_EXCEPTIONS = {
# prevent '3g' to be splitted into ['3', 'g']
"3g": [{ORTH: "3g", LEMMA: "3g"}],
}
spacy.lang.tokenizer_exceptions.BASE_EXCEPTIONS = update_exc(BASE_EXCEPTIONS, CUSTOM_EXCEPTIONS)
Is there a way to add an regexp-keyed exception, to, say, match phone number?
Something like this(highlighted in bold):
CUSTOM_EXCEPTIONS = {
# prevent '3g' to be splitted into ['3', 'g']
"3g": [{ORTH: "3g", LEMMA: "3g"}],
r'([\(]?\+[\(]?\d{2}[\)]?[ ]?\d{2} \d{2} \d{2} \d{2})': [{LEMMA: match_result} for match_result in match_results]
}
The only clue I found is:
https://github.com/explosion/spaCy/issues/840
In that revision of tokenizer_exceptions.py there was some way to use regexps as keys for tokenizer exceptions(however, I haven't found any examples to do so)
But in current revisions, at least initial analysis hasn't shown any ways to do s
So is there a way to solve this task?
(input: regex as a key for exception, output - phone numbers with spaces inside)

Answer 1

No, there's no way to have regular expressions as tokenizer exceptions. The tokenizer only looks for exceptions as exact string matches, mainly for reasons of speed. The other difficulty for this kind of example is that tokenizer exceptions currently can't contain spaces. (Support for spaces is planned for a future version of spacy, but not regexes, which would still be too slow.)
I think the best way to do this would be to add a custom pipeline component at the beginning of the pipeline that retokenizes the document with the retokenizer: https://spacy.io/api/doc#retokenize. You can provide any required attributes like lemmas while retokenizing.