mirror of
https://github.com/explosion/spaCy.git
synced 2024-12-27 18:36:36 +03:00
16 lines
383 B
Python
16 lines
383 B
Python
from ..tokenizer_exceptions import BASE_EXCEPTIONS
|
|
from ...symbols import ORTH
|
|
from ...util import update_exc
|
|
|
|
|
|
_exc = {}
|
|
|
|
for orth in ["n-tosios", "?!"]:
|
|
_exc[orth] = [{ORTH: orth}]
|
|
|
|
mod_base_exceptions = {
|
|
exc: val for exc, val in BASE_EXCEPTIONS.items() if not exc.endswith(".")
|
|
}
|
|
del mod_base_exceptions["8)"]
|
|
TOKENIZER_EXCEPTIONS = update_exc(mod_base_exceptions, _exc)
|