向SpaCy添加单复数组合

custom_vocab = {
    "cat": "cats",
    "dog": "dogs",
    "book": "books",
    # 添加更多的单复数形式
}

import spacy
from spacy.tokens import Token

nlp = spacy.load("en_core_web_sm")

def add_plural_forms(doc):
    for token in doc:
        if token.text.lower() in custom_vocab:
            plural_form = custom_vocab[token.text.lower()]
            token._.set("plural", plural_form)

Token.set_extension("plural", default=None, force=True)
nlp.add_pipe(add_plural_forms, last=True)

text = "I have a cat and a dog."
doc = nlp(text)

for token in doc:
    print(token.text, token._.plural)

I None
have None
a None
cat cats
and None
a None
dog dogs
. None