Extract specific DATE entity from an sentence by the use of spaCy, and calculate the relative time

Question

import spacy
import en_core_web_sm
nlp = en_core_web_sm.load()
doc = nlp('I get cough yesterday, and tomorrow I will go to hostipital')
for t in doc.ents:
    if t.label_ == 'DATE':
        print(t.text)

---

output: yesterday, tomorrow

but I want only 'yesterday' be extracted. how can I optimize my rule to get my expected result.

one more thing, if I already get 'yesterday' and 'Jul 26 2021', how can I get the result of 'Jul 25 2021'.

Sincerely, thanks for help.

Answer 1

You can do this using python datetime and timedelta. Convert the date to a datetime object:

import datetime
from datetime import timedelta

created_date = datetime.datetime.strptime('Jul 25 2021', '%b %d %Y')

Then check if it is "yesterday":

    if t.label_ == 'DATE':
         if t.text.lower() == 'yesterday':
              return_date = created_date + timedelta(days=-1)

    return return_date