Capture substring and send it to a function that modifies it and can replace it in this string

import re


def one_day_or_another_day_relative_to_a_date_func(input_text):
   #print(repr(input_text)) #print what you have captured, and you should replace
   return "aaaaaaaa"


def identify(input_text):
   some_text = r"(?:(?!.s*?n)[^;])*"

   date_capture_pattern = r"([12]d{3}-[01]d-[0-3]d)(D*?)"

   previous_days = r"(d+)s*(?:dias|dia)s*(?:antes|previos|previo|antes|atrás|atras)s*"
   after_days = r"(d+)s*(?:dias|dia)s*(?:después|despues|luego)s*"

   n_patterns = [
   previous_days + r"(?:del|des*el|de|al|a)s*" + some_text + date_capture_pattern + some_text + r"s*(?:,s*o|o)s*" + previous_days,
   after_days + r"(?:del|des*el|de|al|a)s*" + some_text + date_capture_pattern + some_text + r"s*(?:,s*o|o)s*" + previous_days,
   previous_days + r"(?:del|des*el|de|al|a)s*" + some_text + date_capture_pattern + some_text + r"s*(?:,s*o|o)s*" + after_days,
   after_days + r"(?:del|des*el|de|al|a)s*" + some_text + date_capture_pattern + some_text + r"s*(?:,s*o|o)s*" + after_days]

   #Itero la lista de patrones de búsqueda para que el programa intente con uno por uno
   for n_pattern in n_patterns:
       #Este es mi intento de realizar el reemplazo, aunque tiene problemas con modificadores non-greedy
       input_text = re.sub(n_pattern, one_day_or_another_day_relative_to_a_date_func , input_text, re.IGNORECASE)



input_texts = ["8 dias antes o 9 dias antes del 2022-12-22",
           "2 dias despues o 1 dia antes del 2022-12-22, dia en donde ocurrio",
           "a tan solo 2 dias despues de 2022-12-22 o a caso eran 6 dias despues, mmm no recuerdo bien",
           ]


#Testing...
for input_text in input_texts:
   #print(input_text)
   print(one_day_or_another_day_relative_to_a_date_func(input_text))

Incorrect output that I am getting, because if I incorrectly capture the substrings, the replacements will also be incorrect

"aaaaaaaa"
"aaaaaaaa"
"aaaaaaaa"

Having well-defined limits, I don’t understand why this capture pattern try to capture beyond them?

And the output that I need is that:

"aaaaaaaa"
"aaaaaaaa, dia en donde ocurrio"
"a tan solo aaaaaaaa, mmm no recuerdo bien"

Answer

There are several errors in your code, among which:

1. You are printing the result of the one_day_or_another_day_relative_to_a_date_func function. Print the result of identify instead.
2. In the identify function you are not returning the result text. Add return input_text at the end of it.
3. Make the “o…” suffix optional.
4. Use regex alternation instead of multiple patterns, otherwise you may get unexpected results.

Fixed code (I’ve also made it more compact):

import re


def one_day_or_another_day_relative_to_a_date_func(input_text):
   #print(repr(input_text)) #print what you have captured, and you should replace
   return "aaaaaaaa"


def identify(input_text):
   some_text = r"(?:(?!.s*?n)[^;])*"
   date_capture_pattern = r"([12]d{3}-[01]d-[0-3]d)(D*?)"
   previous_days = r"antes|previos|previo|antes|atrás|atras"
   after_days = r"después|despues|luego"
   prev_or_after = r"(d+)s*(?:dias|dia)s*(?:" + previous_days + "|" + after_days + ")s*"
   preposition = r"(?:del|des*el|de|al|a)s*"
   suffix = "(?:" + r"s*(?:,s*o|o)s*" + some_text + prev_or_after + ")?"
   pattern = prev_or_after + some_text + preposition + date_capture_pattern + suffix
   input_text = re.sub(pattern, one_day_or_another_day_relative_to_a_date_func , input_text, re.IGNORECASE)
   return input_text


input_texts = ["8 dias antes o 9 dias antes del 2022-12-22",
           "2 dias despues o 1 dia antes del 2022-12-22, dia en donde ocurrio",
           "a tan solo 2 dias despues de 2022-12-22 o a caso eran 6 dias despues, mmm no recuerdo bien",
           ]


#Testing...
for input_text in input_texts:
   #print(input_text)
   print(identify(input_text))

Result:

aaaaaaaa
aaaaaaaa, dia en donde ocurrio
a tan solo aaaaaaaa, mmm no recuerdo bien