python: How to use for loop in findall

Question

I am trying to use for loop to re.findall() in jupyter notebook. I want to extract all the sentences that contains 'California', 'Colorado', and 'Florida'. I can just write these.

import re

f =open("C:/Users/uib57309/Desktop/test.txt",mode='rt')
lines = f.read()
f.close()

re.findall(r"([^.]*?California[^.]*\.)",lines)

re.findall(r"([^.]*?Colorado[^.]*\.)",lines)

re.findall(r"([^.]*?Florida[^.]*\.)",lines)

But how can I shorten my code with for loop? I tried like these, but this seems to be wrong.

test_list = ['California', 'Colorado', 'Florida'] 

for i in test_list: 

     result = re.findall(r"([^.]*?i[^.]*\.)",lines)

print(result)

Answer 1

In your for loop, result is finding all searches with the literal "i" string character. Use the f-string (for 3.6+); string concatenation or formatting is okay too:

result = re.findall(f"([^.]*?{i}[^.]*\\.)", lines) # works in Python 3.6+

Answer 2

If you really want to do it in a clean way, you must use NLTK to separate sentences. Your code relies on the assumption that a period always separates sentences, but, in general, that is not true.

import nltk
import re

lines = "Hello, California! Hello, e.g., Florida? Bye Massachusetts"

states = ['California', 'Colorado', 'Florida'] 

# Create a regex from the list of states
states_re = re.compile("|".join(states)) 

results = [sent for sent in nltk.sent_tokenize(lines) \
           if states_re.search(sent)] # Check the condition
#['Hello, California!', 'Hello, e.g., Florida?']

Answer 3

you don't need a loop, just create a regex with "|".join

test_list = ['California', 'Colorado', 'Florida']
result = re.findall(r"([^.]*?{}[^.]*\.)".format("|".join(test_list)),lines)

and to make sure the words aren't sub-strings use word boundary (not really necessary with those particular words but for the general case it is. Then the expression uses one more wrapping with r \\b characters:

r"([^.]*?{}[^.]*\.)".format("|".join([r"\b{}\b".format(x) for x in test_list]))

Answer 4

Use word boundary for this task and also make a list to store.

result variable will be overwritten with each iteration of loop.

test_list = ['California', 'Colorado', 'Florida'] 
x = []

for i in test_list: 
    pattern = r"\b"+i+r"\b"
    result = re.findall(pattern,lines)
    x.append(result)

print(x)