Styjun

I'm trying to parse a repetitive string and find multiple matches from it that match the pattern in my findall() function.

Essentially what I want to do is extract the names of the players from the string. They're separated by commas and the last name of the player is preceded by an "and". I tried to get the comma part down but for some reason findall() doesn't seem to repeat the matching pattern even though I added a *.

x = re.findall('Players(?:s([A-Z]+[a-z]+)),*', 'Players Jeremiah, Lou, Martha and Kay,')

I haven't gotten the and part down yet since I'm stuck on the commas but from the above code, I feel like, x should capture Jeremiah, Lou and Martha at the very least. My code only manages to capture Jeremiah.

Your pattern starts with Players... hence it will only match once, because your string has only one Players in it.

This would be the correct pattern which also handles the and part:

• a player name is any character that is non-white space or comma,


• it must be followed either by a comma or white space + and
  

import re

x = re.findall(r'([^s,]+)(?:,|s+and)',
 'Players Jeremiah, Lou, Martha and Kay,')

print(x)

Test run:

$ python3 dummy.py
['Jeremiah', 'Lou', 'Martha', 'Kay']

Update to match further sample sentences given by OP a single regex is no longer enough:

• match a mandatory prefix Players and extract the rest


• match an optional postfix, e.g. are ..., and strip that


• detect player names in the remaining sentence
  
  
  • starts at a word boundary
  
  
  • starts with an uppercase letter
  
  
  • followed by one-or-more lowercase letters
  
  
  • ends at a word boundary
  
  

import re;

for input in (
 'Jeremiah, Lou, Martha and Kay,',
 'Players Jeremiah, Lou, Martha and Kay,',
 'The Players are Martha, Joe, Toby and Kay.',
 'The Players Martha, Joe and Toby are German.',
 'The Players Martha, Joe and Toby are German,',
 ):

 # strip mandatory prefix
 match = re.search('Players(.*)', input)
 if not match:
 continue
 #print(match[1])

 # strip optional postfix
 postfix = re.search('(.*)(?:s+(?:are)s+S+[,.])$', match[1])
 if postfix:
 match = postfix
 #print(match[1])

 result = re.findall(r'(b[A-Z][a-z]+b)', match[1])
 print(input, '->', result)

Test run:

$ python3 dummy.py
Players Jeremiah, Lou, Martha and Kay, -> ['Jeremiah', 'Lou', 'Martha', 'Kay']
The Players are Martha, Joe, Toby and Kay. -> ['Martha', 'Joe', 'Toby', 'Kay']
The Players Martha, Joe and Toby are German. -> ['Martha', 'Joe', 'Toby']
The Players Martha, Joe and Toby are German, -> ['Martha', 'Joe', 'Toby']

I guess you're basically looking for a pattern for proper nouns. In the pattern you're using, it only captures "Jeremiah" because your pattern 'Players(?:s([A-Z]+[a-z]+)),*' specifically looks for a proper noun after the word "Players".

Please try this pattern instead:
(?!^[A-Z]w+)(?<![.]s)([A-Z]w+)

It looks for proper nouns but exclude capitalized words at the beginning of the sentence.

([A-Z]w+) matches a capitalized word

(?<![.]s) ensures that we don't pick anything that follows a full-stop and a space

(?!^[A-Z]w+) leaves out capitalized words at the beginning of the string (in a new line for example)

Warning
A generalized pattern may not be ideal if you require 100% precision in your results. This pattern could potentially understate the number of names if your sentence begins with a name.

Test it out here