I have a string in python that I want to split in a very particular manner. I want to split it into a list containing each separate word, except for the case when a group of words are bordered by a particular character. For example, the following strings would be split as such.

'Jimmy threw his ball through the window.'

becomes

['Jimmy', 'threw', 'his', 'ball', 'through', 'the', 'window.']

However, with a border character I'd want

'Jimmy |threw his ball| through the window.'

to become

['Jimmy', 'threw his ball', 'through', 'the', 'window.']

As an additional component I need - which may appear outside the grouping phrase to appear inside it after splitting up i.e.,

'Jimmy |threw his| ball -|through the| window.'

would become

['Jimmy', 'threw his', 'ball', '-through the', 'window.']

I cannot find a simple, pythonic way to do this without a lot of complicated for loops and if statements. Is there a simple way to handle something like this?

This isn't something with an out-of-the-box solution, but here's a function that's pretty Pythonic that should handle pretty much anything you throw at it.

def extract_groups(s):separator = re.compile("(-?\|[\w ]+\|)")components = separator.split(s)groups = []for component in components:component = component.strip()if len(component) == 0:continueelif component[0] in ['-', '|']:groups.append(component.replace('|', ''))else:groups.extend(component.split(' '))return groups

Using your examples:

>>> extract_groups('Jimmy threw his ball through the window.')
['Jimmy', 'threw', 'his', 'ball', 'through', 'the', 'window.']
>>> extract_groups('Jimmy |threw his ball| through the window.')
['Jimmy', 'threw his ball', 'through the', 'window.']
>>> extract_groups('Jimmy |threw his| ball -|through the| window.')
['Jimmy', 'threw his', 'ball', '-through the', 'window.']