python re, find expression containing an optional group

Question

I have a regular expression that can have either from:

(src://path/to/foldernames canhave spaces/file.xzy)
(src://path/to/foldernames canhave spaces/file.xzy "optional string")

These expressions occur within a much longer string (they are not individual strings). I am having trouble matching both expressions when using re.search or re.findall (as there may be multiple expression in the string).

It's straightforward enough to match either individually but how can I go about matching either case so that two groups are returned, the first with src://path/... and the second with the optional string if it exists or None if not?

I am thinking that I need to somehow specify OR groups---for instance, consider:

The pattern \((.*)( ".*")\) matches the second instance but not the first because it does not contain "...".

r = re.search(r'\((.*)( ".*")\)', '(src://path/to/foldernames canhave spaces/file.xzy)'
r.groups()  # Nothing found
AttributeError: 'NoneType' object has no attribute 'groups'

While \((.*)( ".*")?\) matches the first group but does not individually identify the "optional string" as a group in the second instance.

r = re.search(r'\((.*)( ".*")?\)', '(src://path/to/foldernames canhave spaces/file.xzy "optional string")')
r.groups()
('src://path/to/foldernames canhave spaces/file.xzy "optional string"', None)

Any thoughts, ye' masters of expressions (of the regular variety)?

Answer 1

The simplest way is to make the first * non-greedy:

>>> import re
>>> string = "(src://path/to/foldernames canhave spaces/file.xzy)"
>>> string2 = \
... '(src://path/to/foldernames canhave spaces/file.xzy "optional string")'
>>> re.findall(r'\((.*?)( ".*")?\)', string2)
[('src://path/to/foldernames canhave spaces/file.xzy', ' "optional string"')]
>>> re.findall(r'\((.*?)( ".*")?\)', string)
[('src://path/to/foldernames canhave spaces/file.xzy', '')]

Answer 2

Since " aren't usually allowed to appear in file names, you can simply exclude them from the first group:

r = re.search(r'\(([^"]*)( ".*")?\)', input)

This is generally the preferred alternative to ungreedy repetition, because tends to be a lot more efficient. If your file names can actually contain quotes for some reason, then ungreedy repetition (as in agf's answer) is your best bet.