How to split my strings with re.split?

Question

I'm beginning my adventure with regular expressions. I'm interested in splitting specially formatted strings. If a letter is not inside parentheses it should become a different element of output list. Letters grouped inside parentheses should be put together.

Samples:

my string => wanted list

"ab(hpl)x" => ['a', 'b', 'hpl', 'x']
"(pck)(kx)(sd)" => ['pck', 'kx', 'sd']
"(kx)kxx(kd)" => ['kx', 'k', 'x', 'x', 'kd']
"fghk" => ['f', 'g', 'h', 'k']

How can it be achieved with regular expressions and re.split? Thanks in advance for your help.

Andrew Clark · Accepted Answer · 2011-03-17 19:27:12Z

5

This cannot be done with re.split, as it would require splitting on zero length matches.

From http://docs.python.org/library/re.html#re.split:

Note that split will never split a string on an empty pattern match.

Here is an alternative:

re.findall(r'(\w+(?=\))|\w)', your_string)

And an example:

>>> for s in ("ab(hpl)x", "(pck)(kx)(sd)", "(kx)kxx(kd)", "fghk"):
...     print s, " => ", re.findall(r'(\w+(?=\))|\w)', s)
... 
ab(hpl)x  =>  ['a', 'b', 'hpl', 'x']
(pck)(kx)(sd)  =>  ['pck', 'kx', 'sd']
(kx)kxx(kd)  =>  ['kx', 'k', 'x', 'x', 'kd']
fghk  =>  ['f', 'g', 'h', 'k']

edited Mar 17, 2011 at 19:27

answered Mar 17, 2011 at 18:59

Andrew Clark

210k36 gold badges286 silver badges310 bronze badges

Sign up to request clarification or add additional context in comments.

3 Comments

Steven Rumbalski Over a year ago

@Maciej Ziarko: Please note that this answer, by using '\w' and no lookbehind assertion assumes that parenthesis will always be balanced and that you never have digits and underscores in your data. Given your test data this is a fair assumption, so +1.

Maciej Ziarko Over a year ago

Yes, they will always be balanced. Any other characters in my data are small letters. I like both your answers and I voted them up. BTW: Can you recommend any good regular expressions tutorial/book with nice examples?

Andrew Clark Over a year ago

I mostly used regular-expressions.info for learning, and rubular.com is very useful for quick testing on your regular expressions.

Steven Rumbalski · Accepted Answer · 2011-03-17 19:21:59Z

1

You want findall not split. Use this re: r'(?<=\()[a-z]+(?=\))|[a-z]', which works for all your test cases.

>>> test_cases = ["ab(hpl)x", "(pck)(kx)(sd)", "(kx)kxx(kd)", "fghk"]
>>> pat = re.compile(r'(?<=\()[a-z]+(?=\))|[a-z]')
>>> for test_case in test_cases:
...     print "%-13s  =>  %s" % (test_case, pat.findall(test_case))
...
ab(hpl)x       =>  ['a', 'b', 'hpl', 'x']
(pck)(kx)(sd)  =>  ['pck', 'kx', 'sd']
(kx)kxx(kd)    =>  ['kx', 'k', 'x', 'x', 'kd']
fghk           =>  ['f', 'g', 'h', 'k']

edit:

Replace [a-z] with \w if you want to match upper and lower case letters, numbers, and underscore. You can remove the lookbehind assertion (?<=\() if your parenthesis will never be unbalanced ("abc(def").

edited Mar 17, 2011 at 19:21

answered Mar 17, 2011 at 19:14

Steven Rumbalski

45.8k10 gold badges96 silver badges125 bronze badges

Collectives™ on Stack Overflow

How to split my strings with re.split?

2 Answers 2

3 Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

3 Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related