Reverse string by word with Python

Reverse string by word is a very popular interview question. In python you can solve it easily with code like below.

def reverse_string_by_word(s):
    lst = s.split()  # split by blank space by default
    return ' '.join(lst[::-1])

s = 'Power of Love'
print reverse_string_by_word(s)
# Love of Power

s = 'Hello    World!'
print reverse_string_by_word(s)
# World! Hello

We can see above implementation is good but not enough, in 2nd string we are expecting the ! symbol should be reversed as well, and keep original blank spaces between words. (multiple spaces between Hello and World in the example)

print reverse_string_by_word(s)
# Expected: !World  Hello

To improve the solution, a better choice should be re module. You might want to take a look at re.split() method.

>>> import re
>>> s = 'Hello  World!'

>>> re.split(r'\s+', s)    # will discard blank spaces
['Hello', 'World!']

>>> re.split(r'(\s+)', s)  # will keep spaces as a group
['Hello', '  ', 'World!']

>>> s = '< Welcome to EF.COM! >'

>>> re.split(r'\s+', s)  # split by spaces
['<', 'Welcome', 'to', 'EF.COM!', '>']

>>> re.split(r'(\w+)', s)  # exactly split by word
['< ', 'Welcome', ' ', 'to', ' ', 'EF', '.', 'COM', '! >']

>>> re.split(r'(\s+|\w+)', s)  # split by space and word
['<', ' ', '', 'Welcome', '', ' ', '', 'to', '', ' ', '', 'EF', '.', 'COM', '!', ' ', '>']

>>> ''.join(re.split(r'(\s+|\w+)', s)[::-1])
'> !COM.EF to Welcome <'

>>> ''.join(re.split(r'(\s+)', s)[::-1])
'> EF.COM! to Welcome <'

>>> ''.join(re.split(r'(\w+)', s)[::-1])
'! >COM.EF to Welcome< '

If you would like to increase the readability a little bit, replacing list slicing to reversed() is a choice.

>>> ''.join(reversed(re.split(r'(\s+|\w+)', s)))
'> !COM.EF to Welcome <'

Bingo, so easy!