Reverse String

4. October 2012

Often there are more than one way to do things, even there may be a prefered one. Concerning string reversion, I look at three methods regarding their speed.
A simple way to reverse a given string is the following:

>>> s = 'Reverse' 
>>> s[::-1] 

There is also the possibility to do it with a generator:

>>> s = 'Teststring'
>>> ''.join([char for char in reversed(s)]) #generator

We could also use the shorter form:

>>> s = 'Some other String'
>>> ''.join(reversed(s))
'gnirtS rehto emoS'


If we compare the speed of the three methods in seconds (y-axis), we see that with increasing string length (x-axis), the second and third methods become much slower:

So it’s probably better to use the shorter s[::-1] notation. Note that I just tested this up to strings with 250 characters, as I couldn’t imagine any use case with very big strings.


The short method will work for Unicode too. However, remember to declare the string as unicode string:

>>> s = 'τεσσαρών' # wrong declaration
>>> print(s[::-1])
>>> s = u'τεσσαρών' # correct
>>> print(s[::-1])

An example with combined unicode characters can be found on Rosetta Code.

Code for timing (Python 2.7):

import random
import time
import string
import matplotlib.pyplot as plt

def reverse1(s):
    return s[::-1]

def reverse2(s):
    return ''.join([char for char in reversed(s)])

def reverse3(s):
    return ''.join(reversed(s))

def main():
    t1 = list()
    t2 = list()
    t3 = list()
    for str_len in range(5, 250):
        l = list()
        chars = string.letters + string.digits
        #generate strings
        for i in xrange(0, 100):
            l.append(''.join([random.choice(chars) for i in xrange(str_len)]))

        start = time.clock() #time.time()
        l1 = list()
        for s in l:

        start = time.clock()
        l2 = list()
        for s in l:

        start = time.clock()
        l3 = list()
        for s in l:

    plt.plot(t2, label="generator")
    plt.plot(t3, label="join(reversed(s))")
    plt.plot(t1, label="s[::-1]")
    ax = plt.subplot(111)
    ax.legend(loc='upper center', bbox_to_anchor=(0.5, 1.05), ncol=1, fancybox=True, shadow=True)

if __name__ == "__main__":

Thanks for reading!


Python Algorithms