Thoughts of a programmer

Python 2 and Unicode

If you don’t know what is Unicode, UTF-8 or UTF-16 or if need to refresh your memory on these topics, then before reading this article, check this one What everyone should know about Unicode.

I have divided this article in two parts, one for Python 2 and one for Python 3, since there are some differences in how each handles string and unicode in general. You can either read both of them or read just for the language you are interested in. This article is for Python 2. For Python 3, check out this article.

Python 2 has two built-in types for handling strings. One is the str type and the other one is unicode. All the string literals that you normally create in your python program like

    str_type_string = 'Some random string'

are the str type string literals. These are also called the byte strings as they are merely series of bytes. For example, if you do

    random_byte_string = 'abc\x23\x45\x80\x81'
    print random_byte_string
    print 'Number of bytes =',len(random_byte_string)

you would get the output as follows-

Number of bytes = 7

As you can see from the output above, length of this string is just the number of bytes in the string. One thing to note in the above example is that, the last two bytes are non ASCII characters since they are greater than \x7f but they are also counted as valid bytes. You can use any byte in the range from \x00 to \xff in the byte string.

Now to enter unicode characters, you have several options-

    unicode_string_literal = u'omega\u03a9'

This would give you a unicode type string.

    str_type_string = 'abcd'
    unicode_string = unicode(str_type_string)

By default, unicode function uses the python default encoding scheme which is usually ascii. You can get the python default encoding scheme of your system by

    import sys
    print sys.getdefaultencoding()

You should never change this default encoding since many other programs and modules rely on this default behaviour and may break if you change it to something else. You should instead pass your desired encoding as the argument to the unicode function.

    str_type_string = 'omega\xce\xa9'
    unicode_string = unicode(str_type_string, encoding='utf8')

In the example above, the encoding argument specifies the encoding of the input string given as argument to the unicode function. Here, I am passing ‘utf8’ as I am using \xce\xa9 which is the utf-8 encoding for the Ω character.

    str_type_string = 'omega\xce\xa9'
    unicode_string = str_type_string.decode('utf8')

Finally to convert the unicode string back into normal str or byte string, use the decode counterpart function encode.

    unicode_string = u'\u03a9'
    encoded_byte_string = unicode_string.encode('utf16')

Just pass in the encoding and it will give you the encoded byte string. On my machine it gave the ouput


\xfe\xff is the BOM character followed by the utf-16 encoding of the Ω character \x03\xa9, but in the reverse order since my machine is a little-endian machine. You may get different result on your machine depending upon your machine endianess.

Working with unicode in files

Now suppose, you want to read in a file which is encoded using UTF-8 or UTF-16. You can use the python open() function, but it doesn’t understand about encodings. It will just read the bytes of the file and dump it back to you. So, either you could write your own decoder for decoding those bytes to unicode or instead you can use nice open() function provided by the codecs module.

    import codecs
    unicode_file = 'unicode_test.txt'
    with open(unicode_file, 'r') as file:
        normal_byte_string =

    with, 'r', encoding='utf8') as file:
        unicode_string =

Similary, you can use the file.write() function to write a unicode string to a file and the codecs module will internally convert your string into the proper series of bytes and write to the disk.

General Strategy in working with Unicode

Generally when you have to work with Unicode, what you can do is accept the input from any source, it may be from the user, from a network, from a file, in any encoding, and convert all the data into unicode internally. Then you can safely work on that unicode data, because you know that all the data is in the same format. When the times comes to write or send the data, then convert it back to the original encoding.

Thank you for reading my article. Let me know if you liked my article or any other suggestions for me, in the comments section below. And please, feel free to share :)