Python encoding/decoding problems

Question

How do I decode strings such as this one "weren\xe2\x80\x99t" back to the normal encoding.

So this word is actually weren't and not "weren\xe2\x80\x99t"? For example:

print "\xe2\x80\x9cThings"
string = "\xe2\x80\x9cThings"
print string.decode('utf-8')
print string.encode('ascii', 'ignore')

“Things
“Things
Things

But I actually want to get "Things.

or:

print "weren\xe2\x80\x99t"
string = "weren\xe2\x80\x99t"
print string.decode('utf-8')
print string.encode('ascii', 'ignore')

weren’t
weren’t
werent

But I actually want to get weren't.

How should i do this?

Solution 1

I mapped the most common strange chars so this is pretty much complete answer based on the Oliver W. answer.

This function is by no means ideal,but it is the best place to start with. There are more chars definitions:

http://utf8-chartable.de/unicode-utf8-table.pl?start=8192&number=128&utf8=string
http://www.utf8-chartable.de/unicode-utf8-table.pl?start=128&number=128&names=-&utf8=string-literal

...

def unicodetoascii(text):

    uni2ascii = {
            ord('\xe2\x80\x99'.decode('utf-8')): ord("'"),
            ord('\xe2\x80\x9c'.decode('utf-8')): ord('"'),
            ord('\xe2\x80\x9d'.decode('utf-8')): ord('"'),
            ord('\xe2\x80\x9e'.decode('utf-8')): ord('"'),
            ord('\xe2\x80\x9f'.decode('utf-8')): ord('"'),
            ord('\xc3\xa9'.decode('utf-8')): ord('e'),
            ord('\xe2\x80\x9c'.decode('utf-8')): ord('"'),
            ord('\xe2\x80\x93'.decode('utf-8')): ord('-'),
            ord('\xe2\x80\x92'.decode('utf-8')): ord('-'),
            ord('\xe2\x80\x94'.decode('utf-8')): ord('-'),
            ord('\xe2\x80\x94'.decode('utf-8')): ord('-'),
            ord('\xe2\x80\x98'.decode('utf-8')): ord("'"),
            ord('\xe2\x80\x9b'.decode('utf-8')): ord("'"),

            ord('\xe2\x80\x90'.decode('utf-8')): ord('-'),
            ord('\xe2\x80\x91'.decode('utf-8')): ord('-'),

            ord('\xe2\x80\xb2'.decode('utf-8')): ord("'"),
            ord('\xe2\x80\xb3'.decode('utf-8')): ord("'"),
            ord('\xe2\x80\xb4'.decode('utf-8')): ord("'"),
            ord('\xe2\x80\xb5'.decode('utf-8')): ord("'"),
            ord('\xe2\x80\xb6'.decode('utf-8')): ord("'"),
            ord('\xe2\x80\xb7'.decode('utf-8')): ord("'"),

            ord('\xe2\x81\xba'.decode('utf-8')): ord("+"),
            ord('\xe2\x81\xbb'.decode('utf-8')): ord("-"),
            ord('\xe2\x81\xbc'.decode('utf-8')): ord("="),
            ord('\xe2\x81\xbd'.decode('utf-8')): ord("("),
            ord('\xe2\x81\xbe'.decode('utf-8')): ord(")"),

                            }
    return text.decode('utf-8').translate(uni2ascii).encode('ascii')

print unicodetoascii("weren\xe2\x80\x99t")  

Solution 2

In Python 3 I would do it like this:

string = "\xe2\x80\x9cThings"
bytes_string = bytes(string, encoding="raw_unicode_escape")
happy_result = bytes_string.decode("utf-8", "strict")
print(happy_result)

No translation maps needed, just code :)

Solution 3

You should provide a translation map that maps unicode characters to other unicode characters (the latter should be within the ASCII range if you want to re-encode to it):

uni2ascii = {ord('\xe2\x80\x99'.decode('utf-8')): ord("'")}    
yourstring.decode('utf-8').translate(uni2ascii).encode('ascii')
print(yourstring)  # prints: "weren't"

Comments

[Brana]

I know that i can do this. But is there a ready map that can do this automatically?

[AKMalkadi]

I was looking for this answer!