Python encoding question

Jean-Michel Pichavant jeanmichel at
Fri Feb 25 12:19:33 CET 2011

Marc Muehlfeld wrote:
> Hi,
> I'm doing my first steps with python and I have a problem with 
> understanding an encoding problem I have. My script:
> import os
> os.environ["NLS_LANG"] = "German_Germany.UTF8"
> import cx_Oracle
> connection = cx_Oracle.Connection("username/password at SID")
> cursor = connection.cursor()
> TEST = cursor.fetchone()
> print TEST[0]
> print TEST
> When I run this script It prints me:
> München
> ('M\xc3\xbcnchen',)
> Why is the Umlaut of TEST[0] printed and not from TEST?
> And why are both prints show the wrong encoding, when I switch 
> "fetchone()" to "fetchall()":
> ('M\xc3\xbcnchen',)
> [('M\xc3\xbcnchen',)]
> I'm running Python 2.4.3 on CentOS 5.
> Regards,
> Marc
Nothing related to encoding here. TEST[0] is a string, TEST is a tupple.

s1 = 'aline \n anotherline'

 > print str(s1)

 > print repr(s1)
'aline \n anotherline'

atuple = (s1,)
 > print str(atuple)
('aline \n anotherline',)

 > print repr(atuple)
('aline \n anotherline',)

Read regarding __repr__ 
and __str__.

Basically, __str__ and __repr__ are the same method for tuples, while it 
differs from each other for strings.
If you want a nice representation of tuple elements you have to do it 

print ', '.join([str(elem) for elem in atuple])

In a more general manner only strings will print nicely with carriage 
returns & UTF8 characters. Everyhing else, like tuple, lists, objects 
will using the __repr__ method which displays formal data.

PS :

 > class Foo:
    def __str__(self):
          return 'I am a nice representation of a Foo instance'

 > print Foo()
I am a nice representation of a Foo instance

 > print str(Foo())
I am a nice representation of a Foo instance

 > print repr(Foo())
<__main__.Foo instance at 0xb73a07ac>

More information about the Python-list mailing list