[Jython-bugs] [ jython-Bugs-1252162 ] os.path.abspath does not accept unicode objects

[SourceForge.net <no...@so...>]

Bugs item #1252162, was opened at 2005-08-04 19:48
Message generated for change (Settings changed) made by fwierzbicki
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=112867&aid=1252162&group_id=12867

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: None
>Group: Deferred
Status: Open
Resolution: None
>Priority: 4
Submitted By: leouser (leouserz)
>Assigned to: Nobody/Anonymous (nobody)
Summary: os.path.abspath does not accept unicode objects

Initial Comment:
the problem:
os.path.abspath when fed a unicode object will not
recognize the argument as legal.  It only recognizes
objects of type str.

---------
Here is some text describing the problem:
---------
Hi Jython makers,
 
I guess Im reporting what appears to be a discrepency
between CPython and older versions of Jython in the
os.path.abspath method.  What is happening is that
jyleo in the leoGlobals module is making a call to this
method:
os.path.abspath
 
with a unicode object.
 
It explodes with a stack pointing to:
javapath __tostr line 27
and a message of
abspath() argument must be a string object, not
'unicode' object.
 
let stress at this point Im not looking for 'patch'
ideas for the work Im doing.  Ive got a patch in place.
 But I do believe this is a change in behavior.  Is not
a unicode object a string object in unicode disguise? 
If regular python handles this ok, and other jythons
have, shouldn't the correct behavior when faced with a
unicode object be to process the unicode object?
 
thanks in advance
leouser :)
-------------
here is a reply to the query:

> Is not a unicode object a string object in unicode
disguise?
Yes, in fact it is even worse than that: a Jython 'str'
object is a
wrapper around Java's unicode-based String, and a
Jython 'unicode'
object is implemented as a thin wrapper around the
jython 'str'
object.  We are going through these contortions in an
attempt to make
the CPython standard libraries and unit tests happier,
however it
looks like things are breaking down in this case.

> If regular python handles this ok, and other jythons
have, shouldn't 
the
> correct behavior when faced with a unicode object be
to process the 
unicode
> object?
Yes, that would be the correct behavior.  One of our
primary goals is
to be as consistent as we can sensibly be with CPython.
 Could you
file a bug report so this doesn't get lost?

Thanks,
Frank Wierzbicki


----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=112867&aid=1252162&group_id=12867