shutil.copy Problem

[David Nicolson]

Hi,

I wasn't exactly sure where to send this, I don't know if it is a bug
in Python or not. This is rare, but it has occurred a few times and
seems to be reproducible for those who experience it.

Examine this code:

>>try:
shutil.copy("/file.xml","/Volumes/External/file.xml")
except Exception, err:
print sys.exc_info()[0]
print err

This is the output:
exceptions.UnicodeDecodeError
'ascii' codec can't decode byte 0xd6 in position 26: ordinal not in
range(128)]

What could the possible cause of this be? Shouldn't shutil simply be
reading and writing the bytes and not character decoding them?

Cheers,
David

[Justin Ezequiel]

On Mar 27, 11:10 am, David Nicolson <[e-mail]>
wrote:

Hi,

I wasn't exactly sure where to send this, I don't know if it is a bug
in Python or not. This is rare, but it has occurred a few times and
seems to be reproducible for those who experience it.

Examine this code:

>>try:
>> shutil.copy("/file.xml","/Volumes/External/file.xml")
>>except Exception, err:
>> print sys.exc_info()[0]
>> print err

This is the output:
exceptions.UnicodeDecodeError
'ascii' codec can't decode byte 0xd6 in position 26: ordinal not in
range(128)]

What could the possible cause of this be? Shouldn't shutil simply be
reading and writing the bytes and not character decoding them?

Cheers,
David

what if you try it without the try...except?
perhaps the UnicodeDecodeError is in the print...

[David Nicolson]

Thanks, but it's definitely not the print. In original the code the
print statements are replaced by a call to a log method.

Besides, the exception would be different if it was thrown outside of
the try block.
On 27/03/2007, at 2:42 PM, Justin Ezequiel wrote:

On Mar 27, 11:10 am, David Nicolson <[e-mail]>
wrote:

>Hi,

I wasn't exactly sure where to send this, I don't know if it is a bug
in Python or not. This is rare, but it has occurred a few times and
seems to be reproducible for those who experience it.

Examine this code:

>>>>try:
shutil.copy("/file.xml","/Volumes/External/file.xml")
except Exception, err:
print sys.exc_info()[0]
print err

This is the output:
exceptions.UnicodeDecodeError
'ascii' codec can't decode byte 0xd6 in position 26: ordinal not in
range(128)]

What could the possible cause of this be? Shouldn't shutil simply be
reading and writing the bytes and not character decoding them?

Cheers,
David

what if you try it without the try...except?
perhaps the UnicodeDecodeError is in the print...

--
http://mail.python.org/mailman/listinfo/python-list

[Leo Kislov]

On Mar 26, 8:10 pm, David Nicolson <[e-mail]:

Hi,

I wasn't exactly sure where to send this, I don't know if it is a bug
in Python or not. This is rare, but it has occurred a few times and
seems to be reproducible for those who experience it.

Examine this code:

>>try:
>> shutil.copy("/file.xml","/Volumes/External/file.xml")
>>except Exception, err:
>> print sys.exc_info()[0]
>> print err

This is the output:
exceptions.UnicodeDecodeError
'ascii' codec can't decode byte 0xd6 in position 26: ordinal not in
range(128)]

What could the possible cause of this be?

Show us traceback, without it I doubt anyone can help.

Shouldn't shutil simply be
reading and writing the bytes and not character decoding them?

Yes, shutil.copy copies content verbatim.

-- Leo

[Marc 'BlackJack' Rintsch]

In <ma***************************************@python. org>, David Nicolson
wrote:

Hi,

I wasn't exactly sure where to send this, I don't know if it is a bug
in Python or not. This is rare, but it has occurred a few times and
seems to be reproducible for those who experience it.

Examine this code:

>>try:
>> shutil.copy("/file.xml","/Volumes/External/file.xml")
>>except Exception, err:
>> print sys.exc_info()[0]
>> print err

This is the output:
exceptions.UnicodeDecodeError
'ascii' codec can't decode byte 0xd6 in position 26: ordinal not in
range(128)]

What could the possible cause of this be? Shouldn't shutil simply be
reading and writing the bytes and not character decoding them?

Do you really get this error from the code above or just in the real
program? Are the paths hard coded in the real program too?

Ciao,
Marc 'BlackJack' Rintsch

[Facundo Batista]

David Nicolson wrote:

Thanks, but it's definitely not the print. In original the code the
print statements are replaced by a call to a log method.

Besides, the exception would be different if it was thrown outside of
the try block.

The best you can do is take the piece of code that has the problem, show
it to us, and then copy the traceback.

Regards,

--
.. Facundo
..
Blog: http://www.taniquetil.com.ar/plog/
PyAr: http://www.python.org/ar/

[John Nagle]

Facundo Batista wrote:

David Nicolson wrote:

>>Thanks, but it's definitely not the print. In original the code the
print statements are replaced by a call to a log method.

Besides, the exception would be different if it was thrown outside of
the try block.

The best you can do is take the piece of code that has the problem, show
it to us, and then copy the traceback.

Regards,

There may be some problem here with a file being recognized as Unicode
in binary mode. That shouldn't happen, but looking at the Windows
UNICODE support, it might not be impossible.

Information needed:

- Platform (Windows, Linux, ...)
- Python version
- A hex dump of the first few bytes of the input file.

John Nagle

[David Nicolson]

Hi John,

That was an excellent idea and it was the cause problem. Whether this
is a bug in shutil I'm not sure.

Here is the traceback, Python 2.4.3 on Windows XP:

C:\Documents and Settings\Güstav>C:\python243\python Z:\sh.py
Copying u'C:\\Documents and Settings\\G\xfcstav\\My Documents\\My
Music\\iTunes
\\iTunes Music Library.xml' ...
Traceback (most recent call last):
File "Z:\sh.py", line 12, in ?
shutil.copy(xmlfile,"C:iTunes Music Library.xml")
File "C:\python243\lib\shutil.py", line 81, in copy
copyfile(src, dst)
File "C:\python243\lib\shutil.py", line 41, in copyfile
if _samefile(src, dst):
File "C:\python243\lib\shutil.py", line 36, in _samefile
return (os.path.normcase(os.path.abspath(src)) ==
UnicodeDecodeError: 'ascii' codec can't decode byte 0xfc in
position 27: ordinal
not in range(128)

Here is the code:

import _winreg
import os
import shutil

reg = _winreg.ConnectRegistry(None, _winreg.HKEY_CURRENT_USER)
key = _winreg.OpenKey(reg, r"Software\Microsoft\Windows
\CurrentVersion\Explorer\Shell Folders", 0, _winreg.KEY_READ)
mymusic = _winreg.QueryValueEx(key, "My Music")[0]

xmlfile = os.path.join(os.path.join(mymusic,"iTunes"),"iTune s Music
Library.xml")
print "Copying ",repr(xmlfile),"..."

shutil.copy(xmlfile,"C:\iTunes Music Library.xml")

The shutil line needed to be changed to this to be successful:

shutil.copy(xmlfile.encode("windows-1252"),"C:\iTunes Music
Library.xml"

Regards,
David
On 27/03/2007, at 4:47 PM, John Nagle wrote:

Facundo Batista wrote:

>David Nicolson wrote:

>>Thanks, but it's definitely not the print. In original the code the
print statements are replaced by a call to a log method.

Besides, the exception would be different if it was thrown
outside of
the try block.

The best you can do is take the piece of code that has the
problem, show
it to us, and then copy the traceback.

Regards,

There may be some problem here with a file being recognized as
Unicode
in binary mode. That shouldn't happen, but looking at the Windows
UNICODE support, it might not be impossible.

Information needed:

- Platform (Windows, Linux, ...)
- Python version
- A hex dump of the first few bytes of the input file.

John Nagle
--
http://mail.python.org/mailman/listinfo/python-list

[Leo Kislov]

On Mar 28, 7:01 am, David Nicolson <[e-mail]:

Hi John,

That was an excellent idea and it was the cause problem. Whether this
is a bug inshutilI'm not sure.

Here is the traceback, Python 2.4.3 on Windows XP:

C:\Documents and Settings\Güstav>C:\python243\python Z:\sh.py
Copying u'C:\\Documents and Settings\\G\xfcstav\\My Documents\\My
Music\\iTunes
\\iTunes Music Library.xml' ...
Traceback (most recent call last):
File "Z:\sh.py", line 12, in ?
shutil.copy(xmlfile,"C:iTunes Music Library.xml")

Note, there is no backslash after C:. shutil will try to make an
absolute file name and concatenate it with a current directory name (C:
\Documents and Settings\Güstav) that contains non-ascii characters.
Because of backward compatibility the absolute name won't be unicode.
On the other hand data coming from registry is unicode. When shutil
tries to compare those two file names it fails. To avoid the problem
you need either make both file names unicode or both file names byte-
strings.

However one thing is still mystery to me. Your source code contains
backslash but your traceback doesn't:

shutil.copy(xmlfile,"C:\iTunes Music Library.xml")

Theshutilline needed to be changed to this to be successful:

shutil.copy(xmlfile.encode("windows-1252"),"C:\iTunes Music
Library.xml"

It will work only in some European locales. Using of locale module you
can make it work for 99% of world users, but it will still fail in
cases like German locale and Greek characters in file names. Only
using unicode everywhere in your program is a complete solution. Like

shutil.copy(xmlfile, u"C:\iTunes Music Library.xml")

if you use constant or make sure your file name is unicode:

dest = unicode(....)
shutil.copy(xmlfile, dest)
-- Leo.