Closed GoogleCodeExporter closed 9 years ago
The Unicode handling has been revamped in the Python 3.x line, and I think I
need to bring it over to the 2.x
line.
For reference, can you tell me the value of sys.maxunicode on your build:
import sys
print sys.maxunicode
I suspect it is UCS4, which for some reason, has become a very popular
configuration in the Linux distros.
Also, if I can't reproduce this, can I give you a custom pyd (DLL) to try?
Original comment by mkleehammer
on 16 Apr 2010 at 6:28
Another thing I noticed, the value you want is "unicode_results", not
"unicode_strings".
(I know this seems error prone, but lots of people like to use keywords to pass
the DSN values and it is a DB API
standard, so I allow it. If I don't recognize the keyword, it simply gets
added to the DSN.)
Can you confirm which you were using?
Original comment by mkleehammer
on 16 Apr 2010 at 6:37
The sys.maxunicode == 1114111 == 0x10FFFF.
On my (other) win64 machine maxunicode==65535.
Of course, "unicode_results".
Yes, I will try your custom pyd.
Original comment by av1024@gmail.com
on 18 Apr 2010 at 10:00
I've also encountered this error, except the type I've encountered it on is
nvarchar and it doesn't matter what unicode_results is.
My configuration:
pyodbc 2.1.8
FreeTDS CVS checkout from 11/16/2010
MacOS X 10.6.4
Python 2.6
iODBC driver manager
SQL Server 2008 RTM
I get the same behavior as the reporter with 'select DATALENGTH...' on the
relevant column.
sys.maxunicode = 65535
Original comment by s...@mitre.org
on 17 Nov 2010 at 4:32
I'm closing this because I believe it is fixed in the 3.0.x builds. Also,
Microsoft now has a Linux driver which definitely works.
Original comment by mkleehammer
on 27 Dec 2011 at 1:03
Still encountering this issue on OS/X. iODBC and FreeTDS.
Original comment by gsib...@gmail.com
on 9 Oct 2013 at 3:21
Original issue reported on code.google.com by
av1024@gmail.com
on 13 Apr 2010 at 11:56