Hi Mark, Just committed the REAL/DOUBLE fix, this encoding thingy has to wait until January, sorry. On 22 dec 2009, at 13:43, Mark Bucciarelli wrote:
On Tue, Dec 22, 2009 at 5:30 AM, Gijs Molenaar
wrote: Do you know the behavior of other db API's?
MSQLdb has a use_encoding keyword.
The PostgreSql Django adapter had a set_client_encoding() method on the connection object.
The native SQLite3 Python module just returns Unicode.
How big is the problem to do a manual conversion to unicode in your wrapper?
Painful--I couldn't find a good place to hook in the code.
The problem is that there isn't a real standard way of doing this, at least not defined by the python db API 2.0.
Yup, the PEP is encoding agnostic.
I though the path you started down was a good idea; pass a use_unicode keyword to the connection __init__().
Then if use_unicode = True, just decode("utf8") the string in __strip(), otherwise leave code as is.
m