Markus: Would you be so kind as to zip up and email to me the adonet-dbapi as you now have it patched? I will create a version which does not error out. (May take a few days.) -- Vernon
2009/10/28 Markus Törnqvist <[email protected]> > On Wed, Oct 28, 2009 at 10:53:45PM +0900, Seo Sanghyeon wrote: > >Intro... I, together with Mark Rees, wrote most of adonet-dbapi code in > 2006. > > Hi :) > > >2009/10/28 Markus Törnqvist <[email protected]>: > >> Here's the connect function (in its entirety) from mssql.py > > >> def connect(connstr): > >> relevant_parts = [part for part in connstr.split(';') if not > part.upper().startswith('PROVIDER')] > >> connstr = ';'.join(relevant_parts) > >> #return generic_connect("System.Data", > "System.Data.SqlClient.SqlConnection", connstr) > > >> Way too hacky for my taste but it gets the job kinda done. > > > >Indeed. The question is, what underlying Python DB-API 2 driver does > >django-mssql use? Is it http://pymssql.sourceforge.net/ ? > > It's on adonet-dbapi... > > >Argument to connect() function in Python DB-API 2 compliant drivers is > >not standardized. In case of pymssql, it receives keyword arguments, > >so it should be handled in a manner similar to MySQLdb.py wrapper in > >adonet-dbapi. But from your hack it seems django-mssql's underlying > >Python DB-API 2 driver expects a string argument, so django-mssql is > >building string by itself -- this is ungood. Read psycopg.py wrapper > >in adonet-dbapi to see how to handle this. > > Ok, something like that could be used instead. > > >On the other hand, as I understand, Django does not handle paramstyle > >difference in DB-API 2 drivers, so this is kind of a lost cause. How > >do other Django DB backends deal with the issue? SQLAlchemy, > >fortunately, handles paramstyle difference by itself. > > I'd have to create the param dict from the connect string, very > much like your psycopg :) > > Django is intended to use the dbapi, I don't think there's any other > way of handling this? > > Anyway, I think the StackOverflowException is a bigger issue :) > > -- > mjt > > _______________________________________________ > Users mailing list > [email protected] > http://lists.ironpython.com/listinfo.cgi/users-ironpython.com >
_______________________________________________ Users mailing list [email protected] http://lists.ironpython.com/listinfo.cgi/users-ironpython.com
