[sqlalchemy] Re: Sharding with _BindParamClause

2010-04-02 Thread George V. Reilly
Michael Bayer wrote: check out r0ddd638f1d90 in mercurial.  I've added the function from the example below, plus support for in_op(), to the attribute_shard example. The old ClauseVisitor method is removed and replaced with this more robust method. Very nice! Thanks, Michael. /George --

[sqlalchemy] Re: Sharding with _BindParamClause

2010-04-02 Thread George V. Reilly
On Apr 2, 4:43 pm, George V. Reilly george.v.rei...@gmail.com wrote: Michael Bayer wrote: check out r0ddd638f1d90 in mercurial.  I've added the function from the example below, plus support for in_op(), to the attribute_shard example. The old ClauseVisitor method is removed and replaced

Re: [sqlalchemy] Re: Sharding with _BindParamClause

2010-04-02 Thread Michael Bayer
George V. Reilly wrote: I really don't like the isinstance(other_side, sqlalchemy.sql.expression._BindParamClause)) in the middle of _check_side. Is there a cleaner way to do this? I found that a combination of * the above _check_side and two-sided visit_binary * doing a better job of

Re: [sqlalchemy] Re: Sharding with _BindParamClause

2010-04-02 Thread Michael Bayer
Michael Bayer wrote: George V. Reilly wrote: I really don't like the isinstance(other_side, sqlalchemy.sql.expression._BindParamClause)) in the middle of _check_side. Is there a cleaner way to do this? I found that a combination of * the above _check_side and two-sided visit_binary *

[sqlalchemy] Re: Sharding with _BindParamClause

2010-04-01 Thread George V. Reilly
On Mar 30, 4:42 pm, Michael Bayer mike...@zzzcomputing.com wrote: George V. Reilly wrote: We're using SQLAlchemy sharding to partition accounts across a couple of databases. We want to add more partitions, but first we need to eliminate some unnecessary cross-partition queries. This