On Jan 25, 2014, at 11:48 AM, Jonathan Rogers <jonathanrrog...@gmail.com> wrote:
> I am experimenting with the caching example and have it working as > described. I have a number of mapped classes for tables which are static > lookup tables. The rows in the tables will not change during the > lifetime of application process. So, I would like to automatically apply > a FromCache option to any query which simply returns an instance of a > class mapped to one of the static lookup tables. > > It seems that I should be able to apply a FromCache option to any Query > which has only one entity which is one of the mapped classes I want to > cache. However, I'm not sure what interface to Query would allow me to > determine this. Has anyone solved a similar problem or have any > suggestions for where to start? when i used this I was able to apply a whole list of RelationshipCache options to all queries, that option as written is suited to this since it looks through query._current_path to identify first that it applies. in the case of FromCache, the implementation in the example is more fixed, but I would likely look into further altering CachingQuery.__iter__ to also pull from a set of “fixed” cache options. To see that a Query is against “A” for a given “session.query(A)” you can look in query._entities for exactly one entity which has “.mapper” set to the mapper you’re looking for.
signature.asc
Description: Message signed with OpenPGP using GPGMail
