Added note about result cache changes
This commit is contained in:
parent
adc4840cce
commit
a8052dec37
@ -1,3 +1,17 @@
|
||||
# ResultCache implementation rewritten
|
||||
|
||||
The result cache is completly rewritten and now works on the database result level, not inside the ORM AbstractQuery
|
||||
anymore. This means that for result cached queries the hydration will now always be performed again, regardless of
|
||||
the hydration mode. Affected areas are:
|
||||
|
||||
1. Fixes the problem that entities coming from the result cache were not registered in the UnitOfWork
|
||||
leading to problems during EntityManager#flush. Calls to EntityManager#merge are not necessary anymore.
|
||||
2. Affects the array hydrator which now includes the overhead of hydration compared to caching the final result.
|
||||
|
||||
The API is backwards compatible however most of the getter methods on the `AbstractQuery` object are now
|
||||
deprecated in favor of calling AbstractQuery#getQueryCacheProfile(). This method returns a `Doctrine\DBAL\Cache\QueryCacheProfile`
|
||||
instance with access to result cache driver, lifetime and cache key.
|
||||
|
||||
# EntityManager#getPartialReference() creates read-only entity
|
||||
|
||||
Entities returned from EntityManager#getPartialReference() are now marked as read-only if they
|
||||
|
Loading…
Reference in New Issue
Block a user