#include <InMemoryPool.h>
Inheritance diagram for InMemoryPool::
We might also want a CompactInMemoryPool, optimized for small size.
More importantly, we might want to support CLUSTERING, which in this case means knowing some things about the prevalent schemas which will help with performance. This is also true with the disk version. Basically a set of sets of properties which should be clustered -- if something has one of these properties, it's very likely to have these others. Property-Cluster-Advice. One might also want this to advise of joins -- subrelations... But maybe this is another class (or two, if you count disk). This should undo all the performance loss of reification.
Definition at line 34 of file InMemoryPool.h.
|
Get the first triple which matches the pattern, and then (on following calls) get the next matching triples. max_results is a hint which can improve performance if you know ahead of time that you will not be iterating more than some number of times. Reimplemented from TripleSource. Definition at line 92 of file InMemoryPool.cpp. Referenced by XSBPool::fetch().
|
|
Remove the given triple, if it happens to be present. No operation otherwise. Reimplemented from Pool. Definition at line 112 of file InMemoryPool.cpp. Referenced by XSBPool::remove().
|
|
Replace any occurances of 'old' with 'replacement'. Caller should treat 'old' as if it had been released.
Reimplemented from Pool. Definition at line 117 of file InMemoryPool.cpp. Referenced by XSBPool::replace().
|