Implement copy node rdf #2028
Implement copy node rdf #2028
Conversation
Codecov ReportPatch coverage:
Additional details and impacted files@@ Coverage Diff @@
## master #2028 +/- ##
==========================================
+ Coverage 89.97% 90.08% +0.11%
==========================================
Files 909 919 +10
Lines 32992 33248 +256
==========================================
+ Hits 29685 29953 +268
+ Misses 3307 3295 -12
☔ View full report in Codecov by Sentry. |
| @@ -172,6 +193,68 @@ void HashIndexBuilder<T>::flush() { | |||
| } | |||
| } | |||
|
|
|||
| template<typename T> | |||
| void HashIndexBuilder<T>::rehashSlots(slot_id_t primarySlotId) { | |||
There was a problem hiding this comment.
should we try to cover this code path?
There was a problem hiding this comment.
Looks like the rdf test dataset doesn't trigger the rehash.
There was a problem hiding this comment.
Can we not introduce changes to hash index now? I feel this should be done later in a more careful way.
| @@ -86,7 +86,7 @@ void QueryProcessor::decomposePlanIntoTasks( | |||
| } break; | |||
| case PhysicalOperatorType::READER: { | |||
| auto reader = (Reader*)op; | |||
| if (reader->getContainsSerial()) { | |||
| if (reader->getContainsSerial() || reader->isCopyTurtleFile()) { | |||
There was a problem hiding this comment.
I thought we only need to read in SERIAL mode rather than copy in single thread. If we need to execute in single thread mode I'll refactor this check to front-end
There was a problem hiding this comment.
We need to execute in single thread mode due to the special copy codepath (append to hashIndex first, then append to columnChunk).
acquamarin
force-pushed
the
rdf-copy
branch
from
79b463a
to
f99d35a
Compare
No description provided.