fthevenet commented on PR #12212: URL: https://github.com/apache/lucene/pull/12212#issuecomment-1488891278
I now believe I know what is going on: I initially changed the iterator to advanced on to `baseApproximation` based on my findings that this causes a bug in one of the three hit collection code path, namely the one used when base query is highly constraining vs the drilldowns (`doQueryFirstScoring`). But... while that fixed the problem there, it is causing trouble with the other two (doDrillDownAdvanceScoring and doUnionScoring) which AFAICT are unaware of the "two phase" iterator pattern and always call on `baseIterator`, and so expect that an initial call to `baseIterator::nextDoc` had already been made, hence the failing test cases. So moved the base scorer initialization inside the condition that determines the code path to follow, so that it can be done on either `baseApproximation` or `baseIterator` depending on what the following method needs. It's arguably not the most elegant way to do it but I figured it was the less risky one, as it keeps changes to the rest of the code down to a minimum (and I certainly don't pretend to understand that code well enough to start hacking at it willy-nilly 😅 ) All the randomized test I've ran since that change have passed (as well as the specific test I added), so hopefully its right this time! -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: issues-unsubscr...@lucene.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@lucene.apache.org For additional commands, e-mail: issues-h...@lucene.apache.org
