Mirror/MarginaliaSearch
1
0
Fork 0
mirror of https://github.com/MarginaliaSearch/MarginaliaSearch.git synced 2025-02-23 13:09:00 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

(index) Speed up minDist calculations by excluding large lists

Browse Source
This commit is contained in:
Viktor Lofgren 2024-08-26 13:04:15 +02:00
parent 77efce0673
commit 30bf845c81
2 changed files with 11 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

16
code/index/java/nu/marginalia/index/results/model/PhraseConstraintGroupList.java
View File

@ -1,5 +1,6 @@
package nu.marginalia.index.results.model;
import it.unimi.dsi.fastutil.ints.IntArrayList;
import it.unimi.dsi.fastutil.ints.IntIterator;
import it.unimi.dsi.fastutil.ints.IntList;
import nu.marginalia.index.model.SearchTermsUtil;
@ -142,8 +143,8 @@ public class PhraseConstraintGroupList {
}
public int minDistance(IntList[] positions) {
IntList[] sequences = new IntList[present.cardinality()];
int[] iterOffsets = new int[sequences.length];
List<IntList> sequences = new ArrayList<>(present.cardinality());
IntList iterOffsets = new IntArrayList(present.cardinality());
for (int oi = 0, si = 0; oi < offsets.length; oi++) {
if (!present.get(oi)) {
@ -162,11 +163,16 @@ public class PhraseConstraintGroupList {
if (posForTerm == null) {
return Integer.MAX_VALUE;
}
sequences[si++] = posForTerm;
iterOffsets[si - 1] = -oi;
if (posForTerm.size() > 16) { // heuristic to avoid large sequences, which is expensive and not very useful
continue;
}
sequences.add(posForTerm);
iterOffsets.add(-oi);
}
return SequenceOperations.minDistance(sequences, iterOffsets);
return SequenceOperations.minDistance(sequences.toArray(IntList[]::new), iterOffsets.toIntArray());
}
}
}

1
code/libraries/coded-sequence/java/nu/marginalia/sequence/SequenceOperations.java
View File

@ -158,7 +158,6 @@ public class SequenceOperations {
}
int minDist = Integer.MAX_VALUE;
int maxVal = Integer.MIN_VALUE;
int maxI = 0;