CavalliumDBEngine/src/main/java/it/cavallium/dbengine/lucene/RandomFieldComparator.java

114 lines
2.9 KiB
Java
Raw Normal View History

2021-01-30 22:14:48 +01:00
package it.cavallium.dbengine.lucene;
2020-12-07 22:15:18 +01:00
import it.cavallium.dbengine.utils.LFSR.LFSRIterator;
2020-12-07 22:15:18 +01:00
import java.io.IOException;
import java.math.BigInteger;
import org.apache.lucene.index.LeafReaderContext;
import org.apache.lucene.search.FieldComparator;
import org.apache.lucene.search.LeafFieldComparator;
import org.apache.lucene.search.Scorable;
import org.apache.lucene.search.ScoreCachingWrappingScorer;
import org.jetbrains.annotations.NotNull;
2021-10-13 14:26:54 +02:00
//todo: fix
2020-12-07 22:15:18 +01:00
public class RandomFieldComparator extends FieldComparator<Float> implements LeafFieldComparator {
private final @NotNull LFSRIterator rand;
private final float[] scores;
private float bottom;
private Scorable scorer;
private float topValue;
/** Creates a new comparator based on relevance for {@code numHits}. */
public RandomFieldComparator(@NotNull LFSRIterator rand, int numHits) {
this.rand = rand;
scores = new float[numHits];
}
@Override
public int compare(int slot1, int slot2) {
return Float.compare(scores[slot2], scores[slot1]);
}
@Override
2023-02-22 16:59:35 +01:00
public int compareBottom(int doc) throws IOException {
2020-12-07 22:15:18 +01:00
float score = scorer.score();
assert !Float.isNaN(score);
return Float.compare(score, bottom);
}
@Override
2023-02-22 16:59:35 +01:00
public void copy(int slot, int doc) throws IOException {
2020-12-07 22:15:18 +01:00
scores[slot] = scorer.score();
assert !Float.isNaN(scores[slot]);
}
@Override
public LeafFieldComparator getLeafComparator(LeafReaderContext context) {
return this;
}
@Override
public void setBottom(final int bottom) {
this.bottom = scores[bottom];
}
@Override
public void setTopValue(Float value) {
topValue = Float.MAX_VALUE;
}
@Override
public void setScorer(Scorable scorer) {
// wrap with a ScoreCachingWrappingScorer so that successive calls to
// score() will not incur score computation over and
// over again.
var randomizedScorer = new Scorable() {
@Override
2021-02-03 14:37:02 +01:00
public float score() {
2020-12-07 22:15:18 +01:00
return randomize(scorer.docID());
}
@Override
public int docID() {
return scorer.docID();
}
};
this.scorer = ScoreCachingWrappingScorer.wrap(randomizedScorer);
2020-12-07 22:15:18 +01:00
}
2021-02-03 14:37:02 +01:00
@SuppressWarnings("RedundantCast")
2020-12-07 22:15:18 +01:00
@Override
public Float value(int slot) {
2021-02-03 14:37:02 +01:00
return (float) scores[slot];
2020-12-07 22:15:18 +01:00
}
// Override because we sort reverse of natural Float order:
@Override
public int compareValues(Float first, Float second) {
// Reversed intentionally because relevance by default
// sorts descending:
return second.compareTo(first);
}
@Override
2023-02-22 16:59:35 +01:00
public int compareTop(int doc) throws IOException {
2020-12-07 22:15:18 +01:00
float docValue = scorer.score();
assert !Float.isNaN(docValue);
return Float.compare(docValue, topValue);
}
private float randomize(int num) {
2021-10-13 14:26:54 +02:00
int val = rand.next(BigInteger.valueOf(num)).intValueExact();
int pow24 = 1 << 24;
if (val >= pow24) {
throw new IndexOutOfBoundsException();
}
if (val < 0) {
throw new IndexOutOfBoundsException();
}
2020-12-07 22:15:18 +01:00
return (val & 0x00FFFFFF) / (float)(1 << 24); // only use the lower 24 bits to construct a float from 0.0-1.0
}
}