Mirrors/the-algorithm
2
0
Fork 0
mirror of https://github.com/twitter/the-algorithm.git synced 2024-12-22 18:21:51 +01:00
Code Issues Packages Projects Releases Wiki Activity

Merge 86c4e8c3e5 into fb54d8b549

Browse Source
This commit is contained in:
Rajiv Singh 2023-05-22 17:40:14 -05:00 committed by GitHub
commit 9dfe7772b9
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 22 additions and 23 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

45
src/java/com/twitter/search/earlybird/index/DocValuesHelper.java
View File

@ -1,6 +1,8 @@
package com.twitter.search.earlybird.index; package com.twitter.search.earlybird.index;
import java.io.IOException; import java.io.IOException;
import java.util.Map;
import java.util.concurrent.ConcurrentHashMap;
import org.apache.lucene.index.LeafReader; import org.apache.lucene.index.LeafReader;
import org.apache.lucene.index.Terms; import org.apache.lucene.index.Terms;
@ -9,6 +11,8 @@ import org.apache.lucene.search.DocIdSetIterator;
import org.apache.lucene.util.BytesRef; import org.apache.lucene.util.BytesRef;
public final class DocValuesHelper { public final class DocValuesHelper {
private static final Map<String, TermsEnum> termsEnumCache = new ConcurrentHashMap<>();
private DocValuesHelper() { private DocValuesHelper() {
} }
@ -22,13 +26,10 @@ public final class DocValuesHelper {
*/ */
public static int getFirstDocIdWithValue( public static int getFirstDocIdWithValue(
LeafReader reader, String indexField, BytesRef value) throws IOException { LeafReader reader, String indexField, BytesRef value) throws IOException {
TermsEnum termsEnum = getTermsEnum(reader, indexField); try (DocIdSetIterator docsIterator = getTermsEnum(reader, indexField).postings(null)) {
if (termsEnum == null || !termsEnum.seekExact(value)) { int docId = docsIterator.nextDoc();
return DocIdSetIterator.NO_MORE_DOCS; return docId == DocIdSetIterator.NO_MORE_DOCS ? docId : docId;
} }
DocIdSetIterator docsIterator = termsEnum.postings(null);
return docsIterator.nextDoc();
} }
/** /**
@ -44,27 +45,25 @@ public final class DocValuesHelper {
*/ */
public static int getLargestDocIdWithCeilOfValue( public static int getLargestDocIdWithCeilOfValue(
LeafReader reader, String indexField, BytesRef value) throws IOException { LeafReader reader, String indexField, BytesRef value) throws IOException {
TermsEnum termsEnum = getTermsEnum(reader, indexField); try (DocIdSetIterator docsIterator = getTermsEnum(reader, indexField).postings(null)) {
if (termsEnum == null) { int docId = docsIterator.nextDoc();
return DocIdSetIterator.NO_MORE_DOCS; while (docsIterator.nextDoc() != DocIdSetIterator.NO_MORE_DOCS) {
docId = docsIterator.docID();
}
return docId;
} }
if (termsEnum.seekCeil(value) == TermsEnum.SeekStatus.END) {
return DocIdSetIterator.NO_MORE_DOCS;
}
DocIdSetIterator docsIterator = termsEnum.postings(null);
int docId = docsIterator.nextDoc();
while (docsIterator.nextDoc() != DocIdSetIterator.NO_MORE_DOCS) {
docId = docsIterator.docID();
}
return docId;
} }
private static TermsEnum getTermsEnum(LeafReader reader, String indexField) throws IOException { private static TermsEnum getTermsEnum(LeafReader reader, String indexField) throws IOException {
Terms terms = reader.terms(indexField); TermsEnum termsEnum = termsEnumCache.get(indexField);
if (terms == null) { if (termsEnum == null) {
return null; Terms terms = reader.terms(indexField);
if (terms == null) {
return null;
}
termsEnum = terms.iterator();
termsEnumCache.put(indexField, termsEnum);
} }
return terms.iterator(); return termsEnum;
} }
} }