org.apache.lucene.search.uhighlight.CustomPassageFormatter Maven / Gradle / Ivy
Go to download
Show more of this group Show more artifacts with this name
Show all versions of opensearch Show documentation
Show all versions of opensearch Show documentation
OpenSearch subproject :server
/*
* SPDX-License-Identifier: Apache-2.0
*
* The OpenSearch Contributors require contributions made to
* this file be licensed under the Apache-2.0 license or a
* compatible open source license.
*/
/*
* Licensed to Elasticsearch under one or more contributor
* license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright
* ownership. Elasticsearch licenses this file to you under
* the Apache License, Version 2.0 (the "License"); you may
* not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing,
* software distributed under the License is distributed on an
* "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
* KIND, either express or implied. See the License for the
* specific language governing permissions and limitations
* under the License.
*/
/*
* Modifications Copyright OpenSearch Contributors. See
* GitHub history for details.
*/
package org.apache.lucene.search.uhighlight;
import org.apache.lucene.search.highlight.Encoder;
import org.opensearch.search.fetch.subphase.highlight.HighlightUtils;
/**
* Custom passage formatter that allows us to:
* 1) extract different snippets (instead of a single big string) together with their scores ({@link Snippet})
* 2) use the {@link Encoder} implementations that are already used with the other highlighters
*/
public class CustomPassageFormatter extends PassageFormatter {
private final String preTag;
private final String postTag;
private final Encoder encoder;
public CustomPassageFormatter(String preTag, String postTag, Encoder encoder) {
this.preTag = preTag;
this.postTag = postTag;
this.encoder = encoder;
}
@Override
public Snippet[] format(Passage[] passages, String content) {
Snippet[] snippets = new Snippet[passages.length];
int pos;
for (int j = 0; j < passages.length; j++) {
Passage passage = passages[j];
StringBuilder sb = new StringBuilder();
pos = passage.getStartOffset();
for (int i = 0; i < passage.getNumMatches(); i++) {
int start = passage.getMatchStarts()[i];
assert start >= pos && start < passage.getEndOffset();
// append content before this start
append(sb, content, pos, start);
int end = passage.getMatchEnds()[i];
assert end > start;
// Look ahead to expand 'end' past all overlapping:
while (i + 1 < passage.getNumMatches() && passage.getMatchStarts()[i + 1] < end) {
end = passage.getMatchEnds()[++i];
}
end = Math.min(end, passage.getEndOffset()); // in case match straddles past passage
sb.append(preTag);
append(sb, content, start, end);
sb.append(postTag);
pos = end;
}
// its possible a "term" from the analyzer could span a sentence boundary.
append(sb, content, pos, Math.max(pos, passage.getEndOffset()));
// we remove the paragraph separator if present at the end of the snippet (we used it as separator between values)
if (sb.charAt(sb.length() - 1) == HighlightUtils.PARAGRAPH_SEPARATOR) {
sb.deleteCharAt(sb.length() - 1);
} else if (sb.charAt(sb.length() - 1) == HighlightUtils.NULL_SEPARATOR) {
sb.deleteCharAt(sb.length() - 1);
}
// and we trim the snippets too
snippets[j] = new Snippet(sb.toString().trim(), passage.getScore(), passage.getNumMatches() > 0);
}
return snippets;
}
private void append(StringBuilder dest, String content, int start, int end) {
dest.append(encoder.encodeText(content.substring(start, end)));
}
}