Index: solr/core/src/java/org/apache/solr/highlight/PostingsSolrHighlighter.java =================================================================== --- solr/core/src/java/org/apache/solr/highlight/PostingsSolrHighlighter.java (revision 1464609) +++ solr/core/src/java/org/apache/solr/highlight/PostingsSolrHighlighter.java (working copy) @@ -19,6 +19,7 @@ import java.io.IOException; import java.text.BreakIterator; +import java.util.Arrays; import java.util.Collections; import java.util.Locale; import java.util.Map; @@ -162,9 +163,12 @@ // query-time parameters String[] fieldNames = getHighlightFields(query, req, defaultFields); + // TODO: make this per-field int numSnippets = params.getInt(HighlightParams.SNIPPETS, 1); + int maxPassages[] = new int[fieldNames.length]; + Arrays.fill(maxPassages, numSnippets); - Map snippets = highlighter.highlightFields(fieldNames, query, searcher, docIDs, numSnippets); + Map snippets = highlighter.highlightFields(fieldNames, query, searcher, docIDs, maxPassages); return encodeSnippets(keys, fieldNames, snippets); } else { return null; Index: lucene/highlighter/src/test/org/apache/lucene/search/postingshighlight/TestPostingsHighlighter.java =================================================================== --- lucene/highlighter/src/test/org/apache/lucene/search/postingshighlight/TestPostingsHighlighter.java (revision 1464609) +++ lucene/highlighter/src/test/org/apache/lucene/search/postingshighlight/TestPostingsHighlighter.java (working copy) @@ -503,7 +503,7 @@ int[] docIDs = new int[2]; docIDs[0] = hits[0].doc; docIDs[1] = hits[1].doc; - String snippets[] = highlighter.highlightFields(new String[] {"body"}, query, searcher, docIDs, 1).get("body"); + String snippets[] = highlighter.highlightFields(new String[] {"body"}, query, searcher, docIDs, new int[] { 1 }).get("body"); assertEquals(2, snippets.length); assertEquals("Just a test highlighting from postings. ", snippets[0]); assertEquals("Highlighting the first term. ", snippets[1]); @@ -582,7 +582,7 @@ PostingsHighlighter highlighter = new PostingsHighlighter(); Query query = new TermQuery(new Term("body", "highlighting")); int[] docIDs = new int[] {0}; - String snippets[] = highlighter.highlightFields(new String[] {"body"}, query, searcher, docIDs, 2).get("body"); + String snippets[] = highlighter.highlightFields(new String[] {"body"}, query, searcher, docIDs, new int[] { 2 }).get("body"); assertEquals(1, snippets.length); assertEquals("test this is. another sentence this test has. ", snippets[0]); @@ -618,7 +618,7 @@ }; Query query = new TermQuery(new Term("body", "highlighting")); int[] docIDs = new int[] {0}; - String snippets[] = highlighter.highlightFields(new String[] {"body"}, query, searcher, docIDs, 2).get("body"); + String snippets[] = highlighter.highlightFields(new String[] {"body"}, query, searcher, docIDs, new int[] { 2 }).get("body"); assertEquals(1, snippets.length); assertNull(snippets[0]); @@ -654,7 +654,7 @@ }; Query query = new TermQuery(new Term("body", "highlighting")); int[] docIDs = new int[] {0}; - String snippets[] = highlighter.highlightFields(new String[] {"body"}, query, searcher, docIDs, 2).get("body"); + String snippets[] = highlighter.highlightFields(new String[] {"body"}, query, searcher, docIDs, new int[] { 2 }).get("body"); assertEquals(1, snippets.length); assertEquals("test this is. another sentence this test has. far away is that planet.", snippets[0]); @@ -685,7 +685,7 @@ PostingsHighlighter highlighter = new PostingsHighlighter(); Query query = new TermQuery(new Term("bogus", "highlighting")); int[] docIDs = new int[] {0}; - String snippets[] = highlighter.highlightFields(new String[] {"bogus"}, query, searcher, docIDs, 2).get("bogus"); + String snippets[] = highlighter.highlightFields(new String[] {"bogus"}, query, searcher, docIDs, new int[] { 2 }).get("bogus"); assertEquals(1, snippets.length); assertNull(snippets[0]); @@ -721,7 +721,7 @@ Query query = new TermQuery(new Term("body", "highlighting")); int[] docIDs = new int[1]; docIDs[0] = docID; - String snippets[] = highlighter.highlightFields(new String[] {"body"}, query, searcher, docIDs, 2).get("body"); + String snippets[] = highlighter.highlightFields(new String[] {"body"}, query, searcher, docIDs, new int[] { 2 }).get("body"); assertEquals(1, snippets.length); assertEquals(" ", snippets[0]); @@ -757,7 +757,7 @@ Query query = new TermQuery(new Term("body", "highlighting")); int[] docIDs = new int[1]; docIDs[0] = docID; - String snippets[] = highlighter.highlightFields(new String[] {"body"}, query, searcher, docIDs, 2).get("body"); + String snippets[] = highlighter.highlightFields(new String[] {"body"}, query, searcher, docIDs, new int[] { 2 }).get("body"); assertEquals(1, snippets.length); assertNull(snippets[0]); @@ -814,4 +814,39 @@ ir.close(); dir.close(); } + + public void testMultipleSnippetSizes() throws Exception { + Directory dir = newDirectory(); + IndexWriterConfig iwc = newIndexWriterConfig(TEST_VERSION_CURRENT, new MockAnalyzer(random(), MockTokenizer.SIMPLE, true)); + iwc.setMergePolicy(newLogMergePolicy()); + RandomIndexWriter iw = new RandomIndexWriter(random(), dir, iwc); + + FieldType offsetsType = new FieldType(TextField.TYPE_STORED); + offsetsType.setIndexOptions(IndexOptions.DOCS_AND_FREQS_AND_POSITIONS_AND_OFFSETS); + Field body = new Field("body", "", offsetsType); + Field title = new Field("title", "", offsetsType); + Document doc = new Document(); + doc.add(body); + doc.add(title); + + body.setStringValue("This is a test. Just a test highlighting from postings. Feel free to ignore."); + title.setStringValue("This is a test. Just a test highlighting from postings. Feel free to ignore."); + iw.addDocument(doc); + + IndexReader ir = iw.getReader(); + iw.close(); + + IndexSearcher searcher = newSearcher(ir); + PostingsHighlighter highlighter = new PostingsHighlighter(); + BooleanQuery query = new BooleanQuery(); + query.add(new TermQuery(new Term("body", "test")), BooleanClause.Occur.SHOULD); + query.add(new TermQuery(new Term("title", "test")), BooleanClause.Occur.SHOULD); + Map snippets = highlighter.highlightFields(new String[] { "title", "body" }, query, searcher, new int[] { 0 }, new int[] { 1, 2 }); + String titleHighlight = snippets.get("title")[0]; + String bodyHighlight = snippets.get("body")[0]; + assertEquals("This is a test. ", titleHighlight); + assertEquals("This is a test. Just a test highlighting from postings. ", bodyHighlight); + ir.close(); + dir.close(); + } } Index: lucene/highlighter/src/java/org/apache/lucene/search/postingshighlight/PostingsHighlighter.java =================================================================== --- lucene/highlighter/src/java/org/apache/lucene/search/postingshighlight/PostingsHighlighter.java (revision 1464609) +++ lucene/highlighter/src/java/org/apache/lucene/search/postingshighlight/PostingsHighlighter.java (working copy) @@ -48,6 +48,7 @@ import org.apache.lucene.search.ScoreDoc; import org.apache.lucene.search.TopDocs; import org.apache.lucene.util.BytesRef; +import org.apache.lucene.util.SorterTemplate; import org.apache.lucene.util.UnicodeUtil; /** @@ -194,7 +195,7 @@ * {@link IndexOptions#DOCS_AND_FREQS_AND_POSITIONS_AND_OFFSETS} */ public String[] highlight(String field, Query query, IndexSearcher searcher, TopDocs topDocs, int maxPassages) throws IOException { - Map res = highlightFields(new String[] { field }, query, searcher, topDocs, maxPassages); + Map res = highlightFields(new String[] { field }, query, searcher, topDocs, new int[] { maxPassages }); return res.get(field); } @@ -224,7 +225,9 @@ * {@link IndexOptions#DOCS_AND_FREQS_AND_POSITIONS_AND_OFFSETS} */ public Map highlightFields(String fields[], Query query, IndexSearcher searcher, TopDocs topDocs) throws IOException { - return highlightFields(fields, query, searcher, topDocs, 1); + int maxPassages[] = new int[fields.length]; + Arrays.fill(maxPassages, 1); + return highlightFields(fields, query, searcher, topDocs, maxPassages); } /** @@ -255,7 +258,7 @@ * @throws IllegalArgumentException if field was indexed without * {@link IndexOptions#DOCS_AND_FREQS_AND_POSITIONS_AND_OFFSETS} */ - public Map highlightFields(String fields[], Query query, IndexSearcher searcher, TopDocs topDocs, int maxPassages) throws IOException { + public Map highlightFields(String fields[], Query query, IndexSearcher searcher, TopDocs topDocs, int maxPassages[]) throws IOException { final ScoreDoc scoreDocs[] = topDocs.scoreDocs; int docids[] = new int[scoreDocs.length]; for (int i = 0; i < docids.length; i++) { @@ -269,12 +272,12 @@ * Highlights the top-N passages from multiple fields, * for the provided int[] docids. * - * @param fields field names to highlight. + * @param fieldsIn field names to highlight. * Must have a stored string value and also be indexed with offsets. * @param query query to highlight. * @param searcher searcher that was previously used to execute the query. * @param docidsIn containing the document IDs to highlight. - * @param maxPassages The maximum number of top-N ranked passages per-field used to + * @param maxPassagesIn The maximum number of top-N ranked passages per-field used to * form the highlighted snippets. * @return Map keyed on field name, containing the array of formatted snippets * corresponding to the documents in topDocs. @@ -285,7 +288,13 @@ * @throws IllegalArgumentException if field was indexed without * {@link IndexOptions#DOCS_AND_FREQS_AND_POSITIONS_AND_OFFSETS} */ - public Map highlightFields(String fields[], Query query, IndexSearcher searcher, int[] docidsIn, int maxPassages) throws IOException { + public Map highlightFields(String fieldsIn[], Query query, IndexSearcher searcher, int[] docidsIn, int maxPassagesIn[]) throws IOException { + if (fieldsIn.length < 1) { + throw new IllegalArgumentException("fieldsIn must not be empty"); + } + if (fieldsIn.length != maxPassagesIn.length) { + throw new IllegalArgumentException("invalid number of maxPassagesIn"); + } final IndexReader reader = searcher.getIndexReader(); query = rewrite(query); SortedSet queryTerms = new TreeSet(); @@ -294,20 +303,53 @@ IndexReaderContext readerContext = reader.getContext(); List leaves = readerContext.leaves(); - // Make our own copy because we sort in-place: + // Make our own copies because we sort in-place: int[] docids = new int[docidsIn.length]; System.arraycopy(docidsIn, 0, docids, 0, docidsIn.length); + final String fields[] = new String[fieldsIn.length]; + System.arraycopy(fieldsIn, 0, fields, 0, fieldsIn.length); + final int maxPassages[] = new int[maxPassagesIn.length]; + System.arraycopy(maxPassagesIn, 0, maxPassages, 0, maxPassagesIn.length); // sort for sequential io Arrays.sort(docids); - Arrays.sort(fields); + new SorterTemplate() { + String pivot; + + @Override + protected void swap(int i, int j) { + String tmp = fields[i]; + fields[i] = fields[j]; + fields[j] = tmp; + int tmp2 = maxPassages[i]; + maxPassages[i] = maxPassages[j]; + maxPassages[j] = tmp2; + } + + @Override + protected int compare(int i, int j) { + return fields[i].compareTo(fields[j]); + } + + @Override + protected void setPivot(int i) { + pivot = fields[i]; + } + + @Override + protected int comparePivot(int j) { + return pivot.compareTo(fields[j]); + } + + }.mergeSort(0, fields.length-1); // pull stored data: String[][] contents = loadFieldValues(searcher, fields, docids, maxLength); - Map highlights = new HashMap(); + Map highlights = new HashMap();; for (int i = 0; i < fields.length; i++) { String field = fields[i]; + int numPassages = maxPassages[i]; Term floor = new Term(field, ""); Term ceiling = new Term(field, UnicodeUtil.BIG_TERM); SortedSet fieldTerms = queryTerms.subSet(floor, ceiling); @@ -319,7 +361,7 @@ for(Term term : fieldTerms) { terms[termUpto++] = term.bytes(); } - Map fieldHighlights = highlightField(field, contents[i], getBreakIterator(field), terms, docids, leaves, maxPassages); + Map fieldHighlights = highlightField(field, contents[i], getBreakIterator(field), terms, docids, leaves, numPassages); String[] result = new String[docids.length]; for (int j = 0; j < docidsIn.length; j++) {