List of usage examples for org.apache.lucene.search.highlight Highlighter setTextFragmenter
public void setTextFragmenter(Fragmenter fragmenter)
From source file:com.bugull.mongo.lucene.BuguHighlighter.java
License:Apache License
public String getResult(String fieldName, String fieldValue) throws Exception { BuguIndex index = BuguIndex.getInstance(); QueryParser parser = new QueryParser(index.getVersion(), fieldName, index.getAnalyzer()); Query query = parser.parse(keywords); TokenStream tokens = index.getAnalyzer().tokenStream(fieldName, new StringReader(fieldValue)); QueryScorer scorer = new QueryScorer(query, fieldName); Highlighter highlighter = new Highlighter(formatter, scorer); highlighter.setTextFragmenter(new SimpleSpanFragmenter(scorer)); return highlighter.getBestFragments(tokens, fieldValue, maxFragments, "..."); }
From source file:com.difference.historybook.index.lucene.LuceneIndex.java
License:Apache License
@Override public SearchResultWrapper search(String collection, String query, int offset, int size, boolean includeDebug) throws IndexException { try {//ww w . j a v a 2 s.c o m //TODO: make age be a component in the ranking? BooleanQuery.Builder queryBuilder = new BooleanQuery.Builder(); queryBuilder.add(parser.parse(query), Occur.MUST); queryBuilder.add(new TermQuery(new Term(IndexDocumentAdapter.FIELD_COLLECTION, collection)), Occur.FILTER); Query baseQuery = queryBuilder.build(); FunctionQuery boostQuery = new FunctionQuery( new ReciprocalFloatFunction(new DurationValueSource(new Date().getTime() / 1000, new LongFieldSource(IndexDocumentAdapter.FIELD_TIMESTAMP)), RECIP, 1F, 1F)); Query q = new CustomScoreQuery(baseQuery, boostQuery); QueryScorer queryScorer = new QueryScorer(q, IndexDocumentAdapter.FIELD_SEARCH); Fragmenter fragmenter = new SimpleSpanFragmenter(queryScorer); Highlighter highlighter = new Highlighter(queryScorer); highlighter.setTextFragmenter(fragmenter); GroupingSearch gsearch = new GroupingSearch(IndexDocumentAdapter.FIELD_URL_GROUP).setGroupDocsLimit(1) .setAllGroups(true).setIncludeMaxScore(true); TopGroups<?> groups = gsearch.search(searcher, q, offset, size); ArrayList<SearchResult> results = new ArrayList<>(size); for (int i = offset; i < offset + size && i < groups.groups.length; i++) { ScoreDoc scoreDoc = groups.groups[i].scoreDocs[0]; Document luceneDoc = searcher.doc(scoreDoc.doc); IndexDocumentAdapter doc = new IndexDocumentAdapter(luceneDoc); TokenStream tokenStream = TokenSources.getTokenStream(IndexDocumentAdapter.FIELD_SEARCH, reader.getTermVectors(scoreDoc.doc), luceneDoc.get(IndexDocumentAdapter.FIELD_SEARCH), analyzer, highlighter.getMaxDocCharsToAnalyze() - 1); String[] snippets = highlighter.getBestFragments(tokenStream, luceneDoc.get(IndexDocumentAdapter.FIELD_SEARCH), 3); String snippet = Arrays.asList(snippets).stream().collect(Collectors.joining("\n")); snippet = Jsoup.clean(snippet, Whitelist.simpleText()); String debugInfo = null; if (includeDebug) { Explanation explanation = searcher.explain(q, scoreDoc.doc); debugInfo = explanation.toString(); } results.add(new SearchResult(doc.getKey(), doc.getCollection(), doc.getTitle(), doc.getUrl(), doc.getDomain(), doc.getTimestampText(), snippet, debugInfo, scoreDoc.score)); } SearchResultWrapper wrapper = new SearchResultWrapper().setQuery(query).setOffset(offset) .setMaxResultsRequested(size) .setResultCount(groups.totalGroupCount != null ? groups.totalGroupCount : 0) .setResults(results); if (includeDebug) { wrapper.setDebugInfo(q.toString()); } return wrapper; } catch (IOException | ParseException | InvalidTokenOffsetsException e) { LOG.error(e.getLocalizedMessage()); throw new IndexException(e); } }
From source file:com.doculibre.constellio.lucene.BaseLuceneIndexHelper.java
License:Open Source License
public String highlight(String strToHighlight, String fieldName, Query luceneQuery) { String highlightedText;/*from w w w .ja v a 2s . co m*/ Analyzer analyzer = analyzerProvider.getAnalyzer(Locale.FRENCH); try { Directory directory = FSDirectory.open(indexDir); IndexReader indexReader = DirectoryReader.open(directory); Query rewrittenLuceneQuery = luceneQuery.rewrite(indexReader); QueryScorer luceneScorer = new QueryScorer(rewrittenLuceneQuery); SimpleHTMLFormatter luceneFormatter = new SimpleHTMLFormatter("<span class=\"hit\">", "</span>"); Highlighter luceneHighlighter = new Highlighter(luceneFormatter, luceneScorer); Fragmenter luceneFragmenter; // Si la chaine highlighter est sup 250 carac if (strToHighlight.length() > TAILLE_CHAINE_NON_FRAGMENTEE) { // Cration de best fragments de 100 carac chaque luceneFragmenter = new SimpleFragmenter(TAILLE_FRAGMENT); } else { // Toute la chaine est highlight luceneFragmenter = new SimpleFragmenter(Integer.MAX_VALUE); } luceneHighlighter.setTextFragmenter(luceneFragmenter); TokenStream luceneTokenStream = analyzer.tokenStream(fieldName, new StringReader(strToHighlight)); String fragment = null; if (strToHighlight.length() > TAILLE_CHAINE_NON_FRAGMENTEE) { fragment = luceneHighlighter.getBestFragments(luceneTokenStream, strToHighlight, NB_BEST_FRAGMENT, FRAGMENT_SEP); } else { fragment = luceneHighlighter.getBestFragment(luceneTokenStream, strToHighlight); } if (StringUtils.isBlank(fragment) && fieldName.equalsIgnoreCase("titre")) { fragment = strToHighlight; } indexReader.close(); directory.close(); highlightedText = fragment; } catch (IOException e) { throw new RuntimeException(e); } catch (InvalidTokenOffsetsException e) { throw new RuntimeException(e); } return highlightedText; }
From source file:com.duroty.application.bookmark.manager.BookmarkManager.java
License:Open Source License
/** * DOCUMENT ME!/*from w w w .ja va2 s.c om*/ * * @param repositoryName DOCUMENT ME! * @param token DOCUMENT ME! * @param page DOCUMENT ME! * @param messagesByPage DOCUMENT ME! * @param order DOCUMENT ME! * @param orderType DOCUMENT ME! * * @return DOCUMENT ME! * * @throws BookmarkException DOCUMENT ME! * @throws SearchException DOCUMENT ME! */ public SearchObj search(String repositoryName, String token, int page, int messagesByPage, int order, String orderType, boolean isNotebook) throws BookmarkException { String lucenePath = ""; if (!defaultLucenePath.endsWith(File.separator)) { lucenePath = defaultLucenePath + File.separator + repositoryName + File.separator + Constants.BOOKMARK_LUCENE_BOOKMARK; } else { lucenePath = defaultLucenePath + repositoryName + File.separator + Constants.BOOKMARK_LUCENE_BOOKMARK; } Searcher searcher = null; SearchObj searchObj = new SearchObj(); Highlighter highlighter = null; try { searcher = BookmarkIndexer.getSearcher(lucenePath); Query query = null; Hits hits = null; if (StringUtils.isBlank(token)) { if (isNotebook) { query = SimpleQueryParser.parse("notebook:true", new KeywordAnalyzer()); } else { query = new MatchAllDocsQuery(); } hits = searcher.search(query, new Sort(new SortField[] { SortField.FIELD_SCORE, new SortField(Field_insert_date, SortField.STRING, true) })); } else { query = SimpleQueryParser.parse(token, analyzer); StringBuffer buffer = new StringBuffer(); if (isNotebook) { buffer.append("(" + query.toString() + ") AND "); QueryParser parser = new QueryParser(Field_notebook, new KeywordAnalyzer()); parser.setDefaultOperator(Operator.AND); Query aux = parser.parse(String.valueOf(true)); buffer.append("(" + aux.toString() + ") "); } if (buffer.length() > 0) { QueryParser parser = new QueryParser("", new WhitespaceAnalyzer()); query = parser.parse(buffer.toString()); } hits = searcher.search(query); } Date searchStart = new Date(); Date searchEnd = new Date(); //time in seconds double time = ((double) (searchEnd.getTime() - searchStart.getTime())) / (double) 1000; int hitsLength = hits.length(); if (hitsLength <= 0) { return null; } int start = page * messagesByPage; int end = start + messagesByPage; if (end > 0) { end = Math.min(hitsLength, end); } else { end = hitsLength; } if (start > end) { throw new SearchException("Search index of bound. start > end"); } Vector bookmarks = new Vector(); for (int j = start; j < end; j++) { Document doc = hits.doc(j); if (doc != null) { LuceneBookmark luceneBookmark = new LuceneBookmark(doc); SimpleHTMLFormatter formatter = new SimpleHTMLFormatter("<b>", "</b>"); highlighter = new Highlighter(formatter, new QueryScorer(query)); highlighter.setTextFragmenter(new SimpleFragmenter(150)); BookmarkObj bookmarkObj = new BookmarkObj(); bookmarkObj.setCacheDate(luceneBookmark.getCacheDate()); bookmarkObj.setComments(luceneBookmark.getComments()); bookmarkObj.setContents(luceneBookmark.getCotents()); bookmarkObj.setDepth(luceneBookmark.getDepth()); bookmarkObj.setFlagged(luceneBookmark.isFlagged()); bookmarkObj.setIdint(luceneBookmark.getIdint()); bookmarkObj.setInsertDate(luceneBookmark.getInsertDate()); bookmarkObj.setKeywords(luceneBookmark.getKeywords()); bookmarkObj.setNotebook(luceneBookmark.isNotebook()); bookmarkObj.setParent(Long.parseLong(luceneBookmark.getParent())); bookmarkObj.setTitle(luceneBookmark.getTitle()); bookmarkObj.setTitleHighlight(luceneBookmark.getTitle()); bookmarkObj.setUrl(luceneBookmark.getUrl()); String contents = luceneBookmark.getCotents(); String hcontents = null; if ((contents != null) && (!contents.trim().equals(""))) { contents = contents.replaceAll("\\s+", " "); TokenStream tokenStream = analyzer.tokenStream(Field_contents, new StringReader(contents)); hcontents = highlighter.getBestFragment(tokenStream, contents); if (hcontents != null) { contents = hcontents; } else { contents = null; } } bookmarkObj.setContentsHighlight(contents); String title = luceneBookmark.getTitle(); String htitle = null; if ((title != null) && (!title.trim().equals(""))) { title = title.replaceAll("\\s+", " "); TokenStream tokenStream = analyzer.tokenStream(Field_title, new StringReader(title)); htitle = highlighter.getBestFragment(tokenStream, title); if (htitle != null) { title = htitle; } } bookmarkObj.setTitleHighlight(title); bookmarks.addElement(bookmarkObj); } } searchObj.setHits(hitsLength); searchObj.setTime(time); searchObj.setBookmarks(bookmarks); } catch (Exception ex) { throw new SearchException(ex); } finally { } return searchObj; }
From source file:com.edgenius.wiki.search.service.AbstractSearchService.java
License:Open Source License
private int detach(IndexSearcher searcher, List<SearchResultItem> viewableMatchedResults, TopDocs hits, Query hlQuery, int from, int to, User user) throws IOException { Assert.isTrue(from <= to && from >= 0 && (to >= 0 || to == -1)); //For performance issue, we simply return total result set length without permission filter out. //This means is total length might be larger than the set that user can view, as some result will be filter out //if user doesn't have permission to see. int len = hits.totalHits; if (len > 0 && from < len) { to = to == -1 ? len : (len > to ? to : len); //security filter from return result List<Integer> resultIdx = new ArrayList<Integer>(); for (int idx = from; idx < to; idx++) { //does not include "to" , For example, from:to is 0:10, then return index is 0-9 //TODO: if page includes some result that invisible to user, it is better display message to tell user //some result is hidden for security reason. if (!isAllowView(searcher.doc(hits.scoreDocs[idx].doc), user)) continue; resultIdx.add(idx);//from w w w .j a va2 s . c o m } //create a highlighter for all fragment parser Formatter formatter = new SimpleHTMLFormatter("<span class=\"highlighter\">", "</span>"); Highlighter hl = null; if (hlQuery != null) { Scorer scorer = new QueryScorer(hlQuery); hl = new Highlighter(formatter, scorer); Fragmenter fragmenter = new SimpleFragmenter(FRAGMENT_LEN); hl.setTextFragmenter(fragmenter); } for (int idx : resultIdx) { SearchResultItem item = new SearchResultItem(); Document doc = searcher.doc(hits.scoreDocs[idx].doc); String docType = doc.get(FieldName.DOC_TYPE); //common items in search results item.setType(NumberUtils.toInt(docType)); item.setDatetime(doc.get(FieldName.UPDATE_DATE)); if (userReadingService != null && !new Integer(SharedConstants.SEARCH_USER).toString().equals(docType)) { String username = doc.get(FieldName.CONTRIBUTOR); User contirUser = userReadingService.getUserByName(username); if (contirUser != null) { item.setContributor(contirUser.getFullname()); item.setContributorUsername(username); } } if (Integer.valueOf(SharedConstants.SEARCH_PAGE).toString().equals(docType)) { String content = doc.get(FieldName.PAGE_CONTENT); item.setTitle(doc.get(FieldName.PAGE_TITLE)); item.setSpaceUname(doc.get(FieldName.UNSEARCH_SPACE_UNIXNAME)); //does set item.desc() as content, which maybe very big string. no necessary returned item.setFragment(createFragment(hl, StringUtil.join(" ", item.getTitle(), content))); } else if (Integer.valueOf(SharedConstants.SEARCH_COMMENT).toString().equals(docType)) { String content = doc.get(FieldName.CONTENT); item.setItemUid(doc.get(FieldName.COMMENT_UID)); item.setSpaceUname(doc.get(FieldName.UNSEARCH_SPACE_UNIXNAME)); item.setTitle(doc.get(FieldName.UNSEARCH_PAGE_TITLE)); //does set item.desc() as content, which maybe very big string. no necessary returned item.setFragment(createFragment(hl, content)); } else if (Integer.valueOf(SharedConstants.SEARCH_SPACE).toString().equals(docType)) { String title = doc.get(FieldName.SPACE_NAME); item.setTitle(title); item.setSpaceUname(doc.get(FieldName.SPACE_UNIXNAME)); item.setDesc(doc.get(FieldName.SPACE_DESC)); item.setFragment(createFragment(hl, StringUtil.join(" ", item.getTitle(), item.getDesc()))); } else if (Integer.valueOf(SharedConstants.SEARCH_WIDGET).toString().equals(docType)) { //wTitle-> title; wDesc-> desc; wTitle(could be pageTitle or markup title) ->spaceUname String widgetType = doc.get(FieldName.WIDGET_TYPE); String title = doc.get(FieldName.WIDGET_TITLE); //does content need transfer back?? so far no String content = doc.get(FieldName.WIDGET_CONTENT); if (WidgetModel.TYPE_PAGE_LINKER.equals(widgetType)) { //don't use as Highlighter fragment content = ""; } String desc = doc.get(FieldName.WIDGET_DESC); item.setDesc(desc); item.setTitle(title); //add little confuse field mapping :( item.setSpaceUname(doc.get(FieldName.WIDGET_KEY)); item.setItemUid(widgetType); item.setFragment(createFragment(hl, StringUtil.join(" ", item.getDesc(), content))); } else if (Integer.valueOf(SharedConstants.SEARCH_PAGE_TAG).toString().equals(docType)) { //page tag item.setTitle(doc.get(FieldName.PAGE_TAG_NAME)); item.setSpaceUname(doc.get(FieldName.UNSEARCH_SPACE_UNIXNAME)); item.setFragment(createFragment(hl, item.getTitle())); } else if (Integer.valueOf(SharedConstants.SEARCH_SPACE_TAG).toString().equals(docType)) { //space tag item.setTitle(doc.get(FieldName.SPACE_TAG_NAME)); item.setFragment(createFragment(hl, item.getTitle())); } else if (Integer.valueOf(SharedConstants.SEARCH_USER).toString().equals(docType)) { String username = doc.get(FieldName.USER_NAME); item.setTitle(username); String fullname = doc.get(FieldName.USER_FULLNAME); //hacker - contributor is current user fullname item.setContributor(fullname); if (userReadingService != null) item.setDesc(userReadingService.getUserByName(username).getSetting().getStatus()); item.setFragment(createFragment(hl, fullname)); } else if (Integer.valueOf(SharedConstants.SEARCH_ROLE).toString().equals(docType)) { item.setSpaceUname(doc.get(FieldName.ROLE_NAME)); item.setTitle(doc.get(FieldName.ROLE_DISPLAY_NAME)); item.setDesc(doc.get(FieldName.ROLE_DESC)); //item.setFragment(""); } else if (Integer.valueOf(SharedConstants.SEARCH_ATTACHMENT).toString().equals(docType)) { item.setTitle(doc.get(FieldName.FILE_NAME)); item.setDesc(doc.get(FieldName.FILE_COMMENT)); item.setItemUid(doc.get(FieldName.FILE_NODE_UUID)); item.setSpaceUname(doc.get(FieldName.UNSEARCH_SPACE_UNIXNAME)); String text = doc.get(FieldName.TEXT); //does not mark file content fragment, because it does not store in index String fragment = createFragment(hl, StringUtil.join(" ", item.getDesc(), text)); item.setFragment( (fragment == null || fragment.trim().length() == 0) ? ("Comment: " + item.getDesc()) : fragment); } viewableMatchedResults.add(item); } } return len; }
From source file:com.flaptor.hounder.searcher.SnippetSearcher.java
License:Apache License
/** * Adds snippets to the search results.//from w w w .j a va2 s. c o m * How stuff works: * For each 'group g' in provided GroupedSearchResults: * For each result in 'g': * Use the lucene highlighter to get the terms highlighted on the required field. * Then call getSnippet(...) to get the resulting snippet */ private void addSnippets(GroupedSearchResults res, String snippetOfField, int snippetLength, QueryScorer scorer, Formatter simpleHtmlFormatter) throws IOException { Highlighter highlighter = new Highlighter(simpleHtmlFormatter, scorer); highlighter.setTextFragmenter(NULL_FRAGMENTER); highlighter.setMaxDocCharsToAnalyze(Integer.MAX_VALUE); // make sure the whole text will be analyzed // Here we store every seen phrase. It is used to give less score to // recurrying phrases Set<String> usedSnippets = new HashSet<String>(); for (int j = 0; j < res.groups(); j++) { // for each group Vector<Document> resDocs = res.getGroup(j).last(); int docsLen = resDocs.size(); for (int i = 0; i < docsLen; i++) { // for each document on that group Document doc = resDocs.get(i); // get the document i String text = doc.get(snippetOfField); // text to be snippeted if (null == text) { logger.warn("Asked to snippet an unexisting field: " + snippetOfField); continue; } TokenStream tokenStream = queryParser.tokenStream(snippetOfField, new StringReader(text)); TextFragment[] fragments = highlighter.getBestTextFragments(tokenStream, text, false, 1); String result = null; if (null != fragments && 0 < fragments.length) { result = getSnippet(fragments[0].toString(), snippetLength, scorer, usedSnippets); } if (null == result || 0 == result.length()) { // if (emptySnippetsAllowed) { result = ""; } else { result = text.substring(0, Math.min(text.length(), snippetLength)); } } String snippetF = SNIPPET_FIELDNAME_PREFIX + snippetOfField; doc.add(new Field(snippetF, result.toString(), Field.Store.YES, Field.Index.NO)); } } }
From source file:com.gauronit.tagmata.core.Indexer.java
License:Open Source License
public ArrayList<CardSnapshot> search(String searchText, ArrayList<String> indexNames, boolean searchInTitle, boolean searchInTags, boolean searchInText, boolean superFuzzy) { ArrayList<CardSnapshot> cardSnaps = new ArrayList(); try {/* ww w . j a v a2 s.co m*/ ArrayList<IndexSearcher> searchers = new ArrayList<IndexSearcher>(); for (String indexName : indexNames) { IndexReader reader = IndexReader .open(FSDirectory.open(new File(indexDir + File.separator + indexName), new SimpleFSLockFactory(indexDir + File.separator + indexName))); IndexSearcher searcher = new IndexSearcher(reader); searchers.add(searcher); } BooleanQuery query = new BooleanQuery(); if (searchInTitle) { IndexerUtil.getTokenizedQuery(query, "title", searchText, superFuzzy); } if (searchInTags) { IndexerUtil.getTokenizedQuery(query, "tags", searchText, superFuzzy); } if (searchInText) { IndexerUtil.getTokenizedQuery(query, "text", searchText, superFuzzy); IndexerUtil.getTokenizedQuery(query, "analyzedText", searchText, superFuzzy); } for (IndexSearcher searcher : searchers) { TopScoreDocCollector collector = TopScoreDocCollector.create(10000, false); searcher.search(query, collector); ScoreDoc[] hits = collector.topDocs().scoreDocs; for (ScoreDoc hit : hits) { Document doc = searcher.doc(hit.doc); TokenStream stream = TokenSources.getTokenStream("text", doc.get("analyzedText"), new StandardAnalyzer(Version.LUCENE_20.LUCENE_35)); QueryScorer scorer = new QueryScorer(query, "analyzedText"); Fragmenter fragmenter = new SimpleSpanFragmenter(scorer, 20); Highlighter highlighter = new Highlighter(scorer); highlighter.setTextFragmenter(fragmenter); String[] fragments = highlighter.getBestFragments(stream, doc.get("text"), 5); String highlights = ""; for (String fragment : fragments) { highlights += fragment + "..."; } if (highlights.equals("")) { String text = doc.get("text"); if (text.length() > 100) { highlights += doc.get("text").substring(0, 100); } else { highlights += doc.get("text"); } } cardSnaps.add(new CardSnapshot(highlights, doc)); } searcher.getIndexReader().close(); searcher.close(); searcher = null; } } catch (Exception ex) { ex.printStackTrace(); } return cardSnaps; }
From source file:com.gitblit.LuceneExecutor.java
License:Apache License
/** * //from w w w .j av a 2s . co m * @param analyzer * @param query * @param content * @param result * @return * @throws IOException * @throws InvalidTokenOffsetsException */ private String getHighlightedFragment(Analyzer analyzer, Query query, String content, SearchResult result) throws IOException, InvalidTokenOffsetsException { if (content == null) { content = ""; } int fragmentLength = SearchObjectType.commit == result.type ? 512 : 150; QueryScorer scorer = new QueryScorer(query, "content"); Fragmenter fragmenter = new SimpleSpanFragmenter(scorer, fragmentLength); // use an artificial delimiter for the token String termTag = "!!--["; String termTagEnd = "]--!!"; SimpleHTMLFormatter formatter = new SimpleHTMLFormatter(termTag, termTagEnd); Highlighter highlighter = new Highlighter(formatter, scorer); highlighter.setTextFragmenter(fragmenter); String[] fragments = highlighter.getBestFragments(analyzer, "content", content, 3); if (ArrayUtils.isEmpty(fragments)) { if (SearchObjectType.blob == result.type) { return ""; } // clip commit message String fragment = content; if (fragment.length() > fragmentLength) { fragment = fragment.substring(0, fragmentLength) + "..."; } return "<pre class=\"text\">" + StringUtils.escapeForHtml(fragment, true) + "</pre>"; } // make sure we have unique fragments Set<String> uniqueFragments = new LinkedHashSet<String>(); for (String fragment : fragments) { uniqueFragments.add(fragment); } fragments = uniqueFragments.toArray(new String[uniqueFragments.size()]); StringBuilder sb = new StringBuilder(); for (int i = 0, len = fragments.length; i < len; i++) { String fragment = fragments[i]; String tag = "<pre class=\"text\">"; // resurrect the raw fragment from removing the artificial delimiters String raw = fragment.replace(termTag, "").replace(termTagEnd, ""); // determine position of the raw fragment in the content int pos = content.indexOf(raw); // restore complete first line of fragment int c = pos; while (c > 0) { c--; if (content.charAt(c) == '\n') { break; } } if (c > 0) { // inject leading chunk of first fragment line fragment = content.substring(c + 1, pos) + fragment; } if (SearchObjectType.blob == result.type) { // count lines as offset into the content for this fragment int line = Math.max(1, StringUtils.countLines(content.substring(0, pos))); // create fragment tag with line number and language String lang = ""; String ext = StringUtils.getFileExtension(result.path).toLowerCase(); if (!StringUtils.isEmpty(ext)) { // maintain leading space! lang = " lang-" + ext; } tag = MessageFormat.format("<pre class=\"prettyprint linenums:{0,number,0}{1}\">", line, lang); } sb.append(tag); // replace the artificial delimiter with html tags String html = StringUtils.escapeForHtml(fragment, false); html = html.replace(termTag, "<span class=\"highlight\">").replace(termTagEnd, "</span>"); sb.append(html); sb.append("</pre>"); if (i < len - 1) { sb.append("<span class=\"ellipses\">...</span><br/>"); } } return sb.toString(); }
From source file:com.gitblit.service.LuceneService.java
License:Apache License
/** * * @param analyzer/*w w w . j a v a2 s . c o m*/ * @param query * @param content * @param result * @return * @throws IOException * @throws InvalidTokenOffsetsException */ private String getHighlightedFragment(Analyzer analyzer, Query query, String content, SearchResult result) throws IOException, InvalidTokenOffsetsException { if (content == null) { content = ""; } int tabLength = storedSettings.getInteger(Keys.web.tabLength, 4); int fragmentLength = SearchObjectType.commit == result.type ? 512 : 150; QueryScorer scorer = new QueryScorer(query, "content"); Fragmenter fragmenter = new SimpleSpanFragmenter(scorer, fragmentLength); // use an artificial delimiter for the token String termTag = "!!--["; String termTagEnd = "]--!!"; SimpleHTMLFormatter formatter = new SimpleHTMLFormatter(termTag, termTagEnd); Highlighter highlighter = new Highlighter(formatter, scorer); highlighter.setTextFragmenter(fragmenter); String[] fragments = highlighter.getBestFragments(analyzer, "content", content, 3); if (ArrayUtils.isEmpty(fragments)) { if (SearchObjectType.blob == result.type) { return ""; } // clip commit message String fragment = content; if (fragment.length() > fragmentLength) { fragment = fragment.substring(0, fragmentLength) + "..."; } return "<pre class=\"text\">" + StringUtils.escapeForHtml(fragment, true, tabLength) + "</pre>"; } // make sure we have unique fragments Set<String> uniqueFragments = new LinkedHashSet<String>(); for (String fragment : fragments) { uniqueFragments.add(fragment); } fragments = uniqueFragments.toArray(new String[uniqueFragments.size()]); StringBuilder sb = new StringBuilder(); for (int i = 0, len = fragments.length; i < len; i++) { String fragment = fragments[i]; String tag = "<pre class=\"text\">"; // resurrect the raw fragment from removing the artificial delimiters String raw = fragment.replace(termTag, "").replace(termTagEnd, ""); // determine position of the raw fragment in the content int pos = content.indexOf(raw); // restore complete first line of fragment int c = pos; while (c > 0) { c--; if (content.charAt(c) == '\n') { break; } } if (c > 0) { // inject leading chunk of first fragment line fragment = content.substring(c + 1, pos) + fragment; } if (SearchObjectType.blob == result.type) { // count lines as offset into the content for this fragment int line = Math.max(1, StringUtils.countLines(content.substring(0, pos))); // create fragment tag with line number and language String lang = ""; String ext = StringUtils.getFileExtension(result.path).toLowerCase(); if (!StringUtils.isEmpty(ext)) { // maintain leading space! lang = " lang-" + ext; } tag = MessageFormat.format("<pre class=\"prettyprint linenums:{0,number,0}{1}\">", line, lang); } sb.append(tag); // replace the artificial delimiter with html tags String html = StringUtils.escapeForHtml(fragment, false); html = html.replace(termTag, "<span class=\"highlight\">").replace(termTagEnd, "</span>"); sb.append(html); sb.append("</pre>"); if (i < len - 1) { sb.append("<span class=\"ellipses\">...</span><br/>"); } } return sb.toString(); }
From source file:com.github.wxiaoqi.search.lucene.LuceneDao.java
License:Open Source License
private Highlighter addStringHighlighter(Query query) { QueryScorer scorer = new QueryScorer(query); Fragmenter fragmenter = new SimpleSpanFragmenter(scorer); SimpleHTMLFormatter simpleHTMLFormatter = new SimpleHTMLFormatter("<font color='red'>", "</font>"); Highlighter highlighter = new Highlighter(simpleHTMLFormatter, scorer); highlighter.setTextFragmenter(fragmenter); return highlighter; }