List of usage examples for org.apache.lucene.search.highlight Highlighter Highlighter
public Highlighter(Formatter formatter, Scorer fragmentScorer)
From source file:com.aistor.common.persistence.BaseDaoImpl.java
License:Open Source License
/** * // www .j av a 2s . c o m * @param query * @param list * @param fields ?? */ public List<T> keywordsHighlight(BooleanQuery query, List<T> list, String... fields) { Analyzer analyzer = new IKAnalyzer(); Formatter formatter = new SimpleHTMLFormatter("<span class=\"highlight\">", "</span>"); Highlighter highlighter = new Highlighter(formatter, new QueryScorer(query)); highlighter.setTextFragmenter(new SimpleFragmenter(130)); for (T entity : list) { try { for (String field : fields) { String text = StringUtils.replaceHtml((String) Reflections.invokeGetter(entity, field)); String desciption = highlighter.getBestFragment(analyzer, field, text); if (desciption != null) { Reflections.invokeSetter(entity, fields[0], desciption); break; } Reflections.invokeSetter(entity, fields[0], StringUtils.abbr(text, 130)); } //Reflections.invokeSetter(entity, fields[1], "sdfkjsdlkfjklsdjf"); } catch (IOException e) { e.printStackTrace(); } catch (InvalidTokenOffsetsException e) { e.printStackTrace(); } } return list; }
From source file:com.appeligo.alerts.KeywordAlertThread.java
License:Apache License
/** * @param searchExecutor callback to get the set of hits for the given query. This can be * executed in different ways.//from w w w .j a v a2s. c o m * @return true if we hit too many consecutive exceptions so we broke out of the loop */ private boolean executeKeywordSearch(SearchExecutor searchExecutor, String messagePrefix, boolean groupQueries) { ChunkedResults<KeywordAlert> results = KeywordAlert.getAllInNormalizedQueryOrder(); Hits hits = null; String lastNormalizedQuery = null; Query lastLuceneQuery = null; int consecutiveExceptions = 0; results.beforeFirst(); while (results.next() && isActive()) { KeywordAlert keywordAlert = results.get(); try { if (keywordAlert.isDeleted() || keywordAlert.isDisabled()) { if (log.isDebugEnabled()) log.debug("keyword alert is deleted or disabled"); continue; } User user = keywordAlert.getUser(); if (user == null) { if (log.isDebugEnabled()) log.debug("keyword alert is implicitly deleted (user is null)"); keywordAlert.setDeleted(true); keywordAlert.save(); continue; } if (helper.maxAlertsExceeded(keywordAlert)) { continue; } if (groupQueries) { if ((hits == null) || (!keywordAlert.getNormalizedQuery().equals(lastNormalizedQuery))) { hits = searchExecutor.search(null, keywordAlert.getNormalizedQuery()); lastLuceneQuery = searchExecutor.getLuceneQuery(); } else if (log.isDebugEnabled()) log.debug("Not searching on " + keywordAlert.getNormalizedQuery() + " again"); } else { hits = searchExecutor.search(keywordAlert.getUser().getLineupId(), keywordAlert.getNormalizedQuery()); // Note that I'm searching with the lineup from the user, which will // only ensure that the liveIndex doesn't return shows that don't ever // play for this lineup. However, it does not guarantee that the show // on this user's lineup is playing at the same time (meaning alerts // might tell the user of a show that is only in the future). lastLuceneQuery = searchExecutor.getLuceneQuery(); } lastNormalizedQuery = keywordAlert.getNormalizedQuery(); Highlighter highlighter = new Highlighter(new TermFormatter(), new QueryScorer(lastLuceneQuery)); PorterStemAnalyzer analyzer = new PorterStemAnalyzer(LuceneIndexer.STOP_WORDS); for (int i = 0; i < hits.length(); i++) { Document doc = hits.doc(i); if (!isActive()) { break; } // if (groupQueries && (!"true".equals(doc.get("lineup-"+keywordAlert.getUser().getLineupId())))) { if (groupQueries && (doc.get("lineup-" + keywordAlert.getUser().getLineupId() + "-startTime") == null)) { // This "if" statement checks to make sure the program is or did play on the user's // lineup, which might be on a different station, a different time, past or future. if (log.isDebugEnabled()) log.debug(doc.get("programTitle") + " matched on " + keywordAlert.getNormalizedQuery() + " but it isn't airing on this user's lineup anytime soon."); continue; } Transaction transaction = HibernateUtil.currentSession().beginTransaction(); try { if ((!helper.maxAlertsExceeded(keywordAlert)) && helper.isNewMatch(keywordAlert, doc)) { if (log.isDebugEnabled()) log.debug("KeywordAlertThread found match in " + doc.get("programTitle") + " for " + keywordAlert.getNormalizedQuery() + "... sending messages"); String text = doc.get("text"); String fragments = null; if (text != null) { TokenStream tokenStream = analyzer.tokenStream("text", new StringReader(text)); fragments = highlighter.getBestFragments(tokenStream, text, 3, "..."); } helper.incrementTodaysAlertCount(keywordAlert); helper.sendMessages(keywordAlert, fragments, doc, messagePrefix); } else if (log.isDebugEnabled()) log.debug("KeywordAlertThread found match in " + doc.get("programTitle") + " for " + keywordAlert.getNormalizedQuery() + " but max exceeded or we already matched this one"); } catch (Throwable t) { log.error( "Error processing keyword alerts when searching live lucene index. Rolling back transaction.", t); transaction.rollback(); } finally { if (!transaction.wasRolledBack()) { transaction.commit(); } } } consecutiveExceptions = 0; } catch (Throwable t) { User user = keywordAlert.getUser(); log.error("Caught throwable on keyword " + keywordAlert.getId() + ", " + keywordAlert.getUserQuery() + ", user " + ((user == null) ? null : user.getUsername()), t); consecutiveExceptions++; if (consecutiveExceptions >= maxConsecutiveExceptions) { return true; } } } return false; }
From source file:com.appeligo.search.actions.SearchResults.java
License:Apache License
public List<SearchResult> getSearchResults(int startIndex) { initializeStatics();// w w w . j a v a 2s .co m hasMoreResults = false; try { IndexSearcher searcher = null; try { searcher = newIndexSearcher(); IndexReader reader = searcher.getIndexReader(); Query luceneQuery = generateLuceneQuery(searcher); luceneQuery = luceneQuery.rewrite(reader); Hits hits = searcher.search(luceneQuery); usingSuggestedQuery = false; suggestedQuery = null; if ((didYouMeanParser != null) && ((hits.length() < minimumHits) || (calcScore(searcher, getQuery()) < minimumScore))) { if (log.isDebugEnabled()) { log.debug("Need to suggest because either num hits " + hits.length() + " < " + minimumHits + "\n or top hit score " + (hits.length() > 0 ? hits.score(0) : "[NO HITS]") + " < " + minimumScore); } IndexSearcher compositeSearcher = new IndexSearcher(compositeIndexLocation); try { log.debug("calling suggest() with query=" + getQuery() + " and composite index from " + compositeIndexLocation); //Query didYouMean = didYouMeanParser.suggest(getQuery(), compositeSearcher.getIndexReader()); Query suggestedQueries[] = didYouMeanParser.getSuggestions(getQuery(), compositeSearcher.getIndexReader()); TreeSet<Suggestion> suggestions = new TreeSet<Suggestion>(); if (suggestedQueries != null) { for (int i = 0; i < suggestedQueries.length; i++) { log.debug("trying suggested query: " + suggestedQueries[i].toString(defaultField)); String suggestedQueryString = suggestedQueries[i].toString(defaultField); String constrainedQueryString = suggestedQueryString; if (constrainedQueryString.indexOf('"') < 0 && constrainedQueryString.indexOf('\'') < 0) { constrainedQueryString = "\"" + constrainedQueryString + "\"~5"; // proximity/distance query (within 5 words of each other) } Query suggestedLuceneQuery = generateLuceneQuery(constrainedQueryString, searcher); suggestedLuceneQuery = suggestedLuceneQuery.rewrite(reader); Hits suggestedHits = searcher.search(suggestedLuceneQuery); float score = calcScore(suggestedQueryString, suggestedHits); log.debug("========================================="); log.debug("SCORE = " + score); log.debug("========================================="); suggestions.add( new Suggestion(suggestedQueryString, suggestedLuceneQuery, suggestedHits, score, ((i == 0) ? didYouMeanParser.includesOriginal() : false))); log.debug("hits=" + suggestedHits.length() + ", score=" + score); } } Suggestion best = null; if (suggestions.size() > 0) { best = suggestions.last(); } if (best != null && !best.isOriginal()) { suggestedQuery = best.getQueryString(); if (suggestedQuery != null && suggestedQuery.indexOf('+') >= 0 && getQuery().indexOf('+') < 0) { suggestedQuery = suggestedQuery.replaceAll("\\+", ""); } if (hits.length() == 0) { if (best.getHits().length() > 0) { // Requery probably required because we added proximity before String suggestedQueryString = best.getQueryString(); luceneQuery = generateLuceneQuery(suggestedQueryString, searcher); luceneQuery = luceneQuery.rewrite(reader); hits = searcher.search(luceneQuery); //hits = best.getHits(); //luceneQuery = best.getLuceneQuery(); usingSuggestedQuery = true; } } log.debug("DidYouMeanParser suggested " + suggestedQuery); } else { if (best != null && best.isOriginal()) { log.debug("The suggestion was the original query after all"); } log.debug("DidYouMeanParser did not suggest anything"); } } finally { compositeSearcher.close(); } } /* if (hits.length() == 0 && suggestedQuery != null) { // If we didn't find anything at all, go ahead and show them what the suggested query // will give them Query suggestedLuceneQuery = generateLuceneQuery(suggestedQuery, searcher); suggestedLuceneQuery = suggestedLuceneQuery.rewrite(reader); Hits suggestedHits = searcher.search(suggestedLuceneQuery); if (suggestedHits.length() > 0) { hits = suggestedHits; luceneQuery = suggestedLuceneQuery; usingSuggestedQuery = true; } } */ totalHits = hits.length(); //Get the genere matches: try { BitSetFacetHitCounter facetHitCounter = new BitSetFacetHitCounter(); facetHitCounter.setSearcher(searcher); String baseQueryString = (isUsingSuggestedQuery() ? suggestedQuery : query); String quotedQueryString = baseQueryString; if (quotedQueryString.indexOf('"') == -1 && quotedQueryString.indexOf(' ') > -1) { quotedQueryString = "\"" + quotedQueryString + "\""; } facetHitCounter.setBaseQuery(luceneQuery, baseQueryString); List<HitCount> subQueries = new ArrayList<HitCount>(); for (Map.Entry<String, Query> entry : genreQueries.entrySet()) { subQueries.add( new HitCount(entry.getKey(), entry.getValue(), entry.getValue().toString(), 0)); } facetHitCounter.setSubQueries(subQueries); genreCounts = facetHitCounter.getFacetHitCounts(true); whatMatchedCounts = new ArrayList<HitCount>(); whatMatchedCounts .add(new HitCount("Title", getFieldQuery(baseQueryString, "programTitle", searcher), "programTitle:" + quotedQueryString, 0)); whatMatchedCounts.add( new HitCount("Episode Title", getFieldQuery(baseQueryString, "episodeTitle", searcher), "episodeTitle:" + quotedQueryString, 0)); whatMatchedCounts.add( new HitCount("Description", getFieldQuery(baseQueryString, "description", searcher), "description:" + quotedQueryString, 0)); whatMatchedCounts.add(new HitCount("Content", getFieldQuery(baseQueryString, "text", searcher), "text:" + quotedQueryString, 0)); whatMatchedCounts .add(new HitCount("Credits", getFieldQuery(baseQueryString, "credits", searcher), "credits:" + quotedQueryString, 0)); facetHitCounter.setSubQueries(whatMatchedCounts); whatMatchedCounts = facetHitCounter.getFacetHitCounts(true); //Program Count -- Not sure if there is a better way to do this. HashSet<String> programTitles = new HashSet<String>(); programCounts = new ArrayList<HitCount>(); for (int i = 0; i < hits.length() && programCounts.size() < 5; i++) { String title = hits.doc(i).get("programTitle"); if (!programTitles.contains(title)) { String queryTitle = title; queryTitle = QueryParser.escape(title); if (queryTitle.indexOf('"') > -1) { queryTitle.replace("\"", "\\\""); } if (queryTitle.indexOf(' ') > -1) { queryTitle = "\"" + queryTitle + "\""; } programCounts .add(new HitCount(title, getFieldQuery(queryTitle, "programTitle", searcher), "programTitle:" + queryTitle, 0)); programTitles.add(title); } } facetHitCounter.setSubQueries(programCounts); programCounts = facetHitCounter.getFacetHitCounts(false); } catch (Exception e) { e.printStackTrace(); } results = new ArrayList<SearchResult>(); programToSearchResult.clear(); Query userQuery = getContentQuery(query, searcher); userQuery.rewrite(reader); Highlighter highlighter = new Highlighter(new TermFormatter(), new QueryScorer(userQuery, "text")); log.debug("#hits=" + hits.length()); EPGProvider epgProvider = DefaultEpg.getInstance(); boolean missingWebPaths = false; // We added this to the index midstream, so some do and some don't. // Next index rebuild, and they'll all have it. for (int i = 0; i < pageSize && i + startIndex < hits.length(); i++) { if (hits.doc(i + startIndex).get("webPath") == null) { missingWebPaths = true; break; } } Program[] programs = null; if (missingWebPaths) { List<String> programIds = new ArrayList<String>(pageSize); for (int i = 0; i < pageSize && i + startIndex < hits.length(); i++) { programIds.add(hits.doc(i + startIndex).get("programID")); } programs = DefaultEpg.getInstance().getProgramList(programIds); } for (int i = 0; i < pageSize && i + startIndex < hits.length(); i++) { addDocument(hits.doc(i + startIndex), hits.score(i + startIndex), epgProvider, highlighter, analyzer, null, null, (programs == null ? null : programs[i])); } if (results.size() + startIndex < hits.length()) { hasMoreResults = true; } } finally { if (searcher != null) { searcher.close(); } } } catch (IOException e) { log.error("Error searching index", e); } catch (ParseException e) { log.error("Error searching index", e); } return results; }
From source file:com.bewsia.script.safe.lucene.SEntity.java
License:Open Source License
public String highlight(Query query, String text, String field, int fragmentSize, int maxNumFragments, String separator) throws Exception { Analyzer analyzer = new StandardAnalyzer(Version.LUCENE_36); CachingTokenFilter tokenStream = new CachingTokenFilter( analyzer.tokenStream(field, new StringReader(text))); SimpleHTMLFormatter formatter = new SimpleHTMLFormatter(); Scorer scorer = new org.apache.lucene.search.highlight.QueryScorer(query); Highlighter highlighter = new Highlighter(formatter, scorer); highlighter.setTextFragmenter(new SimpleFragmenter(fragmentSize)); tokenStream.reset();// ww w. j a v a2s . c o m String rv = highlighter.getBestFragments(tokenStream, text, maxNumFragments, separator); return rv.length() == 0 ? text : rv; }
From source file:com.bluedragon.search.search.QueryRun.java
License:Open Source License
private void addRow(IndexSearcher searcher, int docid, float score, int rank, int searchCount, int recordsSearched) throws CorruptIndexException, Exception { DocumentWrap document = new DocumentWrap(searcher.doc(docid)); queryResultData.addRow(1);//from w w w . j a v a 2s . c o m queryResultData.setCurrentRow(queryResultData.getSize()); // Add in the standard columns that we know we have for every search queryResultData.setCell(1, new cfStringData(document.getId())); queryResultData.setCell(2, new cfStringData(document.getName())); queryResultData.setCell(3, new cfNumberData(score)); queryResultData.setCell(4, new cfNumberData(searchCount)); queryResultData.setCell(5, new cfNumberData(recordsSearched)); queryResultData.setCell(6, new cfNumberData(rank + 1)); String uC = queryAttributes.getUniqueColumn(); // Now we do the custom ones List<IndexableField> fields = document.getDocument().getFields(); Iterator<IndexableField> it = fields.iterator(); while (it.hasNext()) { IndexableField fieldable = it.next(); String fieldName = fieldable.name().toLowerCase(); // Check for the unique if (uniqueSet != null && fieldName.equals(uC)) { if (uniqueSet.contains(fieldable.stringValue())) { queryResultData.deleteRow(queryResultData.getSize()); return; } else uniqueSet.add(fieldable.stringValue()); } // Check to see if we have this column if (fieldName.equals("contents") && !queryAttributes.getContentFlag()) continue; if (!activeColumns.containsKey(fieldName)) { int newcolumn = queryResultData.addColumnData(fieldable.name().toUpperCase(), cfArrayData.createArray(1), null); activeColumns.put(fieldName, newcolumn); } int column = activeColumns.get(fieldName); if (column <= 6) continue; queryResultData.setCell(column, new cfStringData(fieldable.stringValue())); } // Do the context stuff if enable if (queryAttributes.getContextPassages() > 0) { Scorer scorer = new QueryScorer(queryAttributes.getQuery()); SimpleHTMLFormatter formatter = new SimpleHTMLFormatter(queryAttributes.getContextHighlightStart(), queryAttributes.getContextHighlightEnd()); Highlighter highlighter = new Highlighter(formatter, scorer); Fragmenter fragmenter = new SimpleFragmenter(queryAttributes.getContextBytes()); highlighter.setTextFragmenter(fragmenter); String nextContext = ""; String contents = document.getAttribute(DocumentWrap.CONTENTS); if (contents != null) { TokenStream tokenStream = AnalyzerFactory.get("simple").tokenStream(DocumentWrap.CONTENTS, new StringReader(contents)); String[] fragments = null; try { fragments = highlighter.getBestFragments(tokenStream, contents, queryAttributes.getContextPassages()); if (fragments.length == 1) { nextContext = fragments[0] + "..."; } else { StringBuilder context = new StringBuilder(); for (int f = 0; f < fragments.length; f++) { context.append("..."); context.append(fragments[f]); } context.append("..."); nextContext = context.toString(); } } catch (Exception e) { } // Add in the context if (!activeColumns.containsKey("context")) { int newcolumn = queryResultData.addColumnData("CONTEXT", cfArrayData.createArray(1), null); activeColumns.put("context", newcolumn); } queryResultData.setCell(activeColumns.get("context"), new cfStringData(nextContext)); } } }
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.doculibre.constellio.lucene.BaseLuceneIndexHelper.java
License:Open Source License
public String highlight(String strToHighlight, String fieldName, Query luceneQuery) { String highlightedText;/*ww w . j a va2s. c o 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!/* w w w .jav a2 s .com*/ * * @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);// w w w.j ava 2 s . c om } //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 va 2s. co 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)); } } }