Examples with EnglishAnalyzer org.apache.lucene.analysis.en.EnglishAnalyzer used on opensource projects

Example 1 with EnglishAnalyzer

use of org.apache.lucene.analysis.en.EnglishAnalyzer in project lucene-solr by apache.

the class TestSolrSynonymParser method testInvalidPositionsInput.

/** parse a syn file with bad syntax */
public void testInvalidPositionsInput() throws Exception {
    String testFile = "testola => the test";
    Analyzer analyzer = new EnglishAnalyzer();
    SolrSynonymParser parser = new SolrSynonymParser(true, true, analyzer);
    expectThrows(ParseException.class, () -> {
        parser.parse(new StringReader(testFile));
    });
    analyzer.close();
}

Example 2 with EnglishAnalyzer

use of org.apache.lucene.analysis.en.EnglishAnalyzer in project lucene-solr by apache.

the class TestSolrSynonymParser method testInvalidPositionsOutput.

/** parse a syn file with bad syntax */
public void testInvalidPositionsOutput() throws Exception {
    String testFile = "the test => testola";
    Analyzer analyzer = new EnglishAnalyzer();
    SolrSynonymParser parser = new SolrSynonymParser(true, true, analyzer);
    expectThrows(ParseException.class, () -> {
        parser.parse(new StringReader(testFile));
    });
    analyzer.close();
}

Example 3 with EnglishAnalyzer

use of org.apache.lucene.analysis.en.EnglishAnalyzer in project Anserini by castorini.

the class SearchWebCollection method main.

public static void main(String[] args) throws Exception {
    SearchArgs searchArgs = new SearchArgs();
    CmdLineParser parser = new CmdLineParser(searchArgs, ParserProperties.defaults().withUsageWidth(90));
    try {
        parser.parseArgument(args);
    } catch (CmdLineException e) {
        System.err.println(e.getMessage());
        parser.printUsage(System.err);
        System.err.println("Example: SearchWebCollection" + parser.printExample(OptionHandlerFilter.REQUIRED));
        return;
    }
    LOG.info("Reading index at " + searchArgs.index);
    Directory dir;
    if (searchArgs.inmem) {
        LOG.info("Using MMapDirectory with preload");
        dir = new MMapDirectory(Paths.get(searchArgs.index));
        ((MMapDirectory) dir).setPreload(true);
    } else {
        LOG.info("Using default FSDirectory");
        dir = FSDirectory.open(Paths.get(searchArgs.index));
    }
    Similarity similarity = null;
    if (searchArgs.ql) {
        LOG.info("Using QL scoring model");
        similarity = new LMDirichletSimilarity(searchArgs.mu);
    } else if (searchArgs.bm25) {
        LOG.info("Using BM25 scoring model");
        similarity = new BM25Similarity(searchArgs.k1, searchArgs.b);
    } else {
        LOG.error("Error: Must specify scoring model!");
        System.exit(-1);
    }
    RerankerCascade cascade = new RerankerCascade();
    boolean useQueryParser = false;
    if (searchArgs.rm3) {
        cascade.add(new Rm3Reranker(new EnglishAnalyzer(), FIELD_BODY, "src/main/resources/io/anserini/rerank/rm3/rm3-stoplist.gov2.txt"));
        useQueryParser = true;
    } else {
        cascade.add(new IdentityReranker());
    }
    FeatureExtractors extractors = null;
    if (searchArgs.extractors != null) {
        extractors = FeatureExtractors.loadExtractor(searchArgs.extractors);
    }
    if (searchArgs.dumpFeatures) {
        PrintStream out = new PrintStream(searchArgs.featureFile);
        Qrels qrels = new Qrels(searchArgs.qrels);
        cascade.add(new WebCollectionLtrDataGenerator(out, qrels, extractors));
    }
    Path topicsFile = Paths.get(searchArgs.topics);
    if (!Files.exists(topicsFile) || !Files.isRegularFile(topicsFile) || !Files.isReadable(topicsFile)) {
        throw new IllegalArgumentException("Topics file : " + topicsFile + " does not exist or is not a (readable) file.");
    }
    TopicReader tr = (TopicReader) Class.forName("io.anserini.search.query." + searchArgs.topicReader + "TopicReader").getConstructor(Path.class).newInstance(topicsFile);
    SortedMap<Integer, String> topics = tr.read();
    final long start = System.nanoTime();
    SearchWebCollection searcher = new SearchWebCollection(searchArgs.index);
    searcher.search(topics, searchArgs.output, similarity, searchArgs.hits, cascade, useQueryParser, searchArgs.keepstop);
    searcher.close();
    final long durationMillis = TimeUnit.MILLISECONDS.convert(System.nanoTime() - start, TimeUnit.NANOSECONDS);
    LOG.info("Total " + topics.size() + " topics searched in " + DurationFormatUtils.formatDuration(durationMillis, "HH:mm:ss"));
}

Example 4 with EnglishAnalyzer

use of org.apache.lucene.analysis.en.EnglishAnalyzer in project Anserini by castorini.

the class IdfPassageScorer method getTermIdfJSON.

@Override
public JSONObject getTermIdfJSON(List<String> sentList) {
    EnglishAnalyzer ea = new EnglishAnalyzer(CharArraySet.EMPTY_SET);
    QueryParser qp = new QueryParser(LuceneDocumentGenerator.FIELD_BODY, ea);
    ClassicSimilarity similarity = new ClassicSimilarity();
    for (String sent : sentList) {
        String[] thisSentence = sent.trim().split("\\s+");
        for (String term : thisSentence) {
            try {
                TermQuery q = (TermQuery) qp.parse(term);
                Term t = q.getTerm();
                double termIDF = similarity.idf(reader.docFreq(t), reader.numDocs());
                termIdfMap.put(term, String.valueOf(termIDF));
            } catch (Exception e) {
                continue;
            }
        }
    }
    return new JSONObject(termIdfMap);
}

Example 5 with EnglishAnalyzer

use of org.apache.lucene.analysis.en.EnglishAnalyzer in project Anserini by castorini.

the class SearchTweets method main.

public static void main(String[] args) throws Exception {
    long initializationTime = System.currentTimeMillis();
    SearchArgs searchArgs = new SearchArgs();
    CmdLineParser parser = new CmdLineParser(searchArgs, ParserProperties.defaults().withUsageWidth(90));
    try {
        parser.parseArgument(args);
    } catch (CmdLineException e) {
        System.err.println(e.getMessage());
        parser.printUsage(System.err);
        System.err.println("Example: SearchTweets" + parser.printExample(OptionHandlerFilter.REQUIRED));
        return;
    }
    LOG.info("Reading index at " + searchArgs.index);
    Directory dir;
    if (searchArgs.inmem) {
        LOG.info("Using MMapDirectory with preload");
        dir = new MMapDirectory(Paths.get(searchArgs.index));
        ((MMapDirectory) dir).setPreload(true);
    } else {
        LOG.info("Using default FSDirectory");
        dir = FSDirectory.open(Paths.get(searchArgs.index));
    }
    IndexReader reader = DirectoryReader.open(dir);
    IndexSearcher searcher = new IndexSearcher(reader);
    if (searchArgs.ql) {
        LOG.info("Using QL scoring model");
        searcher.setSimilarity(new LMDirichletSimilarity(searchArgs.mu));
    } else if (searchArgs.bm25) {
        LOG.info("Using BM25 scoring model");
        searcher.setSimilarity(new BM25Similarity(searchArgs.k1, searchArgs.b));
    } else {
        LOG.error("Error: Must specify scoring model!");
        System.exit(-1);
    }
    RerankerCascade cascade = new RerankerCascade();
    EnglishAnalyzer englishAnalyzer = new EnglishAnalyzer();
    if (searchArgs.rm3) {
        cascade.add(new Rm3Reranker(englishAnalyzer, FIELD_BODY, "src/main/resources/io/anserini/rerank/rm3/rm3-stoplist.twitter.txt"));
        cascade.add(new RemoveRetweetsTemporalTiebreakReranker());
    } else {
        cascade.add(new RemoveRetweetsTemporalTiebreakReranker());
    }
    if (!searchArgs.model.isEmpty() && searchArgs.extractors != null) {
        LOG.debug(String.format("Ranklib model used, modeled loaded from %s", searchArgs.model));
        cascade.add(new RankLibReranker(searchArgs.model, FIELD_BODY, searchArgs.extractors));
    }
    FeatureExtractors extractorChain = null;
    if (searchArgs.extractors != null) {
        extractorChain = FeatureExtractors.loadExtractor(searchArgs.extractors);
    }
    if (searchArgs.dumpFeatures) {
        PrintStream out = new PrintStream(searchArgs.featureFile);
        Qrels qrels = new Qrels(searchArgs.qrels);
        cascade.add(new TweetsLtrDataGenerator(out, qrels, extractorChain));
    }
    MicroblogTopicSet topics = MicroblogTopicSet.fromFile(new File(searchArgs.topics));
    PrintStream out = new PrintStream(new FileOutputStream(new File(searchArgs.output)));
    LOG.info("Writing output to " + searchArgs.output);
    LOG.info("Initialized complete! (elapsed time = " + (System.currentTimeMillis() - initializationTime) + "ms)");
    long totalTime = 0;
    int cnt = 0;
    for (MicroblogTopic topic : topics) {
        long curQueryTime = System.currentTimeMillis();
        // do not cosider the tweets with tweet ids that are beyond the queryTweetTime
        // <querytweettime> tag contains the timestamp of the query in terms of the
        // chronologically nearest tweet id within the corpus
        Query filter = TermRangeQuery.newStringRange(FIELD_ID, "0", String.valueOf(topic.getQueryTweetTime()), true, true);
        Query query = AnalyzerUtils.buildBagOfWordsQuery(FIELD_BODY, englishAnalyzer, topic.getQuery());
        BooleanQuery.Builder builder = new BooleanQuery.Builder();
        builder.add(filter, BooleanClause.Occur.FILTER);
        builder.add(query, BooleanClause.Occur.MUST);
        Query q = builder.build();
        TopDocs rs = searcher.search(q, searchArgs.hits);
        List<String> queryTokens = AnalyzerUtils.tokenize(englishAnalyzer, topic.getQuery());
        RerankerContext context = new RerankerContext(searcher, query, topic.getId(), topic.getQuery(), queryTokens, FIELD_BODY, filter);
        ScoredDocuments docs = cascade.run(ScoredDocuments.fromTopDocs(rs, searcher), context);
        long queryTime = (System.currentTimeMillis() - curQueryTime);
        for (int i = 0; i < docs.documents.length; i++) {
            String qid = topic.getId().replaceFirst("^MB0*", "");
            out.println(String.format("%s Q0 %s %d %f %s", qid, docs.documents[i].getField(FIELD_ID).stringValue(), (i + 1), docs.scores[i], searchArgs.runtag));
        }
        LOG.info("Query " + topic.getId() + " (elapsed time = " + queryTime + "ms)");
        totalTime += queryTime;
        cnt++;
    }
    LOG.info("All queries completed!");
    LOG.info("Total elapsed time = " + totalTime + "ms");
    LOG.info("Average query latency = " + (totalTime / cnt) + "ms");
    reader.close();
    out.close();
}