Examples with IntPoint org.apache.lucene.document.IntPoint used on opensource projects

Example 66 with IntPoint

use of org.apache.lucene.document.IntPoint in project lucene-solr by apache.

the class LineFileDocs method nextDoc.

/** Note: Document instance is re-used per-thread */
public Document nextDoc() throws IOException {
    String line;
    synchronized (this) {
        line = reader.readLine();
        if (line == null) {
            // Always rewind at end:
            if (LuceneTestCase.VERBOSE) {
                System.out.println("TEST: LineFileDocs: now rewind file...");
            }
            reader.close();
            reader = null;
            open(null);
            line = reader.readLine();
        }
    }
    DocState docState = threadDocs.get();
    if (docState == null) {
        docState = new DocState();
        threadDocs.set(docState);
    }
    int spot = line.indexOf(SEP);
    if (spot == -1) {
        throw new RuntimeException("line: [" + line + "] is in an invalid format !");
    }
    int spot2 = line.indexOf(SEP, 1 + spot);
    if (spot2 == -1) {
        throw new RuntimeException("line: [" + line + "] is in an invalid format !");
    }
    docState.body.setStringValue(line.substring(1 + spot2, line.length()));
    final String title = line.substring(0, spot);
    docState.title.setStringValue(title);
    if (docState.titleDV != null) {
        docState.titleDV.setBytesValue(new BytesRef(title));
    }
    docState.titleTokenized.setStringValue(title);
    docState.date.setStringValue(line.substring(1 + spot, spot2));
    final int i = id.getAndIncrement();
    docState.id.setStringValue(Integer.toString(i));
    docState.idNum.setIntValue(i);
    if (docState.idNumDV != null) {
        docState.idNumDV.setLongValue(i);
    }
    if (random.nextInt(5) == 4) {
        // Make some sparse fields
        Document doc = new Document();
        for (IndexableField field : docState.doc) {
            doc.add(field);
        }
        if (random.nextInt(3) == 1) {
            int x = random.nextInt(4);
            doc.add(new IntPoint("docLength" + x, line.length()));
        }
        if (random.nextInt(3) == 1) {
            int x = random.nextInt(4);
            doc.add(new IntPoint("docTitleLength" + x, title.length()));
        }
        if (random.nextInt(3) == 1) {
            int x = random.nextInt(4);
            doc.add(new NumericDocValuesField("docLength" + x, line.length()));
        }
    // TODO: more random sparse fields here too
    }
    return docState.doc;
}

Example 67 with IntPoint

use of org.apache.lucene.document.IntPoint in project lucene-solr by apache.

the class TestIndexWriterExceptions2 method testBasics.

// just one thread, serial merge policy, hopefully debuggable
public void testBasics() throws Exception {
    // disable slow things: we don't rely upon sleeps here.
    Directory dir = newDirectory();
    if (dir instanceof MockDirectoryWrapper) {
        ((MockDirectoryWrapper) dir).setThrottling(MockDirectoryWrapper.Throttling.NEVER);
        ((MockDirectoryWrapper) dir).setUseSlowOpenClosers(false);
    }
    // log all exceptions we hit, in case we fail (for debugging)
    ByteArrayOutputStream exceptionLog = new ByteArrayOutputStream();
    PrintStream exceptionStream = new PrintStream(exceptionLog, true, "UTF-8");
    //PrintStream exceptionStream = System.out;
    // create lots of non-aborting exceptions with a broken analyzer
    final long analyzerSeed = random().nextLong();
    Analyzer analyzer = new Analyzer() {

        @Override
        protected TokenStreamComponents createComponents(String fieldName) {
            MockTokenizer tokenizer = new MockTokenizer(MockTokenizer.SIMPLE, false);
            // TODO: can we turn this on? our filter is probably too evil
            tokenizer.setEnableChecks(false);
            TokenStream stream = tokenizer;
            // emit some payloads
            if (fieldName.contains("payloads")) {
                stream = new MockVariableLengthPayloadFilter(new Random(analyzerSeed), stream);
            }
            stream = new CrankyTokenFilter(stream, new Random(analyzerSeed));
            return new TokenStreamComponents(tokenizer, stream);
        }
    };
    // create lots of aborting exceptions with a broken codec
    // we don't need a random codec, as we aren't trying to find bugs in the codec here.
    Codec inner = RANDOM_MULTIPLIER > 1 ? Codec.getDefault() : new AssertingCodec();
    Codec codec = new CrankyCodec(inner, new Random(random().nextLong()));
    IndexWriterConfig conf = newIndexWriterConfig(analyzer);
    // just for now, try to keep this test reproducible
    conf.setMergeScheduler(new SerialMergeScheduler());
    conf.setCodec(codec);
    int numDocs = atLeast(500);
    IndexWriter iw = new IndexWriter(dir, conf);
    try {
        boolean allowAlreadyClosed = false;
        for (int i = 0; i < numDocs; i++) {
            // TODO: add crankyDocValuesFields, etc
            Document doc = new Document();
            doc.add(newStringField("id", Integer.toString(i), Field.Store.NO));
            doc.add(new NumericDocValuesField("dv", i));
            doc.add(new BinaryDocValuesField("dv2", new BytesRef(Integer.toString(i))));
            doc.add(new SortedDocValuesField("dv3", new BytesRef(Integer.toString(i))));
            doc.add(new SortedSetDocValuesField("dv4", new BytesRef(Integer.toString(i))));
            doc.add(new SortedSetDocValuesField("dv4", new BytesRef(Integer.toString(i - 1))));
            doc.add(new SortedNumericDocValuesField("dv5", i));
            doc.add(new SortedNumericDocValuesField("dv5", i - 1));
            doc.add(newTextField("text1", TestUtil.randomAnalysisString(random(), 20, true), Field.Store.NO));
            // ensure we store something
            doc.add(new StoredField("stored1", "foo"));
            doc.add(new StoredField("stored1", "bar"));
            // ensure we get some payloads
            doc.add(newTextField("text_payloads", TestUtil.randomAnalysisString(random(), 6, true), Field.Store.NO));
            // ensure we get some vectors
            FieldType ft = new FieldType(TextField.TYPE_NOT_STORED);
            ft.setStoreTermVectors(true);
            doc.add(newField("text_vectors", TestUtil.randomAnalysisString(random(), 6, true), ft));
            doc.add(new IntPoint("point", random().nextInt()));
            doc.add(new IntPoint("point2d", random().nextInt(), random().nextInt()));
            if (random().nextInt(10) > 0) {
                // single doc
                try {
                    iw.addDocument(doc);
                    // we made it, sometimes delete our doc, or update a dv
                    int thingToDo = random().nextInt(4);
                    if (thingToDo == 0) {
                        iw.deleteDocuments(new Term("id", Integer.toString(i)));
                    } else if (thingToDo == 1) {
                        iw.updateNumericDocValue(new Term("id", Integer.toString(i)), "dv", i + 1L);
                    } else if (thingToDo == 2) {
                        iw.updateBinaryDocValue(new Term("id", Integer.toString(i)), "dv2", new BytesRef(Integer.toString(i + 1)));
                    }
                } catch (AlreadyClosedException ace) {
                    // OK: writer was closed by abort; we just reopen now:
                    assertTrue(iw.deleter.isClosed());
                    assertTrue(allowAlreadyClosed);
                    allowAlreadyClosed = false;
                    conf = newIndexWriterConfig(analyzer);
                    // just for now, try to keep this test reproducible
                    conf.setMergeScheduler(new SerialMergeScheduler());
                    conf.setCodec(codec);
                    iw = new IndexWriter(dir, conf);
                } catch (Exception e) {
                    if (e.getMessage() != null && e.getMessage().startsWith("Fake IOException")) {
                        exceptionStream.println("\nTEST: got expected fake exc:" + e.getMessage());
                        e.printStackTrace(exceptionStream);
                        allowAlreadyClosed = true;
                    } else {
                        Rethrow.rethrow(e);
                    }
                }
            } else {
                // block docs
                Document doc2 = new Document();
                doc2.add(newStringField("id", Integer.toString(-i), Field.Store.NO));
                doc2.add(newTextField("text1", TestUtil.randomAnalysisString(random(), 20, true), Field.Store.NO));
                doc2.add(new StoredField("stored1", "foo"));
                doc2.add(new StoredField("stored1", "bar"));
                doc2.add(newField("text_vectors", TestUtil.randomAnalysisString(random(), 6, true), ft));
                try {
                    iw.addDocuments(Arrays.asList(doc, doc2));
                    // we made it, sometimes delete our docs
                    if (random().nextBoolean()) {
                        iw.deleteDocuments(new Term("id", Integer.toString(i)), new Term("id", Integer.toString(-i)));
                    }
                } catch (AlreadyClosedException ace) {
                    // OK: writer was closed by abort; we just reopen now:
                    assertTrue(iw.deleter.isClosed());
                    assertTrue(allowAlreadyClosed);
                    allowAlreadyClosed = false;
                    conf = newIndexWriterConfig(analyzer);
                    // just for now, try to keep this test reproducible
                    conf.setMergeScheduler(new SerialMergeScheduler());
                    conf.setCodec(codec);
                    iw = new IndexWriter(dir, conf);
                } catch (Exception e) {
                    if (e.getMessage() != null && e.getMessage().startsWith("Fake IOException")) {
                        exceptionStream.println("\nTEST: got expected fake exc:" + e.getMessage());
                        e.printStackTrace(exceptionStream);
                        allowAlreadyClosed = true;
                    } else {
                        Rethrow.rethrow(e);
                    }
                }
            }
            if (random().nextInt(10) == 0) {
                // trigger flush:
                try {
                    if (random().nextBoolean()) {
                        DirectoryReader ir = null;
                        try {
                            ir = DirectoryReader.open(iw, random().nextBoolean(), false);
                            TestUtil.checkReader(ir);
                        } finally {
                            IOUtils.closeWhileHandlingException(ir);
                        }
                    } else {
                        iw.commit();
                    }
                    if (DirectoryReader.indexExists(dir)) {
                        TestUtil.checkIndex(dir);
                    }
                } catch (AlreadyClosedException ace) {
                    // OK: writer was closed by abort; we just reopen now:
                    assertTrue(iw.deleter.isClosed());
                    assertTrue(allowAlreadyClosed);
                    allowAlreadyClosed = false;
                    conf = newIndexWriterConfig(analyzer);
                    // just for now, try to keep this test reproducible
                    conf.setMergeScheduler(new SerialMergeScheduler());
                    conf.setCodec(codec);
                    iw = new IndexWriter(dir, conf);
                } catch (Exception e) {
                    if (e.getMessage() != null && e.getMessage().startsWith("Fake IOException")) {
                        exceptionStream.println("\nTEST: got expected fake exc:" + e.getMessage());
                        e.printStackTrace(exceptionStream);
                        allowAlreadyClosed = true;
                    } else {
                        Rethrow.rethrow(e);
                    }
                }
            }
        }
        try {
            iw.close();
        } catch (Exception e) {
            if (e.getMessage() != null && e.getMessage().startsWith("Fake IOException")) {
                exceptionStream.println("\nTEST: got expected fake exc:" + e.getMessage());
                e.printStackTrace(exceptionStream);
                try {
                    iw.rollback();
                } catch (Throwable t) {
                }
            } else {
                Rethrow.rethrow(e);
            }
        }
        dir.close();
    } catch (Throwable t) {
        System.out.println("Unexpected exception: dumping fake-exception-log:...");
        exceptionStream.flush();
        System.out.println(exceptionLog.toString("UTF-8"));
        System.out.flush();
        Rethrow.rethrow(t);
    }
    if (VERBOSE) {
        System.out.println("TEST PASSED: dumping fake-exception-log:...");
        System.out.println(exceptionLog.toString("UTF-8"));
    }
}

Example 68 with IntPoint

use of org.apache.lucene.document.IntPoint in project lucene-solr by apache.

the class TestIndexWriterOnDiskFull method addDocWithIndex.

private void addDocWithIndex(IndexWriter writer, int index) throws IOException {
    Document doc = new Document();
    doc.add(newTextField("content", "aaa " + index, Field.Store.NO));
    doc.add(newTextField("id", "" + index, Field.Store.NO));
    doc.add(new NumericDocValuesField("numericdv", 1));
    doc.add(new IntPoint("point", 1));
    doc.add(new IntPoint("point2d", 1, 1));
    writer.addDocument(doc);
}

Example 69 with IntPoint

use of org.apache.lucene.document.IntPoint in project lucene-solr by apache.

the class TestIndexWriterOnDiskFull method addDoc.

// TODO: these are also in TestIndexWriter... add a simple doc-writing method
// like this to LuceneTestCase?
private void addDoc(IndexWriter writer) throws IOException {
    Document doc = new Document();
    doc.add(newTextField("content", "aaa", Field.Store.NO));
    doc.add(new NumericDocValuesField("numericdv", 1));
    doc.add(new IntPoint("point", 1));
    doc.add(new IntPoint("point2d", 1, 1));
    writer.addDocument(doc);
}