Examples with MalformedRecordException org.apache.nifi.serialization.MalformedRecordException used on opensource projects

Example 21 with MalformedRecordException

use of org.apache.nifi.serialization.MalformedRecordException in project nifi by apache.

the class TestJsonTreeRowRecordReader method testSingleJsonElementWithChoiceFields.

@Test
public void testSingleJsonElementWithChoiceFields() throws IOException, MalformedRecordException {
    // Wraps default fields by Choice data type to test mapping to a Choice type.
    final List<RecordField> choiceFields = getDefaultFields().stream().map(f -> new RecordField(f.getFieldName(), RecordFieldType.CHOICE.getChoiceDataType(f.getDataType()))).collect(Collectors.toList());
    final RecordSchema schema = new SimpleRecordSchema(choiceFields);
    try (final InputStream in = new FileInputStream(new File("src/test/resources/json/single-bank-account.json"));
        final JsonTreeRowRecordReader reader = new JsonTreeRowRecordReader(in, Mockito.mock(ComponentLog.class), schema, dateFormat, timeFormat, timestampFormat)) {
        final List<String> fieldNames = schema.getFieldNames();
        final List<String> expectedFieldNames = Arrays.asList(new String[] { "id", "name", "balance", "address", "city", "state", "zipCode", "country" });
        assertEquals(expectedFieldNames, fieldNames);
        final List<RecordFieldType> expectedTypes = Arrays.asList(new RecordFieldType[] { RecordFieldType.INT, RecordFieldType.STRING, RecordFieldType.DOUBLE, RecordFieldType.STRING, RecordFieldType.STRING, RecordFieldType.STRING, RecordFieldType.STRING, RecordFieldType.STRING });
        final List<RecordField> fields = schema.getFields();
        for (int i = 0; i < schema.getFields().size(); i++) {
            assertTrue(fields.get(i).getDataType() instanceof ChoiceDataType);
            final ChoiceDataType choiceDataType = (ChoiceDataType) fields.get(i).getDataType();
            assertEquals(expectedTypes.get(i), choiceDataType.getPossibleSubTypes().get(0).getFieldType());
        }
        final Object[] firstRecordValues = reader.nextRecord().getValues();
        Assert.assertArrayEquals(new Object[] { 1, "John Doe", 4750.89, "123 My Street", "My City", "MS", "11111", "USA" }, firstRecordValues);
        assertNull(reader.nextRecord());
    }
}

Example 22 with MalformedRecordException

use of org.apache.nifi.serialization.MalformedRecordException in project nifi by apache.

the class TestJsonTreeRowRecordReader method testElementWithNestedData.

@Test
public void testElementWithNestedData() throws IOException, MalformedRecordException {
    final DataType accountType = RecordFieldType.RECORD.getRecordDataType(getAccountSchema());
    final List<RecordField> fields = getDefaultFields();
    fields.add(new RecordField("account", accountType));
    fields.remove(new RecordField("balance", RecordFieldType.DOUBLE.getDataType()));
    final RecordSchema schema = new SimpleRecordSchema(fields);
    try (final InputStream in = new FileInputStream(new File("src/test/resources/json/single-element-nested.json"));
        final JsonTreeRowRecordReader reader = new JsonTreeRowRecordReader(in, Mockito.mock(ComponentLog.class), schema, dateFormat, timeFormat, timestampFormat)) {
        final List<RecordFieldType> dataTypes = schema.getDataTypes().stream().map(dt -> dt.getFieldType()).collect(Collectors.toList());
        final List<RecordFieldType> expectedTypes = Arrays.asList(new RecordFieldType[] { RecordFieldType.INT, RecordFieldType.STRING, RecordFieldType.STRING, RecordFieldType.STRING, RecordFieldType.STRING, RecordFieldType.STRING, RecordFieldType.STRING, RecordFieldType.RECORD });
        assertEquals(expectedTypes, dataTypes);
        final Object[] firstRecordValues = reader.nextRecord().getValues();
        final Object[] allButLast = Arrays.copyOfRange(firstRecordValues, 0, firstRecordValues.length - 1);
        Assert.assertArrayEquals(new Object[] { 1, "John Doe", "123 My Street", "My City", "MS", "11111", "USA" }, allButLast);
        final Object last = firstRecordValues[firstRecordValues.length - 1];
        assertTrue(Record.class.isAssignableFrom(last.getClass()));
        final Record record = (Record) last;
        assertEquals(42, record.getValue("id"));
        assertEquals(4750.89, record.getValue("balance"));
        assertNull(reader.nextRecord());
    }
}

Example 23 with MalformedRecordException

use of org.apache.nifi.serialization.MalformedRecordException in project nifi by apache.

the class PublishKafkaRecord_0_10 method onTrigger.

@Override
public void onTrigger(final ProcessContext context, final ProcessSession session) throws ProcessException {
    final List<FlowFile> flowFiles = session.get(FlowFileFilters.newSizeBasedFilter(1, DataUnit.MB, 500));
    if (flowFiles.isEmpty()) {
        return;
    }
    final PublisherPool pool = getPublisherPool(context);
    if (pool == null) {
        context.yield();
        return;
    }
    final String securityProtocol = context.getProperty(KafkaProcessorUtils.SECURITY_PROTOCOL).getValue();
    final String bootstrapServers = context.getProperty(KafkaProcessorUtils.BOOTSTRAP_SERVERS).evaluateAttributeExpressions().getValue();
    final RecordSetWriterFactory writerFactory = context.getProperty(RECORD_WRITER).asControllerService(RecordSetWriterFactory.class);
    final RecordReaderFactory readerFactory = context.getProperty(RECORD_READER).asControllerService(RecordReaderFactory.class);
    final long startTime = System.nanoTime();
    try (final PublisherLease lease = pool.obtainPublisher()) {
        // Send each FlowFile to Kafka asynchronously.
        for (final FlowFile flowFile : flowFiles) {
            if (!isScheduled()) {
                // If stopped, re-queue FlowFile instead of sending it
                session.transfer(flowFile);
                continue;
            }
            final String topic = context.getProperty(TOPIC).evaluateAttributeExpressions(flowFile).getValue();
            final String messageKeyField = context.getProperty(MESSAGE_KEY_FIELD).evaluateAttributeExpressions(flowFile).getValue();
            final Map<String, String> attributes = flowFile.getAttributes();
            try {
                session.read(flowFile, new InputStreamCallback() {

                    @Override
                    public void process(final InputStream rawIn) throws IOException {
                        try (final InputStream in = new BufferedInputStream(rawIn)) {
                            final RecordReader reader = readerFactory.createRecordReader(attributes, in, getLogger());
                            final RecordSet recordSet = reader.createRecordSet();
                            final RecordSchema schema = writerFactory.getSchema(attributes, recordSet.getSchema());
                            lease.publish(flowFile, recordSet, writerFactory, schema, messageKeyField, topic);
                        } catch (final SchemaNotFoundException | MalformedRecordException e) {
                            throw new ProcessException(e);
                        }
                    }
                });
            } catch (final Exception e) {
                // The FlowFile will be obtained and the error logged below, when calling publishResult.getFailedFlowFiles()
                lease.getTracker().fail(flowFile, e);
                continue;
            }
        }
        // Complete the send
        final PublishResult publishResult = lease.complete();
        // Transfer any successful FlowFiles.
        final long transmissionMillis = TimeUnit.NANOSECONDS.toMillis(System.nanoTime() - startTime);
        for (FlowFile success : publishResult.getSuccessfulFlowFiles()) {
            final String topic = context.getProperty(TOPIC).evaluateAttributeExpressions(success).getValue();
            final int msgCount = publishResult.getSuccessfulMessageCount(success);
            success = session.putAttribute(success, MSG_COUNT, String.valueOf(msgCount));
            session.adjustCounter("Messages Sent", msgCount, true);
            final String transitUri = KafkaProcessorUtils.buildTransitURI(securityProtocol, bootstrapServers, topic);
            session.getProvenanceReporter().send(success, transitUri, "Sent " + msgCount + " messages", transmissionMillis);
            session.transfer(success, REL_SUCCESS);
        }
        // Transfer any failures.
        for (final FlowFile failure : publishResult.getFailedFlowFiles()) {
            final int successCount = publishResult.getSuccessfulMessageCount(failure);
            if (successCount > 0) {
                getLogger().error("Failed to send some messages for {} to Kafka, but {} messages were acknowledged by Kafka. Routing to failure due to {}", new Object[] { failure, successCount, publishResult.getReasonForFailure(failure) });
            } else {
                getLogger().error("Failed to send all message for {} to Kafka; routing to failure due to {}", new Object[] { failure, publishResult.getReasonForFailure(failure) });
            }
            session.transfer(failure, REL_FAILURE);
        }
    }
}

Example 24 with MalformedRecordException

use of org.apache.nifi.serialization.MalformedRecordException in project nifi by apache.

the class MockRecordParser method createRecordReader.

@Override
public RecordReader createRecordReader(Map<String, String> variables, InputStream in, ComponentLog logger) throws IOException, SchemaNotFoundException {
    final Iterator<Object[]> itr = records.iterator();
    return new RecordReader() {

        private int recordCount = 0;

        @Override
        public void close() throws IOException {
        }

        @Override
        public Record nextRecord(final boolean coerceTypes, final boolean dropUnknown) throws IOException, MalformedRecordException {
            if (failAfterN >= recordCount) {
                throw new MalformedRecordException("Intentional Unit Test Exception because " + recordCount + " records have been read");
            }
            recordCount++;
            if (!itr.hasNext()) {
                return null;
            }
            final Object[] values = itr.next();
            final Map<String, Object> valueMap = new HashMap<>();
            int i = 0;
            for (final RecordField field : fields) {
                final String fieldName = field.getFieldName();
                valueMap.put(fieldName, values[i++]);
            }
            return new MapRecord(new SimpleRecordSchema(fields), valueMap);
        }

        @Override
        public RecordSchema getSchema() {
            return new SimpleRecordSchema(fields);
        }
    };
}

Example 25 with MalformedRecordException

use of org.apache.nifi.serialization.MalformedRecordException in project nifi by apache.

the class PutElasticsearchHttpRecord method onTrigger.

@Override
public void onTrigger(final ProcessContext context, final ProcessSession session) throws ProcessException {
    FlowFile flowFile = session.get();
    if (flowFile == null) {
        return;
    }
    final RecordReaderFactory readerFactory = context.getProperty(RECORD_READER).asControllerService(RecordReaderFactory.class);
    // Authentication
    final String username = context.getProperty(USERNAME).evaluateAttributeExpressions(flowFile).getValue();
    final String password = context.getProperty(PASSWORD).evaluateAttributeExpressions(flowFile).getValue();
    OkHttpClient okHttpClient = getClient();
    final ComponentLog logger = getLogger();
    final String baseUrl = trimToEmpty(context.getProperty(ES_URL).evaluateAttributeExpressions().getValue());
    HttpUrl.Builder urlBuilder = HttpUrl.parse(baseUrl).newBuilder().addPathSegment("_bulk");
    // Find the user-added properties and set them as query parameters on the URL
    for (Map.Entry<PropertyDescriptor, String> property : context.getProperties().entrySet()) {
        PropertyDescriptor pd = property.getKey();
        if (pd.isDynamic()) {
            if (property.getValue() != null) {
                urlBuilder = urlBuilder.addQueryParameter(pd.getName(), context.getProperty(pd).evaluateAttributeExpressions().getValue());
            }
        }
    }
    final URL url = urlBuilder.build().url();
    final String index = context.getProperty(INDEX).evaluateAttributeExpressions(flowFile).getValue();
    if (StringUtils.isEmpty(index)) {
        logger.error("No value for index in for {}, transferring to failure", new Object[] { flowFile });
        session.transfer(flowFile, REL_FAILURE);
        return;
    }
    final String docType = context.getProperty(TYPE).evaluateAttributeExpressions(flowFile).getValue();
    String indexOp = context.getProperty(INDEX_OP).evaluateAttributeExpressions(flowFile).getValue();
    if (StringUtils.isEmpty(indexOp)) {
        logger.error("No Index operation specified for {}, transferring to failure.", new Object[] { flowFile });
        session.transfer(flowFile, REL_FAILURE);
        return;
    }
    switch(indexOp.toLowerCase()) {
        case "index":
        case "update":
        case "upsert":
        case "delete":
            break;
        default:
            logger.error("Index operation {} not supported for {}, transferring to failure.", new Object[] { indexOp, flowFile });
            session.transfer(flowFile, REL_FAILURE);
            return;
    }
    this.nullSuppression = context.getProperty(SUPPRESS_NULLS).getValue();
    final String id_path = context.getProperty(ID_RECORD_PATH).evaluateAttributeExpressions(flowFile).getValue();
    final RecordPath recordPath = StringUtils.isEmpty(id_path) ? null : recordPathCache.getCompiled(id_path);
    final StringBuilder sb = new StringBuilder();
    try (final InputStream in = session.read(flowFile);
        final RecordReader reader = readerFactory.createRecordReader(flowFile, in, getLogger())) {
        Record record;
        while ((record = reader.nextRecord()) != null) {
            final String id;
            if (recordPath != null) {
                Optional<FieldValue> idPathValue = recordPath.evaluate(record).getSelectedFields().findFirst();
                if (!idPathValue.isPresent() || idPathValue.get().getValue() == null) {
                    throw new IdentifierNotFoundException("Identifier Record Path specified but no value was found, transferring {} to failure.");
                }
                id = idPathValue.get().getValue().toString();
            } else {
                id = null;
            }
            // a missing ID indicates one is to be auto-generated by Elasticsearch
            if (id == null && !indexOp.equalsIgnoreCase("index")) {
                throw new IdentifierNotFoundException("Index operation {} requires a valid identifier value from a flow file attribute, transferring to failure.");
            }
            final StringBuilder json = new StringBuilder();
            ByteArrayOutputStream out = new ByteArrayOutputStream();
            JsonGenerator generator = factory.createJsonGenerator(out);
            writeRecord(record, record.getSchema(), generator);
            generator.flush();
            generator.close();
            json.append(out.toString());
            if (indexOp.equalsIgnoreCase("index")) {
                sb.append("{\"index\": { \"_index\": \"");
                sb.append(index);
                sb.append("\", \"_type\": \"");
                sb.append(docType);
                sb.append("\"");
                if (!StringUtils.isEmpty(id)) {
                    sb.append(", \"_id\": \"");
                    sb.append(id);
                    sb.append("\"");
                }
                sb.append("}}\n");
                sb.append(json);
                sb.append("\n");
            } else if (indexOp.equalsIgnoreCase("upsert") || indexOp.equalsIgnoreCase("update")) {
                sb.append("{\"update\": { \"_index\": \"");
                sb.append(index);
                sb.append("\", \"_type\": \"");
                sb.append(docType);
                sb.append("\", \"_id\": \"");
                sb.append(id);
                sb.append("\" }\n");
                sb.append("{\"doc\": ");
                sb.append(json);
                sb.append(", \"doc_as_upsert\": ");
                sb.append(indexOp.equalsIgnoreCase("upsert"));
                sb.append(" }\n");
            } else if (indexOp.equalsIgnoreCase("delete")) {
                sb.append("{\"delete\": { \"_index\": \"");
                sb.append(index);
                sb.append("\", \"_type\": \"");
                sb.append(docType);
                sb.append("\", \"_id\": \"");
                sb.append(id);
                sb.append("\" }\n");
            }
        }
    } catch (IdentifierNotFoundException infe) {
        logger.error(infe.getMessage(), new Object[] { flowFile });
        flowFile = session.penalize(flowFile);
        session.transfer(flowFile, REL_FAILURE);
        return;
    } catch (final IOException | SchemaNotFoundException | MalformedRecordException e) {
        logger.error("Could not parse incoming data", e);
        flowFile = session.penalize(flowFile);
        session.transfer(flowFile, REL_FAILURE);
        return;
    }
    RequestBody requestBody = RequestBody.create(MediaType.parse("application/json"), sb.toString());
    final Response getResponse;
    try {
        getResponse = sendRequestToElasticsearch(okHttpClient, url, username, password, "PUT", requestBody);
    } catch (final Exception e) {
        logger.error("Routing to {} due to exception: {}", new Object[] { REL_FAILURE.getName(), e }, e);
        flowFile = session.penalize(flowFile);
        session.transfer(flowFile, REL_FAILURE);
        return;
    }
    final int statusCode = getResponse.code();
    if (isSuccess(statusCode)) {
        ResponseBody responseBody = getResponse.body();
        try {
            final byte[] bodyBytes = responseBody.bytes();
            JsonNode responseJson = parseJsonResponse(new ByteArrayInputStream(bodyBytes));
            boolean errors = responseJson.get("errors").asBoolean(false);
            // ES has no rollback, so if errors occur, log them and route the whole flow file to failure
            if (errors) {
                ArrayNode itemNodeArray = (ArrayNode) responseJson.get("items");
                if (itemNodeArray.size() > 0) {
                    // All items are returned whether they succeeded or failed, so iterate through the item array
                    // at the same time as the flow file list, moving each to success or failure accordingly,
                    // but only keep the first error for logging
                    String errorReason = null;
                    for (int i = itemNodeArray.size() - 1; i >= 0; i--) {
                        JsonNode itemNode = itemNodeArray.get(i);
                        int status = itemNode.findPath("status").asInt();
                        if (!isSuccess(status)) {
                            if (errorReason == null) {
                                // Use "result" if it is present; this happens for status codes like 404 Not Found, which may not have an error/reason
                                String reason = itemNode.findPath("//result").asText();
                                if (StringUtils.isEmpty(reason)) {
                                    // If there was no result, we expect an error with a string description in the "reason" field
                                    reason = itemNode.findPath("//error/reason").asText();
                                }
                                errorReason = reason;
                                logger.error("Failed to process {} due to {}, transferring to failure", new Object[] { flowFile, errorReason });
                            }
                        }
                    }
                }
                session.transfer(flowFile, REL_FAILURE);
            } else {
                session.transfer(flowFile, REL_SUCCESS);
                session.getProvenanceReporter().send(flowFile, url.toString());
            }
        } catch (IOException ioe) {
            // Something went wrong when parsing the response, log the error and route to failure
            logger.error("Error parsing Bulk API response: {}", new Object[] { ioe.getMessage() }, ioe);
            session.transfer(flowFile, REL_FAILURE);
            context.yield();
        }
    } else if (statusCode / 100 == 5) {
        // 5xx -> RETRY, but a server error might last a while, so yield
        logger.warn("Elasticsearch returned code {} with message {}, transferring flow file to retry. This is likely a server problem, yielding...", new Object[] { statusCode, getResponse.message() });
        session.transfer(flowFile, REL_RETRY);
        context.yield();
    } else {
        // 1xx, 3xx, 4xx, etc. -> NO RETRY
        logger.warn("Elasticsearch returned code {} with message {}, transferring flow file to failure", new Object[] { statusCode, getResponse.message() });
        session.transfer(flowFile, REL_FAILURE);
    }
    getResponse.close();
}