Examples with ProcessContext org.apache.nifi.processor.ProcessContext used on opensource projects

Example 16 with ProcessContext

use of org.apache.nifi.processor.ProcessContext in project nifi by apache.

the class TestConsumeKafkaRecord_1_0 method setup.

@Before
public void setup() throws InitializationException {
    mockLease = mock(ConsumerLease.class);
    mockConsumerPool = mock(ConsumerPool.class);
    ConsumeKafkaRecord_1_0 proc = new ConsumeKafkaRecord_1_0() {

        @Override
        protected ConsumerPool createConsumerPool(final ProcessContext context, final ComponentLog log) {
            return mockConsumerPool;
        }
    };
    runner = TestRunners.newTestRunner(proc);
    runner.setProperty(KafkaProcessorUtils.BOOTSTRAP_SERVERS, "okeydokey:1234");
    final String readerId = "record-reader";
    final MockRecordParser readerService = new MockRecordParser();
    readerService.addSchemaField("name", RecordFieldType.STRING);
    readerService.addSchemaField("age", RecordFieldType.INT);
    runner.addControllerService(readerId, readerService);
    runner.enableControllerService(readerService);
    final String writerId = "record-writer";
    final RecordSetWriterFactory writerService = new MockRecordWriter("name, age");
    runner.addControllerService(writerId, writerService);
    runner.enableControllerService(writerService);
    runner.setProperty(ConsumeKafkaRecord_1_0.RECORD_READER, readerId);
    runner.setProperty(ConsumeKafkaRecord_1_0.RECORD_WRITER, writerId);
}

Example 17 with ProcessContext

use of org.apache.nifi.processor.ProcessContext in project nifi by apache.

the class TestPublishKafkaRecord_1_0 method setup.

@Before
public void setup() throws InitializationException, IOException {
    mockPool = mock(PublisherPool.class);
    mockLease = mock(PublisherLease.class);
    Mockito.doCallRealMethod().when(mockLease).publish(any(FlowFile.class), any(RecordSet.class), any(RecordSetWriterFactory.class), any(RecordSchema.class), any(String.class), any(String.class));
    when(mockPool.obtainPublisher()).thenReturn(mockLease);
    runner = TestRunners.newTestRunner(new PublishKafkaRecord_1_0() {

        @Override
        protected PublisherPool createPublisherPool(final ProcessContext context) {
            return mockPool;
        }
    });
    runner.setProperty(PublishKafkaRecord_1_0.TOPIC, TOPIC_NAME);
    final String readerId = "record-reader";
    final MockRecordParser readerService = new MockRecordParser();
    readerService.addSchemaField("name", RecordFieldType.STRING);
    readerService.addSchemaField("age", RecordFieldType.INT);
    runner.addControllerService(readerId, readerService);
    runner.enableControllerService(readerService);
    final String writerId = "record-writer";
    final RecordSetWriterFactory writerService = new MockRecordWriter("name, age");
    runner.addControllerService(writerId, writerService);
    runner.enableControllerService(writerService);
    runner.setProperty(PublishKafkaRecord_1_0.RECORD_READER, readerId);
    runner.setProperty(PublishKafkaRecord_1_0.RECORD_WRITER, writerId);
    runner.setProperty(PublishKafka_1_0.DELIVERY_GUARANTEE, PublishKafka_1_0.DELIVERY_REPLICATED);
}

Example 18 with ProcessContext

use of org.apache.nifi.processor.ProcessContext in project nifi by apache.

the class FetchParquetTest method testIOExceptionWhileReadingShouldRouteToRetry.

@Test
public void testIOExceptionWhileReadingShouldRouteToRetry() throws IOException, InitializationException {
    final FetchParquet proc = new FetchParquet() {

        @Override
        public HDFSRecordReader createHDFSRecordReader(ProcessContext context, FlowFile flowFile, Configuration conf, Path path) throws IOException {
            return new HDFSRecordReader() {

                @Override
                public Record nextRecord() throws IOException {
                    throw new IOException("IOException");
                }

                @Override
                public void close() throws IOException {
                }
            };
        }
    };
    configure(proc);
    final File parquetDir = new File(DIRECTORY);
    final File parquetFile = new File(parquetDir, "testFetchParquetToCSV.parquet");
    final int numUsers = 10;
    writeParquetUsers(parquetFile, numUsers);
    final Map<String, String> attributes = new HashMap<>();
    attributes.put(CoreAttributes.PATH.key(), parquetDir.getAbsolutePath());
    attributes.put(CoreAttributes.FILENAME.key(), parquetFile.getName());
    testRunner.enqueue("TRIGGER", attributes);
    testRunner.run();
    testRunner.assertAllFlowFilesTransferred(FetchParquet.REL_RETRY, 1);
    final MockFlowFile flowFile = testRunner.getFlowFilesForRelationship(FetchParquet.REL_RETRY).get(0);
    flowFile.assertContentEquals("TRIGGER");
}

Example 19 with ProcessContext

use of org.apache.nifi.processor.ProcessContext in project nifi by apache.

the class QueryDatabaseTable method onTrigger.

@Override
public void onTrigger(final ProcessContext context, final ProcessSessionFactory sessionFactory) throws ProcessException {
    // Fetch the column/table info once
    if (!setupComplete.get()) {
        super.setup(context);
    }
    ProcessSession session = sessionFactory.createSession();
    final List<FlowFile> resultSetFlowFiles = new ArrayList<>();
    final ComponentLog logger = getLogger();
    final DBCPService dbcpService = context.getProperty(DBCP_SERVICE).asControllerService(DBCPService.class);
    final DatabaseAdapter dbAdapter = dbAdapters.get(context.getProperty(DB_TYPE).getValue());
    final String tableName = context.getProperty(TABLE_NAME).evaluateAttributeExpressions().getValue();
    final String columnNames = context.getProperty(COLUMN_NAMES).evaluateAttributeExpressions().getValue();
    final String maxValueColumnNames = context.getProperty(MAX_VALUE_COLUMN_NAMES).evaluateAttributeExpressions().getValue();
    final String customWhereClause = context.getProperty(WHERE_CLAUSE).evaluateAttributeExpressions().getValue();
    final Integer fetchSize = context.getProperty(FETCH_SIZE).evaluateAttributeExpressions().asInteger();
    final Integer maxRowsPerFlowFile = context.getProperty(MAX_ROWS_PER_FLOW_FILE).evaluateAttributeExpressions().asInteger();
    final Integer outputBatchSizeField = context.getProperty(OUTPUT_BATCH_SIZE).evaluateAttributeExpressions().asInteger();
    final int outputBatchSize = outputBatchSizeField == null ? 0 : outputBatchSizeField;
    final Integer maxFragments = context.getProperty(MAX_FRAGMENTS).isSet() ? context.getProperty(MAX_FRAGMENTS).evaluateAttributeExpressions().asInteger() : 0;
    final JdbcCommon.AvroConversionOptions options = JdbcCommon.AvroConversionOptions.builder().recordName(tableName).maxRows(maxRowsPerFlowFile).convertNames(context.getProperty(NORMALIZE_NAMES_FOR_AVRO).asBoolean()).useLogicalTypes(context.getProperty(USE_AVRO_LOGICAL_TYPES).asBoolean()).defaultPrecision(context.getProperty(DEFAULT_PRECISION).evaluateAttributeExpressions().asInteger()).defaultScale(context.getProperty(DEFAULT_SCALE).evaluateAttributeExpressions().asInteger()).build();
    final StateManager stateManager = context.getStateManager();
    final StateMap stateMap;
    try {
        stateMap = stateManager.getState(Scope.CLUSTER);
    } catch (final IOException ioe) {
        getLogger().error("Failed to retrieve observed maximum values from the State Manager. Will not perform " + "query until this is accomplished.", ioe);
        context.yield();
        return;
    }
    // Make a mutable copy of the current state property map. This will be updated by the result row callback, and eventually
    // set as the current state map (after the session has been committed)
    final Map<String, String> statePropertyMap = new HashMap<>(stateMap.toMap());
    // If an initial max value for column(s) has been specified using properties, and this column is not in the state manager, sync them to the state property map
    for (final Map.Entry<String, String> maxProp : maxValueProperties.entrySet()) {
        String maxPropKey = maxProp.getKey().toLowerCase();
        String fullyQualifiedMaxPropKey = getStateKey(tableName, maxPropKey);
        if (!statePropertyMap.containsKey(fullyQualifiedMaxPropKey)) {
            String newMaxPropValue;
            // but store the new initial max value under the fully-qualified key.
            if (statePropertyMap.containsKey(maxPropKey)) {
                newMaxPropValue = statePropertyMap.get(maxPropKey);
            } else {
                newMaxPropValue = maxProp.getValue();
            }
            statePropertyMap.put(fullyQualifiedMaxPropKey, newMaxPropValue);
        }
    }
    List<String> maxValueColumnNameList = StringUtils.isEmpty(maxValueColumnNames) ? null : Arrays.asList(maxValueColumnNames.split("\\s*,\\s*"));
    final String selectQuery = getQuery(dbAdapter, tableName, columnNames, maxValueColumnNameList, customWhereClause, statePropertyMap);
    final StopWatch stopWatch = new StopWatch(true);
    final String fragmentIdentifier = UUID.randomUUID().toString();
    try (final Connection con = dbcpService.getConnection();
        final Statement st = con.createStatement()) {
        if (fetchSize != null && fetchSize > 0) {
            try {
                st.setFetchSize(fetchSize);
            } catch (SQLException se) {
                // Not all drivers support this, just log the error (at debug level) and move on
                logger.debug("Cannot set fetch size to {} due to {}", new Object[] { fetchSize, se.getLocalizedMessage() }, se);
            }
        }
        String jdbcURL = "DBCPService";
        try {
            DatabaseMetaData databaseMetaData = con.getMetaData();
            if (databaseMetaData != null) {
                jdbcURL = databaseMetaData.getURL();
            }
        } catch (SQLException se) {
        // Ignore and use default JDBC URL. This shouldn't happen unless the driver doesn't implement getMetaData() properly
        }
        final Integer queryTimeout = context.getProperty(QUERY_TIMEOUT).evaluateAttributeExpressions().asTimePeriod(TimeUnit.SECONDS).intValue();
        // timeout in seconds
        st.setQueryTimeout(queryTimeout);
        try {
            logger.debug("Executing query {}", new Object[] { selectQuery });
            final ResultSet resultSet = st.executeQuery(selectQuery);
            int fragmentIndex = 0;
            while (true) {
                final AtomicLong nrOfRows = new AtomicLong(0L);
                FlowFile fileToProcess = session.create();
                try {
                    fileToProcess = session.write(fileToProcess, out -> {
                        // Max values will be updated in the state property map by the callback
                        final MaxValueResultSetRowCollector maxValCollector = new MaxValueResultSetRowCollector(tableName, statePropertyMap, dbAdapter);
                        try {
                            nrOfRows.set(JdbcCommon.convertToAvroStream(resultSet, out, options, maxValCollector));
                        } catch (SQLException | RuntimeException e) {
                            throw new ProcessException("Error during database query or conversion of records to Avro.", e);
                        }
                    });
                } catch (ProcessException e) {
                    // Add flowfile to results before rethrowing so it will be removed from session in outer catch
                    resultSetFlowFiles.add(fileToProcess);
                    throw e;
                }
                if (nrOfRows.get() > 0) {
                    // set attribute how many rows were selected
                    fileToProcess = session.putAttribute(fileToProcess, RESULT_ROW_COUNT, String.valueOf(nrOfRows.get()));
                    fileToProcess = session.putAttribute(fileToProcess, RESULT_TABLENAME, tableName);
                    fileToProcess = session.putAttribute(fileToProcess, CoreAttributes.MIME_TYPE.key(), JdbcCommon.MIME_TYPE_AVRO_BINARY);
                    if (maxRowsPerFlowFile > 0) {
                        fileToProcess = session.putAttribute(fileToProcess, "fragment.identifier", fragmentIdentifier);
                        fileToProcess = session.putAttribute(fileToProcess, "fragment.index", String.valueOf(fragmentIndex));
                    }
                    logger.info("{} contains {} Avro records; transferring to 'success'", new Object[] { fileToProcess, nrOfRows.get() });
                    session.getProvenanceReporter().receive(fileToProcess, jdbcURL, stopWatch.getElapsed(TimeUnit.MILLISECONDS));
                    resultSetFlowFiles.add(fileToProcess);
                    // If we've reached the batch size, send out the flow files
                    if (outputBatchSize > 0 && resultSetFlowFiles.size() >= outputBatchSize) {
                        session.transfer(resultSetFlowFiles, REL_SUCCESS);
                        session.commit();
                        resultSetFlowFiles.clear();
                    }
                } else {
                    // If there were no rows returned, don't send the flowfile
                    session.remove(fileToProcess);
                    context.yield();
                    break;
                }
                fragmentIndex++;
                if (maxFragments > 0 && fragmentIndex >= maxFragments) {
                    break;
                }
            }
            // Even though the maximum value and total count are known at this point, to maintain consistent behavior if Output Batch Size is set, do not store the attributes
            if (outputBatchSize == 0) {
                for (int i = 0; i < resultSetFlowFiles.size(); i++) {
                    // Add maximum values as attributes
                    for (Map.Entry<String, String> entry : statePropertyMap.entrySet()) {
                        // Get just the column name from the key
                        String key = entry.getKey();
                        String colName = key.substring(key.lastIndexOf(NAMESPACE_DELIMITER) + NAMESPACE_DELIMITER.length());
                        resultSetFlowFiles.set(i, session.putAttribute(resultSetFlowFiles.get(i), "maxvalue." + colName, entry.getValue()));
                    }
                    // set count on all FlowFiles
                    if (maxRowsPerFlowFile > 0) {
                        resultSetFlowFiles.set(i, session.putAttribute(resultSetFlowFiles.get(i), "fragment.count", Integer.toString(fragmentIndex)));
                    }
                }
            }
        } catch (final SQLException e) {
            throw e;
        }
        session.transfer(resultSetFlowFiles, REL_SUCCESS);
    } catch (final ProcessException | SQLException e) {
        logger.error("Unable to execute SQL select query {} due to {}", new Object[] { selectQuery, e });
        if (!resultSetFlowFiles.isEmpty()) {
            session.remove(resultSetFlowFiles);
        }
        context.yield();
    } finally {
        session.commit();
        try {
            // Update the state
            stateManager.setState(statePropertyMap, Scope.CLUSTER);
        } catch (IOException ioe) {
            getLogger().error("{} failed to update State Manager, maximum observed values will not be recorded", new Object[] { this, ioe });
        }
    }
}

Example 20 with ProcessContext

use of org.apache.nifi.processor.ProcessContext in project nifi by apache.

the class SplitJson method onTrigger.

@Override
public void onTrigger(final ProcessContext processContext, final ProcessSession processSession) {
    FlowFile original = processSession.get();
    if (original == null) {
        return;
    }
    final ComponentLog logger = getLogger();
    DocumentContext documentContext;
    try {
        documentContext = validateAndEstablishJsonContext(processSession, original);
    } catch (InvalidJsonException e) {
        logger.error("FlowFile {} did not have valid JSON content.", new Object[] { original });
        processSession.transfer(original, REL_FAILURE);
        return;
    }
    final JsonPath jsonPath = JSON_PATH_REF.get();
    Object jsonPathResult;
    try {
        jsonPathResult = documentContext.read(jsonPath);
    } catch (PathNotFoundException e) {
        logger.warn("JsonPath {} could not be found for FlowFile {}", new Object[] { jsonPath.getPath(), original });
        processSession.transfer(original, REL_FAILURE);
        return;
    }
    if (!(jsonPathResult instanceof List)) {
        logger.error("The evaluated value {} of {} was not a JSON Array compatible type and cannot be split.", new Object[] { jsonPathResult, jsonPath.getPath() });
        processSession.transfer(original, REL_FAILURE);
        return;
    }
    List resultList = (List) jsonPathResult;
    Map<String, String> attributes = new HashMap<>();
    final String fragmentId = UUID.randomUUID().toString();
    attributes.put(FRAGMENT_ID.key(), fragmentId);
    attributes.put(FRAGMENT_COUNT.key(), Integer.toString(resultList.size()));
    for (int i = 0; i < resultList.size(); i++) {
        Object resultSegment = resultList.get(i);
        FlowFile split = processSession.create(original);
        split = processSession.write(split, (out) -> {
            String resultSegmentContent = getResultRepresentation(resultSegment, nullDefaultValue);
            out.write(resultSegmentContent.getBytes(StandardCharsets.UTF_8));
        });
        attributes.put(SEGMENT_ORIGINAL_FILENAME.key(), split.getAttribute(CoreAttributes.FILENAME.key()));
        attributes.put(FRAGMENT_INDEX.key(), Integer.toString(i));
        processSession.transfer(processSession.putAllAttributes(split, attributes), REL_SPLIT);
    }
    original = copyAttributesToOriginal(processSession, original, fragmentId, resultList.size());
    processSession.transfer(original, REL_ORIGINAL);
    logger.info("Split {} into {} FlowFiles", new Object[] { original, resultList.size() });
}