mattyb149 commented on a change in pull request #3317: NIFI-6047 Add DetectDuplicateRecord Processor URL: https://github.com/apache/nifi/pull/3317#discussion_r296309936
########## File path: nifi-nar-bundles/nifi-standard-bundle/nifi-standard-processors/src/main/java/org/apache/nifi/processors/standard/DetectDuplicateRecord.java ########## @@ -0,0 +1,620 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ +package org.apache.nifi.processors.standard; + +import com.google.common.base.Joiner; +import com.google.common.hash.BloomFilter; +import com.google.common.hash.Funnels; +import org.apache.commons.codec.binary.Hex; +import org.apache.commons.codec.digest.DigestUtils; +import org.apache.commons.codec.digest.MessageDigestAlgorithms; +import org.apache.nifi.annotation.behavior.*; +import org.apache.nifi.annotation.behavior.InputRequirement.Requirement; +import org.apache.nifi.annotation.documentation.CapabilityDescription; +import org.apache.nifi.annotation.documentation.SeeAlso; +import org.apache.nifi.annotation.documentation.Tags; +import org.apache.nifi.annotation.lifecycle.OnScheduled; +import org.apache.nifi.components.*; +import org.apache.nifi.distributed.cache.client.Deserializer; +import org.apache.nifi.distributed.cache.client.DistributedMapCacheClient; +import org.apache.nifi.distributed.cache.client.Serializer; +import org.apache.nifi.distributed.cache.client.exception.DeserializationException; +import org.apache.nifi.distributed.cache.client.exception.SerializationException; +import org.apache.nifi.expression.AttributeExpression.ResultType; +import org.apache.nifi.expression.ExpressionLanguageScope; +import org.apache.nifi.flowfile.FlowFile; +import org.apache.nifi.flowfile.attributes.CoreAttributes; +import org.apache.nifi.logging.ComponentLog; +import org.apache.nifi.processor.*; +import org.apache.nifi.processor.exception.ProcessException; +import org.apache.nifi.processor.util.StandardValidators; +import org.apache.nifi.record.path.RecordPath; +import org.apache.nifi.record.path.RecordPathResult; +import org.apache.nifi.record.path.util.RecordPathCache; +import org.apache.nifi.record.path.validation.RecordPathPropertyNameValidator; +import org.apache.nifi.record.path.validation.RecordPathValidator; +import org.apache.nifi.schema.access.SchemaNotFoundException; +import org.apache.nifi.serialization.*; +import org.apache.nifi.serialization.record.Record; +import org.apache.nifi.serialization.record.RecordSchema; + +import java.io.*; +import java.nio.charset.Charset; +import java.nio.charset.StandardCharsets; +import java.security.MessageDigest; +import java.util.*; +import java.util.concurrent.TimeUnit; +import java.util.stream.Collectors; + +import static java.util.stream.Collectors.toList; +import static org.apache.commons.codec.binary.StringUtils.getBytesUtf8; +import static org.apache.commons.lang3.StringUtils.*; + +@EventDriven +@SupportsBatching +@InputRequirement(Requirement.INPUT_REQUIRED) +@SystemResourceConsideration(resource = SystemResource.MEMORY, + description = "Caches records from each incoming FlowFile and determines if the cached record has " + + "already been seen. The name of user-defined properties determines the RecordPath values used to " + + "determine if a record is unique. If no user-defined properties are present, the entire record is " + + "used as the input to determine uniqueness. All duplicate records are routed to 'duplicate'. " + + "If the record is not determined to be a duplicate, the Processor routes the record to 'non-duplicate'.") +@Tags({"text", "record", "update", "change", "replace", "modify", "distinct", "unique", + "filter", "hash", "dupe", "duplicate", "dedupe"}) +@CapabilityDescription("Caches records from each incoming FlowFile and determines if the cached record has " + + "already been seen. The name of user-defined properties determines the RecordPath values used to " + + "determine if a record is unique. If no user-defined properties are present, the entire record is " + + "used as the input to determine uniqueness. All duplicate records are routed to 'duplicate'. " + + "If the record is not determined to be a duplicate, the Processor routes the record to 'non-duplicate'." +) +@ReadsAttributes({@ReadsAttribute(attribute="", description="")}) +@WritesAttributes({ + @WritesAttribute(attribute = "record.count", description = "The number of records in the FlowFile") +}) +@DynamicProperty( + name = "RecordPath", + value = "User-defined property values are ignored", + description = "The name of each user-defined property must be a valid RecordPath.") +@SeeAlso(classNames = { + "org.apache.nifi.distributed.cache.client.DistributedMapCacheClientService", + "org.apache.nifi.distributed.cache.server.map.DistributedMapCacheServer", + "org.apache.nifi.processors.standard.DetectDuplicate" +}) +public class DetectDuplicateRecord extends AbstractProcessor { + + private volatile RecordPathCache recordPathCache; + private volatile List<String> recordPaths; + + // VALUES + + static final AllowableValue MD5_ALGORITHM_VALUE = new AllowableValue(MessageDigestAlgorithms.MD5, "MD5", + "The MD5 message-digest algorithm."); + static final AllowableValue SHA1_ALGORITHM_VALUE = new AllowableValue(MessageDigestAlgorithms.SHA_1, "SHA-1", + "The SHA-1 cryptographic hash algorithm."); + static final AllowableValue SHA256_ALGORITHM_VALUE = new AllowableValue(MessageDigestAlgorithms.SHA3_256, "SHA-256", + "The SHA-256 cryptographic hash algorithm."); + static final AllowableValue SHA512_ALGORITHM_VALUE = new AllowableValue(MessageDigestAlgorithms.SHA3_512, "SHA-512", + "The SHA-512 cryptographic hash algorithm."); + +// static final AllowableValue ENTIRE_RECORD_VALUE = new AllowableValue("entire-record", "Entire Record", +// "All field values of a record are used in the order they are listed in the incoming FlowFile to determine if two records are equal."); +// static final AllowableValue RECORD_PATH_VALUE = new AllowableValue("entire-record", "Entire Record", +// "The user-defined RecordPath properties are used in the specified order to determine whether two records are equal. " + +// "If the value of a RecordPath is modified or the ordering of the user-defined RecordPath properties change " + +// "after processing one or more FlowFiles, duplicate detection is effectively reset."); + + static final AllowableValue HASH_SET_VALUE = new AllowableValue("hash-set", "HashSet", + "Exactly matches records seen before with 100% accuracy at the expense of more memory usage. " + + "This is not ideal for large files since a hash of each record is held in memory."); + static final AllowableValue BLOOM_FILTER_VALUE = new AllowableValue("bloom-filter", "BloomFilter", + "Space-efficient data structure ideal for large data sets uses probability to determine if a record was seen before. " + + "False positive matches are possible, but false negatives are not – in other words, a query returns either \"possibly in set\" or \"definitely not in set\". " + + "You should use this option if the FlowFile content is large and you can tolerate some duplication in the data."); + + + // PROPERTIES + + static final PropertyDescriptor RECORD_READER = new PropertyDescriptor.Builder() + .name("record-reader") + .displayName("Record Reader") + .description("Specifies the Controller Service to use for reading incoming data") + .identifiesControllerService(RecordReaderFactory.class) + .required(true) + .build(); + + static final PropertyDescriptor RECORD_WRITER = new PropertyDescriptor.Builder() + .name("record-writer") + .displayName("Record Writer") + .description("Specifies the Controller Service to use for writing out the records") + .identifiesControllerService(RecordSetWriterFactory.class) + .required(true) + .build(); + + static final PropertyDescriptor INCLUDE_ZERO_RECORD_FLOWFILES = new PropertyDescriptor.Builder() + .name("include-zero-record-flowfiles") + .displayName("Include Zero Record FlowFiles") + .description("When converting an incoming FlowFile, if the conversion results in no data, " + + "this property specifies whether or not a FlowFile will be sent to the corresponding relationship") + .expressionLanguageSupported(ExpressionLanguageScope.NONE) + .allowableValues("true", "false") + .defaultValue("true") + .required(true) + .build(); + + static final PropertyDescriptor CACHE_IDENTIFIER = new PropertyDescriptor.Builder() + .name("Cache The Entry Identifier") + .description("When true this cause the processor to check for duplicates and cache the Entry Identifier. When false, " + + "the processor would only check for duplicates and not cache the Entry Identifier, requiring another " + + "processor to add identifiers to the distributed cache.") + .required(true) + .allowableValues("true", "false") + .defaultValue("true") + .build(); + + static final PropertyDescriptor DISTRIBUTED_CACHE_SERVICE = new PropertyDescriptor.Builder() + .name("Distributed Cache Service") + .description("The Controller Service that is used to cache unique records, used to determine duplicates") + .required(false) + .identifiesControllerService(DistributedMapCacheClient.class) + .build(); + + static final PropertyDescriptor CACHE_ENTRY_IDENTIFIER = new PropertyDescriptor.Builder() + .name("Cache Entry Identifier") + .description( + "A FlowFile attribute, or the results of an Attribute Expression Language statement, which will be evaluated " + + "against a FlowFile in order to determine the cached filter type value used to identify duplicates.") + .required(false) + .addValidator(StandardValidators.createAttributeExpressionLanguageValidator(ResultType.STRING, true)) + .defaultValue("${hash.value}") + .expressionLanguageSupported(ExpressionLanguageScope.FLOWFILE_ATTRIBUTES) + .build(); + + static final PropertyDescriptor AGE_OFF_DURATION = new PropertyDescriptor.Builder() + .name("Age Off Duration") + .description("Time interval to age off cached filters. When the cache expires, the entire filter and its values " + + "are destroyed. Leaving this value empty will cause the cached entries to never expire.") + .required(false) + .addValidator(StandardValidators.TIME_PERIOD_VALIDATOR) + .build(); + + static final PropertyDescriptor RECORD_HASHING_ALGORITHM = new PropertyDescriptor.Builder() + .name("Record Hashing Algorithm") + .description("The algorithm used to hash the combined result of RecordPath values in the cache.") + .allowableValues( + MD5_ALGORITHM_VALUE, + SHA1_ALGORITHM_VALUE, + SHA256_ALGORITHM_VALUE, + SHA512_ALGORITHM_VALUE + ) + .defaultValue(SHA1_ALGORITHM_VALUE.getValue()) + .expressionLanguageSupported(ExpressionLanguageScope.NONE) + .required(true) + .build(); + + static final PropertyDescriptor FILTER_TYPE = new PropertyDescriptor.Builder() + .name("Filter Type") + .description("The filter used to determine whether a record has been seen before based on the matching RecordPath criteria.") + .allowableValues( + HASH_SET_VALUE, + BLOOM_FILTER_VALUE + ) + .defaultValue(HASH_SET_VALUE.getValue()) + .required(true) + .build(); + + static final PropertyDescriptor FILTER_CAPACITY_HINT = new PropertyDescriptor.Builder() + .name("Filter Capacity Hint") + .description("An estimation of the total number of unique records to be processed. " + + "The more accurate this number is leads to less resizing operations on a HashSet filter and fewer false negatives on a BloomFilter. " + + "Using the CountText processor block first and then using the ${text.line.count} attribute may be ideal for more dynamic estimates.") + .defaultValue("25000") + .expressionLanguageSupported(ExpressionLanguageScope.NONE) + .addValidator(StandardValidators.INTEGER_VALIDATOR) + .required(true) + .build(); + + static final PropertyDescriptor BLOOM_FILTER_FPP = new PropertyDescriptor.Builder() + .name("BloomFilter Probability") + .description("The desired false positive probability when using the BloomFilter filter type. " + + "Using a value of .05 for example, means we can be 95% certain that the element is in the filter, " + + "and there is a five-percent probability that the result is a false positive. " + + "When the filter returns false, we can be 100% certain that the element is not present.") + .defaultValue("0.10") + .expressionLanguageSupported(ExpressionLanguageScope.NONE) + .addValidator(StandardValidators.NUMBER_VALIDATOR) + .required(false) + .build(); + + + // RELATIONSHIPS + + static final Relationship REL_DUPLICATE = new Relationship.Builder() + .name("duplicate") + .description("Records detected as duplicates in the FlowFile content will be routed to this relationship") + .build(); + + static final Relationship REL_NON_DUPLICATE = new Relationship.Builder() + .name("non-duplicate") + .description("If the record was not found in the cache, it will be routed to this relationship") + .build(); + + static final Relationship REL_FAILURE = new Relationship.Builder() + .name("failure") + .description("If unable to communicate with the cache, the FlowFile will be penalized and routed to this relationship") + .build(); + + private List<PropertyDescriptor> descriptors; + + private Set<Relationship> relationships; + + private final Serializer<String> keySerializer = new StringSerializer(); + private final Serializer<CacheValue> cacheValueSerializer = new CacheValueSerializer(); + private final Deserializer<CacheValue> cacheValueDeserializer = new CacheValueDeserializer(); + + @Override + protected void init(final ProcessorInitializationContext context) { + final List<PropertyDescriptor> descriptors = new ArrayList<>(); + descriptors.add(RECORD_READER); + descriptors.add(RECORD_WRITER); + descriptors.add(INCLUDE_ZERO_RECORD_FLOWFILES); + descriptors.add(CACHE_IDENTIFIER); + descriptors.add(CACHE_ENTRY_IDENTIFIER); + descriptors.add(AGE_OFF_DURATION); + descriptors.add(DISTRIBUTED_CACHE_SERVICE); + descriptors.add(RECORD_HASHING_ALGORITHM); + descriptors.add(FILTER_TYPE); + descriptors.add(FILTER_CAPACITY_HINT); + descriptors.add(BLOOM_FILTER_FPP); + this.descriptors = Collections.unmodifiableList(descriptors); + + final Set<Relationship> relationships = new HashSet<>(); + relationships.add(REL_DUPLICATE); + relationships.add(REL_NON_DUPLICATE); + relationships.add(REL_FAILURE); + this.relationships = Collections.unmodifiableSet(relationships); + } + + @Override + public Set<Relationship> getRelationships() { + return this.relationships; + } + + @Override + public final List<PropertyDescriptor> getSupportedPropertyDescriptors() { + return descriptors; + } + + @Override + protected PropertyDescriptor getSupportedDynamicPropertyDescriptor(final String propertyDescriptorName) { + return new PropertyDescriptor.Builder() + .name(propertyDescriptorName) + .description("Specifies a value to use from the record that match the RecordPath: '" + + propertyDescriptorName + "' which is used together with other specified " + + "record path values to determine the uniqueness of a record. " + + "Expression Language may reference variables 'field.name', 'field.type', and 'field.value' " + + "to access information about the field and the value of the field being evaluated.") + .required(false) + .dynamic(true) + .expressionLanguageSupported(ExpressionLanguageScope.FLOWFILE_ATTRIBUTES) + .addValidator(new RecordPathPropertyNameValidator()) + .build(); + } + + @Override + protected Collection<ValidationResult> customValidate(final ValidationContext validationContext) { + RecordPathValidator recordPathValidator = new RecordPathValidator(); + final List<ValidationResult> validationResults = validationContext.getProperties().keySet().stream() + .filter(property -> property.isDynamic()) + .map(property -> recordPathValidator.validate( + "User-defined Properties", + property.getName(), + validationContext + )).collect(Collectors.toList()); + + if(validationContext.getProperty(BLOOM_FILTER_FPP).isSet()) { + final int falsePositiveProbability = validationContext.getProperty(BLOOM_FILTER_FPP).asInteger(); + if (falsePositiveProbability < 0 || falsePositiveProbability > 1) { + validationResults.add( + new ValidationResult.Builder() + .subject(BLOOM_FILTER_FPP.getName() + " out of range.") + .input(String.valueOf(falsePositiveProbability)) + .explanation("Valid values are 0.0-1.0 inclusive") + .valid(false).build()); + } + } + + if(validationContext.getProperty(CACHE_IDENTIFIER).asBoolean()) { + if(!validationContext.getProperty(DISTRIBUTED_CACHE_SERVICE).isSet()) + validationResults.add(new ValidationResult.Builder() + .subject(DISTRIBUTED_CACHE_SERVICE.getName()) + .explanation(DISTRIBUTED_CACHE_SERVICE.getName() + " is required when " + CACHE_IDENTIFIER.getName() + " is true.") + .valid(false).build()); + + if(!validationContext.getProperty(CACHE_ENTRY_IDENTIFIER).isSet()) + validationResults.add(new ValidationResult.Builder() + .subject(CACHE_ENTRY_IDENTIFIER.getName()) + .explanation(CACHE_ENTRY_IDENTIFIER.getName() + " is required when " + CACHE_IDENTIFIER.getName() + " is true.") + .valid(false).build()); + + if(!validationContext.getProperty(AGE_OFF_DURATION).isSet()) + validationResults.add(new ValidationResult.Builder() + .subject(AGE_OFF_DURATION.getName()) + .explanation(AGE_OFF_DURATION.getName() + " is required when " + CACHE_IDENTIFIER.getName() + " is true.") + .valid(false).build()); + } + + return validationResults; + } + + @OnScheduled + public void compileRecordPaths(final ProcessContext context) { + final List<String> recordPaths = new ArrayList<>(); + + recordPaths.addAll(context.getProperties().keySet().stream() + .filter(property -> property.isDynamic()) + .map(PropertyDescriptor::getName) + .collect(toList())); + + recordPathCache = new RecordPathCache(recordPaths.size()); + this.recordPaths = recordPaths; + } + + @Override + public void onTrigger(final ProcessContext context, final ProcessSession session) throws ProcessException { + final FlowFile flowFile = session.get(); + if (flowFile == null) { + return; + } + + final ComponentLog logger = getLogger(); + final String cacheKey = context.getProperty(CACHE_ENTRY_IDENTIFIER).evaluateAttributeExpressions(flowFile).getValue(); + + if (isBlank(cacheKey)) { + logger.error("FlowFile {} has no attribute for given Cache Entry Identifier", new Object[]{flowFile}); + session.transfer(session.penalize(flowFile), REL_FAILURE); + return; + } + + try { + final long now = System.currentTimeMillis(); + final DistributedMapCacheClient cache = context.getProperty(DISTRIBUTED_CACHE_SERVICE).asControllerService(DistributedMapCacheClient.class); + + final boolean shouldCacheIdentifier = context.getProperty(CACHE_IDENTIFIER).asBoolean(); + final int filterCapacity = context.getProperty(FILTER_CAPACITY_HINT).asInteger(); + Serializable serializableFilter = context.getProperty(FILTER_TYPE).getValue() + .equals(context.getProperty(HASH_SET_VALUE.getValue())) + ? new HashSet<String>(filterCapacity) + : BloomFilter.create( + Funnels.stringFunnel(Charset.defaultCharset()), + filterCapacity, + context.getProperty(BLOOM_FILTER_FPP).asDouble()); + + if(shouldCacheIdentifier && cache.containsKey(cacheKey, keySerializer)) { + CacheValue cacheValue = cache.get(cacheKey, keySerializer, cacheValueDeserializer); + Long durationMS = context.getProperty(AGE_OFF_DURATION).asTimePeriod(TimeUnit.MILLISECONDS); + + if(durationMS != null && (now >= cacheValue.getEntryTimeMS() + durationMS)) { + boolean status = cache.remove(cacheKey, keySerializer); + logger.debug("Removal of expired cached entry with key {} returned {}", new Object[]{cacheKey, status}); + } else { + serializableFilter = cacheValue.getFilter(); + } + } + + final FilterWrapper filter = FilterWrapper.create(serializableFilter); + + final String recordHashingAlgorithm = context.getProperty(RECORD_HASHING_ALGORITHM).getValue(); + final MessageDigest messageDigest = DigestUtils.getDigest(recordHashingAlgorithm); + final Boolean matchWholeRecord = context.getProperties().keySet().stream().noneMatch(p -> p.isDynamic()); + + final RecordReaderFactory readerFactory = context.getProperty(RECORD_READER).asControllerService(RecordReaderFactory.class); + final RecordSetWriterFactory writerFactory = context.getProperty(RECORD_WRITER).asControllerService(RecordSetWriterFactory.class); + final RecordReader reader = readerFactory.createRecordReader(flowFile.getAttributes(), session.read(flowFile), logger); + + FlowFile nonDuplicatesFlowFile = session.create(flowFile); + final FlowFile duplicatesFlowFile = session.create(flowFile); + final RecordSchema writeSchema = writerFactory.getSchema(flowFile.getAttributes(), reader.getSchema()); + final RecordSetWriter nonDuplicatesWriter = writerFactory.createWriter(getLogger(), writeSchema, session.write(nonDuplicatesFlowFile)); + final RecordSetWriter duplicatesWriter = writerFactory.createWriter(getLogger(), writeSchema, session.write(duplicatesFlowFile)); + + nonDuplicatesWriter.beginRecordSet(); + duplicatesWriter.beginRecordSet(); + Record record; + + while ((record = reader.nextRecord()) != null) { + String recordValue; + + if (matchWholeRecord) { + recordValue = record.getSerializedForm().get().getSerialized().toString(); + } else { + final List<String> fieldValues = new ArrayList<>(); + for (final String recordPathText : recordPaths) { + final PropertyValue recordPathPropertyValue = context.getProperty(recordPathText); + final RecordPath recordPath = recordPathCache.getCompiled(recordPathText); + final RecordPathResult result = recordPath.evaluate(record); + fieldValues.addAll(result.getSelectedFields() Review comment: If EL is present in the RecordPath property, this line (specifically the `map()` method) can cause an `IllegalStateException` because `getSelectedFields()` is being called twice, and since it returns a `Stream`, you can't iterate over it more than once. I'd recommend using `getSelectedFields().collect(Collectors.toList())` first, then iterating over that as many times as you need (you can use `List.stream()` here and keep the `forEach()` above). We might want to revisit some of the RecordPath API (maybe just `getSelectedFields()`) to return a `Supplier<Stream>` instead. ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
