amogh-jahagirdar commented on code in PR #12774: URL: https://github.com/apache/iceberg/pull/12774#discussion_r2440744790
########## core/src/main/java/org/apache/iceberg/formats/FormatModelRegistry.java: ########## @@ -0,0 +1,241 @@ +/* + * 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.iceberg.formats; + +import java.util.List; +import java.util.Map; +import java.util.Objects; +import org.apache.iceberg.DataFile; +import org.apache.iceberg.DeleteFile; +import org.apache.iceberg.FileContent; +import org.apache.iceberg.FileFormat; +import org.apache.iceberg.common.DynMethods; +import org.apache.iceberg.deletes.EqualityDeleteWriter; +import org.apache.iceberg.deletes.PositionDelete; +import org.apache.iceberg.deletes.PositionDeleteWriter; +import org.apache.iceberg.encryption.EncryptedOutputFile; +import org.apache.iceberg.io.DataWriter; +import org.apache.iceberg.io.InputFile; +import org.apache.iceberg.relocated.com.google.common.annotations.VisibleForTesting; +import org.apache.iceberg.relocated.com.google.common.base.Preconditions; +import org.apache.iceberg.relocated.com.google.common.collect.ImmutableList; +import org.apache.iceberg.relocated.com.google.common.collect.Maps; +import org.apache.iceberg.util.Pair; +import org.slf4j.Logger; +import org.slf4j.LoggerFactory; + +/** + * A registry that manages file-format-specific readers and writers through a unified object model + * factory interface. + * + * <p>This registry provides access to {@link ReadBuilder}s for data consumption and various writer + * builders: + * + * <ul> + * <li>{@link WriteBuilder} for basic file writing, + * <li>{@link DataWriteBuilder} for data files, + * <li>{@link EqualityDeleteWriteBuilder} for equality deletes, + * <li>{@link PositionDeleteWriteBuilder} for position deletes. + * </ul> + * + * The appropriate builder is selected based on {@link FileFormat} and object model name. + * + * <p>{@link FormatModel} objects are registered through {@link #register(FormatModel)} and used for + * creating readers and writers. Read builders are returned directly from the factory. Write + * builders may be wrapped in specialized content file writer implementations depending on the + * requested builder type. + */ +public final class FormatModelRegistry { + private static final Logger LOG = LoggerFactory.getLogger(FormatModelRegistry.class); + // The list of classes which are used for registering the reader and writer builders + private static final List<String> CLASSES_TO_REGISTER = ImmutableList.of(); + + private static final Map<Pair<FileFormat, Class<?>>, FormatModel<?, ?>> MODELS = + Maps.newConcurrentMap(); + + private FormatModelRegistry() {} + + /** + * Registers an {@link FormatModel} in this registry. + * + * <p>The {@link FormatModel} creates readers and writers for a specific combinations of file + * format (Parquet, ORC, Avro) and object model (for example: "generic", "spark", "flink", etc.). + * Registering custom factories allows integration of new data processing engines for the + * supported file formats with Iceberg's file access mechanisms. + * + * <p>Each factory must be uniquely identified by its combination of file format and object model + * name. This uniqueness constraint prevents ambiguity when selecting factories for read and write + * operations. + * + * @param formatModel the factory implementation to register + * @throws IllegalArgumentException if a factory is already registered for the combination of + * {@link FormatModel#format()} and {@link FormatModel#type()} + */ + public static void register(FormatModel<?, ?> formatModel) { + Pair<FileFormat, Class<?>> key = Pair.of(formatModel.format(), formatModel.type()); + + FormatModel<?, ?> existing = MODELS.get(key); + Preconditions.checkArgument( + existing == null || checkFormatModelEquals(existing, formatModel), + "Cannot register %s: %s is registered for format=%s type=%s schemaType=%s", + formatModel.getClass(), + existing == null ? null : existing.getClass(), + key.first(), + key.second(), + existing == null ? null : existing.schemaType()); + + MODELS.put(key, formatModel); + } + + private static boolean checkFormatModelEquals( Review Comment: `isSameFormatModel(FileFormat<?, ?> first, FormatModel<?, ?> second)` ########## core/src/main/java/org/apache/iceberg/formats/ReadBuilder.java: ########## @@ -0,0 +1,118 @@ +/* + * 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.iceberg.formats; + +import java.nio.ByteBuffer; +import java.util.Map; +import org.apache.iceberg.Schema; +import org.apache.iceberg.expressions.Expression; +import org.apache.iceberg.io.CloseableIterable; +import org.apache.iceberg.mapping.NameMapping; + +/** + * Builder interface for creating file readers across supported data file formats. The {@link + * FormatModel} implementations provides appropriate {@link ReadBuilder} instances + * + * <p>The {@link ReadBuilder} follows the builder pattern to configure and create {@link + * CloseableIterable} instances that read data from source files. Configuration options include + * schema projection, predicate filtering, record batching, and encryption settings. + * + * <p>This interface is directly exposed to users for parameterizing readers. + */ +public interface ReadBuilder { + /** + * Restricts the read to the given range: [start, start + length). + * + * @param newStart the start position for this read + * @param newLength the length of the range this read should scan + */ + ReadBuilder split(long newStart, long newLength); + + /** Set the projection schema. */ + ReadBuilder project(Schema schema); + + /** Sets the expected output schema. If not provided derived from the {@link #project(Schema)}. */ + default ReadBuilder outputSchema(Object schema) { + throw new UnsupportedOperationException("Not implemented yet"); + } + + /** + * Configures whether filtering should be case-sensitive. If the reader supports filtering, it + * must respect this setting. + * + * @param caseSensitive indicates if filtering is case-sensitive + */ + ReadBuilder caseSensitive(boolean caseSensitive); + + /** + * Pushes down the {@link Expression} filter for the reader to prevent reading unnecessary + * records. Some readers might not be able to filter some part of the exception. In this case the + * reader might return unfiltered or partially filtered rows. It is the caller's responsibility to + * apply the filter again. The default implementation sets the filter to be case-sensitive. + * + * @param filter the filter to set + */ + ReadBuilder filter(Expression filter); + + /** + * Sets configuration key/value pairs for the reader. Reader builders should ignore configuration + * keys not known for them. + */ + ReadBuilder set(String key, String value); + + /** + * Sets multiple configuration key/value pairs for the reader. Reader builders should ignore + * configuration keys not known for them. + */ + default ReadBuilder setAll(Map<String, String> properties) { + properties.forEach(this::set); + return this; + } + + /** Enables reusing the containers returned by the reader. Decreases pressure on GC. */ + ReadBuilder reuseContainers(); + + /** Sets the batch size for vectorized readers. */ + ReadBuilder recordsPerBatch(int numRowsPerBatch); Review Comment: This is another one where I know that's on the current ReadBuilder API but I wonder if it makes sense to rename to be a more concise `batchSize` API? ########## core/src/main/java/org/apache/iceberg/formats/ReadBuilder.java: ########## @@ -0,0 +1,118 @@ +/* + * 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.iceberg.formats; + +import java.nio.ByteBuffer; +import java.util.Map; +import org.apache.iceberg.Schema; +import org.apache.iceberg.expressions.Expression; +import org.apache.iceberg.io.CloseableIterable; +import org.apache.iceberg.mapping.NameMapping; + +/** + * Builder interface for creating file readers across supported data file formats. The {@link + * FormatModel} implementations provides appropriate {@link ReadBuilder} instances + * + * <p>The {@link ReadBuilder} follows the builder pattern to configure and create {@link + * CloseableIterable} instances that read data from source files. Configuration options include + * schema projection, predicate filtering, record batching, and encryption settings. + * + * <p>This interface is directly exposed to users for parameterizing readers. + */ +public interface ReadBuilder { + /** + * Restricts the read to the given range: [start, start + length). + * + * @param newStart the start position for this read + * @param newLength the length of the range this read should scan + */ + ReadBuilder split(long newStart, long newLength); Review Comment: `(long start, long length)`? Don't think we need the new prefix Also minor point, but why not call this `range` ? I'm guessing this was just based off the current `ReadBuilder` but with the new interface we don't necessarily need to stick to that. -- 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. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
