Many resources are needed to download a project. Please understand that we have to compensate our server costs. Thank you in advance. Project price only 1 $
You can buy this project and download/modify it how often you want.
/*
* 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.parquet.hadoop;
import java.io.Closeable;
import java.io.IOException;
import java.util.ArrayList;
import java.util.Collections;
import java.util.Iterator;
import java.util.List;
import java.util.Objects;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.FileStatus;
import org.apache.hadoop.fs.FileSystem;
import org.apache.hadoop.fs.Path;
import org.apache.parquet.ParquetReadOptions;
import org.apache.parquet.Preconditions;
import org.apache.parquet.compression.CompressionCodecFactory;
import org.apache.parquet.crypto.FileDecryptionProperties;
import org.apache.parquet.filter.UnboundRecordFilter;
import org.apache.parquet.filter2.compat.FilterCompat;
import org.apache.parquet.filter2.compat.FilterCompat.Filter;
import org.apache.parquet.hadoop.api.ReadSupport;
import org.apache.parquet.hadoop.util.HadoopInputFile;
import org.apache.parquet.HadoopReadOptions;
import org.apache.parquet.hadoop.util.HiddenFileFilter;
import org.apache.parquet.io.InputFile;
/**
* Read records from a Parquet file.
* TODO: too many constructors (https://issues.apache.org/jira/browse/PARQUET-39)
*/
public class ParquetReader implements Closeable {
private final ReadSupport readSupport;
private final Iterator filesIterator;
private final ParquetReadOptions options;
private InternalParquetRecordReader reader;
/**
* @param file the file to read
* @param readSupport to materialize records
* @throws IOException if there is an error while reading
* @deprecated use {@link #builder(ReadSupport, Path)}
*/
@Deprecated
public ParquetReader(Path file, ReadSupport readSupport) throws IOException {
this(new Configuration(), file, readSupport, FilterCompat.NOOP);
}
/**
* @param conf the configuration
* @param file the file to read
* @param readSupport to materialize records
* @throws IOException if there is an error while reading
* @deprecated use {@link #builder(ReadSupport, Path)}
*/
@Deprecated
public ParquetReader(Configuration conf, Path file, ReadSupport readSupport) throws IOException {
this(conf, file, readSupport, FilterCompat.NOOP);
}
/**
* @param file the file to read
* @param readSupport to materialize records
* @param unboundRecordFilter the filter to use to filter records
* @throws IOException if there is an error while reading
* @deprecated use {@link #builder(ReadSupport, Path)}
*/
@Deprecated
public ParquetReader(Path file, ReadSupport readSupport, UnboundRecordFilter unboundRecordFilter) throws IOException {
this(new Configuration(), file, readSupport, FilterCompat.get(unboundRecordFilter));
}
/**
* @param conf the configuration
* @param file the file to read
* @param readSupport to materialize records
* @param unboundRecordFilter the filter to use to filter records
* @throws IOException if there is an error while reading
* @deprecated use {@link #builder(ReadSupport, Path)}
*/
@Deprecated
public ParquetReader(Configuration conf, Path file, ReadSupport readSupport, UnboundRecordFilter unboundRecordFilter) throws IOException {
this(conf, file, readSupport, FilterCompat.get(unboundRecordFilter));
}
private ParquetReader(Configuration conf,
Path file,
ReadSupport readSupport,
FilterCompat.Filter filter) throws IOException {
this(Collections.singletonList((InputFile) HadoopInputFile.fromPath(file, conf)),
HadoopReadOptions.builder(conf, file)
.withRecordFilter(Objects.requireNonNull(filter, "filter cannot be null"))
.build(),
readSupport);
}
private ParquetReader(List files,
ParquetReadOptions options,
ReadSupport readSupport) throws IOException {
this.readSupport = readSupport;
this.options = options;
this.filesIterator = files.iterator();
}
/**
* @return the next record or null if finished
* @throws IOException if there is an error while reading
*/
public T read() throws IOException {
try {
if (reader != null && reader.nextKeyValue()) {
return reader.getCurrentValue();
} else {
initReader();
return reader == null ? null : read();
}
} catch (InterruptedException e) {
throw new IOException(e);
}
}
/**
* @return the row index of the last read row. If no row has been processed, returns -1.
*/
public long getCurrentRowIndex() {
if (reader == null) {
return -1;
}
return reader.getCurrentRowIndex();
}
private void initReader() throws IOException {
if (reader != null) {
reader.close();
reader = null;
}
if (filesIterator.hasNext()) {
InputFile file = filesIterator.next();
ParquetFileReader fileReader = ParquetFileReader.open(file, options);
reader = new InternalParquetRecordReader<>(readSupport, options.getRecordFilter());
reader.initialize(fileReader, options);
}
}
@Override
public void close() throws IOException {
if (reader != null) {
reader.close();
}
}
public static Builder read(InputFile file) throws IOException {
return new Builder<>(file);
}
public static Builder builder(ReadSupport readSupport, Path path) {
return new Builder<>(readSupport, path);
}
public static class Builder {
private final ReadSupport readSupport;
private final InputFile file;
private final Path path;
private Filter filter = null;
protected Configuration conf;
private ParquetReadOptions.Builder optionsBuilder;
@Deprecated
private Builder(ReadSupport readSupport, Path path) {
this.readSupport = Objects.requireNonNull(readSupport, "readSupport cannot be null");
this.file = null;
this.path = Objects.requireNonNull(path, "path cannot be null");
this.conf = new Configuration();
this.optionsBuilder = HadoopReadOptions.builder(conf, path);
}
@Deprecated
protected Builder(Path path) {
this.readSupport = null;
this.file = null;
this.path = Objects.requireNonNull(path, "path cannot be null");
this.conf = new Configuration();
this.optionsBuilder = HadoopReadOptions.builder(conf, path);
}
protected Builder(InputFile file) {
this.readSupport = null;
this.file = Objects.requireNonNull(file, "file cannot be null");
this.path = null;
if (file instanceof HadoopInputFile) {
HadoopInputFile hadoopFile = (HadoopInputFile) file;
this.conf = hadoopFile.getConfiguration();
optionsBuilder = HadoopReadOptions.builder(conf, hadoopFile.getPath());
} else {
this.conf = new Configuration();
optionsBuilder = HadoopReadOptions.builder(conf);
}
}
// when called, resets options to the defaults from conf
public Builder withConf(Configuration conf) {
this.conf = Objects.requireNonNull(conf, "conf cannot be null");
// previous versions didn't use the builder, so may set filter before conf. this maintains
// compatibility for filter. other options are reset by a new conf.
this.optionsBuilder = HadoopReadOptions.builder(conf, path);
if (filter != null) {
optionsBuilder.withRecordFilter(filter);
}
return this;
}
public Builder withFilter(Filter filter) {
this.filter = filter;
optionsBuilder.withRecordFilter(filter);
return this;
}
public Builder useSignedStringMinMax(boolean useSignedStringMinMax) {
optionsBuilder.useSignedStringMinMax(useSignedStringMinMax);
return this;
}
public Builder useSignedStringMinMax() {
optionsBuilder.useSignedStringMinMax();
return this;
}
public Builder useStatsFilter(boolean useStatsFilter) {
optionsBuilder.useStatsFilter(useStatsFilter);
return this;
}
public Builder useStatsFilter() {
optionsBuilder.useStatsFilter();
return this;
}
public Builder useDictionaryFilter(boolean useDictionaryFilter) {
optionsBuilder.useDictionaryFilter(useDictionaryFilter);
return this;
}
public Builder useDictionaryFilter() {
optionsBuilder.useDictionaryFilter();
return this;
}
public Builder useRecordFilter(boolean useRecordFilter) {
optionsBuilder.useRecordFilter(useRecordFilter);
return this;
}
public Builder useRecordFilter() {
optionsBuilder.useRecordFilter();
return this;
}
public Builder useColumnIndexFilter(boolean useColumnIndexFilter) {
optionsBuilder.useColumnIndexFilter(useColumnIndexFilter);
return this;
}
public Builder useColumnIndexFilter() {
optionsBuilder.useColumnIndexFilter();
return this;
}
public Builder usePageChecksumVerification(boolean usePageChecksumVerification) {
optionsBuilder.usePageChecksumVerification(usePageChecksumVerification);
return this;
}
public Builder useBloomFilter(boolean useBloomFilter) {
optionsBuilder.useBloomFilter(useBloomFilter);
return this;
}
public Builder useBloomFilter() {
optionsBuilder.useBloomFilter();
return this;
}
public Builder usePageChecksumVerification() {
optionsBuilder.usePageChecksumVerification();
return this;
}
public Builder withFileRange(long start, long end) {
optionsBuilder.withRange(start, end);
return this;
}
public Builder withCodecFactory(CompressionCodecFactory codecFactory) {
optionsBuilder.withCodecFactory(codecFactory);
return this;
}
public Builder withDecryption(FileDecryptionProperties fileDecryptionProperties) {
optionsBuilder.withDecryption(fileDecryptionProperties);
return this;
}
public Builder set(String key, String value) {
optionsBuilder.set(key, value);
return this;
}
protected ReadSupport getReadSupport() {
// if readSupport is null, the protected constructor must have been used
Preconditions.checkArgument(readSupport != null,
"[BUG] Classes that extend Builder should override getReadSupport()");
return readSupport;
}
public ParquetReader build() throws IOException {
ParquetReadOptions options = optionsBuilder.build();
if (path != null) {
FileSystem fs = path.getFileSystem(conf);
FileStatus stat = fs.getFileStatus(path);
if (stat.isFile()) {
return new ParquetReader<>(
Collections.singletonList((InputFile) HadoopInputFile.fromStatus(stat, conf)),
options,
getReadSupport());
} else {
List files = new ArrayList<>();
for (FileStatus fileStatus : fs.listStatus(path, HiddenFileFilter.INSTANCE)) {
files.add(HadoopInputFile.fromStatus(fileStatus, conf));
}
return new ParquetReader(files, options, getReadSupport());
}
} else {
return new ParquetReader<>(Collections.singletonList(file), options, getReadSupport());
}
}
}
}