[ https://issues.apache.org/jira/browse/HADOOP-17825?focusedWorklogId=633722&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-633722 ]
ASF GitHub Bot logged work on HADOOP-17825: ------------------------------------------- Author: ASF GitHub Bot Created on: 04/Aug/21 18:02 Start Date: 04/Aug/21 18:02 Worklog Time Spent: 10m Work Description: viirya commented on a change in pull request #3250: URL: https://github.com/apache/hadoop/pull/3250#discussion_r682842698 ########## File path: hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/io/compress/zlib/BuiltInGzipCompressor.java ########## @@ -0,0 +1,270 @@ +/* + * 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.hadoop.io.compress.zlib; + +import java.io.IOException; +import java.util.zip.Checksum; +import java.util.zip.Deflater; +import java.util.zip.GZIPOutputStream; + +import org.apache.hadoop.conf.Configuration; +import org.apache.hadoop.io.compress.Compressor; +import org.apache.hadoop.io.compress.DoNotPool; +import org.apache.hadoop.util.DataChecksum; + +/** + * A {@link Compressor} based on the popular gzip compressed file format. + * http://www.gzip.org/ + */ +@DoNotPool +public class BuiltInGzipCompressor implements Compressor { + + /** + * Fixed ten-byte gzip header. See {@link GZIPOutputStream}'s source for + * details. + */ + private static final byte[] GZIP_HEADER = new byte[]{ + 0x1f, (byte) 0x8b, 0x08, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00}; + + private byte[] GZIP_TRAILER = new byte[]{0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00}; + + private final int GZIP_HEADER_LEN = 10; + private final int GZIP_TRAILER_LEN = 8; + + private Deflater deflater; + + private int headerOff = 0; + private int trailerOff = 0; + + private byte[] userBuf = null; + private int userBufOff = 0; + private int userBufLen = 0; + + private int headerBytesWritten = 0; + private int trailerBytesWritten = 0; + + private int currentBufLen = 0; + + private final Checksum crc = DataChecksum.newCrc32(); + + private BuiltInGzipDecompressor.GzipStateLabel state; + + public BuiltInGzipCompressor(Configuration conf) { init(conf); } + + @Override + public boolean finished() { + // Only if the trailer is also written, it is thought as finished. + return deflater.finished() && state == BuiltInGzipDecompressor.GzipStateLabel.FINISHED; + } + + @Override + public boolean needsInput() { + if (state == BuiltInGzipDecompressor.GzipStateLabel.INFLATE_STREAM) { + return deflater.needsInput(); + } + + return (state != BuiltInGzipDecompressor.GzipStateLabel.FINISHED); + } + + @Override + public int compress(byte[] b, int off, int len) throws IOException { + int numAvailBytes = 0; + + // If we are not within uncompressed data yet, output the header. + if (state != BuiltInGzipDecompressor.GzipStateLabel.INFLATE_STREAM) { + if (userBufLen <= 0) { + return numAvailBytes; + } + + int outputHeaderSize = writeHeader(b, off, len); + headerBytesWritten += outputHeaderSize; + + // Completes header output. + if (headerOff == GZIP_HEADER_LEN) { + state = BuiltInGzipDecompressor.GzipStateLabel.INFLATE_STREAM; + } + + numAvailBytes += outputHeaderSize; + + if (outputHeaderSize == len) { + return numAvailBytes; + } + + off += outputHeaderSize; + len -= outputHeaderSize; + } + + if (state == BuiltInGzipDecompressor.GzipStateLabel.INFLATE_STREAM) { + // hand off user data (or what's left of it) to Deflater--but note that + // Deflater may not have consumed all of previous bufferload, in which case + // userBufLen will be zero + if (userBufLen > 0) { + deflater.setInput(userBuf, userBufOff, userBufLen); + + crc.update(userBuf, userBufOff, userBufLen); // CRC-32 is on uncompressed data + + currentBufLen = userBufLen; + userBufOff += userBufLen; + userBufLen = 0; + } + + + // now compress it into b[] + int deflated = deflater.deflate(b, off, len); + + numAvailBytes += deflated; + off += deflated; + len -= deflated; + + // All current input are processed. Going to output trailer. + if (deflater.finished()) { + state = BuiltInGzipDecompressor.GzipStateLabel.TRAILER_CRC; + fillTrailer(); + } else { + return numAvailBytes; + } + } + + numAvailBytes += writeTrailer(b, off, len); + trailerBytesWritten += numAvailBytes; + + if (trailerOff == GZIP_TRAILER_LEN) { + state = BuiltInGzipDecompressor.GzipStateLabel.FINISHED; + currentBufLen = 0; + headerOff = trailerOff = 0; + } + + return numAvailBytes; + } + + @Override + public long getBytesRead() { + return deflater.getTotalIn(); + } + + @Override + public long getBytesWritten() { + return headerBytesWritten + deflater.getTotalOut() + trailerBytesWritten; + } + + @Override + public void end() { + deflater.end(); + } + + @Override + public void finish() { + deflater.finish(); + state = BuiltInGzipDecompressor.GzipStateLabel.FINISHED; + } + + private void init(Configuration conf) { + ZlibCompressor.CompressionLevel level = ZlibFactory.getCompressionLevel(conf); + ZlibCompressor.CompressionStrategy strategy = ZlibFactory.getCompressionStrategy(conf); + + // 'true' (nowrap) => Deflater will handle raw deflate stream only + deflater = new Deflater(level.compressionLevel(), true); + deflater.setStrategy(strategy.compressionStrategy()); + + state = BuiltInGzipDecompressor.GzipStateLabel.HEADER_BASIC; + + GZIP_TRAILER = new byte[]{0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00}; Review comment: > I tried to run the tests locally but found one failure: > > ``` > [INFO] Running org.apache.hadoop.io.compress.TestCodec > [ERROR] Tests run: 28, Failures: 1, Errors: 0, Skipped: 4, Time elapsed: 30.735 s <<< FAILURE! - in org.apache.hadoop.io.compress.TestCodec > [ERROR] testCodecPoolCompressorReinit(org.apache.hadoop.io.compress.TestCodec) Time elapsed: 0.005 s <<< FAILURE! > java.lang.AssertionError: Compressed bytes contrary to configuration > ``` > > Not sure why the CI was all green. This also failed at my side. But it (DefaultCodec) is not touched by this PR. And CI passed it. Maybe it is due to environment difference? -- 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: common-issues-unsubscr...@hadoop.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org Issue Time Tracking ------------------- Worklog Id: (was: 633722) Time Spent: 6h 50m (was: 6h 40m) > Add BuiltInGzipCompressor > ------------------------- > > Key: HADOOP-17825 > URL: https://issues.apache.org/jira/browse/HADOOP-17825 > Project: Hadoop Common > Issue Type: Improvement > Reporter: L. C. Hsieh > Priority: Major > Labels: pull-request-available > Time Spent: 6h 50m > Remaining Estimate: 0h > > Currently, GzipCodec only supports BuiltInGzipDecompressor, if native zlib is > not loaded. So, without Hadoop native codec installed, saving SequenceFile > using GzipCodec will throw exception like "SequenceFile doesn't work with > GzipCodec without native-hadoop code!" > Same as other codecs which we migrated to using prepared packages (lz4, > snappy), it will be better if we support GzipCodec generally without Hadoop > native codec installed. Similar to BuiltInGzipDecompressor, we can use Java > Deflater to support BuiltInGzipCompressor. -- This message was sent by Atlassian Jira (v8.3.4#803005) --------------------------------------------------------------------- To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: common-issues-h...@hadoop.apache.org