Github user vanzin commented on a diff in the pull request:
https://github.com/apache/spark/pull/8880#discussion_r41585457
--- Diff:
core/src/main/scala/org/apache/spark/crypto/CryptoOutputStream.scala ---
@@ -0,0 +1,224 @@
+/*
+ * 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.spark.crypto
+
+import java.io.{IOException, FilterOutputStream, OutputStream}
+import java.nio.ByteBuffer
+import java.security.GeneralSecurityException
+
+import com.google.common.base.Preconditions
+
+import org.apache.spark.Logging
+
+/**
+ * CryptoOutputStream encrypts data. It is not thread-safe. AES CTR mode is
+ * required in order to ensure that the plain text and cipher text have a
1:1
+ * mapping. The encryption is buffer based. The key points of the
encryption are
+ * (1) calculating counter and (2) padding through stream position.
+ * <p/>
+ * counter = base + pos/(algorithm blocksize);
+ * padding = pos%(algorithm blocksize);
+ * <p/>
+ * The underlying stream offset is maintained as state.
+ */
+class CryptoOutputStream(out: OutputStream, codecVal: CryptoCodec,
bufferSizeVal: Int,
+ keyVal: Array[Byte], ivVal: Array[Byte],
streamOffsetVal: Long) extends
+FilterOutputStream(out: OutputStream) with Logging {
+ var codec: CryptoCodec = null
+ var encryptor: Encryptor = null
+ var bufferSize: Int = 0
+ /**
+ * Input data buffer. The data starts at inBuffer.position() and ends at
+ * inBuffer.limit().
+ */
+ var inBuffer: ByteBuffer = null
+ /**
+ * Encrypted data buffer. The data starts at outBuffer.position() and
ends at
+ * outBuffer.limit();
+ */
+ var outBuffer: ByteBuffer = null
+ var streamOffset: Long = 0
+ /**
+ * Padding = pos%(algorithm blocksize); Padding is put into {@link
#inBuffer}
+ * before any other data goes in. The purpose of padding is to put input
data
+ * at proper position.
+ */
+ var padding: Byte = 0
+ var closed: Boolean = false
+ var key: Array[Byte] = null
+ var initIV: Array[Byte] = null
+ var iv: Array[Byte] = null
+ var tmpBuf: Array[Byte] = null
+ val oneByteBuf: Array[Byte] = new Array[Byte](1)
+
+ CryptoStreamUtils.checkCodec(codecVal)
+ bufferSize = CryptoStreamUtils.checkBufferSize(codecVal, bufferSizeVal)
+ codec = codecVal
+ key = keyVal.clone
+ initIV = ivVal.clone
+ iv = ivVal.clone
+ inBuffer = ByteBuffer.allocateDirect(bufferSize)
+ outBuffer = ByteBuffer.allocateDirect(bufferSize)
+ streamOffset = streamOffsetVal
+ try {
+ encryptor = codec.createEncryptor
+ }
+ catch {
+ case e: GeneralSecurityException => {
+ throw e
+ }
+ }
+ updateEncryptor
+
+
+ def this(out: OutputStream, codec: CryptoCodec, bufferSize: Int, key:
Array[Byte],
+ iv: Array[Byte]) {
+ this(out, codec, bufferSize, key, iv, 0)
+ }
+
+ def this(out: OutputStream, codec: CryptoCodec, key: Array[Byte], iv:
Array[Byte], streamOffset:
+ Long) {
+ this(out, codec, CryptoStreamUtils.getBufferSize, key, iv,
streamOffset)
+ }
+
+ def this(out: OutputStream, codec: CryptoCodec, key: Array[Byte], iv:
Array[Byte]) {
+ this(out, codec, key, iv, 0)
+ }
+
+ def getWrappedStream: OutputStream = {
+ out
+ }
+
+ /**
+ * Encryption is buffer based.
+ * If there is enough room in {@link #inBuffer}, then write to this
buffer.
+ * If {@link #inBuffer} is full, then do encryption and write data to the
+ * underlying stream.
+ * @param b the data.
+ * @param offVal the start offset in the data.
+ * @param lenVal the number of bytes to write.
+ * @throws IOException
+ */
+ override def write(b: Array[Byte], offVal: Int, lenVal: Int) {
+ var off = offVal
+ var len = lenVal
+ checkStream
+ if (b == null) {
+ throw new NullPointerException
+ }
+ else if (off < 0 || len < 0 || off > b.length || len > b.length - off)
{
+ throw new IndexOutOfBoundsException
+ }
+ while (len > 0) {
+ val remaining: Int = inBuffer.remaining
+ if (len < remaining) {
+ inBuffer.put(b, off, len)
+ len = 0
+ }
+ else {
+ inBuffer.put(b, off, remaining)
+ off += remaining
+ len -= remaining
+ encrypt
+ }
+ }
+ }
+
+ /**
+ * Do the encryption, input is {@link #inBuffer} and output is
+ * {@link #outBuffer}.
+ */
+ def encrypt {
+ Preconditions.checkState(inBuffer.position >= padding)
+ if (inBuffer.position != padding) {
+ inBuffer.flip
+ outBuffer.clear
+ encryptor.encrypt(inBuffer, outBuffer)
+ inBuffer.clear
+ outBuffer.flip
+ if (padding > 0) {
+ outBuffer.position(padding)
+ padding = 0
+ }
+ val len: Int = outBuffer.remaining
+ val tmp: Array[Byte] = getTmpBuf
+ outBuffer.get(tmp, 0, len)
+ out.write(tmp, 0, len)
+ streamOffset += len
+ if (encryptor.isContextReset) {
+ updateEncryptor
+ }
+ }
+ }
+
+ /** Update the {@link #encryptor}: calculate counter and {@link
#padding}. */
+ def updateEncryptor {
+ val counter: Long = streamOffset / codec.getCipherSuite.algoBlockSize
+ padding = (streamOffset %
codec.getCipherSuite.algoBlockSize).asInstanceOf[Byte]
+ inBuffer.position(padding)
+ codec.calculateIV(initIV, counter, iv)
+ encryptor.init(key, iv)
+ }
+
+ def getTmpBuf: Array[Byte] = {
+ if (tmpBuf == null) {
+ tmpBuf = new Array[Byte](bufferSize)
+ }
+ tmpBuf
+ }
+
+ override def close {
+ this.synchronized {
--- End diff --
There's no other synchronization in this class, which makes me think this
one is unnecessary. The stream is just not thread-safe.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
