/*
 * Copyright (C) 2009-2010, Google Inc. and others
 *
 * This program and the accompanying materials are made available under the
 * terms of the Eclipse Distribution License v. 1.0 which is available at
 * https://www.eclipse.org/org/documents/edl-v10.php.
 *
 * SPDX-License-Identifier: BSD-3-Clause
 */

package org.eclipse.jgit.util.io;

import static java.nio.charset.StandardCharsets.UTF_8;

import java.io.ByteArrayOutputStream;
import java.io.IOException;
import java.io.OutputStream;
import java.io.OutputStreamWriter;
import java.io.Writer;

import org.eclipse.jgit.util.RawParseUtils;

Combines messages from an OutputStream (hopefully in UTF-8) and a Writer.

This class is primarily meant for BaseConnection in contexts where a standard error stream from a command execution, as well as messages from a side-band channel, need to be combined together into a buffer to represent the complete set of messages from a remote repository.

Writes made to the writer are re-encoded as UTF-8 and interleaved into the buffer that getRawStream() also writes to.

toString() returns all written data, after converting it to a String under the assumption of UTF-8 encoding.

Internally RawParseUtils.decode(byte[]) is used by toString() tries to work out a reasonably correct character set for the raw data.

/** * Combines messages from an OutputStream (hopefully in UTF-8) and a Writer. * <p> * This class is primarily meant for {@code BaseConnection} in contexts where a * standard error stream from a command execution, as well as messages from a * side-band channel, need to be combined together into a buffer to represent * the complete set of messages from a remote repository. * <p> * Writes made to the writer are re-encoded as UTF-8 and interleaved into the * buffer that {@link #getRawStream()} also writes to. * <p> * {@link #toString()} returns all written data, after converting it to a String * under the assumption of UTF-8 encoding. * <p> * Internally {@link org.eclipse.jgit.util.RawParseUtils#decode(byte[])} is used * by {@code toString()} tries to work out a reasonably correct character set * for the raw data. */
public class MessageWriter extends Writer { private final ByteArrayOutputStream buf; private final OutputStreamWriter enc;
Create an empty writer.
/** * Create an empty writer. */
public MessageWriter() { buf = new ByteArrayOutputStream(); enc = new OutputStreamWriter(getRawStream(), UTF_8); }
{@inheritDoc}
/** {@inheritDoc} */
@Override public void write(char[] cbuf, int off, int len) throws IOException { synchronized (buf) { enc.write(cbuf, off, len); enc.flush(); } }
Get the underlying byte stream that character writes to this writer drop into.
Returns:the underlying byte stream that character writes to this writer drop into. Writes to this stream should should be in UTF-8.
/** * Get the underlying byte stream that character writes to this writer drop * into. * * @return the underlying byte stream that character writes to this writer * drop into. Writes to this stream should should be in UTF-8. */
public OutputStream getRawStream() { return buf; }
{@inheritDoc}
/** {@inheritDoc} */
@Override public void close() throws IOException { // Do nothing, we are buffered with no resources. }
{@inheritDoc}
/** {@inheritDoc} */
@Override public void flush() throws IOException { // Do nothing, we are buffered with no resources. } /** @return string version of all buffered data. */
{@inheritDoc}
/** {@inheritDoc} */
@Override public String toString() { return RawParseUtils.decode(buf.toByteArray()); } }