multipartforminputstream.java

Wicket一个开发Java Web应用程序框架。它使得开发web应用程序变得容易而轻松。 Wicket利用一个POJO data beans组件使得它可以与任何持久层技术相结合。

/*
 * 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.wicket.util.upload;

import java.io.ByteArrayOutputStream;
import java.io.IOException;
import java.io.InputStream;
import java.io.OutputStream;
import java.io.UnsupportedEncodingException;

/**
 * <p>
 * Low level API for processing file uploads.
 * 
 * <p>
 * This class can be used to process data streams conforming to MIME 'multipart' format as defined
 * in <a href="http://www.ietf.org/rfc/rfc1867.txt">RFC 1867</a>. Arbitrarily large amounts of data
 * in the stream can be processed under constant memory usage.
 * 
 * <p>
 * The format of the stream is defined in the following way:<br>
 * 
 * <code>
 *   multipart-body := preamble 1*encapsulation close-delimiter epilogue<br>
 *   encapsulation := delimiter body CRLF<br>
 *   delimiter := "--" boundary CRLF<br>
 *   close-delimiter := "--" boundary "--"<br>
 *   preamble := &lt;ignore&gt;<br>
 *   epilogue := &lt;ignore&gt;<br>
 *   body := header-part CRLF body-part<br>
 *   header-part := 1*header CRLF<br>
 *   header := header-name ":" header-value<br>
 *   header-name := &lt;printable ascii characters except ":"&gt;<br>
 *   header-value := &lt;any ascii characters except CR & LF&gt;<br>
 *   body-data := &lt;arbitrary data&gt;<br>
 * </code>
 * 
 * <p>
 * Note that body-data can contain another multipart entity. There is limited support for single
 * pass processing of such nested streams. The nested stream is <strong>required</strong> to have a
 * boundary token of the same length as the parent stream (see {@link #setBoundary(byte[])}).
 * 
 * <p>
 * Here is an example of usage of this class.<br>
 * 
 * <pre>
 *      try {
 *          MultipartStream multipartStream = new MultipartStream(input,
 *                                                                boundary);
 *          boolean nextPart = malitPartStream.skipPreamble();
 *          OutputStream output;
 *          while(nextPart) {
 *              header = chunks.readHeader();
 *              // process headers
 *              // create some output stream
 *              multipartStream.readBodyPart(output);
 *              nextPart = multipartStream.readBoundary();
 *          }
 *      } catch(MultipartStream.MalformedStreamException e) {
 *            // the stream failed to follow required syntax
 *      } catch(IOException) {
 *            // a read or write error occurred
 *      }
 * 
 * 
 * </pre>
 * 
 * @author <a href="mailto:[e-mail]">Rafal Krzewski</a>
 * @author <a href="mailto:[e-mail]">Martin Cooper</a>
 * @author Sean C. Sullivan
 * 
 * @version $Id: MultipartFormInputStream.java 584925 2007-10-15 21:21:25Z gseitz $
 */
public class MultipartFormInputStream
{

	// ----------------------------------------------------- Manifest constants


	/**
	 * The Carriage Return ASCII character value.
	 */
	public static final byte CR = 0x0D;


	/**
	 * The Line Feed ASCII character value.
	 */
	public static final byte LF = 0x0A;


	/**
	 * The dash (-) ASCII character value.
	 */
	public static final byte DASH = 0x2D;


	/**
	 * The maximum length of <code>header-part</code> that will be processed (10 kilobytes = 10240
	 * bytes.).
	 */
	public static final int HEADER_PART_SIZE_MAX = 10240;


	/**
	 * The default length of the buffer used for processing a request.
	 */
	protected static final int DEFAULT_BUFSIZE = 4096;


	/**
	 * A byte sequence that marks the end of <code>header-part</code> (<code>CRLFCRLF</code>).
	 */
	protected static final byte[] HEADER_SEPARATOR = { CR, LF, CR, LF };


	/**
	 * A byte sequence that that follows a delimiter that will be followed by an encapsulation (<code>CRLF</code>).
	 */
	protected static final byte[] FIELD_SEPARATOR = { CR, LF };


	/**
	 * A byte sequence that that follows a delimiter of the last encapsulation in the stream (<code>--</code>).
	 */
	protected static final byte[] STREAM_TERMINATOR = { DASH, DASH };


	// ----------------------------------------------------------- Data members


	/**
	 * The input stream from which data is read.
	 */
	private InputStream input;


	/**
	 * The length of the boundary token plus the leading <code>CRLF--</code>.
	 */
	private int boundaryLength;


	/**
	 * The amount of data, in bytes, that must be kept in the buffer in order to detect delimiters
	 * reliably.
	 */
	private int keepRegion;


	/**
	 * The byte sequence that partitions the stream.
	 */
	private byte[] boundary;


	/**
	 * The length of the buffer used for processing the request.
	 */
	private int bufSize;


	/**
	 * The buffer used for processing the request.
	 */
	private byte[] buffer;


	/**
	 * The index of first valid character in the buffer. <br>
	 * 0 <= head < bufSize
	 */
	private int head;


	/**
	 * The index of last valid character in the buffer + 1. <br>
	 * 0 <= tail <= bufSize
	 */
	private int tail;


	/**
	 * The content encoding to use when reading headers.
	 */
	private String headerEncoding;


	// ----------------------------------------------------------- Constructors


	/**
	 * Default constructor.
	 * 
	 * @see #MultipartFormInputStream(InputStream, byte[], int)
	 * @see #MultipartFormInputStream(InputStream, byte[])
	 * 
	 */
	public MultipartFormInputStream()
	{
	}


	/**
	 * <p>
	 * Constructs a <code>MultipartStream</code> with a custom size buffer.
	 * 
	 * <p>
	 * Note that the buffer must be at least big enough to contain the boundary string, plus 4
	 * characters for CR/LF and double dash, plus at least one byte of data. Too small a buffer size
	 * setting will degrade performance.
	 * 
	 * @param input
	 *            The <code>InputStream</code> to serve as a data source.
	 * @param boundary
	 *            The token used for dividing the stream into <code>encapsulations</code>.
	 * @param bufSize
	 *            The size of the buffer to be used, in bytes.
	 * 
	 * 
	 * @see #MultipartFormInputStream()
	 * @see #MultipartFormInputStream(InputStream, byte[])
	 * 
	 */
	public MultipartFormInputStream(InputStream input, byte[] boundary, int bufSize)
	{
		this.input = input;
		this.bufSize = bufSize;
		buffer = new byte[bufSize];

		// We prepend CR/LF to the boundary to chop trailing CR/LF from
		// body-data tokens.
		this.boundary = new byte[boundary.length + 4];
		boundaryLength = boundary.length + 4;
		keepRegion = boundary.length + 3;
		this.boundary[0] = CR;
		this.boundary[1] = LF;
		this.boundary[2] = DASH;
		this.boundary[3] = DASH;
		System.arraycopy(boundary, 0, this.boundary, 4, boundary.length);

		head = 0;
		tail = 0;
	}


	/**
	 * <p>
	 * Constructs a <code>MultipartStream</code> with a default size buffer.
	 * 
	 * @param input
	 *            The <code>InputStream</code> to serve as a data source.
	 * @param boundary
	 *            The token used for dividing the stream into <code>encapsulations</code>.
	 * 
	 * @see #MultipartFormInputStream()
	 * @see #MultipartFormInputStream(InputStream, byte[], int)
	 * 
	 */
	public MultipartFormInputStream(InputStream input, byte[] boundary)
	{
		this(input, boundary, DEFAULT_BUFSIZE);
	}


	// --------------------------------------------------------- Public methods


	/**
	 * Retrieves the character encoding used when reading the headers of an individual part. When
	 * not specified, or <code>null</code>, the platform default encoding is used.
	 * 
	 * 
	 * @return The encoding used to read part headers.
	 */
	public String getHeaderEncoding()
	{
		return headerEncoding;
	}


	/**
	 * Specifies the character encoding to be used when reading the headers of individual parts.
	 * When not specified, or <code>null</code>, the platform default encoding is used.
	 * 
	 * @param encoding
	 *            The encoding used to read part headers.
	 */
	public void setHeaderEncoding(String encoding)
	{
		headerEncoding = encoding;
	}


	/**
	 * Reads a byte from the <code>buffer</code>, and refills it as necessary.
	 * 
	 * @return The next byte from the input stream.
	 * 
	 * @exception IOException
	 *                if there is no more data available.
	 */
	public byte readByte() throws IOException
	{
		// Buffer depleted ?
		if (head == tail)
		{
			head = 0;
			// Refill.
			tail = input.read(buffer, head, bufSize);
			if (tail == -1)
			{
				// No more data available.
				throw new IOException("No more data is available");
			}
		}
		return buffer[head++];
	}


	/**
	 * Skips a <code>boundary</code> token, and checks whether more <code>encapsulations</code>
	 * are contained in the stream.
	 * 
	 * @return <code>true</code> if there are more encapsulations in this stream;
	 *         <code>false</code> otherwise.
	 * 
	 * @exception MalformedStreamException
	 *                if the stream ends unexpectedly or fails to follow required syntax.
	 */
	public boolean readBoundary() throws MalformedStreamException
	{
		byte[] marker = new byte[2];
		boolean nextChunk = false;

		head += boundaryLength;
		try
		{
			marker[0] = readByte();
			if (marker[0] == LF)
			{
				// Work around IE5 Mac bug with input type=image.
				// Because the boundary delimiter, not including the trailing
				// CRLF, must not appear within any file (RFC 2046, section
				// 5.1.1), we know the missing CR is due to a buggy browser
				// rather than a file containing something similar to a
				// boundary.
				return true;
			}

			marker[1] = readByte();
			if (arrayequals(marker, STREAM_TERMINATOR, 2))
			{
				nextChunk = false;
			}
			else if (arrayequals(marker, FIELD_SEPARATOR, 2))
			{
				nextChunk = true;
			}
			else
			{
				throw new MalformedStreamException("Unexpected characters follow a boundary");
			}
		}
		catch (IOException e)
		{
			throw new MalformedStreamException("Stream ended unexpectedly");
		}
		return nextChunk;
	}


	/**
	 * <p>
	 * Changes the boundary token used for partitioning the stream.
	 * 
	 * <p>
	 * This method allows single pass processing of nested multipart streams.
	 * 
	 * <p>
	 * The boundary token of the nested stream is <code>required</code> to be of the same length
	 * as the boundary token in parent stream.
	 * 
	 * <p>
	 * Restoring the parent stream boundary token after processing of a nested stream is left to the
	 * application.
	 * 
	 * @param boundary
	 *            The boundary to be used for parsing of the nested stream.
	 * 
	 * @exception IllegalBoundaryException
	 *                if the <code>boundary</code> has a different length than the one being
	 *                currently parsed.
	 */
	public void setBoundary(byte[] boundary) throws IllegalBoundaryException
	{
		if (boundary.length != boundaryLength - 4)
		{
			throw new IllegalBoundaryException("The length of a boundary token can not be changed");
		}
		System.arraycopy(boundary, 0, this.boundary, 4, boundary.length);
	}

	/**
	 * <p>
	 * Reads the <code>header-part</code> of the current <code>encapsulation</code>.
	 * <p>
	 * Headers are returned verbatim to the input stream, including the trailing <code>CRLF</code>
	 * marker. Parsing is left to the application.
	 * 
	 * @param maxSize
	 *            The maximum amount to read before giving up
	 * 
	 * @return The <code>header-part</code> of the current encapsulation.
	 * 
	 * @exception MalformedStreamException
	 *                if the stream ends unexpectedly.
	 */
	public String readHeaders(final int maxSize) throws MalformedStreamException
	{
		int i = 0;
		byte[] b = new byte[1];