org.apache.coyote.http11.filters

Class ChunkedInputFilter

java.lang.Object

org.apache.coyote.http11.filters.ChunkedInputFilter

All Implemented Interfaces:

InputFilter, InputBuffer, ApplicationBufferHandler

public class ChunkedInputFilter
extends java.lang.Object
implements InputFilter, ApplicationBufferHandler

Chunked input filter. Parses chunked data according to http://www.w3.org/Protocols/rfc2616/rfc2616-sec3.html#sec3.6.1

Author:

Remy Maucherat

Field Summary

Fields

Modifier and Type	Field and Description
protected InputBuffer	buffer Next buffer in the pipeline.
protected static ByteChunk	ENCODING
protected static java.lang.String	ENCODING_NAME
protected boolean	endChunk Flag set to true when the end chunk has been read.
protected boolean	needCRLFParse Flag set to true if the next call to doRead() must parse a CRLF pair before doing anything else.
protected java.nio.ByteBuffer	readChunk Byte chunk used to read bytes.
protected int	remaining Number of bytes remaining in the current chunk.
protected ByteChunk	trailingHeaders Byte chunk used to store trailing headers.

Constructor Summary

Constructors

Constructor and Description
ChunkedInputFilter(int maxTrailerSize, java.util.Set<java.lang.String> allowedTrailerHeaders, int maxExtensionSize, int maxSwallowSize)

Method Summary

Modifier and Type	Method and Description
int	available() Amount of bytes still available in a buffer.
int	doRead(ApplicationBufferHandler handler) Read from the input stream into the ByteBuffer provided by the ApplicationBufferHandler.
long	end() End the current request.
void	expand(int size)
java.nio.ByteBuffer	getByteBuffer()
ByteChunk	getEncodingName() Return the name of the associated encoding; Here, the value is "identity".
boolean	isFinished() Has the request body been read fully?
protected boolean	parseChunkHeader() Parse the header of a chunk.
protected void	parseCRLF(boolean tolerant) Parse CRLF at end of chunk.
protected void	parseEndChunk() Parse end chunk data.
protected int	readBytes() Read bytes from the previous buffer.
void	recycle() Make the filter ready to process the next request.
void	setBuffer(InputBuffer buffer) Set the next buffer in the filter pipeline.
void	setByteBuffer(java.nio.ByteBuffer buffer)
void	setRequest(Request request) Read the content length from the request.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

ENCODING_NAME

protected static final java.lang.String ENCODING_NAME

See Also:

Constant Field Values

ENCODING

protected static final ByteChunk ENCODING

buffer

protected InputBuffer buffer

Next buffer in the pipeline.

remaining

protected int remaining

Number of bytes remaining in the current chunk.

readChunk

protected java.nio.ByteBuffer readChunk

Byte chunk used to read bytes.

endChunk

protected boolean endChunk

Flag set to true when the end chunk has been read.

trailingHeaders

protected final ByteChunk trailingHeaders

Byte chunk used to store trailing headers.

needCRLFParse

protected boolean needCRLFParse

Flag set to true if the next call to doRead() must parse a CRLF pair before doing anything else.

Constructor Detail

ChunkedInputFilter

public ChunkedInputFilter(int maxTrailerSize,
                          java.util.Set<java.lang.String> allowedTrailerHeaders,
                          int maxExtensionSize,
                          int maxSwallowSize)

Method Detail

doRead

public int doRead(ApplicationBufferHandler handler)
           throws java.io.IOException

Description copied from interface: InputBuffer

Read from the input stream into the ByteBuffer provided by the ApplicationBufferHandler. IMPORTANT: the current model assumes that the protocol will 'own' the ByteBuffer and return a pointer to it.

Specified by:

doRead in interface InputBuffer

Parameters:

handler - ApplicationBufferHandler that provides the buffer to read data into.

Returns:

The number of bytes that have been added to the buffer or -1 for end of stream

Throws:

java.io.IOException - If an I/O error occurs reading from the input stream

setRequest

public void setRequest(Request request)

Read the content length from the request.

Specified by:

setRequest in interface InputFilter

Parameters:

request - The request to be associated with this filter

end

public long end()
         throws java.io.IOException

End the current request.

Specified by:

end in interface InputFilter

Returns:

0 is the expected return value. A positive value indicates that too many bytes were read. This method is allowed to use buffer.doRead to consume extra bytes. The result of this method can't be negative (if an error happens, an IOException should be thrown instead).

Throws:

java.io.IOException - If an error happens

available

public int available()

Amount of bytes still available in a buffer.

Specified by:

available in interface InputFilter

Returns:

The number of bytes in the buffer

setBuffer

public void setBuffer(InputBuffer buffer)

Set the next buffer in the filter pipeline.

Specified by:

setBuffer in interface InputFilter

Parameters:

buffer - The next buffer

recycle

public void recycle()

Make the filter ready to process the next request.

Specified by:

recycle in interface InputFilter

getEncodingName

public ByteChunk getEncodingName()

Return the name of the associated encoding; Here, the value is "identity".

Specified by:

getEncodingName in interface InputFilter

Returns:

The encoding name as a byte chunk to facilitate comparison with the value read from the HTTP headers which will also be a ByteChunk

isFinished

public boolean isFinished()

Description copied from interface: InputFilter

Has the request body been read fully?

Specified by:

isFinished in interface InputFilter

Returns:

true if the request body has been fully read, otherwise false

readBytes

protected int readBytes()
                 throws java.io.IOException

Read bytes from the previous buffer.

Returns:

The byte count which has been read

Throws:

java.io.IOException - Read error

parseChunkHeader

protected boolean parseChunkHeader()
                            throws java.io.IOException

Parse the header of a chunk. A chunk header can look like one of the following:
A10CRLF
F23;chunk-extension to be ignoredCRLF

The letters before CRLF or ';' (whatever comes first) must be valid hex digits. We should not parse F23IAMGONNAMESSTHISUP34CRLF as a valid header according to the spec.

Returns:

true if the chunk header has been successfully parsed

Throws:

java.io.IOException - Read error

parseCRLF

protected void parseCRLF(boolean tolerant)
                  throws java.io.IOException

Parse CRLF at end of chunk.

Parameters:

tolerant - Should tolerant parsing (LF and CRLF) be used? This is recommended (RFC2616, section 19.3) for message headers.

Throws:

java.io.IOException - An error occurred parsing CRLF

parseEndChunk

protected void parseEndChunk()
                      throws java.io.IOException

Parse end chunk data.

Throws:

java.io.IOException - Error propagation

setByteBuffer

public void setByteBuffer(java.nio.ByteBuffer buffer)

Specified by:

setByteBuffer in interface ApplicationBufferHandler

getByteBuffer

public java.nio.ByteBuffer getByteBuffer()

Specified by:

getByteBuffer in interface ApplicationBufferHandler

expand

public void expand(int size)

Specified by:

expand in interface ApplicationBufferHandler