DCMTK  Version 3.6.1 20170228
OFFIS DICOM Toolkit
 All Classes Files Functions Variables Typedefs Enumerations Enumerator Friends Macros Modules Pages
Public Member Functions | Private Member Functions | Private Attributes | List of all members
DcmRLEEncoder Class Reference

this class implements an RLE compressor conforming to the DICOM standard. More...

Public Member Functions

 DcmRLEEncoder (int doPad)
 default constructor More...
 
 ~DcmRLEEncoder ()
 destructor
 
void add (unsigned char ch)
 this method adds one byte to the byte stream to be compressed with the RLE compressor. More...
 
void add (const unsigned char *buf, size_t bufcount)
 this method adds a block of bytes to the byte stream to be compressed with the RLE compressor. More...
 
void flush ()
 this method finalizes the compressed RLE stream, i.e. More...
 
size_t size () const
 returns the size of compressed RLE stream in bytes. More...
 
OFBool fail () const
 returns true if the RLE compressor has run out of memory. More...
 
void write (void *target) const
 copies the compressed RLE byte stream into a target array of at least size() bytes. More...
 
void write (DcmEncoderOutputStream &os) const
 copies the compressed RLE byte stream into an output stream More...
 

Private Member Functions

 DcmRLEEncoder (const DcmRLEEncoder &)
 private undefined copy constructor
 
DcmRLEEncoderoperator= (const DcmRLEEncoder &)
 private undefined copy assignment operator
 
void move (size_t numberOfBytes)
 this method moves the given number of bytes from buff_ to currentBlock_ and "flushes" currentBlock_ to blockList_ if necessary. More...
 

Private Attributes

int fail_
 this flag indicates a failure of the RLE codec. More...
 
int pad_
 this flag indicates whether the RLE codec must pad encoded data to an even number of bytes (as required by DICOM). More...
 
unsigned char * currentBlock_
 this member points to a block of size DcmRLEEncoder_BLOCKSIZE (unless fail_ is true). More...
 
size_t offset_
 contains the number of bytes already written the the memory block pointed to by currentBlock_. More...
 
OFList< unsigned char * > blockList_
 this member contains a list of memory blocks of size DcmRLEEncoder_BLOCKSIZE which already have been filled with encoded RLE data. More...
 
unsigned char * RLE_buff_
 this member points to a buffer of 132 bytes that is used by the RLE encoding algorithm.
 
int RLE_prev_
 value of the last byte fed to the RLE compressor. More...
 
int RLE_pcount_
 repeat counter, for RLE compressor may temporarily become negative, guaranteed to be >= 0 between method calls.
 
unsigned int RLE_bindex_
 index of next unused byte in RLE_buff_.
 

Detailed Description

this class implements an RLE compressor conforming to the DICOM standard.

The class is loosely based on an implementation by Phil Norman.

Constructor & Destructor Documentation

DcmRLEEncoder::DcmRLEEncoder ( int  doPad)
inline

default constructor

Parameters
doPadif true, RLE codec will pad output data to even number of bytes

Member Function Documentation

void DcmRLEEncoder::add ( unsigned char  ch)
inline

this method adds one byte to the byte stream to be compressed with the RLE compressor.

Parameters
chbyte to be added
void DcmRLEEncoder::add ( const unsigned char *  buf,
size_t  bufcount 
)
inline

this method adds a block of bytes to the byte stream to be compressed with the RLE compressor.

Parameters
bufbuffer to be added
bufcountnumber of bytes in buffer
OFBool DcmRLEEncoder::fail ( ) const
inline

returns true if the RLE compressor has run out of memory.

In this case, no output has been created.

void DcmRLEEncoder::flush ( )
inline

this method finalizes the compressed RLE stream, i.e.

flushes all pending literal or repeat runs. This method can be called at any time; however, it must be called before size() or write() can be used. Intermediate calls should be avoided since they possibly decrease the compression ratio.

void DcmRLEEncoder::move ( size_t  numberOfBytes)
inlineprivate

this method moves the given number of bytes from buff_ to currentBlock_ and "flushes" currentBlock_ to blockList_ if necessary.

Parameters
numberOfBytesnumber of bytes to copy
size_t DcmRLEEncoder::size ( ) const
inline

returns the size of compressed RLE stream in bytes.

The size is guaranteed to be an even number of bytes (padded with a trailing zero byte as required by DICOM if necessary). This method may only be called after flush() has been executed to finalize the compressed stream.

Returns
size of compressed stream, in bytes
void DcmRLEEncoder::write ( void *  target) const
inline

copies the compressed RLE byte stream into a target array of at least size() bytes.

Parameters
targetpointer to array of at least size() bytes, must not be NULL.
void DcmRLEEncoder::write ( DcmEncoderOutputStream os) const
inline

copies the compressed RLE byte stream into an output stream

Parameters
osoutput stream

References DcmEncoderOutputStream::write().

Member Data Documentation

OFList<unsigned char *> DcmRLEEncoder::blockList_
private

this member contains a list of memory blocks of size DcmRLEEncoder_BLOCKSIZE which already have been filled with encoded RLE data.

The current block (pointed to by currentBlock_) is not contained in this list.

unsigned char* DcmRLEEncoder::currentBlock_
private

this member points to a block of size DcmRLEEncoder_BLOCKSIZE (unless fail_ is true).

This is the current block of data to which the RLE stream is written

int DcmRLEEncoder::fail_
private

this flag indicates a failure of the RLE codec.

Once a failure is flagged, the codec will consume all input and not produce any more output. A failure status can only be caused by an out-of-memory condition.

size_t DcmRLEEncoder::offset_
private

contains the number of bytes already written the the memory block pointed to by currentBlock_.

Value is always less than DcmRLEEncoder_BLOCKSIZE.

int DcmRLEEncoder::pad_
private

this flag indicates whether the RLE codec must pad encoded data to an even number of bytes (as required by DICOM).

True if padding is required, false otherwise

int DcmRLEEncoder::RLE_prev_
private

value of the last byte fed to the RLE compressor.

This byte is not yet stored in the RLE_buff_ buffer. Type is int because this allows an "impossible" -1 as default value


The documentation for this class was generated from the following file:


Generated on Tue Feb 28 2017 for DCMTK Version 3.6.1 20170228 by Doxygen 1.8.8