RakNet::BitStream Class Reference

#include <BitStream.h>

List of all members.

Public Member Functions

 BitStream ()
 Default Constructor.
 BitStream (const unsigned int initialBytesToAllocate)
 Create the bitstream, with some number of bytes to immediately allocate.
 BitStream (unsigned char *_data, const unsigned int lengthInBytes, bool _copyData)
 Initialize the BitStream, immediately setting the data it contains to a predefined pointer.
void Reset (void)
 Resets the bitstream for reuse.
template<class templateType >
bool Serialize (bool writeToBitstream, templateType &inOutTemplateVar)
 Bidirectional serialize/deserialize any integral type to/from a bitstream.
template<class templateType >
bool SerializeDelta (bool writeToBitstream, templateType &inOutCurrentValue, const templateType &lastValue)
 Bidirectional serialize/deserialize any integral type to/from a bitstream.
template<class templateType >
bool SerializeDelta (bool writeToBitstream, templateType &inOutCurrentValue)
 Bidirectional version of SerializeDelta when you don't know what the last value is, or there is no last value.
template<class templateType >
bool SerializeCompressed (bool writeToBitstream, templateType &inOutTemplateVar)
 Bidirectional serialize/deserialize any integral type to/from a bitstream.
template<class templateType >
bool SerializeCompressedDelta (bool writeToBitstream, templateType &inOutCurrentValue, const templateType &lastValue)
 Bidirectional serialize/deserialize any integral type to/from a bitstream.
template<class templateType >
bool SerializeCompressedDelta (bool writeToBitstream, templateType &inOutTemplateVar)
 Save as SerializeCompressedDelta(templateType &currentValue, const templateType &lastValue) when we have an unknown second parameter.
bool Serialize (bool writeToBitstream, char *inOutByteArray, const unsigned int numberOfBytes)
 Bidirectional serialize/deserialize an array or casted stream or raw data. This does NOT do endian swapping.
bool SerializeFloat16 (bool writeToBitstream, float &inOutFloat, float floatMin, float floatMax)
 Serialize a float into 2 bytes, spanning the range between floatMin and floatMax.
template<class serializationType , class sourceType >
bool SerializeCasted (bool writeToBitstream, sourceType &value)
template<class templateType >
bool SerializeBitsFromIntegerRange (bool writeToBitstream, templateType &value, const templateType minimum, const templateType maximum, bool allowOutsideRange=false)
template<class templateType >
bool SerializeBitsFromIntegerRange (bool writeToBitstream, templateType &value, const templateType minimum, const templateType maximum, const int requiredBits, bool allowOutsideRange=false)
template<class templateType >
bool SerializeNormVector (bool writeToBitstream, templateType &x, templateType &y, templateType &z)
 Bidirectional serialize/deserialize a normalized 3D vector, using (at most) 4 bytes + 3 bits instead of 12-24 bytes.
template<class templateType >
bool SerializeVector (bool writeToBitstream, templateType &x, templateType &y, templateType &z)
 Bidirectional serialize/deserialize a vector, using 10 bytes instead of 12.
template<class templateType >
bool SerializeNormQuat (bool writeToBitstream, templateType &w, templateType &x, templateType &y, templateType &z)
 Bidirectional serialize/deserialize a normalized quaternion in 6 bytes + 4 bits instead of 16 bytes. Slightly lossy.
template<class templateType >
bool SerializeOrthMatrix (bool writeToBitstream, templateType &m00, templateType &m01, templateType &m02, templateType &m10, templateType &m11, templateType &m12, templateType &m20, templateType &m21, templateType &m22)
 Bidirectional serialize/deserialize an orthogonal matrix by creating a quaternion, and writing 3 components of the quaternion in 2 bytes each.
bool SerializeBits (bool writeToBitstream, unsigned char *inOutByteArray, const BitSize_t numberOfBitsToSerialize, const bool rightAlignedBits=true)
 Bidirectional serialize/deserialize numberToSerialize bits to/from the input.
template<class templateType >
void Write (const templateType &inTemplateVar)
 Write any integral type to a bitstream.
template<class templateType >
void WritePtr (templateType *inTemplateVar)
 Write the dereferenced pointer to any integral type to a bitstream.
template<class templateType >
void WriteDelta (const templateType &currentValue, const templateType &lastValue)
 Write any integral type to a bitstream.
template<class templateType >
void WriteDelta (const templateType &currentValue)
 WriteDelta when you don't know what the last value is, or there is no last value.
template<class templateType >
void WriteCompressed (const templateType &inTemplateVar)
 Write any integral type to a bitstream.
template<class templateType >
void WriteCompressedDelta (const templateType &currentValue, const templateType &lastValue)
 Write any integral type to a bitstream.
template<class templateType >
void WriteCompressedDelta (const templateType &currentValue)
 Save as WriteCompressedDelta(const templateType &currentValue, const templateType &lastValue) when we have an unknown second parameter.
template<class templateType >
bool Read (templateType &outTemplateVar)
 Read any integral type from a bitstream.
template<class templateType >
bool ReadDelta (templateType &outTemplateVar)
 Read any integral type from a bitstream.
template<class templateType >
bool ReadCompressed (templateType &outTemplateVar)
 Read any integral type from a bitstream.
template<class templateType >
bool ReadCompressedDelta (templateType &outTemplateVar)
 Read any integral type from a bitstream.
bool Read (BitStream *bitStream, BitSize_t numberOfBits)
 Read one bitstream to another.
void Write (const char *inputByteArray, const unsigned int numberOfBytes)
 Write an array or casted stream or raw data. This does NOT do endian swapping.
void Write (BitStream *bitStream, BitSize_t numberOfBits)
 Write one bitstream to another.
void WriteFloat16 (float x, float floatMin, float floatMax)
 Write a float into 2 bytes, spanning the range between floatMin and floatMax.
template<class serializationType , class sourceType >
void WriteCasted (const sourceType &value)
template<class templateType >
void WriteBitsFromIntegerRange (const templateType value, const templateType minimum, const templateType maximum, bool allowOutsideRange=false)
template<class templateType >
void WriteBitsFromIntegerRange (const templateType value, const templateType minimum, const templateType maximum, const int requiredBits, bool allowOutsideRange=false)
template<class templateType >
void WriteNormVector (templateType x, templateType y, templateType z)
 Write a normalized 3D vector, using (at most) 4 bytes + 3 bits instead of 12-24 bytes.
template<class templateType >
void WriteVector (templateType x, templateType y, templateType z)
 Write a vector, using 10 bytes instead of 12.
template<class templateType >
void WriteNormQuat (templateType w, templateType x, templateType y, templateType z)
 Write a normalized quaternion in 6 bytes + 4 bits instead of 16 bytes. Slightly lossy.
template<class templateType >
void WriteOrthMatrix (templateType m00, templateType m01, templateType m02, templateType m10, templateType m11, templateType m12, templateType m20, templateType m21, templateType m22)
 Write an orthogonal matrix by creating a quaternion, and writing 3 components of the quaternion in 2 bytes each.
bool Read (char *output, const unsigned int numberOfBytes)
 Read an array or casted stream of byte.
bool ReadFloat16 (float &outFloat, float floatMin, float floatMax)
 Read a float into 2 bytes, spanning the range between floatMin and floatMax.
template<class serializationType , class sourceType >
bool ReadCasted (sourceType &value)
template<class templateType >
bool ReadBitsFromIntegerRange (templateType &value, const templateType minimum, const templateType maximum, bool allowOutsideRange=false)
template<class templateType >
bool ReadBitsFromIntegerRange (templateType &value, const templateType minimum, const templateType maximum, const int requiredBits, bool allowOutsideRange=false)
template<class templateType >
bool ReadNormVector (templateType &x, templateType &y, templateType &z)
 Read a normalized 3D vector, using (at most) 4 bytes + 3 bits instead of 12-24 bytes.
template<class templateType >
bool ReadVector (templateType &x, templateType &y, templateType &z)
 Read 3 floats or doubles, using 10 bytes, where those float or doubles comprise a vector.
template<class templateType >
bool ReadNormQuat (templateType &w, templateType &x, templateType &y, templateType &z)
 Read a normalized quaternion in 6 bytes + 4 bits instead of 16 bytes.
template<class templateType >
bool ReadOrthMatrix (templateType &m00, templateType &m01, templateType &m02, templateType &m10, templateType &m11, templateType &m12, templateType &m20, templateType &m21, templateType &m22)
 Read an orthogonal matrix from a quaternion, reading 3 components of the quaternion in 2 bytes each and extrapolatig the 4th.
void ResetReadPointer (void)
 Sets the read pointer back to the beginning of your data.
void ResetWritePointer (void)
 Sets the write pointer back to the beginning of your data.
void AssertStreamEmpty (void)
 This is good to call when you are done with the stream to make sure you didn't leave any data left over void.
void PrintBits (char *out) const
 RAKNET_DEBUG_PRINTF the bits in the stream. Great for debugging.
void IgnoreBits (const BitSize_t numberOfBits)
 Ignore data we don't intend to read.
void IgnoreBytes (const unsigned int numberOfBytes)
 Ignore data we don't intend to read.
void SetWriteOffset (const BitSize_t offset)
 Move the write pointer to a position on the array.
BitSize_t GetNumberOfBitsUsed (void) const
 Returns the length in bits of the stream.
BitSize_t GetNumberOfBytesUsed (void) const
 Returns the length in bytes of the stream.
BitSize_t GetReadOffset (void) const
 Returns the number of bits into the stream that we have read.
void SetReadOffset (const BitSize_t newReadOffset)
 Sets the read bit index.
BitSize_t GetNumberOfUnreadBits (void) const
 Returns the number of bits left in the stream that haven't been read.
BitSize_t CopyData (unsigned char **_data) const
 Makes a copy of the internal data for you _data will point to the stream. Partial bytes are left aligned.
unsigned char * GetData (void) const
void WriteBits (const unsigned char *inByteArray, BitSize_t numberOfBitsToWrite, const bool rightAlignedBits=true)
 Write numberToWrite bits from the input source.
void WriteAlignedBytes (const unsigned char *inByteArray, const unsigned int numberOfBytesToWrite)
 Align the bitstream to the byte boundary and then write the specified number of bits.
void WriteAlignedBytesSafe (const char *inByteArray, const unsigned int inputLength, const unsigned int maxBytesToWrite)
 Aligns the bitstream, writes inputLength, and writes input. Won't write beyond maxBytesToWrite.
bool ReadAlignedBytes (unsigned char *inOutByteArray, const unsigned int numberOfBytesToRead)
 Read bits, starting at the next aligned bits.
bool ReadAlignedBytesSafe (char *inOutByteArray, int &inputLength, const int maxBytesToRead)
 Reads what was written by WriteAlignedBytesSafe.
bool ReadAlignedBytesSafeAlloc (char **outByteArray, int &inputLength, const unsigned int maxBytesToRead)
 Same as ReadAlignedBytesSafe() but allocates the memory for you using new, rather than assuming it is safe to write to.
void AlignWriteToByteBoundary (void)
 Align the next write and/or read to a byte boundary.
void AlignReadToByteBoundary (void)
 Align the next write and/or read to a byte boundary.
bool ReadBits (unsigned char *inOutByteArray, BitSize_t numberOfBitsToRead, const bool alignBitsToRight=true)
 Read numberOfBitsToRead bits to the output source.
void Write0 (void)
 Write a 0.
void Write1 (void)
 Write a 1.
bool ReadBit (void)
 Reads 1 bit and returns true if that bit is 1 and false if it is 0.
void AssertCopyData (void)
 If we used the constructor version with copy data off, this makes sure it is set to on and the data pointed to is copied.
void SetNumberOfBitsAllocated (const BitSize_t lengthInBits)
 Use this if you pass a pointer copy to the constructor (_copyData==false) and want to overallocate to prevent reallocation.
void AddBitsAndReallocate (const BitSize_t numberOfBitsToWrite)
 Reallocates (if necessary) in preparation of writing numberOfBitsToWrite.
bool Read (char *varString)
 Read strings, non reference.
void PadWithZeroToByteLength (unsigned int bytes)
 Write zeros until the bitstream is filled up to bytes.
template<>
void Write (const bool &inTemplateVar)
 Write a bool to a bitstream.
template<>
void Write (const SystemAddress &inTemplateVar)
 Write a systemAddress to a bitstream.
template<>
void Write (const RakString &inTemplateVar)
 Write a string to a bitstream.
template<>
void WriteDelta (const bool &currentValue, const bool &lastValue)
 Write a bool delta. Same thing as just calling Write.
template<>
void WriteCompressed (const float &inTemplateVar)
 For values between -1 and 1.
template<>
void WriteCompressed (const double &inTemplateVar)
 For values between -1 and 1.
template<>
void WriteCompressed (const RakString &inTemplateVar)
 Compress the string.
template<>
void WriteCompressedDelta (const bool &currentValue, const bool &lastValue)
 Write a bool delta. Same thing as just calling Write.
template<>
void WriteCompressedDelta (const bool &currentValue)
 Save as WriteCompressedDelta(bool currentValue, const templateType &lastValue) when we have an unknown second bool.
template<>
bool Read (bool &outTemplateVar)
 Read a bool from a bitstream.
template<>
bool Read (SystemAddress &outTemplateVar)
 Read a systemAddress from a bitstream.
template<>
bool ReadDelta (bool &outTemplateVar)
 Read a bool from a bitstream.
template<>
bool ReadCompressed (float &outTemplateVar)
 For values between -1 and 1.
template<>
bool ReadCompressed (double &outTemplateVar)
 For values between -1 and 1.
template<>
bool ReadCompressed (RakString &outTemplateVar)
 For strings.
template<>
bool ReadCompressedDelta (bool &outTemplateVar)
 Read a bool from a bitstream.

Static Public Member Functions

static int NumberOfLeadingZeroes (uint8_t x)
static bool DoEndianSwap (void)
 ---- Member function template specialization declarations ----


Detailed Description

This class allows you to write and read native types as a string of bits. BitStream is used extensively throughout RakNet and is designed to be used by users as well.
See also:
BitStreamSample.txt

Constructor & Destructor Documentation

RakNet::BitStream::BitStream ( const unsigned int  initialBytesToAllocate  ) 

Create the bitstream, with some number of bytes to immediately allocate.

There is no benefit to calling this, unless you know exactly how many bytes you need and it is greater than BITSTREAM_STACK_ALLOCATION_SIZE. In that case all it does is save you one or more realloc calls.

Parameters:
[in] initialBytesToAllocate the number of bytes to pre-allocate.

RakNet::BitStream::BitStream ( unsigned char *  _data,
const unsigned int  lengthInBytes,
bool  _copyData 
)

Initialize the BitStream, immediately setting the data it contains to a predefined pointer.

Set _copyData to true if you want to make an internal copy of the data you are passing. Set it to false to just save a pointer to the data. You shouldn't call Write functions with _copyData as false, as this will write to unallocated memory 99% of the time you will use this function to cast Packet::data to a bitstream for reading, in which case you should write something as follows:

 RakNet::BitStream bs(packet->data, packet->length, false);
Parameters:
[in] _data An array of bytes.
[in] lengthInBytes Size of the _data.
[in] _copyData true or false to make a copy of _data or not.


Member Function Documentation

void RakNet::BitStream::AlignReadToByteBoundary ( void   )  [inline]

Align the next write and/or read to a byte boundary.

This can be used to 'waste' bits to byte align for efficiency reasons It can also be used to force coalesced bitstreams to start on byte boundaries so so WriteAlignedBits and ReadAlignedBits both calculate the same offset when aligning.

void RakNet::BitStream::AlignWriteToByteBoundary ( void   )  [inline]

Align the next write and/or read to a byte boundary.

This can be used to 'waste' bits to byte align for efficiency reasons It can also be used to force coalesced bitstreams to start on byte boundaries so so WriteAlignedBits and ReadAlignedBits both calculate the same offset when aligning.

BitSize_t RakNet::BitStream::CopyData ( unsigned char **  _data  )  const

Makes a copy of the internal data for you _data will point to the stream. Partial bytes are left aligned.

Parameters:
[out] _data The allocated copy of GetData()
Returns:
The length in bits of the stream.

unsigned char* RakNet::BitStream::GetData ( void   )  const [inline]

Gets the data that BitStream is writing to / reading from. Partial bytes are left aligned.

Returns:
A pointer to the internal state

void RakNet::BitStream::IgnoreBits ( const BitSize_t  numberOfBits  ) 

Ignore data we don't intend to read.

Parameters:
[in] numberOfBits The number of bits to ignore

void RakNet::BitStream::IgnoreBytes ( const unsigned int  numberOfBytes  ) 

Ignore data we don't intend to read.

Parameters:
[in] numberOfBits The number of bytes to ignore

static int RakNet::BitStream::NumberOfLeadingZeroes ( uint8_t  x  )  [static]

Get the number of leading zeros for a number

Parameters:
[in] x Number to test

template<>
bool RakNet::BitStream::Read ( SystemAddress outTemplateVar  )  [inline]

Read a systemAddress from a bitstream.

Parameters:
[in] outTemplateVar The value to read

template<>
bool RakNet::BitStream::Read ( bool &  outTemplateVar  )  [inline]

Read a bool from a bitstream.

Parameters:
[in] outTemplateVar The value to read

bool RakNet::BitStream::Read ( char *  output,
const unsigned int  numberOfBytes 
)

Read an array or casted stream of byte.

The array is raw data. There is no automatic endian conversion with this function

Parameters:
[in] output The result byte array. It should be larger than numberOfBytes.
[in] numberOfBytes The number of byte to read
Returns:
true on success false if there is some missing bytes.

bool RakNet::BitStream::Read ( BitStream bitStream,
BitSize_t  numberOfBits 
)

Read one bitstream to another.

Parameters:
[in] numberOfBits bits to read
bitStream the bitstream to read into from
Returns:
true on success, false on failure.

template<class templateType >
bool RakNet::BitStream::Read ( templateType &  outTemplateVar  )  [inline]

Read any integral type from a bitstream.

Read any integral type from a bitstream. Define __BITSTREAM_NATIVE_END if you need endian swapping.

Define __BITSTREAM_NATIVE_END if you need endian swapping.

Parameters:
[in] outTemplateVar The value to read
Returns:
true on success, false on failure.
Parameters:
[in] outTemplateVar The value to read

bool RakNet::BitStream::ReadAlignedBytes ( unsigned char *  inOutByteArray,
const unsigned int  numberOfBytesToRead 
)

Read bits, starting at the next aligned bits.

Note that the modulus 8 starting offset of the sequence must be the same as was used with WriteBits. This will be a problem with packet coalescence unless you byte align the coalesced packets.

Parameters:
[in] inOutByteArray The byte array larger than numberOfBytesToRead
[in] numberOfBytesToRead The number of byte to read from the internal state
Returns:
true if there is enough byte.

bool RakNet::BitStream::ReadAlignedBytesSafe ( char *  inOutByteArray,
int &  inputLength,
const int  maxBytesToRead 
)

Reads what was written by WriteAlignedBytesSafe.

Parameters:
[in] inOutByteArray The data
[in] maxBytesToRead Maximum number of bytes to read
Returns:
true on success, false on failure.

bool RakNet::BitStream::ReadAlignedBytesSafeAlloc ( char **  outByteArray,
int &  inputLength,
const unsigned int  maxBytesToRead 
)

Same as ReadAlignedBytesSafe() but allocates the memory for you using new, rather than assuming it is safe to write to.

Parameters:
[in] outByteArray outByteArray will be deleted if it is not a pointer to 0
Returns:
true on success, false on failure.

bool RakNet::BitStream::ReadBits ( unsigned char *  inOutByteArray,
BitSize_t  numberOfBitsToRead,
const bool  alignBitsToRight = true 
)

Read numberOfBitsToRead bits to the output source.

alignBitsToRight should be set to true to convert internal bitstream data to userdata. It should be false if you used WriteBits with rightAlignedBits false

Parameters:
[in] inOutByteArray The resulting bits array
[in] numberOfBitsToRead The number of bits to read
[in] alignBitsToRight if true bits will be right aligned.
Returns:
true if there is enough bits to read

template<class templateType >
bool RakNet::BitStream::ReadBitsFromIntegerRange ( templateType &  value,
const templateType  minimum,
const templateType  maximum,
const int  requiredBits,
bool  allowOutsideRange = false 
) [inline]

Parameters:
[in] requiredBits Primarily for internal use, called from above function() after calculating number of bits needed to represent maximum-minimum

template<class templateType >
bool RakNet::BitStream::ReadBitsFromIntegerRange ( templateType &  value,
const templateType  minimum,
const templateType  maximum,
bool  allowOutsideRange = false 
) [inline]

Given the minimum and maximum values for an integer type, figure out the minimum number of bits to represent the range Then read only those bits

Note:
A static is used so that the required number of bits for (maximum-minimum) is only calculated once. This does require that minimum and are fixed values for a given line of code for the life of the program
Parameters:
[in] value Integer value to read, which should be between minimum and maximum
[in] minimum Minimum value of value
[in] maximum Maximum value of value
[in] allowOutsideRange If true, all sends will take an extra bit, however value can deviate from outside minimum and maximum. If false, will assert if the value deviates. This should match the corresponding value passed to Write().

template<class serializationType , class sourceType >
bool RakNet::BitStream::ReadCasted ( sourceType &  value  )  [inline]

Read one type serialized to another (smaller) type, to save bandwidth serializationType should be uint8_t, uint16_t, uint24_t, or uint32_t Example: int num; ReadCasted<uint8_t>(num); would read 1 bytefrom the stream, and put the value in an integer

Parameters:
[in] value The value to write

template<class templateType >
bool RakNet::BitStream::ReadCompressed ( templateType &  outTemplateVar  )  [inline]

Read any integral type from a bitstream.

Undefine __BITSTREAM_NATIVE_END if you need endian swapping. For floating point, this is lossy, using 2 bytes for a float and 4 for a double. The range must be between -1 and +1. For non-floating point, this is lossless, but only has benefit if you use less than half the bits of the type If you are not using __BITSTREAM_NATIVE_END the opposite is true for types larger than 1 byte

Parameters:
[in] outTemplateVar The value to read
Returns:
true on success, false on failure.
Undefine __BITSTREAM_NATIVE_END if you need endian swapping. For floating point, this is lossy, using 2 bytes for a float and 4 for a double. The range must be between -1 and +1. For non-floating point, this is lossless, but only has benefit if you use less than half the bits of the type If you are not using __BITSTREAM_NATIVE_END the opposite is true for types larger than 1 byte
Parameters:
[in] outTemplateVar The value to read

template<>
bool RakNet::BitStream::ReadCompressedDelta ( bool &  outTemplateVar  )  [inline]

Read a bool from a bitstream.

Parameters:
[in] outTemplateVar The value to read

template<class templateType >
bool RakNet::BitStream::ReadCompressedDelta ( templateType &  outTemplateVar  )  [inline]

Read any integral type from a bitstream.

If the written value differed from the value compared against in the write function, var will be updated. Otherwise it will retain the current value. the current value will be updated. For floating point, this is lossy, using 2 bytes for a float and 4 for a double. The range must be between -1 and +1. For non-floating point, this is lossless, but only has benefit if you use less than half the bits of the type If you are not using __BITSTREAM_NATIVE_END the opposite is true for types larger than 1 byte ReadCompressedDelta is only valid from a previous call to WriteDelta

Parameters:
[in] outTemplateVar The value to read
Returns:
true on success, false on failure.
If the written value differed from the value compared against in the write function, var will be updated. Otherwise it will retain the current value. the current value will be updated. For floating point, this is lossy, using 2 bytes for a float and 4 for a double. The range must be between -1 and +1. For non-floating point, this is lossless, but only has benefit if you use less than half the bits of the type If you are not using __BITSTREAM_NATIVE_END the opposite is true for types larger than 1 byte ReadCompressedDelta is only valid from a previous call to WriteDelta
Parameters:
[in] outTemplateVar The value to read

template<>
bool RakNet::BitStream::ReadDelta ( bool &  outTemplateVar  )  [inline]

Read a bool from a bitstream.

Parameters:
[in] outTemplateVar The value to read

template<class templateType >
bool RakNet::BitStream::ReadDelta ( templateType &  outTemplateVar  )  [inline]

Read any integral type from a bitstream.

If the written value differed from the value compared against in the write function, var will be updated. Otherwise it will retain the current value. ReadDelta is only valid from a previous call to WriteDelta

Parameters:
[in] outTemplateVar The value to read
Returns:
true on success, false on failure.
If the written value differed from the value compared against in the write function, var will be updated. Otherwise it will retain the current value. ReadDelta is only valid from a previous call to WriteDelta
Parameters:
[in] outTemplateVar The value to read

bool RakNet::BitStream::ReadFloat16 ( float &  outFloat,
float  floatMin,
float  floatMax 
)

Read a float into 2 bytes, spanning the range between floatMin and floatMax.

Parameters:
[in] outFloat The float to read
[in] floatMin Predetermined minimum value of f
[in] floatMax Predetermined maximum value of f

template<class templateType >
bool RakNet::BitStream::ReadNormQuat ( templateType &  w,
templateType &  x,
templateType &  y,
templateType &  z 
) [inline]

Read a normalized quaternion in 6 bytes + 4 bits instead of 16 bytes.

Parameters:
[in] w w
[in] x x
[in] y y
[in] z z
Returns:
true on success, false on failure.

template<class templateType >
bool RakNet::BitStream::ReadNormVector ( templateType &  x,
templateType &  y,
templateType &  z 
) [inline]

Read a normalized 3D vector, using (at most) 4 bytes + 3 bits instead of 12-24 bytes.

Will further compress y or z axis aligned vectors. Accurate to 1/32767.5.

Parameters:
[in] x x
[in] y y
[in] z z
Returns:
true on success, false on failure.

template<class templateType >
bool RakNet::BitStream::ReadOrthMatrix ( templateType &  m00,
templateType &  m01,
templateType &  m02,
templateType &  m10,
templateType &  m11,
templateType &  m12,
templateType &  m20,
templateType &  m21,
templateType &  m22 
) [inline]

Read an orthogonal matrix from a quaternion, reading 3 components of the quaternion in 2 bytes each and extrapolatig the 4th.

Use 6 bytes instead of 36 Lossy, although the result is renormalized

Returns:
true on success, false on failure.

template<class templateType >
bool RakNet::BitStream::ReadVector ( templateType &  x,
templateType &  y,
templateType &  z 
) [inline]

Read 3 floats or doubles, using 10 bytes, where those float or doubles comprise a vector.

Loses accuracy to about 3/10ths and only saves 2 bytes, so only use if accuracy is not important.

Parameters:
[in] x x
[in] y y
[in] z z
Returns:
true on success, false on failure.

bool RakNet::BitStream::Serialize ( bool  writeToBitstream,
char *  inOutByteArray,
const unsigned int  numberOfBytes 
) [inline]

Bidirectional serialize/deserialize an array or casted stream or raw data. This does NOT do endian swapping.

Parameters:
[in] writeToBitstream true to write from your data to this bitstream. False to read from this bitstream and write to your data
[in] inOutByteArray a byte buffer
[in] numberOfBytes the size of input in bytes
Returns:
true if writeToBitstream is true. true if writeToBitstream is false and the read was successful. false if writeToBitstream is false and the read was not successful.

template<class templateType >
bool RakNet::BitStream::Serialize ( bool  writeToBitstream,
templateType &  inOutTemplateVar 
) [inline]

Bidirectional serialize/deserialize any integral type to/from a bitstream.

Undefine __BITSTREAM_NATIVE_END if you need endian swapping.

Parameters:
[in] writeToBitstream true to write from your data to this bitstream. False to read from this bitstream and write to your data
[in] inOutTemplateVar The value to write
Returns:
true if writeToBitstream is true. true if writeToBitstream is false and the read was successful. false if writeToBitstream is false and the read was not successful.

bool RakNet::BitStream::SerializeBits ( bool  writeToBitstream,
unsigned char *  inOutByteArray,
const BitSize_t  numberOfBitsToSerialize,
const bool  rightAlignedBits = true 
) [inline]

Bidirectional serialize/deserialize numberToSerialize bits to/from the input.

Right aligned data means in the case of a partial byte, the bits are aligned from the right (bit 0) rather than the left (as in the normal internal representation) You would set this to true when writing user data, and false when copying bitstream data, such as writing one bitstream to another

Parameters:
[in] writeToBitstream true to write from your data to this bitstream. False to read from this bitstream and write to your data
[in] inOutByteArray The data
[in] numberOfBitsToSerialize The number of bits to write
[in] rightAlignedBits if true data will be right aligned
Returns:
true if writeToBitstream is true. true if writeToBitstream is false and the read was successful. false if writeToBitstream is false and the read was not successful.

template<class templateType >
bool RakNet::BitStream::SerializeBitsFromIntegerRange ( bool  writeToBitstream,
templateType &  value,
const templateType  minimum,
const templateType  maximum,
const int  requiredBits,
bool  allowOutsideRange = false 
) [inline]

Parameters:
[in] requiredBits Primarily for internal use, called from above function() after calculating number of bits needed to represent maximum-minimum

template<class templateType >
bool RakNet::BitStream::SerializeBitsFromIntegerRange ( bool  writeToBitstream,
templateType &  value,
const templateType  minimum,
const templateType  maximum,
bool  allowOutsideRange = false 
) [inline]

Given the minimum and maximum values for an integer type, figure out the minimum number of bits to represent the range Then serialize only those bits

Note:
A static is used so that the required number of bits for (maximum-minimum) is only calculated once. This does require that minimum and are fixed values for a given line of code for the life of the program
Parameters:
[in] writeToBitstream true to write from your data to this bitstream. False to read from this bitstream and write to your data
[in] value Integer value to write, which should be between minimum and maximum
[in] minimum Minimum value of value
[in] maximum Maximum value of value
[in] allowOutsideRange If true, all sends will take an extra bit, however value can deviate from outside minimum and maximum. If false, will assert if the value deviates

template<class serializationType , class sourceType >
bool RakNet::BitStream::SerializeCasted ( bool  writeToBitstream,
sourceType &  value 
) [inline]

Serialize one type casted to another (smaller) type, to save bandwidth serializationType should be uint8_t, uint16_t, uint24_t, or uint32_t Example: int num=53; SerializeCasted<uint8_t>(true, num); would use 1 byte to write what would otherwise be an integer (4 or 8 bytes)

Parameters:
[in] writeToBitstream true to write from your data to this bitstream. False to read from this bitstream and write to your data
[in] value The value to serialize

template<class templateType >
bool RakNet::BitStream::SerializeCompressed ( bool  writeToBitstream,
templateType &  inOutTemplateVar 
) [inline]

Bidirectional serialize/deserialize any integral type to/from a bitstream.

Undefine __BITSTREAM_NATIVE_END if you need endian swapping. If you are not using __BITSTREAM_NATIVE_END the opposite is true for types larger than 1 byte For floating point, this is lossy, using 2 bytes for a float and 4 for a double. The range must be between -1 and +1. For non-floating point, this is lossless, but only has benefit if you use less than half the bits of the type

Parameters:
[in] writeToBitstream true to write from your data to this bitstream. False to read from this bitstream and write to your data
[in] inOutTemplateVar The value to write
Returns:
true if writeToBitstream is true. true if writeToBitstream is false and the read was successful. false if writeToBitstream is false and the read was not successful.

template<class templateType >
bool RakNet::BitStream::SerializeCompressedDelta ( bool  writeToBitstream,
templateType &  inOutTemplateVar 
) [inline]

Save as SerializeCompressedDelta(templateType &currentValue, const templateType &lastValue) when we have an unknown second parameter.

Returns:
true on data read. False on insufficient data in bitstream

template<class templateType >
bool RakNet::BitStream::SerializeCompressedDelta ( bool  writeToBitstream,
templateType &  inOutCurrentValue,
const templateType &  lastValue 
) [inline]

Bidirectional serialize/deserialize any integral type to/from a bitstream.

If the current value is different from the last value the current value will be written. Otherwise, a single bit will be written For floating point, this is lossy, using 2 bytes for a float and 4 for a double. The range must be between -1 and +1. For non-floating point, this is lossless, but only has benefit if you use less than half the bits of the type If you are not using __BITSTREAM_NATIVE_END the opposite is true for types larger than 1 byte

Parameters:
[in] writeToBitstream true to write from your data to this bitstream. False to read from this bitstream and write to your data
[in] inOutCurrentValue The current value to write
[in] lastValue The last value to compare against. Only used if writeToBitstream is true.
Returns:
true if writeToBitstream is true. true if writeToBitstream is false and the read was successful. false if writeToBitstream is false and the read was not successful.

template<class templateType >
bool RakNet::BitStream::SerializeDelta ( bool  writeToBitstream,
templateType &  inOutCurrentValue 
) [inline]

Bidirectional version of SerializeDelta when you don't know what the last value is, or there is no last value.

Parameters:
[in] writeToBitstream true to write from your data to this bitstream. False to read from this bitstream and write to your data
[in] inOutCurrentValue The current value to write
Returns:
true if writeToBitstream is true. true if writeToBitstream is false and the read was successful. false if writeToBitstream is false and the read was not successful.

template<class templateType >
bool RakNet::BitStream::SerializeDelta ( bool  writeToBitstream,
templateType &  inOutCurrentValue,
const templateType &  lastValue 
) [inline]

Bidirectional serialize/deserialize any integral type to/from a bitstream.

If the current value is different from the last value the current value will be written. Otherwise, a single bit will be written

Parameters:
[in] writeToBitstream true to write from your data to this bitstream. False to read from this bitstream and write to your data
[in] inOutCurrentValue The current value to write
[in] lastValue The last value to compare against. Only used if writeToBitstream is true.
Returns:
true if writeToBitstream is true. true if writeToBitstream is false and the read was successful. false if writeToBitstream is false and the read was not successful.

bool RakNet::BitStream::SerializeFloat16 ( bool  writeToBitstream,
float &  inOutFloat,
float  floatMin,
float  floatMax 
)

Serialize a float into 2 bytes, spanning the range between floatMin and floatMax.

Parameters:
[in] writeToBitstream true to write from your data to this bitstream. False to read from this bitstream and write to your data
[in] inOutFloat The float to write
[in] floatMin Predetermined minimum value of f
[in] floatMax Predetermined maximum value of f

template<class templateType >
bool RakNet::BitStream::SerializeNormQuat ( bool  writeToBitstream,
templateType &  w,
templateType &  x,
templateType &  y,
templateType &  z 
) [inline]

Bidirectional serialize/deserialize a normalized quaternion in 6 bytes + 4 bits instead of 16 bytes. Slightly lossy.

Parameters:
[in] writeToBitstream true to write from your data to this bitstream. False to read from this bitstream and write to your data
[in] w w
[in] x x
[in] y y
[in] z z
Returns:
true if writeToBitstream is true. true if writeToBitstream is false and the read was successful. false if writeToBitstream is false and the read was not successful.

template<class templateType >
bool RakNet::BitStream::SerializeNormVector ( bool  writeToBitstream,
templateType &  x,
templateType &  y,
templateType &  z 
) [inline]

Bidirectional serialize/deserialize a normalized 3D vector, using (at most) 4 bytes + 3 bits instead of 12-24 bytes.

Will further compress y or z axis aligned vectors. Accurate to 1/32767.5.

Parameters:
[in] writeToBitstream true to write from your data to this bitstream. False to read from this bitstream and write to your data
[in] x x
[in] y y
[in] z z
Returns:
true if writeToBitstream is true. true if writeToBitstream is false and the read was successful. false if writeToBitstream is false and the read was not successful.

template<class templateType >
bool RakNet::BitStream::SerializeOrthMatrix ( bool  writeToBitstream,
templateType &  m00,
templateType &  m01,
templateType &  m02,
templateType &  m10,
templateType &  m11,
templateType &  m12,
templateType &  m20,
templateType &  m21,
templateType &  m22 
) [inline]

Bidirectional serialize/deserialize an orthogonal matrix by creating a quaternion, and writing 3 components of the quaternion in 2 bytes each.

Use 6 bytes instead of 36 Lossy, although the result is renormalized

Returns:
true on success, false on failure.

template<class templateType >
bool RakNet::BitStream::SerializeVector ( bool  writeToBitstream,
templateType &  x,
templateType &  y,
templateType &  z 
) [inline]

Bidirectional serialize/deserialize a vector, using 10 bytes instead of 12.

Loses accuracy to about 3/10ths and only saves 2 bytes, so only use if accuracy is not important.

Parameters:
[in] writeToBitstream true to write from your data to this bitstream. False to read from this bitstream and write to your data
[in] x x
[in] y y
[in] z z
Returns:
true if writeToBitstream is true. true if writeToBitstream is false and the read was successful. false if writeToBitstream is false and the read was not successful.

void RakNet::BitStream::SetWriteOffset ( const BitSize_t  offset  ) 

Move the write pointer to a position on the array.

Parameters:
[in] offset the offset from the start of the array.
Attention:
Dangerous if you don't know what you are doing! For efficiency reasons you can only write mid-stream if your data is byte aligned.

template<>
void RakNet::BitStream::Write ( const RakString inTemplateVar  )  [inline]

Write a string to a bitstream.

Parameters:
[in] var The value to write

template<>
void RakNet::BitStream::Write ( const SystemAddress inTemplateVar  )  [inline]

Write a systemAddress to a bitstream.

Parameters:
[in] inTemplateVar The value to write

template<>
void RakNet::BitStream::Write ( const bool &  inTemplateVar  )  [inline]

Write a bool to a bitstream.

Parameters:
[in] inTemplateVar The value to write

void RakNet::BitStream::Write ( BitStream bitStream,
BitSize_t  numberOfBits 
)

Write one bitstream to another.

Parameters:
[in] numberOfBits bits to write
bitStream the bitstream to copy from

void RakNet::BitStream::Write ( const char *  inputByteArray,
const unsigned int  numberOfBytes 
)

Write an array or casted stream or raw data. This does NOT do endian swapping.

Parameters:
[in] inputByteArray a byte buffer
[in] numberOfBytes the size of input in bytes

template<class templateType >
void RakNet::BitStream::Write ( const templateType &  inTemplateVar  )  [inline]

Write any integral type to a bitstream.

Undefine __BITSTREAM_NATIVE_END if you need endian swapping.

Parameters:
[in] inTemplateVar The value to write

void RakNet::BitStream::WriteAlignedBytes ( const unsigned char *  inByteArray,
const unsigned int  numberOfBytesToWrite 
)

Align the bitstream to the byte boundary and then write the specified number of bits.

This is faster than WriteBits but wastes the bits to do the alignment and requires you to call ReadAlignedBits at the corresponding read position.

Parameters:
[in] inByteArray The data
[in] numberOfBytesToWrite The size of input.

void RakNet::BitStream::WriteAlignedBytesSafe ( const char *  inByteArray,
const unsigned int  inputLength,
const unsigned int  maxBytesToWrite 
)

Aligns the bitstream, writes inputLength, and writes input. Won't write beyond maxBytesToWrite.

Parameters:
[in] inByteArray The data
[in] inputLength The size of input.
[in] maxBytesToWrite Max bytes to write

void RakNet::BitStream::WriteBits ( const unsigned char *  inByteArray,
BitSize_t  numberOfBitsToWrite,
const bool  rightAlignedBits = true 
)

Write numberToWrite bits from the input source.

Right aligned data means in the case of a partial byte, the bits are aligned from the right (bit 0) rather than the left (as in the normal internal representation) You would set this to true when writing user data, and false when copying bitstream data, such as writing one bitstream to another.

Parameters:
[in] inByteArray The data
[in] numberOfBitsToWrite The number of bits to write
[in] rightAlignedBits if true data will be right aligned

template<class templateType >
void RakNet::BitStream::WriteBitsFromIntegerRange ( const templateType  value,
const templateType  minimum,
const templateType  maximum,
const int  requiredBits,
bool  allowOutsideRange = false 
) [inline]

Parameters:
[in] requiredBits Primarily for internal use, called from above function() after calculating number of bits needed to represent maximum-minimum

template<class templateType >
void RakNet::BitStream::WriteBitsFromIntegerRange ( const templateType  value,
const templateType  minimum,
const templateType  maximum,
bool  allowOutsideRange = false 
) [inline]

Given the minimum and maximum values for an integer type, figure out the minimum number of bits to represent the range Then write only those bits

Note:
A static is used so that the required number of bits for (maximum-minimum) is only calculated once. This does require that minimum and are fixed values for a given line of code for the life of the program
Parameters:
[in] value Integer value to write, which should be between minimum and maximum
[in] minimum Minimum value of value
[in] maximum Maximum value of value
[in] allowOutsideRange If true, all sends will take an extra bit, however value can deviate from outside minimum and maximum. If false, will assert if the value deviates. This should match the corresponding value passed to Read().

template<class destinationType , class sourceType >
void RakNet::BitStream::WriteCasted ( const sourceType &  value  )  [inline]

Write one type serialized as another (smaller) type, to save bandwidth serializationType should be uint8_t, uint16_t, uint24_t, or uint32_t Example: int num=53; WriteCasted<uint8_t>(num); would use 1 byte to write what would otherwise be an integer (4 or 8 bytes)

Parameters:
[in] value The value to write

template<class templateType >
void RakNet::BitStream::WriteCompressed ( const templateType &  inTemplateVar  )  [inline]

Write any integral type to a bitstream.

Undefine __BITSTREAM_NATIVE_END if you need endian swapping. If you are not using __BITSTREAM_NATIVE_END the opposite is true for types larger than 1 byte For floating point, this is lossy, using 2 bytes for a float and 4 for a double. The range must be between -1 and +1. For non-floating point, this is lossless, but only has benefit if you use less than half the bits of the type

Parameters:
[in] inTemplateVar The value to write
Undefine __BITSTREAM_NATIVE_END if you need endian swapping. For floating point, this is lossy, using 2 bytes for a float and 4 for a double. The range must be between -1 and +1. For non-floating point, this is lossless, but only has benefit if you use less than half the bits of the type If you are not using __BITSTREAM_NATIVE_END the opposite is true for types larger than 1 byte
Parameters:
[in] inTemplateVar The value to write

template<>
void RakNet::BitStream::WriteCompressedDelta ( const bool &  currentValue,
const bool &  lastValue 
) [inline]

Write a bool delta. Same thing as just calling Write.

Parameters:
[in] currentValue The current value to write
[in] lastValue The last value to compare against

template<class templateType >
void RakNet::BitStream::WriteCompressedDelta ( const templateType &  currentValue  )  [inline]

template<class templateType >
void RakNet::BitStream::WriteCompressedDelta ( const templateType &  currentValue,
const templateType &  lastValue 
) [inline]

Write any integral type to a bitstream.

If the current value is different from the last value the current value will be written. Otherwise, a single bit will be written For floating point, this is lossy, using 2 bytes for a float and 4 for a double. The range must be between -1 and +1. For non-floating point, this is lossless, but only has benefit if you use less than half the bits of the type If you are not using __BITSTREAM_NATIVE_END the opposite is true for types larger than 1 byte

Parameters:
[in] currentValue The current value to write
[in] lastValue The last value to compare against
If the current value is different from the last value the current value will be written. Otherwise, a single bit will be written For floating point, this is lossy, using 2 bytes for a float and 4 for a double. The range must be between -1 and +1. For non-floating point, this is lossless, but only has benefit if you use less than half the bits of the type If you are not using __BITSTREAM_NATIVE_END the opposite is true for types larger than 1 byte
Parameters:
[in] currentValue The current value to write
[in] lastValue The last value to compare against

template<>
void RakNet::BitStream::WriteDelta ( const bool &  currentValue,
const bool &  lastValue 
) [inline]

Write a bool delta. Same thing as just calling Write.

Parameters:
[in] currentValue The current value to write
[in] lastValue The last value to compare against

template<class templateType >
void RakNet::BitStream::WriteDelta ( const templateType &  currentValue  )  [inline]

WriteDelta when you don't know what the last value is, or there is no last value.

Parameters:
[in] currentValue The current value to write

template<class templateType >
void RakNet::BitStream::WriteDelta ( const templateType &  currentValue,
const templateType &  lastValue 
) [inline]

Write any integral type to a bitstream.

If the current value is different from the last value the current value will be written. Otherwise, a single bit will be written

Parameters:
[in] currentValue The current value to write
[in] lastValue The last value to compare against
If the current value is different from the last value the current value will be written. Otherwise, a single bit will be written
Parameters:
[in] currentValue The current value to write
[in] lastValue The last value to compare against

void RakNet::BitStream::WriteFloat16 ( float  x,
float  floatMin,
float  floatMax 
)

Write a float into 2 bytes, spanning the range between floatMin and floatMax.

Parameters:
[in] x The float to write
[in] floatMin Predetermined minimum value of f
[in] floatMax Predetermined maximum value of f

template<class templateType >
void RakNet::BitStream::WriteNormQuat ( templateType  w,
templateType  x,
templateType  y,
templateType  z 
) [inline]

Write a normalized quaternion in 6 bytes + 4 bits instead of 16 bytes. Slightly lossy.

Parameters:
[in] w w
[in] x x
[in] y y
[in] z z

template<class templateType >
void RakNet::BitStream::WriteNormVector ( templateType  x,
templateType  y,
templateType  z 
) [inline]

Write a normalized 3D vector, using (at most) 4 bytes + 3 bits instead of 12-24 bytes.

Will further compress y or z axis aligned vectors. Accurate to 1/32767.5.

Parameters:
[in] x x
[in] y y
[in] z z

template<class templateType >
void RakNet::BitStream::WriteOrthMatrix ( templateType  m00,
templateType  m01,
templateType  m02,
templateType  m10,
templateType  m11,
templateType  m12,
templateType  m20,
templateType  m21,
templateType  m22 
) [inline]

Write an orthogonal matrix by creating a quaternion, and writing 3 components of the quaternion in 2 bytes each.

Use 6 bytes instead of 36 Lossy, although the result is renormalized

template<class templateType >
void RakNet::BitStream::WritePtr ( templateType *  inTemplateVar  )  [inline]

Write the dereferenced pointer to any integral type to a bitstream.

Undefine __BITSTREAM_NATIVE_END if you need endian swapping.

Parameters:
[in] inTemplateVar The value to write

template<class templateType >
void RakNet::BitStream::WriteVector ( templateType  x,
templateType  y,
templateType  z 
) [inline]

Write a vector, using 10 bytes instead of 12.

Loses accuracy to about 3/10ths and only saves 2 bytes, so only use if accuracy is not important.

Parameters:
[in] x x
[in] y y
[in] z z


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

Generated on Wed Feb 1 13:33:47 2012 for RakNet by  doxygen 1.5.7.1