namespace for dmlc More...

Namespaces
	io

	lua_stack

	parameter

	serializer
	internal namespace for serializers

Classes
class	array_view
	Read only data structure to reference continuous memory region of array. Provide unified view for vector, array and C style array. This data structure do not guarantee aliveness of referenced array. More...

class	BlockingQueueThread
	Blocking queue thread class. More...

class	ConcurrentBlockingQueue
	Cocurrent blocking queue. More...

class	Config
	class for config parser More...

class	DataIter
	data iterator interface this is not a C++ style iterator, but nice for data pulling:) This interface is used to pull in the data The system can do some useful tricks for you like pre-fetching from disk and pre-computation. More...

class	FunctionRegEntryBase
	Common base class for function registry. More...

struct	has_saveload
	whether a type have save/load function More...

struct	IfThenElseType
	template to select type based on condition For example, IfThenElseType<true, int, float>::Type will give int More...

class	InputSplit
	input split creates that allows reading of records from split of data, independent part that covers all the dataset More...

class	InputSplitShuffle
	class to construct input split with global shuffling More...

struct	is_arithmetic
	whether a type is arithemetic type More...

struct	is_floating_point
	whether a type is floating point type More...

struct	is_integral
	whether a type is integer type More...

struct	is_pod
	whether a type is pod type More...

class	istream
	a std::istream class that can can wrap Stream objects, can use istream with that output to underlying Stream More...

class	JSONObjectReadHelper
	Helper class to read JSON into a class or struct object. More...

class	JSONReader
	Lightweight JSON Reader to read any STL compositions and structs. The user need to know the schema of the. More...

class	JSONWriter
	Lightweight json to write any STL compositions. More...

class	LuaRef
	an reference to lua object More...

class	LuaState
	A Lua state. More...

class	ManualEvent
	Simple manual-reset event gate which remains open after signalled. More...

struct	MemoryFixedSizeStream
	A Stream that operates on fixed region of memory This class allows us to read/write from/to a fixed memory region. More...

class	MemoryPool
	A memory pool that allocate memory of fixed size and alignment. More...

struct	MemoryStringStream
	A in memory stream that is backed by std::string. This class allows us to read/write from/to a std::string. More...

struct	nullopt_t
	dummy type for assign null to optional More...

class	OMPException
	OMP Exception class catches, saves and rethrows exception from OMP blocks. More...

class	optional
	c++17 compatible optional class. More...

class	ostream
	a std::ostream class that can can wrap Stream objects, can use ostream with that output to underlying Stream More...

class	Parser
	parser interface that parses input data used to load dmlc data format into your own data format Difference between RowBlockIter and Parser: RowBlockIter caches the data internally that can be used to iterate the dataset multiple times, Parser holds very limited internal state and was usually used to read data only once More...

struct	ParserFactoryReg
	registry entry of parser factory More...

class	RecordIOChunkReader
	reader of binary recordio from Blob returned by InputSplit This class divides the blob into several independent parts specified by caller, and read from one segment. The part reading can be used together with InputSplit::NextChunk for multi-threaded parsing(each thread take a RecordIOChunkReader) More...

class	RecordIOReader
	reader of binary recordio to reads in record from stream More...

class	RecordIOWriter
	writer of binary recordio binary format for recordio recordio format: magic lrecord data pad More...

class	Registry
	Registry class. Registry can be used to register global singletons. The most commonly use case are factory functions. More...

class	Row
	one row of training instance More...

struct	RowBlock
	a block of data, containing several rows in sparse matrix This is useful for (streaming-sxtyle) algorithms that scans through rows of data examples include: SGD, GD, L-BFGS, kmeans More...

class	RowBlockIter
	Data structure that holds the data Row block iterator interface that gets RowBlocks Difference between RowBlockIter and Parser: RowBlockIter caches the data internally that can be used to iterate the dataset multiple times, Parser holds very limited internal state and was usually used to read data only once. More...

class	ScopedThread
	Wrapper class to manage std::thread; uses RAII pattern to automatically join std::thread upon destruction. More...

class	SeekStream
	interface of i/o stream that support seek More...

class	Serializable
	interface for serializable objects More...

class	Spinlock
	Simple userspace spinlock implementation. More...

class	Str2T
	Interface class that defines a single method get() to convert a string into type T. Define template specialization of this class to define the conversion method for a particular type. More...

class	Str2T< double >
	Template specialization of Str2T<> interface for double type. More...

class	Str2T< float >
	Template specialization of Str2T<> interface for float type. More...

class	Str2T< int32_t >
	Template specialization of Str2T<> interface for signed 32-bit integer. More...

class	Str2T< int64_t >
	Template specialization of Str2T<> interface for signed 64-bit integer. More...

class	Str2T< uint32_t >
	Template specialization of Str2T<> interface for unsigned 32-bit integer. More...

class	Str2T< uint64_t >
	Template specialization of Str2T<> interface for unsigned 64-bit integer. More...

class	Stream
	interface of stream I/O for serialization More...

class	TemporaryDirectory
	Manager class for temporary directories. Whenever a new TemporaryDirectory object is constructed, a temporary directory is created. The directory is deleted when the object is deleted or goes out of scope. Note: no symbolic links are allowed inside the temporary directory. More...

class	ThreadedIter
	a iterator that was backed by a thread to pull data eagerly from a single producer into a bounded buffer the consumer can pull the data at its own rate More...

class	ThreadGroup
	Thread lifecycle management group. More...

class	ThreadlocalAllocator
	A thread local allocator that get memory from a threadlocal memory pool. This is suitable to allocate objects that do not cross thread. More...

struct	ThreadlocalSharedPtr
	a shared pointer like type that allocate object from a threadlocal object pool. This object is not thread-safe but can be faster than shared_ptr in certain usecases. More...

class	ThreadLocalStore
	A threadlocal store to store threadlocal variables. Will return a thread local singleton of type T. More...

class	TimerThread
	Managed timer thread. More...

struct	type_name_helper
	helper class to construct a string that represents type name More...

struct	type_name_helper< mxnet::Tuple< T > >

struct	type_name_helper< nnvm::Tuple< T > >

Typedefs
typedef float	real_t
	this defines the float point that will be used to store feature values More...

typedef unsigned	index_t
	this defines the unsigned integer type that can normally be used to store feature index More...

using	SharedMutex = std::recursive_mutex
	Standard mutex for C++ < 14. More...

using	WriteLock = std::unique_lock< SharedMutex >
	Standard unique lock for C++ < 14. More...

using	ReadLock = std::unique_lock< SharedMutex >
	Standard unique lock for C++ < 14. More...

Enumerations
enum	ConcurrentQueueType { ConcurrentQueueType::kFIFO, ConcurrentQueueType::kPriority }
	type of concurrent queue More...

Functions
template<typename T >
T *	BeginPtr (std::vector< T > &vec)
	safely get the beginning address of a vector More...

template<typename T >
const T *	BeginPtr (const std::vector< T > &vec)
	get the beginning address of a const vector More...

char *	BeginPtr (std::string &str)
	get the beginning address of a string More...

const char *	BeginPtr (const std::string &str)
	get the beginning address of a const string More...

std::vector< std::string >	Split (const std::string &s, char delim)
	Split a string by delimiter. More...

template<typename T >
size_t	HashCombine (size_t key, const T &value)
	hash an object and combines the key with previous keys More...

template<>
size_t	HashCombine< size_t > (size_t key, const size_t &value)
	specialize for size_t More...

void	ByteSwap (void *data, size_t elem_bytes, size_t num_elems)
	A generic inplace byte swapping function. More...

template<typename T >
std::ostream &	operator<< (std::ostream &os, const optional< T > &t)
	serialize an optional object to string. More...

template<typename T >
std::istream &	operator>> (std::istream &is, optional< T > &t)
	parse a string object into optional<T> More...

std::istream &	operator>> (std::istream &is, optional< bool > &t)
	specialization of '>>' istream parsing for optional<bool> More...

	DMLC_DECLARE_TYPE_NAME (optional< int >,"int or None")
	description for optional int More...

	DMLC_DECLARE_TYPE_NAME (optional< bool >,"boolean or None")
	description for optional bool More...

	DMLC_DECLARE_TYPE_NAME (optional< float >,"float or None")
	description for optional float More...

	DMLC_DECLARE_TYPE_NAME (optional< double >,"double or None")
	description for optional double More...

bool	isspace (char c)
	Inline implementation of isspace(). Tests whether the given character is a whitespace letter. More...

bool	isblank (char c)
	Inline implementation of isblank(). Tests whether the given character is a space or tab character. More...

bool	isdigit (char c)
	Inline implementation of isdigit(). Tests whether the given character is a decimal digit. More...

bool	isalpha (char c)
	Inline implementation of isalpha(). Tests whether the given character is an alphabet letter. More...

bool	isdigitchars (char c)
	Tests whether the given character is a valid letter in the string representation of a floating-point value, i.e. decimal digits, signs (+/-), decimal point (.), or exponent marker (e/E). More...

template<typename FloatType , bool CheckRange = false>
FloatType	ParseFloat (const char nptr, char *endptr)
	Common implementation for dmlc::strtof() and dmlc::strtod() TODO: the current version does not support hex number. More...

float	strtof (const char nptr, char *endptr)
	A faster implementation of strtof(). See documentation of std::strtof() for more information. Note that this function does not check for overflow. Use strtof_check_range() to check for overflow. TODO: the current version does not support hex number TODO: the current version does not handle long decimals: you may only have up to 19 digits after the decimal point, and you cannot have too many digits before the decimal point either. More...

float	strtof_check_range (const char nptr, char *endptr)
	A faster implementation of strtof(). See documentation of std::strtof() for more information. This function will check for overflow. If the converted value is outside the range for the float type, errno is set to ERANGE and HUGE_VALF is returned. TODO: the current version does not support hex number TODO: the current version does not handle long decimals: you may only have up to 19 digits after the decimal point, and you cannot have too many digits before the decimal point either. More...

double	strtod (const char nptr, char *endptr)
	A faster implementation of strtod(). See documentation of std::strtof() for more information. Note that this function does not check for overflow. Use strtod_check_range() to check for overflow. TODO: the current version does not support hex number TODO: the current version does not handle long decimals: you may only have up to 19 digits after the decimal point, and you cannot have too many digits before the decimal point either. More...

double	strtod_check_range (const char nptr, char *endptr)
	A faster implementation of strtod(). See documentation of std::strtod() for more information. This function will check for overflow. If the converted value is outside the range for the double type, errno is set to ERANGE and HUGE_VAL is returned. TODO: the current version does not support hex number TODO: the current version does not handle long decimals: you may only have up to 19 digits after the decimal point, and you cannot have too many digits before the decimal point either. More...

template<typename SignedIntType >
SignedIntType	ParseSignedInt (const char nptr, char *endptr, int base)
	A fast string-to-integer convertor, for signed integers TODO: the current version supports only base <= 10. More...

template<typename UnsignedIntType >
UnsignedIntType	ParseUnsignedInt (const char nptr, char *endptr, int base)
	A fast string-to-integer convertor, for unsigned integers TODO: the current version supports only base <= 10. More...

uint64_t	strtoull (const char nptr, char *endptr, int base)
	A faster implementation of strtoull(). See documentation of std::strtoull() for more information. Note that this function does not check for overflow. TODO: the current version supports only base <= 10. More...

long	atol (const char *p)
	A faster implementation of atol(). See documentation of std::atol() for more information. This function will use base 10. Note that this function does not check for overflow. More...

float	atof (const char *nptr)
	A faster implementation of atof(). Unlike std::atof(), this function returns float type. Note that this function does not check for overflow. TODO: the current version does not support hex number TODO: the current version does not handle long decimals: you may only have up to 19 digits after the decimal point, and you cannot have too many digits before the decimal point either. More...

float	stof (const std::string &value, size_t *pos=nullptr)
	A faster implementation of stof(). See documentation of std::stof() for more information. This function will test for overflow and invalid arguments. TODO: the current version does not support hex number TODO: the current version does not handle long decimals: you may only have up to 19 digits after the decimal point, and you cannot have too many digits before the decimal point either. More...

double	stod (const std::string &value, size_t *pos=nullptr)
	A faster implementation of stod(). See documentation of std::stod() for more information. This function will test for overflow and invalid arguments. TODO: the current version does not support hex number TODO: the current version does not handle long decimals: you may only have up to 19 digits after the decimal point, and you cannot have too many digits before the decimal point either. More...

template<typename T >
T	Str2Type (const char begin, const char end)
	Convenience function for converting string into type T. More...

template<typename T1 , typename T2 >
int	ParsePair (const char begin, const char end, const char **endptr, T1 &v1, T2 &v2)
	Parse colon seperated pair v1[:v2]. More...

template<typename T1 , typename T2 , typename T3 >
int	ParseTriple (const char begin, const char end, const char **endptr, T1 &v1, T2 &v2, T3 &v3)
	Parse colon seperated triple v1:v2[:v3]. More...

template<typename Duration , typename TimerFunction >
bool	CreateTimer (const std::string &timer_name, const Duration &duration, ThreadGroup *owner, TimerFunction timer_function)
	Utility function to easily create a timer. More...

double	GetTime (void)
	return time in seconds More...

template<typename T >
std::string	type_name ()
	the string representation of type name More...

	DMLC_DECLARE_TRAITS (has_saveload, mxnet::NDArray, true)
	traits More...

	DMLC_DECLARE_TYPE_NAME (mxnet::TShape,"Shape(tuple)")

	DMLC_DECLARE_TYPE_NAME (mxnet::Tuple< int >,"Shape(tuple)")

	DMLC_DECLARE_TYPE_NAME (mxnet::Tuple< dmlc::optional< int >>,"Shape(tuple)")

	DMLC_DECLARE_TYPE_NAME (nnvm::Tuple< int >,"Shape(tuple)")

	DMLC_DECLARE_TYPE_NAME (nnvm::Tuple< dmlc::optional< int >>,"Shape(tuple)")

	DMLC_DECLARE_TYPE_NAME (optional< mxnet::TShape >,"Shape or None")
	description for optional TShape More...

	DMLC_DECLARE_TYPE_NAME (optional< mxnet::Tuple< int >>,"Shape or None")

	DMLC_DECLARE_TYPE_NAME (optional< nnvm::TShape >,"Shape or None")
	description for optional TShape More...

Variables
constexpr const nullopt_t	nullopt = nullopt_t(0)

const int	kStrtofMaxDigits = 19
	Maximum number of decimal digits dmlc::strtof() / dmlc::strtod() will process. Trailing digits will be ignored. More...

Detailed Description

namespace for dmlc

Typedef Documentation

typedef unsigned dmlc::index_t

this defines the unsigned integer type that can normally be used to store feature index

using dmlc::ReadLock = typedef std::unique_lock<SharedMutex>

Standard unique lock for C++ < 14.

typedef float dmlc::real_t

this defines the float point that will be used to store feature values

using dmlc::SharedMutex = typedef std::recursive_mutex

Standard mutex for C++ < 14.

using dmlc::WriteLock = typedef std::unique_lock<SharedMutex>

Standard unique lock for C++ < 14.

Enumeration Type Documentation

enum dmlc::ConcurrentQueueType

strong

type of concurrent queue

Enumerator
kFIFO	FIFO queue.
kPriority	queue with priority

Function Documentation

float dmlc::atof ( const char * nptr )

inline

A faster implementation of atof(). Unlike std::atof(), this function returns float type. Note that this function does not check for overflow. TODO: the current version does not support hex number TODO: the current version does not handle long decimals: you may only have up to 19 digits after the decimal point, and you cannot have too many digits before the decimal point either.

Parameters

nptr	Beginning of the string that's to be converted into float

Returns: Converted value, in float type

long dmlc::atol ( const char * p )

inline

A faster implementation of atol(). See documentation of std::atol() for more information. This function will use base 10. Note that this function does not check for overflow.

Parameters

p	Beginning of the string that's to be converted into integer of type long

Returns: Converted value, as long integer (width is system-dependent)

template<typename T >

T* dmlc::BeginPtr ( std::vector< T > & vec )

inline

safely get the beginning address of a vector

Parameters

vec	input vector

Returns: beginning address of a vector

template<typename T >

const T* dmlc::BeginPtr ( const std::vector< T > & vec )

inline

get the beginning address of a const vector

Parameters

vec	input vector

Returns: beginning address of a vector

char* dmlc::BeginPtr ( std::string & str )

inline

get the beginning address of a string

Parameters

str	input string

Returns: beginning address of a string

const char* dmlc::BeginPtr ( const std::string & str )

inline

get the beginning address of a const string

Parameters

str	input string

Returns: beginning address of a string

void dmlc::ByteSwap	(	void *	data,
		size_t	elem_bytes,
		size_t	num_elems
	)

inline

A generic inplace byte swapping function.

Parameters

data	The data pointer.
elem_bytes	The number of bytes of the data elements
num_elems	Number of elements in the data.

Note: Always try pass in constant elem_bytes to enable compiler optimization

template<typename Duration , typename TimerFunction >

bool dmlc::CreateTimer	(	const std::string &	timer_name,
		const Duration &	duration,
		ThreadGroup *	owner,
		TimerFunction	timer_function
	)

inline

Utility function to easily create a timer.

Template Parameters

Duration	Duration type (i.e. std::chrono::milliseconds)
TimerFunction	Function to call each time the timer expires

Parameters

timer_name	Name of the timer. Must be unique per ThreadGroup object
duration	Duration of the timer between calls to timer_function
owner	ThreadGroup owner of the timer
timer_function	Function to call each time the timer expires

Returns: true if the timer was successfully created

dmlc::DMLC_DECLARE_TRAITS	(	has_saveload	,
		mxnet::NDArray	,
		true
	)

traits

dmlc::DMLC_DECLARE_TYPE_NAME	(	optional< int >	,
		"int or None"
	)

description for optional int

dmlc::DMLC_DECLARE_TYPE_NAME	(	optional< bool >	,
		"boolean or None"
	)

description for optional bool

dmlc::DMLC_DECLARE_TYPE_NAME	(	optional< float >	,
		"float or None"
	)

description for optional float

dmlc::DMLC_DECLARE_TYPE_NAME	(	optional< double >	,
		"double or None"
	)

description for optional double

dmlc::DMLC_DECLARE_TYPE_NAME	(	mxnet::TShape	,
		"Shape(tuple)"
	)

dmlc::DMLC_DECLARE_TYPE_NAME	(	mxnet::Tuple< int >	,
		"Shape(tuple)"
	)

dmlc::DMLC_DECLARE_TYPE_NAME	(	mxnet::Tuple< dmlc::optional< int >>	,
		"Shape(tuple)"
	)

dmlc::DMLC_DECLARE_TYPE_NAME	(	nnvm::Tuple< int >	,
		"Shape(tuple)"
	)

dmlc::DMLC_DECLARE_TYPE_NAME	(	nnvm::Tuple< dmlc::optional< int >>	,
		"Shape(tuple)"
	)

dmlc::DMLC_DECLARE_TYPE_NAME	(	optional< nnvm::TShape >	,
		"Shape or None"
	)

description for optional TShape

dmlc::DMLC_DECLARE_TYPE_NAME	(	optional< mxnet::TShape >	,
		"Shape or None"
	)

description for optional TShape

dmlc::DMLC_DECLARE_TYPE_NAME	(	optional< mxnet::Tuple< int >>	,
		"Shape or None"
	)

double dmlc::GetTime ( void )

inline

return time in seconds

template<typename T >

size_t dmlc::HashCombine	(	size_t	key,
		const T &	value
	)

inline

hash an object and combines the key with previous keys

template<>

size_t dmlc::HashCombine< size_t >	(	size_t	key,
		const size_t &	value
	)

inline

specialize for size_t

bool dmlc::isalpha ( char c )

inline

Inline implementation of isalpha(). Tests whether the given character is an alphabet letter.

Parameters

c	Character to test

Returns: Result of the test

bool dmlc::isblank ( char c )

inline

Inline implementation of isblank(). Tests whether the given character is a space or tab character.

Parameters

c	Character to test

Returns: Result of the test

bool dmlc::isdigit ( char c )

inline

Inline implementation of isdigit(). Tests whether the given character is a decimal digit.

Parameters

c	Character to test

Returns: Result of the test

bool dmlc::isdigitchars ( char c )

inline

Tests whether the given character is a valid letter in the string representation of a floating-point value, i.e. decimal digits, signs (+/-), decimal point (.), or exponent marker (e/E).

Parameters

c	Character to test

Returns: Result of the test

bool dmlc::isspace ( char c )

inline

Inline implementation of isspace(). Tests whether the given character is a whitespace letter.

Parameters

c	Character to test

Returns: Result of the test

template<typename T >

std::ostream& dmlc::operator<<	(	std::ostream &	os,
		const optional< T > &	t
	)

serialize an optional object to string.

dmlc::optional<int> x;
std::cout << x;  // None
x = 0;
std::cout << x;  // 0

Parameters

os	output stream
t	source optional<T> object

Returns: output stream

template<typename T >

std::istream& dmlc::operator>>	(	std::istream &	is,
		optional< T > &	t
	)

parse a string object into optional<T>

dmlc::optional<int> x;
std::string s1 = "1";
std::istringstream is1(s1);
s1 >> x;  // x == optional<int>(1)
std::string s2 = "None";
std::istringstream is2(s2);
s2 >> x;  // x == optional<int>()

Parameters

is	input stream
t	target optional<T> object

Returns: input stream

std::istream& dmlc::operator>>	(	std::istream &	is,
		optional< bool > &	t
	)

inline

specialization of '>>' istream parsing for optional<bool>

Permits use of generic parameter FieldEntry<DType> class to create FieldEntry<optional<bool>> without explicit specialization.

dmlc::optional<bool> x;
std::string s1 = "true";
std::istringstream is1(s1);
s1 >> x;  // x == optional<bool>(true)
std::string s2 = "None";
std::istringstream is2(s2);
s2 >> x;  // x == optional<bool>()

Parameters

is	input stream
t	target optional<bool> object

Returns: input stream

template<typename FloatType , bool CheckRange = false>

FloatType dmlc::ParseFloat	(	const char *	nptr,
		char **	endptr
	)

inline

Common implementation for dmlc::strtof() and dmlc::strtod() TODO: the current version does not support hex number.

Parameters

nptr	Beginning of the string that's to be converted into a floating-point number
endptr	After the conversion, this pointer will be set to point one past the last character used in the conversion.

Returns: Converted floating-point value, in FloatType

Template Parameters

FloatType	Type of floating-point number to be obtained. This must be either float or double.
CheckRange	Whether to check for overflow. If set to true, an out- of-range value will cause errno to be set to ERANGE and ParseFloat() to return HUGE_VAL / HUGE_VALF; otherwise, all out-of-range vlaues will be silently clipped.

template<typename T1 , typename T2 >

int dmlc::ParsePair	(	const char *	begin,
		const char *	end,
		const char **	endptr,
		T1 &	v1,
		T2 &	v2
	)

inline

Parse colon seperated pair v1[:v2].

Parameters

begin	pointer to string
end	one past end of string
endptr	After conversion, will be set to one past of parsed string
v1	first value in the pair
v2	second value in the pair

Returns: number of values parsed

Template Parameters

T1	type of v1
T2	type of v2

template<typename SignedIntType >

SignedIntType dmlc::ParseSignedInt	(	const char *	nptr,
		char **	endptr,
		int	base
	)

inline

A fast string-to-integer convertor, for signed integers TODO: the current version supports only base <= 10.

Parameters

nptr	Beginning of the string that's to be converted into a signed integer
endptr	After the conversion, this pointer will be set to point one past the last character used in the conversion.
base	Base to use for integer conversion

Returns: Converted value, in SignedIntType

Template Parameters

SignedIntType Type of signed integer to be obtained.

template<typename T1 , typename T2 , typename T3 >

int dmlc::ParseTriple	(	const char *	begin,
		const char *	end,
		const char **	endptr,
		T1 &	v1,
		T2 &	v2,
		T3 &	v3
	)

inline

Parse colon seperated triple v1:v2[:v3].

Parameters

begin	pointer to string
end	one past end of string
endptr	After conversion, will be set to one past of parsed string
v1	first value in the triple
v2	second value in the triple
v3	third value in the triple

Returns: number of values parsed

Template Parameters

T1	type of v1
T2	type of v2
T3	type of v3

template<typename UnsignedIntType >

UnsignedIntType dmlc::ParseUnsignedInt	(	const char *	nptr,
		char **	endptr,
		int	base
	)

inline

A fast string-to-integer convertor, for unsigned integers TODO: the current version supports only base <= 10.

Parameters

nptr	Beginning of the string that's to be converted into an unsigned integer
endptr	After the conversion, this pointer will be set to point one past the last character used in the conversion.
base	Base to use for integer conversion

Returns: Converted value, in UnsignedIntType

Template Parameters

UnsignedIntType Type of unsigned integer to be obtained.

std::vector<std::string> dmlc::Split	(	const std::string &	s,
		char	delim
	)

inline

Split a string by delimiter.

Parameters

s	String to be splitted.
delim	The delimiter.

Returns: a splitted vector of strings.

double dmlc::stod	(	const std::string &	value,
		size_t *	pos = `nullptr`
	)

inline

A faster implementation of stod(). See documentation of std::stod() for more information. This function will test for overflow and invalid arguments. TODO: the current version does not support hex number TODO: the current version does not handle long decimals: you may only have up to 19 digits after the decimal point, and you cannot have too many digits before the decimal point either.

Parameters

value	The string to convert into double
pos	If not null, it will store the number of characters processed

Returns: Converted value, in double type

Exceptions

std::out_of_range	If the converted value would fall out of the range of the double type
std::invalid_argument	If no conversion could be performed

float dmlc::stof	(	const std::string &	value,
		size_t *	pos = `nullptr`
	)

inline

A faster implementation of stof(). See documentation of std::stof() for more information. This function will test for overflow and invalid arguments. TODO: the current version does not support hex number TODO: the current version does not handle long decimals: you may only have up to 19 digits after the decimal point, and you cannot have too many digits before the decimal point either.

Parameters

value	The string to convert into float
pos	If not null, it will store the number of characters processed

Returns: Converted value, in float type

Exceptions

std::out_of_range	If the converted value would fall out of the range of the double type
std::invalid_argument	If no conversion could be performed

template<typename T >

T dmlc::Str2Type	(	const char *	begin,
		const char *	end
	)

inline

Convenience function for converting string into type T.

Parameters

begin	Beginning of the string to convert
end	End of the string to convert

Returns: Converted value, in type T

Template Parameters

Type	of converted value

double dmlc::strtod	(	const char *	nptr,
		char **	endptr
	)

inline

A faster implementation of strtod(). See documentation of std::strtof() for more information. Note that this function does not check for overflow. Use strtod_check_range() to check for overflow. TODO: the current version does not support hex number TODO: the current version does not handle long decimals: you may only have up to 19 digits after the decimal point, and you cannot have too many digits before the decimal point either.

Parameters

nptr	Beginning of the string that's to be converted into double
endptr	After the conversion, this pointer will be set to point one past the last character used in the conversion.

Returns: Converted floating-point value, in double type

double dmlc::strtod_check_range	(	const char *	nptr,
		char **	endptr
	)

inline

A faster implementation of strtod(). See documentation of std::strtod() for more information. This function will check for overflow. If the converted value is outside the range for the double type, errno is set to ERANGE and HUGE_VAL is returned. TODO: the current version does not support hex number TODO: the current version does not handle long decimals: you may only have up to 19 digits after the decimal point, and you cannot have too many digits before the decimal point either.

Parameters

nptr	Beginning of the string that's to be converted into double
endptr	After the conversion, this pointer will be set to point one past the last character used in the conversion.

Returns: Converted floating-point value, in float type

float dmlc::strtof	(	const char *	nptr,
		char **	endptr
	)

inline

A faster implementation of strtof(). See documentation of std::strtof() for more information. Note that this function does not check for overflow. Use strtof_check_range() to check for overflow. TODO: the current version does not support hex number TODO: the current version does not handle long decimals: you may only have up to 19 digits after the decimal point, and you cannot have too many digits before the decimal point either.

Parameters

nptr	Beginning of the string that's to be converted into float
endptr	After the conversion, this pointer will be set to point one past the last character used in the conversion.

Returns: Converted floating-point value, in float type

float dmlc::strtof_check_range	(	const char *	nptr,
		char **	endptr
	)

inline

A faster implementation of strtof(). See documentation of std::strtof() for more information. This function will check for overflow. If the converted value is outside the range for the float type, errno is set to ERANGE and HUGE_VALF is returned. TODO: the current version does not support hex number TODO: the current version does not handle long decimals: you may only have up to 19 digits after the decimal point, and you cannot have too many digits before the decimal point either.

Parameters

nptr	Beginning of the string that's to be converted into float
endptr	After the conversion, this pointer will be set to point one past the last character used in the conversion.

Returns: Converted floating-point value, in float type

uint64_t dmlc::strtoull	(	const char *	nptr,
		char **	endptr,
		int	base
	)

inline

A faster implementation of strtoull(). See documentation of std::strtoull() for more information. Note that this function does not check for overflow. TODO: the current version supports only base <= 10.

Parameters

nptr	Beginning of the string that's to be converted into integer of type unsigned long long
endptr	After the conversion, this pointer will be set to point one past the last character used in the conversion.
base	Base to use for integer conversion

Returns: Converted value, as unsigned 64-bit integer

template<typename T >

std::string dmlc::type_name ( )

inline

the string representation of type name

Template Parameters

T	the type to query

Returns: a const string of typename.

Variable Documentation

const int dmlc::kStrtofMaxDigits = 19

Maximum number of decimal digits dmlc::strtof() / dmlc::strtod() will process. Trailing digits will be ignored.

constexpr const nullopt_t dmlc::nullopt = nullopt_t(0)

Assign null to optional: optional<T> x = nullopt;

Namespaces

Classes

Typedefs

Enumerations

Functions

Variables

Detailed Description

Typedef Documentation

Enumeration Type Documentation

Function Documentation

Variable Documentation