A Wavelet Tree class for byte sequences. More...

#include <wt_huff.hpp>

Public Types
typedef int_vector::size_type	size_type
typedef unsigned char	value_type
typedef BitVector	bit_vector_type
typedef RankSupport	rank_1_type
typedef SelectSupport	select_1_type
typedef SelectSupportZero	select_0_type
Public Member Functions
template<typename RandomAccessContainer >
	wt_huff (const RandomAccessContainer &rac, size_type size)
	Constructor.
template<uint8_t w>
	wt_huff (const int_vector< w > &rac)
template<typename RandomAccessContainer >
void	construct (const RandomAccessContainer &rac, size_type size)
template<class size_type_class >
	wt_huff (int_vector_file_buffer< 8, size_type_class > &rac, size_type size)
template<class size_type_class >
void	construct (int_vector_file_buffer< 8, size_type_class > &rac, size_type size)
	Construct the wavelet tree from a random access container.
	wt_huff (const wt_huff &wt)
	Copy constructor.
wt_huff &	operator= (const wt_huff &wt)
	Assignment operator.
void	swap (wt_huff &wt)
	Swap operator.
size_type	size () const
	Returns the size of the original vector.
bool	empty () const
	Returns whether the wavelet tree contains no data.
value_type	operator[] (size_type i) const
	Recovers the ith symbol of the original vector.
size_type	rank (size_type i, value_type c) const
	Calculates how many symbols c are in the prefix [0..i-1] of the supported vector.
size_type	rank_ith_symbol (size_type i, value_type &c) const
	Calculates how many occurrences of symbol wt[i] are in the prefix [0..i-1] of the original sequence.
size_type	select (size_type i, value_type c) const
	Calculates the ith occurrence of the symbol c in the supported vector.
void	interval_symbols (size_type i, size_type j, size_type &k, std::vector< unsigned char > &cs, std::vector< size_type > &rank_c_i, std::vector< size_type > &rank_c_j) const
	Calculates for each symbol c in wt[i..j-1], how many times c occurs in wt[0..i-1] and wt[0..j-1].
size_type	serialize (std::ostream &out, structure_tree_node *v=NULL, std::string name="") const
	Serializes the data structure into the given ostream.
void	load (std::istream &in)
	Loads the data structure from the given istream.
Public Attributes
const size_type &	sigma
const bit_vector_type &	tree

Detailed Description

template<class BitVector = bit_vector, class RankSupport = typename BitVector::rank_1_type, class SelectSupport = typename BitVector::select_1_type, class SelectSupportZero = typename BitVector::select_0_type, bool dfs_shape = 0>
class sdsl::wt_huff< BitVector, RankSupport, SelectSupport, SelectSupportZero, dfs_shape >

A Wavelet Tree class for byte sequences.

A wavelet tree is build for a vector of characters over the alphabet $\Sigma$ . This class should be used only for small alphabets $\Sigma \ll n$ (see int_wavelet_tree for a wavelet tree for big alphabets). The wavelet tree $wt$ consists of a tree of bitvectors and provides three efficient methods:

The "[]"-operator: returns the ith symbol of vector for which the wavelet tree was build for.
The rank method: returns the number of occurences of symbol in the prefix [0..i-1] in the vector for which the wavelet tree was build for.
The select method: returns the index $i\in [0..size()-1]$ of the jth occurence of symbol .

The idea of using a Huffman shaped wavelet was first mentioned on page 17 of the following technical report: Veli Mäkinen and Gonzalo Navarro: Succinct Suffix Arrays based on Run-Length Encoding. Available under: http://swp.dcc.uchile.cl/TR/2005/TR_DCC-2005-004.pdf

  \par Space complexity

$\Order{n H_0 + 2|\Sigma|\log n}$ bits, where $n$ is the size of the vector the wavelet tree was build for.

Constructor & Destructor Documentation

template<class BitVector = bit_vector, class RankSupport = typename BitVector::rank_1_type, class SelectSupport = typename BitVector::select_1_type, class SelectSupportZero = typename BitVector::select_0_type, bool dfs_shape = 0>

template<typename RandomAccessContainer >

sdsl::wt_huff< BitVector, RankSupport, SelectSupport, SelectSupportZero, dfs_shape >::wt_huff	(	const RandomAccessContainer &	rac,
		size_type	size
	)		`[inline]`

Constructor.

   \param rac Reference to the vector (or unsigned char array) for which the wavelet tree should be build.
   \param size Size of the prefix of the vector (or unsigned char array) for which the wavelet tree should be build.
   \par Time complexity

$\Order{n\log|\Sigma|}$ , where $n=size$

Member Function Documentation

template<class BitVector = bit_vector, class RankSupport = typename BitVector::rank_1_type, class SelectSupport = typename BitVector::select_1_type, class SelectSupportZero = typename BitVector::select_0_type, bool dfs_shape = 0>

template<class size_type_class >

void sdsl::wt_huff< BitVector, RankSupport, SelectSupport, SelectSupportZero, dfs_shape >::construct	(	int_vector_file_buffer< 8, size_type_class > &	rac,
		size_type	size
	)		`[inline]`

Construct the wavelet tree from a random access container.

Parameters:

rac	A random access container
size	The length of the prefix of the random access container, for which the wavelet tree should be build

template<class BitVector = bit_vector, class RankSupport = typename BitVector::rank_1_type, class SelectSupport = typename BitVector::select_1_type, class SelectSupportZero = typename BitVector::select_0_type, bool dfs_shape = 0>

void sdsl::wt_huff< BitVector, RankSupport, SelectSupport, SelectSupportZero, dfs_shape >::interval_symbols	(	size_type	i,
		size_type	j,
		size_type &	k,
		std::vector< unsigned char > &	cs,
		std::vector< size_type > &	rank_c_i,
		std::vector< size_type > &	rank_c_j
	)		const `[inline]`

Calculates for each symbol c in wt[i..j-1], how many times c occurs in wt[0..i-1] and wt[0..j-1].

   \param i The start index (inclusive) of the interval.
   \param j The end index (exclusive) of the interval.
   \param k Reference that will contain the number of different symbols in wt[i..j-1].

Parameters:

cs	Reference to a vector of size k that will contain all symbols that occur in wt[i..j-1] in arbitrary order.
rank_c_i	Reference to a vector which equals rank_c_i[p] = rank(i,cs[p]), for $0 \leq p < k$
rank_c_j	Reference to a vector which equals rank_c_j[p] = rank(j,cs[p]), for $0 \leq p < k$

Time complexity: $\Order{\min{\sigma, k \log \sigma}}$

Precondition: $i\leq j$ $cs.size() \geq \sigma$ $rank_{c_i}.size() \geq \sigma$ $rank_{c_j}.size() \geq \sigma$

template<class BitVector = bit_vector, class RankSupport = typename BitVector::rank_1_type, class SelectSupport = typename BitVector::select_1_type, class SelectSupportZero = typename BitVector::select_0_type, bool dfs_shape = 0>

value_type sdsl::wt_huff< BitVector, RankSupport, SelectSupport, SelectSupportZero, dfs_shape >::operator[] ( size_type i ) const [inline]

Recovers the ith symbol of the original vector.

Parameters:

i	The index of the symbol in the original vector. $i \in [0..size()-1]$

Returns:: The ith symbol of the original vector.

Time complexity: $\Order{H_0}$ on average, where is the zero order entropy of the sequence.

template<class BitVector = bit_vector, class RankSupport = typename BitVector::rank_1_type, class SelectSupport = typename BitVector::select_1_type, class SelectSupportZero = typename BitVector::select_0_type, bool dfs_shape = 0>

size_type sdsl::wt_huff< BitVector, RankSupport, SelectSupport, SelectSupportZero, dfs_shape >::rank	(	size_type	i,
		value_type	c
	)		const `[inline]`

Calculates how many symbols c are in the prefix [0..i-1] of the supported vector.

Parameters:

i	The exclusive index of the prefix range [0..i-1], so $i\in[0..size()]$ .
c	The symbol to count the occurrences in the prefix.

Returns:: The number of occurrences of symbol c in the prefix [0..i-1] of the supported vector.

Time complexity: $\Order{H_0}$

template<class BitVector = bit_vector, class RankSupport = typename BitVector::rank_1_type, class SelectSupport = typename BitVector::select_1_type, class SelectSupportZero = typename BitVector::select_0_type, bool dfs_shape = 0>

size_type sdsl::wt_huff< BitVector, RankSupport, SelectSupport, SelectSupportZero, dfs_shape >::rank_ith_symbol	(	size_type	i,
		value_type &	c
	)		const `[inline]`

Calculates how many occurrences of symbol wt[i] are in the prefix [0..i-1] of the original sequence.

   \param i The index of the symbol.

Parameters:

c	Reference that will contain symbol wt[i].

Returns:: The number of occurrences of symbol wt[i] in the prefix [0..i-1]

Time complexity: $\Order{H_0}$

template<class BitVector = bit_vector, class RankSupport = typename BitVector::rank_1_type, class SelectSupport = typename BitVector::select_1_type, class SelectSupportZero = typename BitVector::select_0_type, bool dfs_shape = 0>

size_type sdsl::wt_huff< BitVector, RankSupport, SelectSupport, SelectSupportZero, dfs_shape >::select	(	size_type	i,
		value_type	c
	)		const `[inline]`

Calculates the ith occurrence of the symbol c in the supported vector.

Parameters:

i	The ith occurrence. $i\in [1..rank(size(),c)]$ .
c	The symbol c.

Time complexity: $\Order{H_0}$

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

sdsl/include/sdsl/wt_huff.hpp

Public Types

Public Member Functions

Public Attributes

Detailed Description

Constructor & Destructor Documentation

Member Function Documentation