utils.h

/*
  Copyright (c) 2018, 2019 Jouni Siren
  Copyright (c) 2015, 2016, 2017 Genome Research Ltd.

  Author: Jouni Siren <jouni.siren@iki.fi>

  Permission is hereby granted, free of charge, to any person obtaining a copy
  of this software and associated documentation files (the "Software"), to deal
  in the Software without restriction, including without limitation the rights
  to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
  copies of the Software, and to permit persons to whom the Software is
  furnished to do so, subject to the following conditions:

  The above copyright notice and this permission notice shall be included in all
  copies or substantial portions of the Software.

  THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
  IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
  FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
  AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
  LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
  OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
  SOFTWARE.
*/

#ifndef GCSA_UTILS_H
#define GCSA_UTILS_H

#include <algorithm>
#include <cassert>
#include <cstdint>
#include <fstream>
#include <iostream>
#include <vector>

#include <sdsl/wavelet_trees.hpp>

// TODO Later: Get rid of OpenMP.
#include <omp.h>

// Parallel sorting is only available with libstdc++ parallel mode.
#ifdef __GLIBCXX__
#include <parallel/algorithm>
#endif

namespace gcsa
{

/*
  utils.h: Common utility methods.
*/

//------------------------------------------------------------------------------

typedef std::uint64_t size_type;
typedef std::uint8_t  char_type;
typedef std::uint8_t  comp_type;
typedef std::uint8_t  byte_type;

constexpr size_type BYTE_BITS = 8;
constexpr size_type WORD_BITS = 64;

constexpr size_type KILOBYTE     = 1024;
constexpr size_type MILLION      = 1000000;
constexpr size_type MEGABYTE     = KILOBYTE * KILOBYTE;
constexpr size_type GIGABYTE     = KILOBYTE * MEGABYTE;

constexpr double KILOBYTE_DOUBLE = 1024.0;
constexpr double MILLION_DOUBLE  = 1000000.0;
constexpr double MEGABYTE_DOUBLE = KILOBYTE_DOUBLE * KILOBYTE_DOUBLE;
constexpr double GIGABYTE_DOUBLE = KILOBYTE_DOUBLE * MEGABYTE_DOUBLE;

// Exit code for exceeding parameter-defined size limit.
constexpr int EXIT_SIZE_LIMIT_EXCEEDED = 0x2BAD512E;

//------------------------------------------------------------------------------

/*
  range_type stores a closed range [first, second]. Empty ranges are indicated by
  first > second. The emptiness check uses +1 to handle the common special case
  [0, -1].
*/

typedef std::pair<size_type, size_type> range_type;

struct Range
{
  inline static size_type length(range_type range)
  {
    return range.second + 1 - range.first;
  }

  inline static bool empty(range_type range)
  {
    return (range.first + 1 > range.second + 1);
  }

  inline static bool empty(size_type sp, size_type ep)
  {
    return (sp + 1 > ep + 1);
  }

  inline static size_type bound(size_type value, range_type bounds)
  {
    return bound(value, bounds.first, bounds.second);
  }

  inline static size_type bound(size_type value, size_type low, size_type high)
  {
    return std::max(std::min(value, high), low);
  }

  inline static range_type empty_range()
  {
    return range_type(1, 0);
  }
};

template<class A, class B>
std::ostream& operator<<(std::ostream& stream, const std::pair<A, B>& data)
{
  return stream << "(" << data.first << ", " << data.second << ")";
}

//------------------------------------------------------------------------------

/*
  Global verbosity setting for index construction. Used in conditions of type
  if(Verbosity::level >= Verbosity::THRESHOLD). While the level can be set directly,
  Verbosity::set() does a few sanity checks.

  SILENT    no status information
  BASIC     basic progress information and statistics on the input and the final index
  EXTENDED  add intermediate statistics
  FULL      add technical information on processing individual files
*/

struct Verbosity
{
  static size_type level;

  static void set(size_type new_level);
  static std::string levelName();

  constexpr static size_type SILENT   = 0;
  constexpr static size_type BASIC    = 1;
  constexpr static size_type EXTENDED = 2;
  constexpr static size_type DEFAULT  = 3;
  constexpr static size_type FULL     = 3;
};

//------------------------------------------------------------------------------

struct Version
{
  static std::string str(bool verbose = false);
  static void print(std::ostream& out, const std::string& tool_name, bool verbose = false, size_type new_lines = 2);

  constexpr static size_type MAJOR_VERSION = 1;
  constexpr static size_type MINOR_VERSION = 3;
  constexpr static size_type PATCH_VERSION = 0;

  constexpr static size_type GCSA_VERSION  = 3;
  constexpr static size_type LCP_VERSION   = 1;
};

//------------------------------------------------------------------------------

template<class IntegerType>
inline size_type
bit_length(IntegerType val)
{
  return sdsl::bits::hi(val) + 1;
}

//------------------------------------------------------------------------------

/*
  Thomas Wang's integer hash function. In many implementations, std::hash
  is identity function for integers, which leads to performance issues.
*/

inline size_type
wang_hash_64(size_type key)
{
  key = (~key) + (key << 21); // key = (key << 21) - key - 1;
  key = key ^ (key >> 24);
  key = (key + (key << 3)) + (key << 8); // key * 265
  key = key ^ (key >> 14);
  key = (key + (key << 2)) + (key << 4); // key * 21
  key = key ^ (key >> 28);
  key = key + (key << 31);
  return key;
}

//------------------------------------------------------------------------------

inline double
inMegabytes(size_type bytes)
{
  return bytes / MEGABYTE_DOUBLE;
}

inline double
inGigabytes(size_type bytes)
{
  return bytes / GIGABYTE_DOUBLE;
}

inline double
inBPC(size_type bytes, size_type size)
{
  return (8.0 * bytes) / size;
}

inline double
inMicroseconds(double seconds)
{
  return seconds * MILLION_DOUBLE;
}

const size_type DEFAULT_INDENT = 18;

void printHeader(const std::string& header, size_type indent = DEFAULT_INDENT);
void printTime(const std::string& header, size_type queries, double seconds, size_type indent = DEFAULT_INDENT);

//------------------------------------------------------------------------------

double readTimer();       // Seconds from an arbitrary time point.
size_type memoryUsage();  // Peak memory usage in bytes.

size_type readVolume();   // Only for GCSA construction.
size_type writeVolume();  // Only for GCSA construction.

//------------------------------------------------------------------------------

/*
  Temporary file names have the pattern "prefix_hostname_pid_counter", where
  - prefix is given as an argument to getName();
  - hostname is the name of the host;
  - pid is the process id; and
  - counter is a running counter starting from 0.

  The generated names are stored until the file is deleted with remove(). All
  remaining temporary files are deleted when the program exits (normally or
  with std::exit()).

  TempFile is not thread-safe!
*/

namespace TempFile
{
  extern const std::string DEFAULT_TEMP_DIR;
  extern std::string temp_dir;

  void setDirectory(const std::string& directory);
  std::string getName(const std::string& name_part);
  void remove(std::string& filename);  // Also clears the filename.
  // Forget about current temporary files so that they aren't deleted.
  void forget();
}

// Returns the total length of the rows, excluding line ends.
size_type readRows(const std::string& filename, std::vector<std::string>& rows, bool skip_empty_rows);

size_type fileSize(std::ifstream& file);
size_type fileSize(std::ofstream& file);

//------------------------------------------------------------------------------

/*
  parallelQuickSort() uses less working space than parallelMergeSort(). Calling omp_set_nested(1)
  improves the speed of parallelQuickSort().
*/

template<class Iterator, class Comparator>
void
parallelQuickSort(Iterator first, Iterator last, const Comparator& comp)
{
#ifdef __GLIBCXX__
  int nested = omp_get_nested();
  omp_set_nested(1);
  __gnu_parallel::sort(first, last, comp, __gnu_parallel::balanced_quicksort_tag());
  omp_set_nested(nested);
#else
  std::sort(first, last, comp);
#endif
}

template<class Iterator>
void
parallelQuickSort(Iterator first, Iterator last)
{
#ifdef __GLIBCXX__
  int nested = omp_get_nested();
  omp_set_nested(1);
  __gnu_parallel::sort(first, last, __gnu_parallel::balanced_quicksort_tag());
  omp_set_nested(nested);
#else
  std::sort(first, last);
#endif
}

template<class Iterator, class Comparator>
void
parallelMergeSort(Iterator first, Iterator last, const Comparator& comp)
{
#ifdef __GLIBCXX__
  __gnu_parallel::sort(first, last, comp, __gnu_parallel::multiway_mergesort_tag());
#else
  std::sort(first, last, comp);
#endif
}

template<class Iterator>
void
parallelMergeSort(Iterator first, Iterator last)
{
#ifdef __GLIBCXX__
  __gnu_parallel::sort(first, last, __gnu_parallel::multiway_mergesort_tag());
#else
  std::sort(first, last);
#endif
}

template<class Iterator, class Comparator>
void
sequentialSort(Iterator first, Iterator last, const Comparator& comp)
{
#ifdef __GLIBCXX__
  __gnu_parallel::sort(first, last, comp, __gnu_parallel::sequential_tag());
#else
  std::sort(first, last, comp);
#endif
}

template<class Iterator>
void
sequentialSort(Iterator first, Iterator last)
{
#ifdef __GLIBCXX__
  __gnu_parallel::sort(first, last, __gnu_parallel::sequential_tag());
#else
  std::sort(first, last);
#endif
}

template<class Element>
void
removeDuplicates(std::vector<Element>& vec, bool parallel)
{
  if(parallel) { parallelQuickSort(vec.begin(), vec.end()); }
  else         { sequentialSort(vec.begin(), vec.end()); }
  vec.resize(std::unique(vec.begin(), vec.end()) - vec.begin());
}

template<class Element, class Random>
void
deterministicShuffle(std::vector<Element>& vec, Random& rng, bool parallel)
{
  // We sort to get a deterministic shuffle of the elements instead of the offsets.
  if(parallel) { parallelQuickSort(vec.begin(), vec.end()); }
  else         { sequentialSort(vec.begin(), vec.end()); }
  for(size_type i = vec.size(); i > 0; i--)
  {
    std::swap(vec[i - 1], vec[rng() % i]);
  }
}

//------------------------------------------------------------------------------

/*
  Split the vector approximately evenly between the given number of threads. The comparator
  should return true when it is safe to split between the first argument and the second argument.
*/

template<class Element, class Comparator>
std::vector<range_type>
getBounds(const std::vector<Element>& vec, size_type threads, const Comparator& comp)
{
  std::vector<range_type> bounds(threads);
  for(size_type thread = 0, start = 0; thread < threads; thread++)
  {
    bounds[thread].first = start;
    if(start < vec.size())
    {
      start += std::max((size_type)1, (vec.size() - start) / (threads - thread));
      while(start < vec.size() && !comp(vec[start - 1], vec[start])) { start++; }
    }
    bounds[thread].second = start - 1;
  }
  return bounds;
}

// Get a good chunk size for dynamic scheduling.
size_type getChunkSize(size_type n, size_type min_size);

//------------------------------------------------------------------------------

/*
  GCSA uses a contiguous byte alphabet [0, sigma - 1] internally. Array C is based on the
  number of occurrences of each character in the BWT.
*/

template<class AlphabetType>
inline bool
hasChar(const AlphabetType& alpha, comp_type comp)
{
  return (alpha.C[comp + 1] > alpha.C[comp]);
}

template<class AlphabetType>
inline range_type
charRange(const AlphabetType& alpha, comp_type comp)
{
  return range_type(alpha.C[comp], alpha.C[comp + 1] - 1);
}

template<class AlphabetType>
comp_type
findChar(const AlphabetType& alpha, size_type bwt_pos)
{
  comp_type comp = 0;
  while(alpha.C[comp + 1] <= bwt_pos) { comp++; }
  return comp;
}

template<class BWTType, class AlphabetType>
inline size_type
LF(const BWTType& bwt, const AlphabetType& alpha, size_type i, comp_type comp)
{
  return alpha.C[comp] + bwt.rank(i, comp);
}

template<class BWTType, class AlphabetType>
inline range_type
LF(const BWTType& bwt, const AlphabetType& alpha, range_type range, comp_type comp)
{
  return range_type(LF(bwt, alpha, range.first, comp), LF(bwt, alpha, range.second + 1, comp) - 1);
}

//------------------------------------------------------------------------------

} // namespace gcsa

#endif // GCSA_UTILS_H