Files
libreoffice/tools/source/generic/fract.cxx
Caolán McNamara d430865f35 nothing to see here, move along
Change-Id: I1dc03bb80562c4234e8a44bca0253dd3041a5151
2017-09-21 09:36:26 +01:00

549 lines
15 KiB
C++

/* -*- Mode: C++; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4 -*- */
/*
* This file is part of the LibreOffice project.
*
* This Source Code Form is subject to the terms of the Mozilla Public
* License, v. 2.0. If a copy of the MPL was not distributed with this
* file, You can obtain one at http://mozilla.org/MPL/2.0/.
*
* This file incorporates work covered by the following license notice:
*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed
* with this work for additional information regarding copyright
* ownership. The ASF licenses this file to you under the Apache
* License, Version 2.0 (the "License"); you may not use this file
* except in compliance with the License. You may obtain a copy of
* the License at http://www.apache.org/licenses/LICENSE-2.0 .
*/
#include <tools/fract.hxx>
#include <tools/debug.hxx>
#include <tools/lineend.hxx>
#include <tools/stream.hxx>
#include <o3tl/safeint.hxx>
#include <rtl/ustring.hxx>
#include <sal/log.hxx>
#include <osl/diagnose.h>
#include <limits.h>
#include <algorithm>
#include <cmath>
#include <boost/math/common_factor_rt.hpp>
#include <boost/rational.hpp>
static boost::rational<sal_Int32> rational_FromDouble(double dVal);
static void rational_ReduceInaccurate(boost::rational<sal_Int32>& rRational, unsigned nSignificantBits);
struct Fraction::Impl
{
bool valid;
boost::rational<sal_Int32> value;
Impl()
: valid(false)
{
}
Impl(const Impl&) = delete;
Impl& operator=(const Impl&) = delete;
};
Fraction::Fraction() : mpImpl(new Impl)
{
mpImpl->valid = true;
}
Fraction::Fraction( const Fraction& rFrac ) : mpImpl(new Impl)
{
mpImpl->valid = rFrac.mpImpl->valid;
if (mpImpl->valid)
mpImpl->value.assign( rFrac.mpImpl->value.numerator(), rFrac.mpImpl->value.denominator() );
}
Fraction::Fraction( Fraction&& rFrac ) : mpImpl(std::move(rFrac.mpImpl))
{
}
// Initialized by setting nNum as nominator and nDen as denominator
// Negative values in the denominator are invalid and cause the
// inversion of both nominator and denominator signs
// in order to return the correct value.
Fraction::Fraction( sal_Int64 nNum, sal_Int64 nDen ) : mpImpl(new Impl)
{
assert( nNum >= std::numeric_limits<sal_Int32>::min() );
assert( nNum <= std::numeric_limits<sal_Int32>::max( ));
assert( nDen >= std::numeric_limits<sal_Int32>::min() );
assert( nDen <= std::numeric_limits<sal_Int32>::max( ));
if ( nDen == 0 )
{
mpImpl->valid = false;
SAL_WARN( "tools.fraction", "'Fraction(" << nNum << ",0)' invalid fraction created" );
return;
}
mpImpl->value.assign( nNum, nDen);
mpImpl->valid = true;
}
Fraction::Fraction( double dVal ) : mpImpl(new Impl)
{
try
{
mpImpl->value = rational_FromDouble( dVal );
if ( HasOverflowValue() )
throw boost::bad_rational();
mpImpl->valid = true;
}
catch (const boost::bad_rational&)
{
mpImpl->valid = false;
SAL_WARN( "tools.fraction", "'Fraction(" << dVal << ")' invalid fraction created" );
}
}
Fraction::~Fraction()
{
}
bool Fraction::HasOverflowValue()
{
//coverity[result_independent_of_operands]
return mpImpl->value.numerator() < std::numeric_limits<sal_Int32>::min() ||
mpImpl->value.numerator() > std::numeric_limits<sal_Int32>::max() ||
mpImpl->value.denominator() < std::numeric_limits<sal_Int32>::min() ||
mpImpl->value.denominator() > std::numeric_limits<sal_Int32>::max();
}
Fraction::operator double() const
{
if (!mpImpl->valid)
{
SAL_WARN( "tools.fraction", "'double()' on invalid fraction" );
return 0.0;
}
return boost::rational_cast<double>(mpImpl->value);
}
// This methods first validates both values.
// If one of the arguments is invalid, the whole operation is invalid.
// After computation detect if result overflows a sal_Int32 value
// which cause the operation to be marked as invalid
Fraction& Fraction::operator += ( const Fraction& rVal )
{
if ( !rVal.mpImpl->valid )
mpImpl->valid = false;
if ( !mpImpl->valid )
{
SAL_WARN( "tools.fraction", "'operator +=' with invalid fraction" );
return *this;
}
mpImpl->value += rVal.mpImpl->value;
if ( HasOverflowValue() )
{
mpImpl->valid = false;
SAL_WARN( "tools.fraction", "'operator +=' detected overflow" );
}
return *this;
}
Fraction& Fraction::operator -= ( const Fraction& rVal )
{
if ( !rVal.mpImpl->valid )
mpImpl->valid = false;
if ( !mpImpl->valid )
{
SAL_WARN( "tools.fraction", "'operator -=' with invalid fraction" );
return *this;
}
mpImpl->value -= rVal.mpImpl->value;
if ( HasOverflowValue() )
{
mpImpl->valid = false;
SAL_WARN( "tools.fraction", "'operator -=' detected overflow" );
}
return *this;
}
namespace
{
template<typename T> bool checked_multiply_by(boost::rational<T>& i, const boost::rational<T>& r)
{
// Protect against self-modification
T num = r.numerator();
T den = r.denominator();
// Avoid overflow and preserve normalization
T gcd1 = boost::math::gcd(i.numerator(), den);
T gcd2 = boost::math::gcd(num, i.denominator());
bool fail = false;
fail |= o3tl::checked_multiply(i.numerator() / gcd1, num / gcd2, num);
fail |= o3tl::checked_multiply(i.denominator() / gcd2, den / gcd1, den);
i.assign(num, den);
return fail;
}
}
Fraction& Fraction::operator *= ( const Fraction& rVal )
{
if ( !rVal.mpImpl->valid )
mpImpl->valid = false;
if ( !mpImpl->valid )
{
SAL_WARN( "tools.fraction", "'operator *=' with invalid fraction" );
return *this;
}
bool bFail = checked_multiply_by(mpImpl->value, rVal.mpImpl->value);
if (bFail || HasOverflowValue())
{
mpImpl->valid = false;
}
return *this;
}
Fraction& Fraction::operator /= ( const Fraction& rVal )
{
if ( !rVal.mpImpl->valid )
mpImpl->valid = false;
if ( !mpImpl->valid )
{
SAL_WARN( "tools.fraction", "'operator /=' with invalid fraction" );
return *this;
}
mpImpl->value /= rVal.mpImpl->value;
if ( HasOverflowValue() )
{
mpImpl->valid = false;
SAL_WARN( "tools.fraction", "'operator /=' detected overflow" );
}
return *this;
}
/** Inaccurate cancellation for a fraction.
Clip both nominator and denominator to said number of bits. If
either of those already have equal or less number of bits used,
this method does nothing.
@param nSignificantBits denotes, how many significant binary
digits to maintain, in both nominator and denominator.
@example ReduceInaccurate(8) has an error <1% [1/2^(8-1)] - the
largest error occurs with the following pair of values:
binary 1000000011111111111111111111111b/1000000000000000000000000000000b
= 1082130431/1073741824
= approx. 1.007812499
A ReduceInaccurate(8) yields 1/1.
*/
void Fraction::ReduceInaccurate( unsigned nSignificantBits )
{
if ( !mpImpl->valid )
{
SAL_WARN( "tools.fraction", "'ReduceInaccurate' on invalid fraction" );
return;
}
if ( !mpImpl->value.numerator() )
return;
rational_ReduceInaccurate(mpImpl->value, nSignificantBits);
}
sal_Int32 Fraction::GetNumerator() const
{
if ( !mpImpl->valid )
{
SAL_WARN( "tools.fraction", "'GetNumerator()' on invalid fraction" );
return 0;
}
return mpImpl->value.numerator();
}
sal_Int32 Fraction::GetDenominator() const
{
if ( !mpImpl->valid )
{
SAL_WARN( "tools.fraction", "'GetDenominator()' on invalid fraction" );
return -1;
}
return mpImpl->value.denominator();
}
Fraction& Fraction::operator=( const Fraction& rFrac )
{
if (this == &rFrac)
return *this;
Fraction tmp(rFrac);
std::swap(mpImpl, tmp.mpImpl);
return *this;
}
Fraction& Fraction::operator=( Fraction&& rFrac )
{
mpImpl = std::move(rFrac.mpImpl);
return *this;
}
bool Fraction::IsValid() const
{
return mpImpl->valid;
}
Fraction::operator sal_Int32() const
{
if ( !mpImpl->valid )
{
SAL_WARN( "tools.fraction", "'operator sal_Int32()' on invalid fraction" );
return 0;
}
return boost::rational_cast<sal_Int32>(mpImpl->value);
}
Fraction operator+( const Fraction& rVal1, const Fraction& rVal2 )
{
Fraction aErg( rVal1 );
aErg += rVal2;
return aErg;
}
Fraction operator-( const Fraction& rVal1, const Fraction& rVal2 )
{
Fraction aErg( rVal1 );
aErg -= rVal2;
return aErg;
}
Fraction operator*( const Fraction& rVal1, const Fraction& rVal2 )
{
Fraction aErg( rVal1 );
aErg *= rVal2;
return aErg;
}
Fraction operator/( const Fraction& rVal1, const Fraction& rVal2 )
{
Fraction aErg( rVal1 );
aErg /= rVal2;
return aErg;
}
bool operator !=( const Fraction& rVal1, const Fraction& rVal2 )
{
return !(rVal1 == rVal2);
}
bool operator <=( const Fraction& rVal1, const Fraction& rVal2 )
{
return !(rVal1 > rVal2);
}
bool operator >=( const Fraction& rVal1, const Fraction& rVal2 )
{
return !(rVal1 < rVal2);
}
bool operator == ( const Fraction& rVal1, const Fraction& rVal2 )
{
if ( !rVal1.mpImpl->valid || !rVal2.mpImpl->valid )
{
SAL_WARN( "tools.fraction", "'operator ==' with an invalid fraction" );
return false;
}
return rVal1.mpImpl->value == rVal2.mpImpl->value;
}
bool operator < ( const Fraction& rVal1, const Fraction& rVal2 )
{
if ( !rVal1.mpImpl->valid || !rVal2.mpImpl->valid )
{
SAL_WARN( "tools.fraction", "'operator <' with an invalid fraction" );
return false;
}
return rVal1.mpImpl->value < rVal2.mpImpl->value;
}
bool operator > ( const Fraction& rVal1, const Fraction& rVal2 )
{
if ( !rVal1.mpImpl->valid || !rVal2.mpImpl->valid )
{
SAL_WARN( "tools.fraction", "'operator >' with an invalid fraction" );
return false;
}
return rVal1.mpImpl->value > rVal2.mpImpl->value;
}
SvStream& ReadFraction( SvStream& rIStream, Fraction const & rFract )
{
sal_Int32 num(0), den(0);
rIStream.ReadInt32( num );
rIStream.ReadInt32( den );
if ( den <= 0 )
{
SAL_WARN( "tools.fraction", "'ReadFraction()' read an invalid fraction" );
rFract.mpImpl->valid = false;
}
else
{
rFract.mpImpl->value.assign( num, den );
rFract.mpImpl->valid = true;
}
return rIStream;
}
SvStream& WriteFraction( SvStream& rOStream, const Fraction& rFract )
{
if ( !rFract.mpImpl->valid )
{
SAL_WARN( "tools.fraction", "'WriteFraction()' write an invalid fraction" );
rOStream.WriteInt32( 0 );
rOStream.WriteInt32( -1 );
} else {
rOStream.WriteInt32( rFract.mpImpl->value.numerator() );
rOStream.WriteInt32( rFract.mpImpl->value.denominator() );
}
return rOStream;
}
// If dVal > LONG_MAX or dVal < LONG_MIN, the rational throws a boost::bad_rational.
// Otherwise, dVal and denominator are multiplied by 10, until one of them
// is larger than (LONG_MAX / 10).
//
// NOTE: here we use 'sal_Int32' due that only values in sal_Int32 range are valid.
static boost::rational<sal_Int32> rational_FromDouble(double dVal)
{
if ( dVal > std::numeric_limits<sal_Int32>::max() ||
dVal < std::numeric_limits<sal_Int32>::min() )
throw boost::bad_rational();
const sal_Int32 nMAX = std::numeric_limits<sal_Int32>::max() / 10;
sal_Int32 nDen = 1;
while ( std::abs( dVal ) < nMAX && nDen < nMAX ) {
dVal *= 10;
nDen *= 10;
}
return boost::rational<sal_Int32>( sal_Int32(dVal), nDen );
}
// Similar to clz_table that can be googled
const char nbits_table[32] =
{
32, 1, 23, 2, 29, 24, 14, 3,
30, 27, 25, 18, 20, 15, 10, 4,
31, 22, 28, 13, 26, 17, 19, 9,
21, 12, 16, 8, 11, 7, 6, 5
};
static int impl_NumberOfBits( sal_uInt32 nNum )
{
// http://en.wikipedia.org/wiki/De_Bruijn_sequence
// background paper: Using de Bruijn Sequences to Index a 1 in a
// Computer Word (1998) Charles E. Leiserson,
// Harald Prokop, Keith H. Randall
// (e.g. http://citeseer.ist.psu.edu/leiserson98using.html)
const sal_uInt32 nDeBruijn = 0x7DCD629;
if ( nNum == 0 )
return 0;
// Get it to form like 0000001111111111b
nNum |= ( nNum >> 1 );
nNum |= ( nNum >> 2 );
nNum |= ( nNum >> 4 );
nNum |= ( nNum >> 8 );
nNum |= ( nNum >> 16 );
sal_uInt32 nNumber;
int nBonus;
nNumber = nNum;
nBonus = 0;
// De facto shift left of nDeBruijn using multiplication (nNumber
// is all ones from topmost bit, thus nDeBruijn + (nDeBruijn *
// nNumber) => nDeBruijn * (nNumber+1) clears all those bits to
// zero, sets the next bit to one, and thus effectively shift-left
// nDeBruijn by lg2(nNumber+1). This generates a distinct 5bit
// sequence in the msb for each distinct position of the last
// leading 0 bit - that's the property of a de Bruijn number.
nNumber = nDeBruijn + ( nDeBruijn * nNumber );
// 5-bit window indexes the result
return ( nbits_table[nNumber >> 27] ) + nBonus;
}
/** Inaccurate cancellation for a fraction.
Clip both nominator and denominator to said number of bits. If
either of those already have equal or less number of bits used,
this method does nothing.
@param nSignificantBits denotes, how many significant binary
digits to maintain, in both nominator and denominator.
@example ReduceInaccurate(8) has an error <1% [1/2^(8-1)] - the
largest error occurs with the following pair of values:
binary 1000000011111111111111111111111b/1000000000000000000000000000000b
= 1082130431/1073741824
= approx. 1.007812499
A ReduceInaccurate(8) yields 1/1.
*/
static void rational_ReduceInaccurate(boost::rational<sal_Int32>& rRational, unsigned nSignificantBits)
{
if ( !rRational )
return;
// http://www.boost.org/doc/libs/release/libs/rational/rational.html#Internal%20representation
const bool bNeg = ( rRational.numerator() < 0 );
sal_Int32 nMul = bNeg? -rRational.numerator(): rRational.numerator();
sal_Int32 nDiv = rRational.denominator();
DBG_ASSERT(nSignificantBits<65, "More than 64 bit of significance is overkill!");
// How much bits can we lose?
const int nMulBitsToLose = std::max( ( impl_NumberOfBits( nMul ) - int( nSignificantBits ) ), 0 );
const int nDivBitsToLose = std::max( ( impl_NumberOfBits( nDiv ) - int( nSignificantBits ) ), 0 );
const int nToLose = std::min( nMulBitsToLose, nDivBitsToLose );
// Remove the bits
nMul >>= nToLose;
nDiv >>= nToLose;
if ( !nMul || !nDiv ) {
// Return without reduction
OSL_FAIL( "Oops, we reduced too much..." );
return;
}
rRational.assign( bNeg ? -nMul : nMul, nDiv );
}
/* vim:set shiftwidth=4 softtabstop=4 expandtab: */