Math::Fraction - To Manipulate Exact Fractions (v.53b, Beta Release)


Fraction documentation Contained in the Fraction distribution.

Index


Code Index:

NAME

Top

Math::Fraction - To Manipulate Exact Fractions (v.53b, Beta Release)

SYNOPSIS

Top

    use Math::Fraction;

    $a = frac(1,2); $b = frac(6,7);

    print "$a + $b = ", $a + $b, "$a * $b = ", $a * $b;
    print $a->num;

DESCRIPTION

Top

This program is meant to replace the old bigrat perl library. It can do everything it can do and a lot more.

Some of its features include:

Usage

 frac(FRACTION, TAGS) || Math::Fraction->new(FRACTION, TAGS)
 ex $f1=frac(2,3); $f2=frac(7,3,MIXED); $f3=frac("-10 3/4");

FRACTION can equal any of the following:

(Numerator, Denominator)
(Number, Numerator, Denominator)
Decimal
"Numerator/Denominator"
"Number Numerator/Denominator"

Any of these numbers can be any real number however if you enter in negative numbers in anything but the First Number for the Mixed (3 numbers) the negative will be ignored.

TAGS can equal 0, one or more of the following

NORMAL|MIXED|MIXED_RAW|RAW|DEF_MIXED

Controls How the fraction is displayed:

NORMAL

display it in the #/# form

MIXED_RAW

display it in the # #/# form

MIXED

the same as MIXED_RAW but if one part is equal to 0 it will leave it off

RAW

the same as NORAML but always includes the sign

DEF_MIXED

will let it be what ever the default value is at the time

REDUCE|NO_REDUCE|DEF_REDUCE

Controls the automatic reduction of fraction after operations are performed on it.

AUTO|NO_AUTO|DEF_AUTO

Set rether to automatically convert between BIG and SMALL as needed see Notes on the AUTO tag.

SMALL|BIG|DEF_BIG

When the AUTO tag is NOT set it will set whether to use Arbitrary-Length numbers using the Math::BigInt and Math::BigFloat package. (Not the ** operator will not work however due to limitations of the packages.) When the AUTO tag is set these tags will have NO effect. (Note the default tags are NORMAL REDUCE AUTO and SMALL)

Methods

reduce

returns a reduced fraction but leaves the original object untouched.

string(NORMAL|MIXED|MIXED_RAW|RAW)

returns the fraction as a string. if no parameters are given the objects default display method will be used.

decimal|num

returned the decimal value of the fraction

list|list(MIXED)

returns a list containing the fraction if MIXED is used than a 3 item list is returned.

is_tag

returns 1 is that tag exists in the fraction undef if is set to the default -1 otherwise

is_tag(INC_DEF)

returns 1 is that tag exists in the fraction -1 if is tag does not exist but the default is set to that, 0 otherwise.

tags

returns a list of the objects tags

tags(INC_DEF)

returns a list of the objects tags if a particular tag is set to read a default the default tag is returned instead.

All of the above methods may all be exported so that they can be used as functions with a fraction as their first parameter. The string, decimal|num functions can be imported with the tag STR_NUM instead of having to list each one.

modify

modifies the object. Works almost the same as the new method but it doesn't return anything and preserves the objects tags unless overridden by new entries

modify_reduce

same as reduce but it modifies the object instead of returning a fraction.

modify_num(Numerator)

modifies the fraction's numerator.

modify_den(Denominator)

modified the fraction's denominator.

modify_tag(TAGS)

modifies the fraction tags.

The is_tag, tags, and modify_tags methods can be used on the class its self to get at or modify the default tags.

The following methods will always modify or read the Class defaults

digits

returns the default number of digests to return when doing floating point operations with BIG numbers, if set to undef Math::BigFloat will decide.

  modify_digits(NUM)

Dafault Sets

Default sets are way of modifying the defaults with out effecting other functions. Functions that relay on the default values or modify the default should start with a $set_id = temp_set and end with a temp_set($set_id).

See "EXAMPLES" for examples of how default sets work.

The following methods are met to manage default sets and will always work on the Class defaults.

sets

returns a list of all the sets;

name_set

return the name of the current set.

name_set(NAME)

name the current set

save_set

saves the current sent based on its name as given above

save_set(NAME)

save the current set as NAME

save_set(RAND)

save the current set using a unique name

load_set(NAME)

loads a set.

copy_set(NAME_ORG, NAME_NEW)

copies a set. Returns true if successful.

del_set(NAME)

deletes a set.

exists_set(NAME)

returns true if the set exists.

use_set(NAME)

uses a set, that is any changes you make to the used set will also change the original set, like a link.

temp_set

loads a temp set using the default default values and returns a unique id you need to keep.

temp_set(ID)

restores the original set based on the id you should of kept.

Unless otherwise specified all the set methods will return the name of the set being worked on if it was successful, false otherwise

tags(SET)

lists all the tags in SET.

is_tag(TAG, SET)

returns true if TAG exists in SET

digits(SET)

returns what digits is set to in SET;

Overloaded Operators

The following operations have been overridden and will return a fraction:

  +  -  /  *  +  +=  -=  *=  /=  ++  -- abs

The following operations have also been overridden:

 <=> == != < <= > >=

The following operations have also been overridden however they may spit out nasty fractions.

  ** sqrt

Whenever you try to access a fraction as a string the string method will be called and when try to access it as a number the decimal method will be called.

This means that almost all other operations will work however some might return decimals like the sin and cos;

Notes on the AUTO tag

With the AUTO tag set Fractions will be converted between SMALL and BIG as needed. The BIG and SMALL tag will be *ignorded* unless you explicitly specify NO_AUTO in auto to control how the fraction is stored.

When you give it a number it will decide if it is small enough to be stored as a SMALL or if the fraction needs to converted to BIG. However, in order for it to recognize a big fraction the number needs to be in quotes, thus frac(7823495784957895478,781344567825678454) will still be stored as a small with some of the digits lost.

When calculating to SMALL numbers that results in a number that is to big for SMALL the calculation is done AGAIN but this time with BIG numbers (so that it will calculate all the digits correctly) and the new fraction will become a BIG.

When calculating to BIG numbers that results in a number that is small enough to be a SMALL the new fraction will become a SMALL.

Normally, the AUTO tag will save time as calculating with BIG numbers can be quite time consuming however it might slow thinks down if it constantly converts between the two thus in some cases it may be wise to turn it off.

EXAMPLES

Top

This is a small demonstration of what the fraction module can do.

It is run for the most part with these two functions.

 sub pevel {print ">$_[0]\n"; $ans = eval $_[0]; print " $ans\n"; }
 sub evelp {print ">$_[0]\n"; eval $_[0]; } 

You can see it for yourself my typing in perl -e "use Math::FractionDemo; frac_calc;" then frac_demo.

 >frac(1, 3)
  1/3
 >frac(4, 3, MIXED)
  1 1/3
 >frac(1, 1, 3)
  4/3
 >frac(1, 1, 3, MIXED)
  1 1/3
 >frac(10)
  10/1
 >frac(10, MIXED)
  10
 >frac(.66667)
  2/3
 >frac(1.33333, MIXED)
  1 1/3
 >frac("5/6")
  5/6
 >frac("1 2/3")
  5/3
 >frac(10, 20, NO_REDUCE)
  10/20

 >$f1=frac(2,3); $f2=frac(4,5);
 >$f1 + $f2
  22/15
 >$f1 * $f2
  8/15
 >$f1 + 1.6667
  7/3
 >$f2->modify_tag(MIXED)
 >$f2 + 10
  10 4/5
 >frac($ans, NORMAL) # trick to create new fraction with different tags
  54/5
 >$f1 + $f2          # Add two unlikes it goes to default mode
  22/15
 >$f1**1.2
  229739670999407/373719281884655
 >$f1->num**1.2
  0.614738607654485
 >frac(1,2)+frac(2,5)
  9/10

 >$f1=frac(5,3,NORMAL); $f2=frac(7,5);
 >"$f1  $f2"
  5/3  7/5
 >Math::Fraction->modify_tag(MIXED)
 >"$f1  $f2"
  5/3  1 2/5
 >$f1 = frac("3267893629762/32678632179820")
  3267893629762/32678632179820
 >$f2 = frac("5326875886785/76893467996910")
  5326875886785/76893467996910
 >$f1->is_tag(BIG).",".$f2->is_tag(BIG) # Notice how neither is BIG
  0,0
 >$f1+$f2
  21267734600460495169085706/125638667885089122116217810
 >$ans->is_tag(BIG)                     # But this answer is.
  1
 >$f1*$f2
  1740766377695750621849517/251277335770178244232435620
 >$ans->is_tag(BIG)                     # And so is this one.
  1

 >$f1 = frac("3267893629762/32678632179820", BIG)
  3267893629762/32678632179820
 >$f1->is_tag(BIG)   # Notice how the big tag had no effect.
  0
 >$f1->modify_tag(NO_AUTO, BIG)
 >$f1->is_tag(BIG)   # But now it does.  You have to turn off AUTO.
  1
 >$f1->num
  .10000093063197482237806917498797382196606
 >Math::Fraction->modify_digits(15)
 >$f1->num
  .1000009306319748
 >$f1 = frac("0.1231231234564564564564564564561234567891234567891234")
  13680347037037036999999999999963000037/
                             111111111000000000000000000000000000000
 >Math::Fraction->modify_digits(65)
 >$f1->num
  .123123123456456456456456456456123456789123456789123456789123456789

 >$f1 = frac(7,5);
 >$f2 = frac("3267893629762/32678632179820", NO_AUTO, BIG)
 >Math::Fraction->modify_tag(MIXED); Math::Fraction->modify_digits(60)
 >"$f1 ".$f2->num
  1 2/5 .1000009306319748223780691749879738219660647769485035912494771
 >Math::Fraction->load_set(DEFAULT)
 >"$f1 ".$f2->num
  7/5 .10000093063197482237806917498797382196606
 >Math::Fraction->modify_digits(25)
 >"$f1 ".$f2->num
  7/5 .10000093063197482237806917
 >$s = Math::Fraction->temp_set
 >Math::Fraction->modify_tag(MIXED); Math::Fraction->modify_digits(15)
 >"$f1 ".$f2->num
  1 2/5 .1000009306319748
 >Math::Fraction->temp_set($s)
 >Math::Fraction->exists_set($s)

 >"$f1 ".$f2->num  # Notice how it goes back to the previous settings.
  7/5 .10000093063197482237806917

 >Math::Fraction->name_set('temp1')
 >Math::Fraction->modify_tag(MIXED, NO_AUTO)
 >Math::Fraction->modify_digits(60)
 >&s(Math::Fraction->tags, Math::Fraction->digits)
  MIXED REDUCE SMALL NO_AUTO 60
 >Math::Fraction->save_set  # If no name is given it will be saved via
 >                          # its given name
 >Math::Fraction->load_set(DEFAULT)
 >&s(Math::Fraction->tags, Math::Fraction->digits)
  NORMAL REDUCE SMALL AUTO undef
 >&s(Math::Fraction->tags('temp1'), Math::Fraction->digits('temp1'))
  MIXED REDUCE SMALL NO_AUTO 60
 >  # ^^ Notice how this lets you preview other sets w/out loading them.
 >Math::Fraction->load_set(DEFAULT)
 >Math::Fraction->use_set('temp1')
 >Math::Fraction->modify_tag(NO_REDUCE)
 >&s(Math::Fraction->tags, Math::Fraction->digits)
  MIXED NO_REDUCE SMALL NO_AUTO 60
 >&s(Math::Fraction->tags('temp1'), Math::Fraction->digits('temp1'))
  MIXED NO_REDUCE SMALL NO_AUTO 60
 >  # ^^ Notice how this also modifies the temp1 tag becuase it is
 >  #    being used if it was just loaded it would not do this
 >  #    becuase there is no link.

NOTES

Top

Beta Release

Originally Developed with Perl v 5.003_37 for Win32.

Has been testing on Perl Ver 5.003 on a solaris machine and perl 5.004 on Windows 95.

Built on a Linux 2 machine with perl v5.003.

Please send me feedback at kevina@clark.net

This is my first real attempt at writing a Perl Module and at Object-Oriented Programming. (Although I know this is not a true-true Object-Oriented Module as I cheated a little). I mainly wrote it to teach my self how to program Object-Oriently in Perl and for the challenge.

If you know of any faster or simpler way of doing any this please let me know.

SEE ALSO

Top

Math::FractionDemo, perl(1b)

AUTHOR and COPYRIGHT

Top


Fraction documentation Contained in the Fraction distribution.

package Math::Fraction;

# Purpose: To Manipulate Exact Fractions
#
# Copyright 1997 by Kevin Atkinson (kevina@cark.net)
# Version .53b  (2 Feb 1998)
# Beta Release
# Originally Developed with Perl v 5.003_37 for Win32.
# Has been testing on Perl Ver 5.003 on a solaris machine and Perl 5.004
# on Windows 95
# Built on a Linux 2 machine with perl v5.003
#
# Please send me feedback at kevina@clark.net

use vars qw($VERSION @ISA @EXPORT);

require Exporter;
$VERSION = "0.53";
@ISA = qw(Exporter);
@EXPORT = qw(frac);
@EXPORT_OK = qw(reduce string decimal num list is_tag);
%EXPORT_TAGS = (
  STR_NUM => [qw(string decimal num)],
);

use Carp;
use strict;
use Math::BigInt;
use Math::BigFloat;
use overload
   "+"   => "add",
   "-"   => "sub",
   "*"   => "mul",
   "/"   => "div",
   "abs" => "abs",
   "**"  => "pow",
   "sqrt"=> "sqrt",
   "<=>" => "cmp",
   '""'  => "string",
   "0+"  => "decimal",
   "fallback" => 1;

my %DEF = (
  CURRENT => {TAGS => ['NORMAL','REDUCE','SMALL','AUTO'], DIGITS => undef, SYSTEM => 1, NAME => 'DEFAULT'},
  DEFAULT => {TAGS => ['NORMAL','REDUCE','SMALL','AUTO'], DIGITS => undef, READONLY=>1, SYSTEM=>1},
  BLANK   => {TAGS => ['','','']                 , DIGITS => ''   , READONLY=>1, SYSTEM=>1},
);

my ($OUTFORMAT, $REDUCE, $SIZE, $AUTO, $INTERNAL, $RED_STATE) = (0..5);
my $TAG_END = 3;          #Last index of tags ment to be kept.

my %TAGS = (
  NORMAL     => [$OUTFORMAT, 'NORMAL'],
  MIXED      => [$OUTFORMAT, 'MIXED'],
  MIXED_RAW  => [$OUTFORMAT, 'MIXED_RAW'],
  RAW        => [$OUTFORMAT, 'RAW'],
  DEF_MIXED  => [$OUTFORMAT, undef],
  REDUCE     => [$REDUCE, 'REDUCE'],
  NO_REDUCE  => [$REDUCE, 'NO_REDUCE'],
  DEF_REDUCE => [$REDUCE, undef],
  SMALL      => [$SIZE, 'SMALL'],
  BIG        => [$SIZE, 'BIG'],
  DEF_BIG    => [$SIZE, undef],
  AUTO       => [$AUTO, 'AUTO'],
  NO_AUTO    => [$AUTO, 'NO_AUTO'],
  DEF_AUTO   => [$AUTO, undef],
  CONVERTED  => [$INTERNAL, 'CONVERTED'],
  IS_REDUCED => [$RED_STATE, 'IS_REDUCED'],
);

my @DEF_TAG = qw(DEF_MIXED DEF_REDUCE DEF_BIG DEF_AUTO);

my $ID = 01;

sub new {
  my $proto = shift;
  my $class = ref($proto) || $proto;
  my ($self, @frac, @tags, $tag, $decimal, $p1, $p2, $p3);
  if (&_is_decimal($_[0]) and &_is_decimal($_[1]) and &_is_decimal($_[2]) ) {
      my $sign = $_[0]/abs($_[0]);
      @tags = &_tags(@_[3..$#_]);
      ($decimal, $p1, $p2, $p3) = &_fix_num(\@tags, @_[0..2]);
      ($p1, $p2, $p3) = (abs($p1),abs($p2),abs($p3) );
      @frac = ($p1*$p3+$p2, $sign*$p3);
      @frac = &_de_decimal(@frac, \@tags) if $decimal;
  } elsif (&_is_decimal($_[0]) and &_is_decimal($_[1]) ) {
      @tags = &_tags(@_[2..$#_]);
      ($decimal, @frac) = &_fix_num(\@tags, @_[0..1]);
      @frac = &_de_decimal(@frac, \@tags) if $decimal;
      @frac = &_simplify_sign(@frac);
  } elsif (&_is_decimal($_[0]) ) {
    {
      @tags = &_tags(@_[1..$#_]);
      ($decimal, $p1) = &_fix_num(\@tags, $_[0]);
      @frac=($p1,1), last if not $decimal;
      (@frac[0..1], $tag) = &_from_decimal($p1);
      @tags = &_tags(@tags, $tag);
      ($decimal,@frac) = &_fix_num(\@tags, @frac);
      @frac = &_de_decimal(@frac, \@tags) if $decimal;
    }
  } elsif ($_[0] =~ /\s*([\+\-]?)\s*([0-9e\.\+\-]+)\s+([0-9e\.\+\-]+)\s*\/\s*([0-9e\.\+\-]+)/) {
      my $sign = $1.'1';
      @tags = &_tags(@_[1..$#_]);
      ($decimal, $p1, $p2, $p3) = &_fix_num(\@tags, $2, $3, $4);
      ($p1, $p2, $p3) = (abs($p1),abs($p2),abs($p3) );
      @frac = ($p1*$p3+$p2, $sign*$p3);
      @frac = &_de_decimal($p1*$p3+$p2, $sign*$p3, \@tags) if $decimal;
  } elsif ($_[0] =~ /\s*([0-9e\.\+\-]+)\s*\/\s*([0-9e\.\+\-]+)/) {
      @tags = &_tags(@_[1..$#_]);
      ($decimal, @frac) = &_fix_num(\@tags, $1, $2);
      @frac = &_de_decimal(@frac, \@tags) if $decimal;
      @frac = &_simplify_sign(@frac);
  } else {
      croak("\"$_[0]\" is of unknown format");
  }
  croak ("Can not have 0 as the denominator") if $frac[1] == 0;

  if ( &_tag($REDUCE, \@tags) ne 'NO_REDUCE'
       and &_tag($RED_STATE, \@tags) ne 'IS_REDUCED' )
  {
    my $not_reduced;
    ($not_reduced, @frac) = &_reduce(@frac);
    @frac = &_fix_auto('DOWN',\@tags, @frac) if $not_reduced
                                       and &_tag($AUTO, \@tags) eq 'AUTO';
  }
                         
  @tags[$RED_STATE] = undef if &_tag($RED_STATE, \@tags) eq 'IS_REDUCED';

  $self->{'frac'}=\@frac;
  $self->{'tags'}=\@tags;
  bless ($self, $class);
  return $self;
}

# The following functions are met to be exported as shortcuts to method
# operations.

sub frac {
  #special exported function to simplify defining fractions
  return Math::Fraction->new(@_);
}

# Now are the methodes

sub string {
  my $self = shift;
  my @frac;
  my $mixed = &_tag ($OUTFORMAT, [$_[0]], $self->{'tags'} );
  if ($mixed eq 'MIXED') {
      @frac = $self->list('MIXED');
      my $string = "";
      $string .= "$frac[0]"           if $frac[0] != 0;
      $string .= " "                  if $frac[0] != 0 and $frac[1] !=0;
      $string .= "$frac[1]/$frac[2]"  if $frac[1] != 0;
      $string  = "0"                  if $string eq '';
      return $string;
  } elsif ($mixed eq 'MIXED_RAW') {
      @frac = $self->list('MIXED');
      return "$frac[0] $frac[1]/$frac[2]";
  } elsif ($mixed eq 'RAW') {
      @frac = $self->list;
      return ($frac[0] >= 0 ? '+':'')."$frac[0]/$frac[1]";
  } else {
      @frac = $self->list;
      return "$frac[0]/$frac[1]";
  }
}

sub list {
  my $self = shift;
  my @frac = @{$self->{'frac'}};
  if ($_[0] eq "MIXED") {
    my $whole=$frac[0]/$frac[1];
    $whole=int($whole) if not ref($frac[0]);
    $frac[0] = abs($frac[0] - $frac[1]*$whole);
    @frac = ($whole, @frac);
  }
  foreach (@frac) {s/^\+//;};
  return @frac;
}

sub reduce {
  my $self = shift;
  my ($undef, @frac) = &_reduce(@{$self->{'frac'}});
  return Math::Fraction->new(@frac, @{$self->{'tags'}});
}


sub decimal {
  my $self = shift;
  my @frac = @{$self->{'frac'}};
  return $frac[0]/$frac[1] if not ref($frac[0]);
  return Math::BigFloat->new(Math::BigFloat::fdiv($frac[0], $frac[1], $DEF{CURRENT}{DIGITS}) ) if ref($frac[0]);
}

sub num {
  my $self = shift;
  my @frac = @{$self->{'frac'}};
  return $frac[0]/$frac[1] if not ref($frac[0]);
  return Math::BigFloat->new(Math::BigFloat::fdiv($frac[0], $frac[1], $DEF{CURRENT}{DIGITS}) ) if ref($frac[0]);
}

## For the next three methods:
# If used on the object use the tags of the object
# If given a class use the dafault tags,
# .... if a default set is specified then return for that set.

sub is_tag {
  my $self = shift;
  my $tag = shift;
  my $default = 1 if $_[0] eq 'INC_DEF';
  my $is_tag = 0;
  my @tags;
  {
    $is_tag = 0, last if not $TAGS{$tag}; #if there is no such tag ret=0
    my ($num, $tag) = @{$TAGS{$tag}};
    if (ref($self) eq "Math::Fraction") {
      @tags = @{$self->{'tags'}};
      $is_tag = 1    , last if $tags[$num] eq $tag;
      $is_tag = undef, last if $tags[$num] eq undef and not $default;
      $is_tag = -1   , last if $DEF{CURRENT}{TAGS}[$num] eq $tag
                             and $tags[$num] eq undef and $default;
      $is_tag = 0; 
    } else {
      my $set;
      $set = 'CURRENT' unless $set = $_[0];
      $set = 'BLANK'   unless exists $DEF{$set};
      $is_tag = 1   , last if $DEF{$set}{TAGS}[$num] eq $tag;
      $is_tag = 0;
    }
  }
  return $is_tag;
}

sub tags {
  my $self = shift;
  my @tags;
  if (ref($self) eq "Math::Fraction") {
    my $inc_def = 1 if $_[0] eq 'INC_DEF';
    @tags = @{$self->{'tags'}}[0..$TAG_END];
    my $num;
    foreach $num (0 .. $#tags) {
      $tags[$num] = $DEF_TAG[$num]  if $tags[$num] eq undef and not $inc_def;
      $tags[$num] = $DEF{CURRENT}{TAGS}[$num] if $tags[$num] eq undef and $inc_def;
    }
  } elsif (ref($self) ne "Math::Fraction") {
    my $set;
    $set = 'CURRENT' unless $set = $_[0];
    $set = 'BLANK'   unless exists $DEF{$set};
    @tags = @{$DEF{$set}{TAGS}};
  }
  return @tags;
}

sub digits {
  my $self = shift;
  my $set;
  $set = 'CURRENT' unless $set = $_[0];
  $set = 'BLANK'   unless exists $DEF{$set};
  return $DEF{$set}{DIGITS};
}

##
# These mehods are used form managing default sets.

sub sets {
  my $self = shift;
  return keys %DEF;
}

sub name_set {
  shift; 
  return $DEF{CURRENT}{NAME}  if not $_[0];
  $DEF{CURRENT}{NAME} = $_[0] if     $_[0];
}

sub exists_set {
  return exists $DEF{$_[1]};
}

sub use_set {
  my $self = shift;
  my $name = shift;
  if (exists $DEF{$name} and not $DEF{$name}{READONLY}) {
    $DEF{CURRENT} = $DEF{$name};
    return $name;
  } else {
    return undef;
  }
}

sub temp_set {
  my $self = shift;
  my $name = shift;
  if (not $name) {
    $ID++;
    $name = "\cI\cD$ID";
    $self->copy_set('CURRENT', $name);
    $self->copy_set('DEFAULT', 'CURRENT');
    return $name;
  } else { #if $name;
    my $return = $self->copy_set($name, 'CURRENT');
    $self->del_set($name);
    return $return
  }
}


sub load_set {
  my $self = shift;
  if (exists $DEF{$_[0]}) {
    $self->copy_set($_[0],'CURRENT') if exists $DEF{$_[0]};
    return $_[0]
  } else {
    return undef;
  }
}

sub save_set {
  my $self = shift;
  my $name;
  $name = $DEF{CURRENT}{NAME} unless $name = shift;
  ++$ID, $name = "\cI\cD:$ID" if not $name or $name eq 'RAND';
  return $self->copy_set('CURRENT', $name) && $name;
}

sub copy_set {
  shift;
  my ($name1, $name2) = @_;
  if ($DEF{$name2}{READONLY} or $name2 eq 'BLANK' or not exists $DEF{$name1}) {
    return 0;
  } else {
    $DEF{$name2} = {};                         # kill any links from use;
    $DEF{$name2}{TAGS} = [@{$DEF{$name1}{TAGS}}];
    $DEF{$name2}{DIGITS} = $DEF{$name1}{DIGITS};
    $DEF{$name2}{NAME} = $name2 unless $name2 eq 'CURRENT';
    $DEF{$name2}{NAME} = $name1   if   $name2 eq 'CURRENT';
    return 1;
  }
}

sub del_set {
  if (exists $DEF{$_[1]} and not $DEF{$_[1]}{SYSTEM}) {
    delete $DEF{$_[1]};
    return $_[1];
  }
}

# All of the modify methods are not meant to return anything, they modify
# the object being referenced too.

sub modify {
  # This method works almost like the new method except that it takes an
  # object as an argement and will modify it instead of creating a new
  # object, also any tags assosated with the object are left in tact
  # unless a new tag is given to override the old.

  my $me = shift;
  my $self;
  my @tags = @{$me->{'tags'}};
  $self = Math::Fraction->new(@_, @tags, @_);  # The extra @_ is their to override tags
  $me->{'frac'} = $self->{'frac'};
  $me->{'tags'} = $self->{'tags'};
}

sub modify_digits {
  my $self = shift;
  $DEF{CURRENT}{DIGITS} = shift;
}

sub modify_reduce {
  my $me = shift;
  my $self = $me->reduce;
  $me->{'frac'} = $self->{'frac'};
  $me->{'tags'} = $self->{'tags'};
}


sub modify_num {
  my $self = shift;
  $self->[0] = $_[0]
}

sub modify_den {
  my $self = shift;
  $self->[1] = $_[0]
}

sub modify_tag {
  my $self = shift;
  my ($return, @return);
  my $newtag;
 foreach $newtag (@_) {
  my $tagnum = &_tagnum($newtag);
  if ($tagnum == -1) {
    push @return, undef;
  } elsif (ref($self) eq "Math::Fraction") {
    my @frac = @{$self->{'frac'}};
    my @tags = @{$self->{'tags'}};
    my @newtags = &_tags(@tags,$newtag);
    # Now transform the Fraction based on the new tag.
    if ($tagnum == $SIZE) {
      my $newtag = &_tag($SIZE, \@newtags);
      @frac = map { "$_"+0 } @frac                if $newtag eq 'SMALL';
      @frac = map { Math::BigInt->new($_) } @frac if $newtag eq 'BIG';
    } elsif ($tagnum == $REDUCE) {
      (undef, @frac) = &_reduce(@frac) if &_tag($REDUCE, \@newtags) eq 'REDUCE';
    }
    # Finally Modify the Fraction
    $self->{'frac'} = \@frac; 
    $self->{'tags'} = \@newtags;
  } else {  
    $DEF{CURRENT}{TAGS}[$tagnum] = $newtag;
  }
  push @return, $newtag;
 }
 return @return;
}
    
# These methods are meant to be called with the overload operators.

sub add {
  my @frac1 = @{$_[0]->{'frac'}};
  my @tags1 = @{$_[0]->{'tags'}};
  my (@frac2, @frac, @tags2, $frac);
  my $skipauto = 0;
  @frac2 = @{$_[1]->{'frac'}}, @tags2 = @{$_[1]->{'tags'}} if ref($_[1]) eq "Math::Fraction";
  @frac2 = &_from_decimal($_[1]), $tags2[$INTERNAL] = 'CONVERTED' if ref($_[1]) ne "Math::Fraction";
  my @tags = &_tags_preserve([@tags1],[@tags2]);

 LOOP: {
  if (&_tag($REDUCE, \@tags) eq 'NO_REDUCE') {
    @frac = ($frac1[0]*$frac2[1]+$frac2[0]*$frac1[1],$frac1[1]*$frac2[1]);
  } else {
    # Taken from Knuth v2 (rev 2), p313.
    # It will always return a reduced fraction.
    my $gcd1 = &_gcd($frac1[1],$frac2[1]);
    my $tmp = $frac1[0]*($frac2[1]/$gcd1) + $frac2[0]*($frac1[1]/$gcd1);
    my $gcd2 = &_gcd($tmp,$gcd1);
    @frac = ( $tmp/$gcd2, ($frac1[1]/$gcd1)*($frac2[1]/$gcd2) );
    $tags[$RED_STATE] = 'IS_REDUCED';
  }
  if ( (&_tag($AUTO, \@tags) eq 'AUTO') and (not $skipauto) and
     ($tags[$SIZE] eq 'SMALL') and ($frac[0]=~/[eE]/ or $frac[1]=~/[eE]/) )
  {
    (@frac1[0..1], @frac2[0..1]) = map { Math::BigInt->new($_) } (@frac1, @frac2);
    $tags[$SIZE] = 'BIG';
    $skipauto = 1;
    redo LOOP;
  }
 }
  return Math::Fraction->new(@frac, @tags);
}

sub sub {
  my ($frac1, $frac2) = ($_[$_[2]], $_[not $_[2]]);  # swap if needed
  $frac1 = Math::Fraction->new($frac1, 'CONVERTED')  if ref($frac1) ne "Math::Fraction";
  $frac2 = Math::Fraction->new($frac2, 'CONVERTED')  if ref($frac2) ne "Math::Fraction";

  $frac2 = Math::Fraction->new($frac2->{'frac'}[0], -$frac2->{'frac'}[1], @{$frac2->{'tags'}});

  return $frac1 + $frac2;
}

sub mul {
  my @frac1 = @{$_[0]{'frac'}};
  my @tags1 = @{$_[0]{'tags'}};
  my (@frac2, @frac, @tags2);
  @frac2 = @{$_[1]->{'frac'}}, @tags2 = @{$_[1]->{'tags'}} if ref($_[1]) eq "Math::Fraction";
  @frac2 = (&_from_decimal($_[1])), $tags2[$INTERNAL] = 'CONVERTED' if ref($_[1]) ne "Math::Fraction";
  my @tags = &_tags_preserve([@tags1],[@tags2]);
  my $skipauto = 0;
 LOOP: {
  if (&_tag($REDUCE, \@tags) eq 'NO_REDUCE') {
    @frac = ($frac1[0]*$frac2[0],$frac1[1]*$frac2[1]);
  } else {
    my($gcd1, $gcd2)=(&_gcd($frac1[0],$frac2[1]),&_gcd($frac2[0],$frac1[1]));
    $frac[0] = ($frac1[0]/$gcd1)*($frac2[0]/$gcd2);
    $frac[1] = ($frac1[1]/$gcd2)*($frac2[1]/$gcd1);
    $tags[$RED_STATE] =  'IS_REDUCED';
  }
  if ( (&_tag($AUTO, \@tags) eq 'AUTO') and (not $skipauto) and
       ($tags[$SIZE] eq 'SMALL') and ($frac[0]=~/[eE]/ or $frac[1]=~/[eE]/) )
  {
    (@frac1[0..1], @frac2[0..1]) = map { Math::BigInt->new($_) } (@frac1, @frac2);
    $tags[$SIZE] = 'BIG';
    $skipauto = 1;
    redo LOOP;
  }
 }
  return Math::Fraction->new(@frac, @tags);
}

sub div {
  my ($frac1, $frac2) = ($_[$_[2]], $_[not $_[2]]);  # swap if needed
  $frac1 = Math::Fraction->new($frac1, 'CONVERTED')  if ref($frac1) ne "Math::Fraction";
  $frac2 = Math::Fraction->new($frac2, 'CONVERTED')  if ref($frac2) ne "Math::Fraction";

  $frac2 = Math::Fraction->new($frac2->{'frac'}[1], $frac2->{'frac'}[0], @{$frac2->{'tags'}});
      #Makes a copy of the fraction with the num and den switched.

  return $frac1 * $frac2;
}

sub pow {
  my (@frac, @frac1, @tags1);
  @frac1 = @{$_[$_[2]]->{'frac'}}, @tags1 = @{$_[$_[2]]->{'tags'}} if ref($_[$_[2]]) eq "Math::Fraction";
  @frac1 = &_from_decimal($_[$_[2]])                       if ref($_[$_[2]]) ne "Math::Fraction";
  my $frac2;
  $frac2 = $_[not $_[2]]->decimal        if ref($_[not $_[2]]) eq "Math::Fraction";
  $frac2 = $_[not $_[2]]                 if ref($_[not $_[2]]) ne "Math::Fraction";
  my @tags = @tags1;
  my $skipauto = 0;

 LOOP: { 
  @frac = ($frac1[0]**$frac2,$frac1[1]**$frac2);

  if ( (&_tag($AUTO, \@tags) eq 'AUTO') and (not $skipauto) and
     ($tags[$SIZE] eq 'SMALL') and ($frac[0]=~/[eE]/ or $frac[1]=~/[eE]/) )
  {
    @frac1 = map { Math::BigInt->new($_) } @frac1;
    $tags[$SIZE] = 'BIG';
    $skipauto = 1;
    redo LOOP;
  }
 }

  return Math::Fraction->new(@frac, @tags);
}

sub sqrt {
  my $self = shift;
  my @frac = @{$self->{'frac'}};
  my @tags = @{$self->{'tags'}};
  my $ans;
  if ( ref($frac[0]) ) {
    $frac[0] = Math::BigFloat->new( Math::BigFloat::fsqrt($frac[0], $DEF{CURRENT}{DIGITS}) );
    $frac[1] = Math::BigFloat->new( Math::BigFloat::fsqrt($frac[1], $DEF{CURRENT}{DIGITS}) );
  } else {
    @frac = (sqrt($frac[0]) , sqrt($frac[1]));
  }
  return Math::Fraction->new(@frac, @tags);
}


sub abs {
  my $self = shift;
  my @frac = @{$self->{'frac'}};
  my @tags = @{$self->{'tags'}};
  return Math::Fraction->new(abs($frac[0]),abs($frac[1]),@tags,'IS_REDUCED');
}

sub cmp {
  my @frac1 = @{$_[0]->{'frac'}};
  my @tags1 = @{$_[0]->{'tags'}};
  my (@frac2, @frac, @tags2, $x, $y);
  @frac2 = @{$_[1]->{'frac'}}, @tags2 = @{$_[1]->{'tags'}} if ref($_[1]) eq "Math::Fraction";
  @frac2 = &_from_decimal($_[1]), @tags2 = qw(CONVERTED)   if ref($_[1]) ne "Math::Fraction";
  my @tags = &_tags_preserve([@tags1],[@tags2]);
  if (&_tag($REDUCE, \@tags) == 'NO_REDUCE') {
    $x = $frac1[0]*$frac2[1];
    $y = $frac2[0]*$frac1[1];
  } else {
    my $gcd1 = &_gcd($frac1[1],$frac2[1]);
    $x = $frac1[0]*($frac2[1]/$gcd1);
    $y = $frac2[0]*($frac1[1]/$gcd1);
  }
  return $x <=> $y;
}

# These function are that functions and not ment to be used as methods

sub _fix_num {
  my $tagsref = shift;
  my @return = @_;
  my $auto = &_tag($AUTO, $tagsref) eq 'AUTO';
  $tagsref->[$SIZE] = &_tag($SIZE, $tagsref); 
  $tagsref->[$SIZE] = 'SMALL'  if $auto;
  my $num;
  my $decimal = 0;
  foreach $num (@return) {
    if (ref($num) eq "Math::BigFloat") {
      $tagsref->[$SIZE] = 'BIG' unless $auto;
      $decimal = 1;
    } elsif (ref($num) eq "Math::BigInt") {
      $tagsref->[$SIZE] = 'BIG' unless $auto;
    } elsif (ref($num)) {
      # do nothing
    } elsif ($num =~ /[\.\e\E]/) {
      $decimal = 1;
    }
    if ($auto) {
      $num =~ /[\+\-]?\s*0*([0-9]*)\s*\.?\s*([0-9]*)0*/;
      my $length = length($1)+length($2);
      $tagsref->[$SIZE] = 'BIG' if $length > 15;
    }
  }
  if ($tagsref->[$SIZE] eq 'BIG') {
    @return = map {Math::BigInt->new("$_")}   @return  if not $decimal;
    @return = map {Math::BigFloat->new("$_")} @return  if     $decimal;
  }
  if ($tagsref->[$SIZE] eq 'SMALL' and $auto) {
    @return = map {"$_"+0} @return;
  }
  return ($decimal, @return);
}

sub _fix_auto {
  my $direction = shift;
  my $tagsref = shift;
  my @return = @_;
  $tagsref->[$SIZE] = 'SMALL';
  my $num;
  foreach $num (@return) {
    $num =~ /[\+\-]?\s*0*([0-9]*)\s*\.?\s*([0-9]*)0*/;
    my $length = length($1)+length($2);
    $tagsref->[$SIZE] = 'BIG' if $length > 15;
  }
  if ($tagsref->[$SIZE] eq 'BIG' and $direction eq 'BOTH') {
    @return = map {Math::BigInt->new("$_")} @return;
  } elsif ($tagsref->[$SIZE] eq 'SMALL') {
    @return = map {"$_"+0} @return;
  }
  return (@return);
}

sub _is_decimal {
  my $return = $_[0] =~ /^\s*[\+\-0-9eE\.]+\s*$/;
  return $return;
}

sub _reduce {
  my @frac = @_;
  my $gcd = &_gcd(@frac);
  if ($gcd == 1 ) {
    return (0, @frac)
  } else {
    return (1, $frac[0]/$gcd, $frac[1]/$gcd);
  }
}  

sub _simplify_sign {
  my @frac = @_;
  my $sign = 1;
  $sign = ($frac[0]/abs($frac[0]))*($frac[1]/abs($frac[1])) if $frac[0];
  @frac = ($sign*abs($frac[0]), abs($frac[1]) );
  return @frac;
}

sub _tags {
  my @return = (undef, undef);
  my ($NUM, $VALUE) = (0, 1);

  foreach (@_) {
    next if not $TAGS{$_};
    my ($num, $value) = @{$TAGS{$_}};
    $return[$num] = $value;
  }

  return @return;
}


sub _tag {
  my $item = shift;
  my $return;
  my $ref;
  foreach $ref (@_, $DEF{CURRENT}{TAGS}) {
    last if $return = ${$ref}[$item];
  }
  return $return
}

sub _tagnum {
  my $item = shift;
  if (exists $TAGS{$item}) {
    return $TAGS{$item}[0];
  } else {
    return -1;
  }
}

sub _tags_preserve {
  my @tags1 = @{$_[0]};
  my @tags2 = @{$_[1]};
  my @tags;
  if ($tags1[$INTERNAL] eq 'CONVERTED') {
    @tags = @tags2;
  } elsif ($tags2[$INTERNAL] eq 'CONVERTED') {
    @tags = @tags1;
  } else {
    @tags = map {$tags1[$_] eq $tags2[$_] and $tags1[$_]} (0 .. $#tags1) ;
  }
  return @tags;
}

sub _gcd {
  # Using Euclid's method found in Knuth v2 (rev 2) p320 brought to my
  # attention from the BigInt module

  my ($x, $y) = (abs($_[0]), abs($_[1]));
  if ( ref($x) ) {
    $x = Math::BigInt->new( $x->bgcd($y) );
  } else {
    {
      $x=1, last if $y > 1e17; # If this is so % will thinks its a zero so if
                               # $y>1e17 will simply will basicly give up and
                               # have it return 1 as the GCD.
      my ($x0);
      while ($y != 0) {
        $x0 = $x;
        ($x, $y) = ($y, $x % $y);
        # Note $x0 = $x, $x = $y, $y= $x % $y   Before the Swith
        $x=1, last  if ($x0>99999999 or $x>999999999) and int($x0/$x)*$x+$y != $x0;
        # This is to see if the mod operater through up on us when dealing with
        # large numbers.  If it did set the gcd = 1 and quit.
      }
    }
  }
  return $x;
}

sub _de_decimal {
    my @frac = @_;
    my @return;
    my $big = &_tag($SIZE, $_[2]);
    my (@int_part, @decimal_part);
    if ($big eq "BIG") {
      my @digits = (1,1);
      ($int_part[0], $digits[0]) = $frac[0]->fnorm =~ /(\d+)E\-(\d+)/;
      ($int_part[1], $digits[1]) = $frac[1]->fnorm =~ /(\d+)E\-(\d+)/;
      @digits = sort {$a <=> $b} @digits;
      my $factor = 10**$digits[1];
      @frac = (($_[0]*$factor),($_[1]*$factor));
      chop $frac[0]; chop $frac[1];
      @frac = (Math::BigInt->new($frac[0]), Math::BigInt->new($frac[1]) );
   } else {
      ($int_part[0], $decimal_part[0]) = $frac[0] =~ /(\d+)\.(\d+)/;
      ($int_part[1], $decimal_part[1]) = $frac[1] =~ /(\d+)\.(\d+)/;
      @decimal_part = sort {$a <=> $b} (length($decimal_part[0]),length($decimal_part[1]) );
      my $factor = 10**$decimal_part[1];
      @frac = ($_[0]*$factor, $_[1]*$factor);
   }
   return @frac;
}

sub _from_decimal {
  my $decimal = shift;       # the decimal (1.312671267127)
  my $big = 'BIG' if ref($decimal);
  my ($repeat);              # flag to keep track if it is repeating or not
  my ($sign);
  my ($factor, $int_factor);
  my ($factor2);
  my ($whole_num, $whole_num_len);
  my ($int_part);                        # integer part (1)
  my ($decimal_part, $decimal_part_len); # decimal part (312671267127)
  my ($decimal_part2);               # decimal part - last bit \/ (312671267)
  my ($pat, $pat_len);               # repeating pat (1267)
  my ($pat_lastb);                   # last bit of repeating pat (127)
  my ($beg_part, $beg_part_len);       # non-repeating part (3)
  my ($other_part, $other_part_len);   # repeating part     (1267126712127)
  my ($frac1, $frac2, $frac3);

  my $rnd_mode = $Math::BigFloat::rnd_mode;  # to avoid problems with incon.
  $Math::BigFloat::rnd_mode = 'trunc';       # rounding

  $decimal = "$decimal";
  $decimal =~ s/\s//g;
  ($sign, $int_part, $decimal_part) = $decimal =~ /([\+\-]?)\s*(\d*)\.(\d+)$/;
  $sign .= '1';
  $decimal_part_len = length($decimal_part);
  $int_part = "" unless $int_part;
  $factor = '1'.'0'x(length($decimal_part));
  $factor = Math::BigFloat->new($factor) if $big;
     # Make it a BigFloat now to simplfy latter
  $int_factor = '1'.'0'x(length($int_part));
  $beg_part_len = 0;
 OuterBlock:
  while ($beg_part_len < $decimal_part_len) {
    $beg_part = substr($decimal_part, 0, $beg_part_len);
    $other_part = substr($decimal_part, $beg_part_len);
    $other_part_len = length($other_part);
    my $i;
    for ($i = 1; $i < ($other_part_len/2+1); $i++) {
      $pat = substr($other_part, 0, $i);
      $pat_len = $i;
      local $_ = $other_part;
      $repeat = undef;
      while (1) {
        ($_) = /^$pat(.*)/;
        my $length = length($_);

        if ( $length <= $pat_len) {
          last unless $length;
          $pat_lastb = substr($pat, 0, $length);
          $repeat=1 ,last OuterBlock if $pat_lastb eq $_;
          if ($pat_lastb eq $_ - 1) {
             # this is needed to see if it really is the repeating fracton
             # we intented it to be.  If we don't do this 1.1212 would become
             # 1120/999 = 1.1211211211.
             # The first three lines converts it to a fraction and the
             # rests tests it to the actual repeating decimal/
             # The NO_REDUCE flag is their to save time as reducing large
             # fraction can take a bit of time which is unnecessary as we will
             # be converting it to a decimal.
            $decimal_part2 = substr($decimal_part, 0, $decimal_part_len - length($pat_lastb));
            $factor2 = '1'.'0'x(length($decimal_part2));
            $frac1 = Math::Fraction->new('0'.$beg_part,"1"."0"x$beg_part_len, 'NO_REDUCE', $big);
            $frac2 = Math::Fraction->new('0'.$pat,"9"x$pat_len."0"x$beg_part_len, 'NO_REDUCE', $big);
            $frac3 = $frac1 + $frac2;
            my $what_i_get = $frac3->decimal;
            my $places = length($what_i_get);
            my $decimal_p_tmp = $decimal_part2                      if not $big;
               $decimal_p_tmp = Math::BigFloat->new($decimal_part2) if  $big;
            my $what_i_should_get = (($decimal_p_tmp)/$factor2)."$pat"x($places);
              # The rest of this is doing nothing more but trying to compare
              # the what_i_get and what_i_should_get but becuse the stupid
              # BigFloat module is so pragmentic all this hopla is nessary
            $what_i_should_get = Math::BigFloat->new($what_i_should_get)           if $big;
            $what_i_should_get = $what_i_should_get->fround(length($what_i_get)-1) if $big;
            $what_i_should_get = Math::BigFloat->new($what_i_should_get)           if $big;
              # ^^ Needed because the dam fround method does not return a
              #    BigFloat object!!!!!!
            my $pass = "$what_i_get" eq "$what_i_should_get" if $big;
               $pass = $what_i_get == $what_i_should_get  if  not $big;
            $repeat=1, last OuterBlock if ($pass);
          }
        }
      }
    }
    $beg_part_len++;
  }
  if ($repeat) {
    $frac1 = Math::Fraction->new('0'.$beg_part,"1"."0"x$beg_part_len, $big);
    $frac2 = Math::Fraction->new('0'.$pat,"9"x$pat_len."0"x$beg_part_len, $big);
    $int_part = Math::Fraction->new('0'.$int_part, 1, 'BIG') if $big;
    $frac3 = $sign*($int_part + $frac1 + $frac2);
    return @{$frac3->{'frac'}};
  } else {
    return ($decimal*$factor, $factor, $big);
  }
  $Math::BigFloat::rnd_mode = $rnd_mode;   # set it back to what it was.
}

1;

__END__