gcc/ada/urealp.ads

   1 ------------------------------------------------------------------------------
   2 --                                                                          --
   3 --                         GNAT COMPILER COMPONENTS                         --
   4 --                                                                          --
   5 --                               U R E A L P                                --
   6 --                                                                          --
   7 --                                 S p e c                                  --
   8 --                                                                          --
   9 --          Copyright (C) 1992-2009  Free Software Foundation, Inc.         --
  10 --                                                                          --
  11 -- GNAT is free software;  you can  redistribute it  and/or modify it under --
  12 -- terms of the  GNU General Public License as published  by the Free Soft- --
  13 -- ware  Foundation;  either version 3,  or (at your option) any later ver- --
  14 -- sion.  GNAT is distributed in the hope that it will be useful, but WITH- --
  15 -- OUT ANY WARRANTY;  without even the  implied warranty of MERCHANTABILITY --
  16 -- or FITNESS FOR A PARTICULAR PURPOSE.                                     --
  17 --                                                                          --
  18 -- As a special exception under Section 7 of GPL version 3, you are granted --
  19 -- additional permissions described in the GCC Runtime Library Exception,   --
  20 -- version 3.1, as published by the Free Software Foundation.               --
  21 --                                                                          --
  22 -- You should have received a copy of the GNU General Public License and    --
  23 -- a copy of the GCC Runtime Library Exception along with this program;     --
  24 -- see the files COPYING3 and COPYING.RUNTIME respectively.  If not, see    --
  25 -- <http://www.gnu.org/licenses/>.                                          --
  26 --                                                                          --
  27 -- GNAT was originally developed  by the GNAT team at  New York University. --
  28 -- Extensive contributions were provided by Ada Core Technologies Inc.      --
  29 --                                                                          --
  30 ------------------------------------------------------------------------------
  31
  32 --  Support for universal real arithmetic
  33
  34 with Types; use Types;
  35 with Uintp; use Uintp;
  36
  37 package Urealp is
  38
  39    ---------------------------------------
  40    -- Representation of Universal Reals --
  41    ---------------------------------------
  42
  43    --  A universal real value is represented by a single value (which is
  44    --  an index into an internal table). These values are not hashed, so
  45    --  the equality operator should not be used on Ureal values (instead
  46    --  use the UR_Eq function).
  47
  48    --  A Ureal value represents an arbitrary precision universal real value,
  49    --  stored internally using four components
  50
  51    --    the numerator (Uint, always non-negative)
  52    --    the denominator (Uint, always non-zero, always positive if base = 0)
  53    --    a real base (Nat, either zero, or in the range 2 .. 16)
  54    --    a sign flag (Boolean), set if negative
  55
  56    --  If the base is zero, then the absolute value of the Ureal is simply
  57    --  numerator/denominator. If the base is non-zero, then the absolute
  58    --  value is num / (rbase ** den).
  59
  60    --  Negative numbers are represented by the sign of the numerator being
  61    --  negative. The denominator is always positive.
  62
  63    --  A normalized Ureal value has base = 0, and numerator/denominator
  64    --  reduced to lowest terms, with zero itself being represented as 0/1.
  65    --  This is a canonical format, so that for normalized Ureal values it
  66    --  is the case that two equal values always have the same denominator
  67    --  and numerator values.
  68
  69    --  Note: a value of minus zero is legitimate, and the operations in
  70    --  Urealp preserve the handling of signed zeroes in accordance with
  71    --  the rules of IEEE P754 ("IEEE floating point").
  72
  73    ------------------------------
  74    -- Types for Urealp Package --
  75    ------------------------------
  76
  77    type Ureal is private;
  78    --  Type used for representation of universal reals
  79
  80    No_Ureal : constant Ureal;
  81    --  Constant used to indicate missing or unset Ureal value
  82
  83    ---------------------
  84    -- Ureal Constants --
  85    ---------------------
  86
  87    function Ureal_0 return Ureal;
  88    --  Returns value 0.0
  89
  90    function Ureal_M_0 return Ureal;
  91    --  Returns value -0.0
  92
  93    function Ureal_Tenth return Ureal;
  94    --  Returns value 0.1
  95
  96    function Ureal_Half return Ureal;
  97    --  Returns value 0.5
  98
  99    function Ureal_1 return Ureal;
 100    --  Returns value 1.0
 101
 102    function Ureal_2 return Ureal;
 103    --  Returns value 2.0
 104
 105    function Ureal_10 return Ureal;
 106    --  Returns value 10.0
 107
 108    function Ureal_100 return Ureal;
 109    --  Returns value 100.0
 110
 111    function Ureal_2_80 return Ureal;
 112    --  Returns value 2.0 ** 80
 113
 114    function Ureal_2_M_80 return Ureal;
 115    --  Returns value 2.0 ** (-80)
 116
 117    function Ureal_2_128 return Ureal;
 118    --  Returns value 2.0 ** 128
 119
 120    function Ureal_2_M_128 return Ureal;
 121    --  Returns value 2.0 ** (-128)
 122
 123    function Ureal_10_36 return Ureal;
 124    --  Returns value 10.0 ** 36
 125
 126    function Ureal_M_10_36 return Ureal;
 127    --  Returns value -(10.0
 128
 129    -----------------
 130    -- Subprograms --
 131    -----------------
 132
 133    procedure Initialize;
 134    --  Initialize Ureal tables. Note that Initialize must not be called if
 135    --  Tree_Read is used. Note also that there is no Lock routine in this
 136    --  unit. These tables are among the few tables that can be expanded
 137    --  during Gigi processing.
 138
 139    procedure Tree_Read;
 140    --  Initializes internal tables from current tree file using the relevant
 141    --  Table.Tree_Read routines. Note that Initialize should not be called if
 142    --  Tree_Read is used. Tree_Read includes all necessary initialization.
 143
 144    procedure Tree_Write;
 145    --  Writes out internal tables to current tree file using the relevant
 146    --  Table.Tree_Write routines.
 147
 148    function Rbase (Real : Ureal) return Nat;
 149    --  Return the base of the universal real
 150
 151    function Denominator (Real : Ureal) return Uint;
 152    --  Return the denominator of the universal real
 153
 154    function Numerator (Real : Ureal) return Uint;
 155    --  Return the numerator of the universal real
 156
 157    function Norm_Den (Real : Ureal) return Uint;
 158    --  Return the denominator of the universal real after a normalization
 159
 160    function Norm_Num (Real : Ureal) return Uint;
 161    --  Return the numerator of the universal real after a normalization
 162
 163    function UR_From_Uint (UI : Uint) return Ureal;
 164    --  Returns real corresponding to universal integer value
 165
 166    function UR_To_Uint (Real : Ureal) return Uint;
 167    --  Return integer value obtained by accurate rounding of real value.
 168    --  The rounding of values half way between two integers is away from
 169    --  zero, as required by normal Ada 95 rounding semantics.
 170
 171    function UR_Trunc (Real : Ureal) return Uint;
 172    --  Return integer value obtained by a truncation of real towards zero
 173
 174    function UR_Ceiling (Real : Ureal) return Uint;
 175    --  Return value of smallest integer not less than the given value
 176
 177    function UR_Floor (Real : Ureal) return Uint;
 178    --  Return value of smallest integer not greater than the given value
 179
 180    --  Conversion table for above four functions
 181
 182    --    Input    To_Uint    Trunc    Ceiling    Floor
 183    --     1.0        1         1         1         1
 184    --     1.2        1         1         2         1
 185    --     1.5        2         1         2         1
 186    --     1.7        2         1         2         1
 187    --     2.0        2         2         2         2
 188    --    -1.0       -1        -1        -1        -1
 189    --    -1.2       -1        -1        -1        -2
 190    --    -1.5       -2        -1        -1        -2
 191    --    -1.7       -2        -1        -1        -2
 192    --    -2.0       -2        -2        -2        -2
 193
 194    function UR_From_Components
 195      (Num      : Uint;
 196       Den      : Uint;
 197       Rbase    : Nat := 0;
 198       Negative : Boolean := False)
 199       return     Ureal;
 200    --  Builds real value from given numerator, denominator and base. The
 201    --  value is negative if Negative is set to true, and otherwise is
 202    --  non-negative.
 203
 204    function UR_Add (Left : Ureal; Right : Ureal) return Ureal;
 205    function UR_Add (Left : Ureal; Right : Uint)  return Ureal;
 206    function UR_Add (Left : Uint;  Right : Ureal) return Ureal;
 207    --  Returns real sum of operands
 208
 209    function UR_Div (Left : Ureal; Right : Ureal) return Ureal;
 210    function UR_Div (Left : Uint;  Right : Ureal) return Ureal;
 211    function UR_Div (Left : Ureal; Right : Uint)  return Ureal;
 212    --  Returns real quotient of operands. Fatal error if Right is zero
 213
 214    function UR_Mul (Left : Ureal; Right : Ureal) return Ureal;
 215    function UR_Mul (Left : Uint;  Right : Ureal) return Ureal;
 216    function UR_Mul (Left : Ureal; Right : Uint)  return Ureal;
 217    --  Returns real product of operands
 218
 219    function UR_Sub (Left : Ureal; Right : Ureal) return Ureal;
 220    function UR_Sub (Left : Uint;  Right : Ureal) return Ureal;
 221    function UR_Sub (Left : Ureal; Right : Uint)  return Ureal;
 222    --  Returns real difference of operands
 223
 224    function UR_Exponentiate (Real  : Ureal; N : Uint) return  Ureal;
 225    --  Returns result of raising Ureal to Uint power.
 226    --  Fatal error if Left is 0 and Right is negative.
 227
 228    function UR_Abs (Real : Ureal) return Ureal;
 229    --  Returns abs function of real
 230
 231    function UR_Negate (Real : Ureal) return Ureal;
 232    --  Returns negative of real
 233
 234    function UR_Eq (Left, Right : Ureal) return Boolean;
 235    --  Compares reals for equality
 236
 237    function UR_Max (Left, Right : Ureal) return Ureal;
 238    --  Returns the maximum of two reals
 239
 240    function UR_Min (Left, Right : Ureal) return Ureal;
 241    --  Returns the minimum of two reals
 242
 243    function UR_Ne (Left, Right : Ureal) return Boolean;
 244    --  Compares reals for inequality
 245
 246    function UR_Lt (Left, Right : Ureal) return Boolean;
 247    --  Compares reals for less than
 248
 249    function UR_Le (Left, Right : Ureal) return Boolean;
 250    --  Compares reals for less than or equal
 251
 252    function UR_Gt (Left, Right : Ureal) return Boolean;
 253    --  Compares reals for greater than
 254
 255    function UR_Ge (Left, Right : Ureal) return Boolean;
 256    --  Compares reals for greater than or equal
 257
 258    function UR_Is_Zero (Real : Ureal) return Boolean;
 259    --  Tests if real value is zero
 260
 261    function UR_Is_Negative (Real : Ureal) return Boolean;
 262    --  Tests if real value is negative, note that negative zero gives true
 263
 264    function UR_Is_Positive (Real : Ureal) return Boolean;
 265    --  Test if real value is greater than zero
 266
 267    procedure UR_Write (Real : Ureal);
 268    --  Writes value of Real to standard output. Used only for debugging and
 269    --  tree/source output. If the result is easily representable as a standard
 270    --  Ada literal, it will be given that way, but as a result of evaluation
 271    --  of static expressions, it is possible to generate constants (e.g. 1/13)
 272    --  which have no such representation. In such cases (and in cases where it
 273    --  is too much work to figure out the Ada literal), the string that is
 274    --  output is of the form [numerator/denominator].
 275
 276    procedure pr (Real : Ureal);
 277    pragma Export (Ada, pr);
 278    --  Writes value of Real to standard output with a terminating line return,
 279    --  using UR_Write as described above. This is for use from the debugger.
 280
 281    ------------------------
 282    -- Operator Renamings --
 283    ------------------------
 284
 285    function "+" (Left : Ureal; Right : Ureal) return Ureal renames UR_Add;
 286    function "+" (Left : Uint;  Right : Ureal) return Ureal renames UR_Add;
 287    function "+" (Left : Ureal; Right : Uint)  return Ureal renames UR_Add;
 288
 289    function "/" (Left : Ureal; Right : Ureal) return Ureal renames UR_Div;
 290    function "/" (Left : Uint;  Right : Ureal) return Ureal renames UR_Div;
 291    function "/" (Left : Ureal; Right : Uint)  return Ureal renames UR_Div;
 292
 293    function "*" (Left : Ureal; Right : Ureal) return Ureal renames UR_Mul;
 294    function "*" (Left : Uint;  Right : Ureal) return Ureal renames UR_Mul;
 295    function "*" (Left : Ureal; Right : Uint)  return Ureal renames UR_Mul;
 296
 297    function "-" (Left : Ureal; Right : Ureal) return Ureal renames UR_Sub;
 298    function "-" (Left : Uint;  Right : Ureal) return Ureal renames UR_Sub;
 299    function "-" (Left : Ureal; Right : Uint)  return Ureal renames UR_Sub;
 300
 301    function "**"  (Real  : Ureal; N : Uint) return Ureal
 302                                                      renames UR_Exponentiate;
 303
 304    function "abs" (Real : Ureal) return Ureal renames UR_Abs;
 305
 306    function "-"   (Real : Ureal) return Ureal renames UR_Negate;
 307
 308    function "="   (Left, Right : Ureal) return Boolean renames UR_Eq;
 309
 310    function "<"   (Left, Right : Ureal) return Boolean renames UR_Lt;
 311
 312    function "<="  (Left, Right : Ureal) return Boolean renames UR_Le;
 313
 314    function ">="  (Left, Right : Ureal) return Boolean renames UR_Ge;
 315
 316    function ">"   (Left, Right : Ureal) return Boolean renames UR_Gt;
 317
 318    -----------------------------
 319    -- Mark/Release Processing --
 320    -----------------------------
 321
 322    --  The space used by Ureal data is not automatically reclaimed. However,
 323    --  a mark-release regime is implemented which allows storage to be
 324    --  released back to a previously noted mark. This is used for example
 325    --  when doing comparisons, where only intermediate results get stored
 326    --  that do not need to be saved for future use.
 327
 328    type Save_Mark is private;
 329
 330    function Mark return Save_Mark;
 331    --  Note mark point for future release
 332
 333    procedure Release (M : Save_Mark);
 334    --  Release storage allocated since mark was noted
 335
 336    ------------------------------------
 337    -- Representation of Ureal Values --
 338    ------------------------------------
 339
 340 private
 341
 342    type Ureal is new Int range Ureal_Low_Bound .. Ureal_High_Bound;
 343    for Ureal'Size use 32;
 344
 345    No_Ureal : constant Ureal := Ureal'First;
 346
 347    type Save_Mark is new Int;
 348
 349    pragma Inline (Denominator);
 350    pragma Inline (Mark);
 351    pragma Inline (Norm_Num);
 352    pragma Inline (Norm_Den);
 353    pragma Inline (Numerator);
 354    pragma Inline (Rbase);
 355    pragma Inline (Release);
 356    pragma Inline (Ureal_0);
 357    pragma Inline (Ureal_M_0);
 358    pragma Inline (Ureal_Tenth);
 359    pragma Inline (Ureal_Half);
 360    pragma Inline (Ureal_1);
 361    pragma Inline (Ureal_2);
 362    pragma Inline (Ureal_10);
 363    pragma Inline (UR_From_Components);
 364
 365 end Urealp;