| /* Decimal number arithmetic module header for the decNumber C Library. |
| Copyright (C) 2005, 2007 Free Software Foundation, Inc. |
| Contributed by IBM Corporation. Author Mike Cowlishaw. |
| |
| This file is part of GCC. |
| |
| GCC is free software; you can redistribute it and/or modify it under |
| the terms of the GNU General Public License as published by the Free |
| Software Foundation; either version 2, or (at your option) any later |
| version. |
| |
| In addition to the permissions in the GNU General Public License, |
| the Free Software Foundation gives you unlimited permission to link |
| the compiled version of this file into combinations with other |
| programs, and to distribute those combinations without any |
| restriction coming from the use of this file. (The General Public |
| License restrictions do apply in other respects; for example, they |
| cover modification of the file, and distribution when not linked |
| into a combine executable.) |
| |
| GCC is distributed in the hope that it will be useful, but WITHOUT ANY |
| WARRANTY; without even the implied warranty of MERCHANTABILITY or |
| FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License |
| for more details. |
| |
| You should have received a copy of the GNU General Public License |
| along with GCC; see the file COPYING. If not, write to the Free |
| Software Foundation, 51 Franklin Street, Fifth Floor, Boston, MA |
| 02110-1301, USA. */ |
| |
| /* ------------------------------------------------------------------ */ |
| /* Decimal Number arithmetic module header */ |
| /* ------------------------------------------------------------------ */ |
| |
| #ifndef DECNUMBER_H |
| #define DECNUMBER_H |
| |
| #define DECNAME "decNumber" /* Short name */ |
| #define DECFULLNAME "Decimal Number Module" /* Verbose name */ |
| #define DECAUTHOR "Mike Cowlishaw" /* Who to blame */ |
| |
| #include "libdecnumber/decContext.h" |
| |
| /* Bit settings for decNumber.bits */ |
| #define DECNEG 0x80 /* Sign; 1=negative, 0=positive or zero */ |
| #define DECINF 0x40 /* 1=Infinity */ |
| #define DECNAN 0x20 /* 1=NaN */ |
| #define DECSNAN 0x10 /* 1=sNaN */ |
| /* The remaining bits are reserved; they must be 0 */ |
| #define DECSPECIAL (DECINF|DECNAN|DECSNAN) /* any special value */ |
| |
| /* Define the decNumber data structure. The size and shape of the */ |
| /* units array in the structure is determined by the following */ |
| /* constant. This must not be changed without recompiling the */ |
| /* decNumber library modules. */ |
| |
| #define DECDPUN 3 /* DECimal Digits Per UNit [must be >0 */ |
| /* and <10; 3 or powers of 2 are best]. */ |
| |
| /* DECNUMDIGITS is the default number of digits that can be held in */ |
| /* the structure. If undefined, 1 is assumed and it is assumed */ |
| /* that the structure will be immediately followed by extra space, */ |
| /* as required. DECNUMDIGITS is always >0. */ |
| #if !defined(DECNUMDIGITS) |
| #define DECNUMDIGITS 1 |
| #endif |
| |
| /* The size (integer data type) of each unit is determined by the */ |
| /* number of digits it will hold. */ |
| #if DECDPUN<=2 |
| #define decNumberUnit uint8_t |
| #elif DECDPUN<=4 |
| #define decNumberUnit uint16_t |
| #else |
| #define decNumberUnit uint32_t |
| #endif |
| /* The number of units needed is ceil(DECNUMDIGITS/DECDPUN) */ |
| #define DECNUMUNITS ((DECNUMDIGITS+DECDPUN-1)/DECDPUN) |
| |
| /* The data structure... */ |
| typedef struct { |
| int32_t digits; /* Count of digits in the coefficient; >0 */ |
| int32_t exponent; /* Unadjusted exponent, unbiased, in */ |
| /* range: -1999999997 through 999999999 */ |
| uint8_t bits; /* Indicator bits (see above) */ |
| /* Coefficient, from least significant unit */ |
| decNumberUnit lsu[DECNUMUNITS]; |
| } decNumber; |
| |
| /* Notes: */ |
| /* 1. If digits is > DECDPUN then there will one or more */ |
| /* decNumberUnits immediately following the first element of lsu.*/ |
| /* These contain the remaining (more significant) digits of the */ |
| /* number, and may be in the lsu array, or may be guaranteed by */ |
| /* some other mechanism (such as being contained in another */ |
| /* structure, or being overlaid on dynamically allocated */ |
| /* storage). */ |
| /* */ |
| /* Each integer of the coefficient (except potentially the last) */ |
| /* contains DECDPUN digits (e.g., a value in the range 0 through */ |
| /* 99999999 if DECDPUN is 8, or 0 through 999 if DECDPUN is 3). */ |
| /* */ |
| /* 2. A decNumber converted to a string may need up to digits+14 */ |
| /* characters. The worst cases (non-exponential and exponential */ |
| /* formats) are -0.00000{9...}# and -9.{9...}E+999999999# */ |
| /* (where # is '\0') */ |
| |
| |
| /* ---------------------------------------------------------------- */ |
| /* decNumber public functions and macros */ |
| /* ---------------------------------------------------------------- */ |
| |
| |
| /* Conversions */ |
| decNumber * decNumberFromInt32(decNumber *, int32_t); |
| decNumber * decNumberFromUInt32(decNumber *, uint32_t); |
| decNumber *decNumberFromInt64(decNumber *, int64_t); |
| decNumber *decNumberFromUInt64(decNumber *, uint64_t); |
| decNumber * decNumberFromString(decNumber *, const char *, decContext *); |
| char * decNumberToString(const decNumber *, char *); |
| char * decNumberToEngString(const decNumber *, char *); |
| uint32_t decNumberToUInt32(const decNumber *, decContext *); |
| int32_t decNumberToInt32(const decNumber *, decContext *); |
| int64_t decNumberIntegralToInt64(const decNumber *dn, decContext *set); |
| uint8_t * decNumberGetBCD(const decNumber *, uint8_t *); |
| decNumber * decNumberSetBCD(decNumber *, const uint8_t *, uint32_t); |
| |
| /* Operators and elementary functions */ |
| decNumber * decNumberAbs(decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberAdd(decNumber *, const decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberAnd(decNumber *, const decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberCompare(decNumber *, const decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberCompareSignal(decNumber *, const decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberCompareTotal(decNumber *, const decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberCompareTotalMag(decNumber *, const decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberDivide(decNumber *, const decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberDivideInteger(decNumber *, const decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberExp(decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberFMA(decNumber *, const decNumber *, const decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberInvert(decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberLn(decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberLogB(decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberLog10(decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberMax(decNumber *, const decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberMaxMag(decNumber *, const decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberMin(decNumber *, const decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberMinMag(decNumber *, const decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberMinus(decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberMultiply(decNumber *, const decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberNormalize(decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberOr(decNumber *, const decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberPlus(decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberPower(decNumber *, const decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberQuantize(decNumber *, const decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberReduce(decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberRemainder(decNumber *, const decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberRemainderNear(decNumber *, const decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberRescale(decNumber *, const decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberRotate(decNumber *, const decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberSameQuantum(decNumber *, const decNumber *, const decNumber *); |
| decNumber * decNumberScaleB(decNumber *, const decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberShift(decNumber *, const decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberSquareRoot(decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberSubtract(decNumber *, const decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberToIntegralExact(decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberToIntegralValue(decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberXor(decNumber *, const decNumber *, const decNumber *, decContext *); |
| |
| /* Utilities */ |
| enum decClass decNumberClass(const decNumber *, decContext *); |
| const char * decNumberClassToString(enum decClass); |
| decNumber * decNumberCopy(decNumber *, const decNumber *); |
| decNumber * decNumberCopyAbs(decNumber *, const decNumber *); |
| decNumber * decNumberCopyNegate(decNumber *, const decNumber *); |
| decNumber * decNumberCopySign(decNumber *, const decNumber *, const decNumber *); |
| decNumber * decNumberNextMinus(decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberNextPlus(decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberNextToward(decNumber *, const decNumber *, const decNumber *, decContext *); |
| decNumber * decNumberTrim(decNumber *); |
| const char * decNumberVersion(void); |
| decNumber * decNumberZero(decNumber *); |
| |
| /* Functions for testing decNumbers (normality depends on context) */ |
| int32_t decNumberIsNormal(const decNumber *, decContext *); |
| int32_t decNumberIsSubnormal(const decNumber *, decContext *); |
| |
| /* Macros for testing decNumber *dn */ |
| #define decNumberIsCanonical(dn) (1) /* All decNumbers are saintly */ |
| #define decNumberIsFinite(dn) (((dn)->bits&DECSPECIAL)==0) |
| #define decNumberIsInfinite(dn) (((dn)->bits&DECINF)!=0) |
| #define decNumberIsNaN(dn) (((dn)->bits&(DECNAN|DECSNAN))!=0) |
| #define decNumberIsNegative(dn) (((dn)->bits&DECNEG)!=0) |
| #define decNumberIsQNaN(dn) (((dn)->bits&(DECNAN))!=0) |
| #define decNumberIsSNaN(dn) (((dn)->bits&(DECSNAN))!=0) |
| #define decNumberIsSpecial(dn) (((dn)->bits&DECSPECIAL)!=0) |
| #define decNumberIsZero(dn) (*(dn)->lsu==0 \ |
| && (dn)->digits==1 \ |
| && (((dn)->bits&DECSPECIAL)==0)) |
| #define decNumberRadix(dn) (10) |
| |
| #endif |