* ************************************************ * ** Math Lib C Interface - ASCII Conversion ** * ** ** * ************************************************ *********************************************************************** ** * ** Copyright 1984, Amiga Computer Inc. All rights reserved. * ** No part of this program may be reproduced, transmitted, * ** transcribed, stored in retrieval system, or translated into * ** any language or computer language, in any form or by any * ** means, electronic, mechanical, magnetic, optical, chemical, * ** manual or otherwise, without the prior written permission of * ** Amiga Computer Incorporated, 3350 Scott Blvd, Bld #7, * ** Santa Clara, CA 95051 * ** * *********************************************************************** xdef _afp,_fpa,_dbf,_arnd * C entries for conversion math functions xref FFPAFP,FFPFPA,FFPDBF,FFPARND ******* amiga.lib/afp ************************************************** * * NAME * afp - Convert ASCII string variable into fast floating point * * SYNOPSIS * ffp_value = afp(string); * * FUNCTION * Accepts the address of the ASCII string in C format that is * converted into an FFP floating point number. * * The string is expected in this Format: * {S}{digits}{'.'}{digits}{'E'}{S}{digits} * <*******MANTISSA*******><***EXPONENT***> * * * Syntax rules: * Both signs are optional and are '+' or '-'. The mantissa must be * present. The exponent need not be present. The mantissa may lead * with a decimal point. The mantissa need not have a decimal point. * Examples: All of these values represent the number fourty-two. * 42 .042e3 * 42. +.042e+03 * +42. 0.000042e6 * 0000042.00 420000e-4 * 420000.00e-0004 * * Floating point range: * Fast floating point supports the value zero and non-zero values * within the following bounds - * 18 20 * 9.22337177 x 10 > +number > 5.42101070 x 10 * 18 -20 * -9.22337177 x 10 > -number > -2.71050535 x 10 * * Precision: * This conversion results in a 24 bit precision with guaranteed * error less than or equal to one-half least significant bit. * * INPUTS * string - Pointer to the ASCII string to be converted. * * * OUTPUTS * string - points to the character which terminated the scan * equ - fast floating point equivalent ***************************************************************************** _afp: movem.l d3-d7,-(sp) * Save registers used move.l 24(sp),a0 * Put ASCII string address in A0 jsr FFPAFP * Execute Motorola function bra.s fpal3 * Set up functional result and exit ******* amiga.lib/fpa ************************************************** * * NAME * fpa - convert fast floating point into ASCII string equivalent * * SYNOPSIS * exp = fpa(fnum, &string[0]); * * FUNCTION * Accepts an FFP number and the address of the ASCII string where it's * converted output is to be stored. The number is converted to a NULL * terminated ASCII string in and stored at the address provided. * Additionally, the base ten (10) exponent in binary form is returned. * * INPUTS * fnum - Motorola Fast Floating Point number * &string[0] - address for output of converted ASCII character string * (16 bytes) * * RESULT * &string[0] - converted ASCII character string * exp - integer exponent value in binary form * * BUGS * None ***************************************************************************** * Docs recovered 9/1/88 from Rob Peck's sun - Bryce _fpa: movem.l d3-d7,-(sp) * Save registers used move.l 24(sp),d7 * Put FFP number to convert in D7 jsr FFPFPA * Execute Motorola function move.l 42(sp),a0 * Put ASCII string address in A0 fpal1: moveq #6,d1 * Set up for 14 byte data transfer fpal2: move (sp)+,(a0)+ * Move one (1) word at a time dbf d1,fpal2 * Loop until all 14 bytes moved fpal3: move.l d7,d0 * Get exponent as functional value movem.l (sp)+,d3-d7 * Restore registers used rts ******* amiga.lib/dbf ************************************************** * * NAME * dbf - convert FFP dual-binary number to FFP format * * SYNOPSIS * fnum = dbf(exp, mant); * * FUNCTION * Accepts a dual-binary format (described below) floating point * number and converts it to an FFP format floating point number. * The dual-binary format is defined as: * * exp bit 16 = sign (0=>positive, 1=>negative) * exp bits 15-0 = binary integer representing the base * ten (10) exponent * man = binary integer mantissa * * INPUTS * exp - binary integer representing sign and exponent * mant - binary integer representing the mantissa * * RESULT * fnum - converted FFP floating point format number * * BUGS * None ***************************************************************************** _dbf: movem.l d3-d7,-(sp) * Save registers used moveq #0,d6 * Insure D6 is cleared move.l 28(sp),d7 * D7 = mantissa tst.l d7 bpl dbfl1 bset #16,d6 * Set negative mantissa indicator neg.l d7 * Convert to positive mantissa value dbfl1: move 26(sp),d6 * Set up actual exponent value jsr FFPDBF * Execute Motorola function bra.s fpal3 * Set up functional result and exit ******* amiga.lib/arnd ************************************************* * * NAME * arnd - ASCII round of the provided floating point string * * SYNOPSIS * arnd(place, exp, &string[0]); * * FUNCTION * Accepts an ASCII string representing an FFP floating point * number, the binary representation of the exponent of said * floating point number and the number of places to round to. * A rounding process is initiated, either to the left or right * of the decimal place and the result placed back at the * input address defined by &string[0]. * * INPUTS * place - integer representing number of decimal places to round to * exp - integer representing exponent value of the ASCII string * &string[0] - address where rounded ASCII string is to be placed * (16 bytes) * * RESULT * &string[0] - rounded ASCII string * * BUGS * None ***************************************************************************** _arnd: movem.l d3-d7,-(sp) * Save registers used movem.l 24(sp),d6-d7 * Set up rounding magnitude and exponent move.l 32(sp),a0 * Get address of ASCII string lea -14(sp),a1 * Point A1 to stack area move.l a1,sp * Point SP there too moveq #6,d1 * Set up for 7 word mem-mem move arndl1: move (a0)+,(a1)+ * Move one (1) word at a time dbf d1,arndl1 * Loop until all 7 words moved jsr FFPARND * Execute Motorola function move.l 46(sp),a0 * Get string address for result bra.s fpal1 * Move the string and get out ****i** amiga.lib/fpbcd ************************************************ * * NAME * fpbcd - convert FFP floating point number to BCD format * * SYNOPSIS * fpbcd(fnum, &string[0]); * * FUNCTION * Accepts a floating point number and the address where the * converted BCD data is to be stored. The FFP number is * converted and stored at the specified address in an ASCII * form in accordance with the following format: * * MMMM S E S B * * Where: M = Four bytes of BCD, each with two (2) digits of * the mantissa (8 digits) * S = Sign of mantissa (0x00 = positive, 0xFF = negative) * E = BCD byte for two (2) digit exponent * S = Sign of exponent (0x00 = positive, 0xFF = negative) * B = One (1) byte binary two's compliment representation * of the exponent * * INPUTS * fnum - floating point number * &string[0] - address where converted BCD data is to be placed * * RESULT * &string[0] - converted BCD data ***************************************************************************** ;_fpbcd: ; movem.l d3-d7,-(sp) * Save registers used ; move.l 24(sp),d7 * Put FFP value in D7 ; jsr FFPFPBCD * Execute Motorola function ; move.l 40(sp),a0 * Put ASCII string address in A0 ; moveq #3,d1 * Set up for 8 byte data transfer ; bra.s fpal2 * Move the string and get out end