version 1.29, 1999/12/03 18:24:22

version 1.61, 2011/10/07 17:10:15

Line 1

Line 1

\ High level floating point 14jan94py 
\ High level floating point 14jan94py 


\ Copyright (C) 1995,1997 Free Software Foundation, Inc. 
\ Copyright (C) 1995,1997,2003,2004,2005,2006,2007,2009,2010 Free Software Foundation, Inc. 


\ This file is part of Gforth. 
\ This file is part of Gforth. 


\ Gforth is free software; you can redistribute it and/or 
\ Gforth is free software; you can redistribute it and/or 
\ modify it under the terms of the GNU General Public License 
\ modify it under the terms of the GNU General Public License 
\ as published by the Free Software Foundation; either version 2 
\ as published by the Free Software Foundation, either version 3 
\ of the License, or (at your option) any later version. 
\ of the License, or (at your option) any later version. 


\ This program is distributed in the hope that it will be useful, 
\ This program is distributed in the hope that it will be useful, 
Line 15

Line 15

\ GNU General Public License for more details. 
\ GNU General Public License for more details. 


\ You should have received a copy of the GNU General Public License 
\ You should have received a copy of the GNU General Public License 
\ along with this program; if not, write to the Free Software 
\ along with this program. If not, see http://www.gnu.org/licenses/. 
\ Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. 



\ 1 cells 4 = [IF] 
\ 1 cells 4 = [IF] 
\ ' cells Alias sfloats 
\ ' cells Alias sfloats 
Line 51

Line 50

\G enough space to align it. 
\G enough space to align it. 
here dup dfaligned swap ?DO bl c, LOOP ; 
here dup dfaligned swap ?DO bl c, LOOP ; 


1 sfloats constant sfloat+ ( sfaddr1  sfaddr2 ) \ floatext sfloatplus 
(Field) sfloat+ ( sfaddr1  sfaddr2 ) \ floatext sfloatplus 
\G Increment @i{sfaddr1} by the number of address units corresponding to the size of 
\G @code{1 sfloats +}. 
\G a singleprecision IEEE floatingpoint number, to give @i{sfaddr2}."" 
1 sfloats , 
dofield: lastxt codeaddress! \ change the constant into a field 


(Field) dfloat+ ( dfaddr1  dfaddr2 ) \ floatext dfloatplus 
1 dfloats constant dfloat+ ( dfaddr1  dfaddr2 ) \ floatext dfloatplus 
\G @code{1 dfloats +}. 
\G Increment @i{dfaddr1} by the number of address units corresponding to the size of 
1 dfloats , 
\G a doubleprecision IEEE floatingpoint number, to give @i{dfaddr2}."" 

dofield: lastxt codeaddress! \ change the constant into a field 



: f, ( f  ) \ gforth 
: f, ( f  ) \ gforth 
\G Reserve data space for one floatingpoint number and store 
\G Reserve data space for one floatingpoint number and store 
\G @i{f} in the space. 
\G @i{f} in the space. 
Line 76 DOES> (  r )

Line 73 DOES> (  r )

\G floatingpoint stack. 
\G floatingpoint stack. 
fp0 @ fp@  [ 1 floats ] Literal / ; 
fp0 @ fp@  [ 1 floats ] Literal / ; 


: FLit (  r ) r> dup f@ float+ >r ; 

: FLiteral ( compilation r  ; runtime  r ) \ float fliteral 
: FLiteral ( compilation r  ; runtime  r ) \ float fliteral 
\G Compile appropriate code such that, at runtime, @i{r} is placed 
\G Compile appropriate code such that, at runtime, @i{r} is placed 
\G on the (floatingpoint) stack. Interpretation semantics are undefined. 
\G on the (floatingpoint) stack. Interpretation semantics are undefined. 
BEGIN here cell+ dup faligned <> WHILE postpone noop REPEAT 
BEGIN here cell+ cell+ dup faligned <> WHILE postpone noop REPEAT 
postpone FLit f, ; immediate 
postpone ahead here >r f, postpone then 

r> postpone literal postpone f@ ; immediate 


&15 Value precision (  u ) \ floatext 
&15 Value precision (  u ) \ floatext 
\G @i{u} is the number of significant digits currently used by 
\G @i{u} is the number of significant digits currently used by 
Line 104 DOES> (  r )

Line 101 DOES> (  r )

IF ' emit THEN ; 
IF ' emit THEN ; 


: f. ( r  ) \ floatext fdot 
: f. ( r  ) \ floatext fdot 
\G Display (the floatingpoint number) @i{r} using fixedpoint notation, 
\G Display (the floatingpoint number) @i{r} without exponent, 
\G followed by a space. 
\G followed by a space. 
f$ dup >r 0< 
f$ dup >r 0<= 
IF '0 emit 
IF '0 emit 
ELSE scratch r@ min type r@ precision  zeros THEN 
ELSE scratch r@ min type r@ precision  zeros THEN 
'. emit r@ negate zeros 
'. emit r@ negate zeros 
Line 114 DOES> (  r )

Line 111 DOES> (  r )

\ I'm afraid this does not really implement ansi semantics wrt precision. 
\ I'm afraid this does not really implement ansi semantics wrt precision. 
\ Shouldn't precision indicate the number of places shown after the point? 
\ Shouldn't precision indicate the number of places shown after the point? 



\ Why do you think so? ANS Forth appears ambiguous on this point. anton. 


: fe. ( r  ) \ floatext fedot 
: fe. ( r  ) \ floatext fedot 
\G Display @i{r} using engineering notation, followed by a space. 
\G Display @i{r} using engineering notation (with exponent dividable 

\G by 3), followed by a space. 
f$ 1 s>d 3 fm/mod 3 * >r 1+ >r 
f$ 1 s>d 3 fm/mod 3 * >r 1+ >r 
scratch r@ min type '. emit scratch r> /string type 
scratch r@ tuck min tuck  >r type r> zeros 

'. emit scratch r> /string type 
'E emit r> . ; 
'E emit r> . ; 


: fs. ( r  ) \ floatext fsdot 
: fs. ( r  ) \ floatext fsdot 
\G Display @i{r} using scientific notation, followed by a space. 
\G Display @i{r} using scientific notation (with exponent), followed 

\G by a space. 
f$ 1 
f$ 1 
scratch over c@ emit '. emit 1 /string type 
scratch over c@ emit '. emit 1 /string type 
'E emit . ; 
'E emit . ; 


require debugs.fs 



: sfnumber ( caddr u  r true  false ) 
: sfnumber ( caddr u  r true  false ) 
2dup [CHAR] e scan ( caddr u caddr2 u2 ) 
2dup [CHAR] e scan ( caddr u caddr2 u2 ) 
dup 0= 
dup 0= 
Line 141 require debugs.fs

Line 141 require debugs.fs

2drop false 
2drop false 
THEN ; 
THEN ; 


:noname ( caddr u  ) 
[ifdef] recognizer: 

' noop 

:noname postpone Fliteral ; 

dup 

recognizer: r:fnumber 



: fnumrecognizer ( addr u  float inttable  addr u r:fail ) 

2dup sfnumber 

IF 

2drop r:fnumber EXIT 

THEN 

r:fail ; 



' fnumrecognizer 

forthrecognizer getrecognizers 

1+ forthrecognizer setrecognizers 

[else] 

[ifundef] compilernotfound1 

defer compilernotfound1 

' no.extensions IS compilernotfound1 



:noname compilernotfound1 execute ; is compilernotfound 



defer interpreternotfound1 

' no.extensions IS interpreternotfound1 



:noname interpreternotfound1 execute ; is interpreternotfound 

[then] 



:noname ( caddr u  ... xt ) 
2dup sfnumber 
2dup sfnumber 
IF 
IF 
2drop POSTPONE FLiteral 
2drop [comp'] FLiteral 
ELSE 
ELSE 
defers compilernotfound 
defers compilernotfound1 
ENDIF ; 
ENDIF ; 
IS compilernotfound 
IS compilernotfound1 


:noname ( caddr u  r ) 
:noname ( caddr u  ... xt ) 
2dup sfnumber 
2dup sfnumber 
IF 
IF 
2drop 
2drop ['] noop 
ELSE 
ELSE 
defers interpreternotfound 
defers interpreternotfound1 
ENDIF ; 
ENDIF ; 
IS interpreternotfound 
IS interpreternotfound1 

[then] 


: fvariable ( "name"  ) \ float fvariable 
: fvariable ( "name"  ) \ float fvariable 
Create 0.0E0 f, ; 
Create 0.0E0 f, ; 
Line 179 IS interpreternotfound

Line 209 IS interpreternotfound

\G Divide 1.0e0 by @i{r1}. 
\G Divide 1.0e0 by @i{r1}. 
1.0e0 fswap f/ ; 
1.0e0 fswap f/ ; 



getcurrent environmentwordlist setcurrent 

1.7976931348623157e308 FConstant maxfloat 

setcurrent 


\ We now have primitives for these, so we need not define them 
\ We now have primitives for these, so we need not define them 


Line 203 IS interpreternotfound

Line 236 IS interpreternotfound

f fabs frot frot f* f< ; 
f fabs frot frot f* f< ; 


: f~ ( r1 r2 r3  flag ) \ floatext fproximate 
: f~ ( r1 r2 r3  flag ) \ floatext fproximate 
\G ANS Forth medley: r3>0: @code{f~abs}; r3=0: r1=r2; r3<0: @code{fnegate f~abs}. 
\G ANS Forth medley for comparing r1 and r2 for equality: r3>0: 

\G @code{f~abs}; r3=0: bitwise comparison; r3<0: @code{fnegate f~rel}. 
fdup f0= 
fdup f0= 
IF 
IF \ bitwise comparison 
fdrop f= \ !! this does not work, because 0=0 with f= on LinuxIntel 
fp@ float+ 1 floats over float+ over str= 
\ the standard says they should compare unequal 
fdrop fdrop fdrop 
\ the comparison should be done with COMPARE 

EXIT 
EXIT 
THEN 
THEN 
fdup f0> 
fdup f0> 
Line 218 IS interpreternotfound

Line 251 IS interpreternotfound

fnegate f~rel 
fnegate f~rel 
THEN ; 
THEN ; 


: f.s (  ) \ gforth fdots 
0e 8 0 [do] fp@ [i] + c@ $80 = [if] [i] constant fsignoffset [then] [loop] 
\G Display the number of items on the floatingpoint stack, 

\G followed by a list of the items; TOS is the rightmost item. 
: fcopysign ( r1 r2  r3 ) \ gforth 
." <" fdepth 0 .r ." > " fdepth 0 max maxdepth.s @ min dup 0 
\G r3 takes its absolute value from r1 and its sign from r2 
?DO dup i  1 floats fp@ + f@ f. LOOP drop ; 
\ !! implementation relies on IEEE DP format 

fp@ fsignoffset + dup c@ $80 and >r ( r1 r2 addrr1sign ) 

float+ dup c@ $7f and r> or swap c! 

fdrop ; 



\ proposals from Krishna Myeni in <cjsp2d$47l$1@ngspoold02.news.aol.com> 

\ not sure if they are a good idea 



: ftrunc ( r1  r2 ) \ X:ftrunc 

\ round towards 0 

fdup fabs floor fswap fcopysign ; 



: FMOD ( r1 r2  r ) 

\ remainder of r1/r2 

FOVER FOVER F/ ftrunc F* F ; 

