mirror of
https://gitlab.com/freepascal.org/fpc/source.git
synced 2025-05-20 20:32:30 +02:00

delta for which 1.0 and 1.0+delta is different, rather than some power-of-10 ballpark equivalent (fixes mantis #11308) * print the same number of digits for doubles on systems which support extended as on those which don't (i.e., one digit less on the former). This solves regressions after the previous change and is Delphi-compatible. * adapted tests for the previous change git-svn-id: trunk@11025 -
460 lines
15 KiB
PHP
460 lines
15 KiB
PHP
{
|
|
This file is part of the Free Pascal run time library.
|
|
Copyright (c) 1999-2000 by Michael Van Canneyt,
|
|
member of the Free Pascal development team
|
|
|
|
See the file COPYING.FPC, included in this distribution,
|
|
for details about the copyright.
|
|
|
|
This program 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.
|
|
|
|
**********************************************************************}
|
|
|
|
type
|
|
{ See symconst.pas tfloattype }
|
|
treal_type = (
|
|
rt_s32real,rt_s64real,rt_s80real,
|
|
rt_c64bit,rt_currency,rt_s128real
|
|
);
|
|
{ corresponding to single double extended fixed comp for i386 }
|
|
|
|
Procedure str_real (len,f : longint; d : ValReal; real_type :treal_type; out s : string);
|
|
{$ifdef SUPPORT_EXTENDED}
|
|
type
|
|
TSplitExtended = packed record
|
|
case byte of
|
|
0: (bytes: Array[0..9] of byte);
|
|
1: (words: Array[0..4] of word);
|
|
2: (cards: Array[0..1] of cardinal; w: word);
|
|
end;
|
|
const
|
|
maxDigits = 17;
|
|
{$else}
|
|
{$ifdef SUPPORT_DOUBLE}
|
|
type
|
|
TSplitDouble = packed record
|
|
case byte of
|
|
0: (bytes: Array[0..7] of byte);
|
|
1: (words: Array[0..3] of word);
|
|
2: (cards: Array[0..1] of cardinal);
|
|
end;
|
|
const
|
|
maxDigits = 15;
|
|
{$else}
|
|
{$ifdef SUPPORT_SINGLE}
|
|
type
|
|
TSplitSingle = packed record
|
|
case byte of
|
|
0: (bytes: Array[0..3] of byte);
|
|
1: (words: Array[0..1] of word);
|
|
2: (cards: Array[0..0] of cardinal);
|
|
end;
|
|
const
|
|
maxDigits = 9;
|
|
{$endif SUPPORT_SINGLE}
|
|
{$endif SUPPORT_DOUBLE}
|
|
{$endif SUPPORT_EXTENDED}
|
|
|
|
type
|
|
{ the value in the last position is used for rounding }
|
|
TIntPartStack = array[1..maxDigits+1] of valReal;
|
|
|
|
var
|
|
roundCorr, corrVal, factor : valReal;
|
|
spos, endpos, fracCount: longint;
|
|
correct, currprec: longint;
|
|
temp : string;
|
|
power : string[10];
|
|
sign : boolean;
|
|
dot : byte;
|
|
fraczero, expMaximal: boolean;
|
|
|
|
|
|
maxlen : longint; { Maximal length of string for float }
|
|
minlen : longint; { Minimal length of string for float }
|
|
explen : longint; { Length of exponent, including E and sign.
|
|
Must be strictly larger than 2 }
|
|
const
|
|
maxexp = 1e+35; { Maximum value for decimal expressions }
|
|
minexp = 1e-35; { Minimum value for decimal expressions }
|
|
zero = '0000000000000000000000000000000000000000';
|
|
|
|
procedure RoundStr(var s: string; lastPos: byte);
|
|
var carry: longint;
|
|
begin
|
|
carry := 1;
|
|
repeat
|
|
s[lastPos] := chr(ord(s[lastPos])+carry);
|
|
carry := 0;
|
|
if s[lastPos] > '9' then
|
|
begin
|
|
s[lastPos] := '0';
|
|
carry := 1;
|
|
end;
|
|
dec(lastPos);
|
|
until carry = 0;
|
|
end;
|
|
|
|
procedure getIntPart(d: valreal);
|
|
var
|
|
intPartStack: TIntPartStack;
|
|
intPart, stackPtr, endStackPtr, digits: longint;
|
|
overflow: boolean;
|
|
begin
|
|
{$ifdef DEBUG_NASM}
|
|
writeln(stderr,'getintpart(d) entry');
|
|
{$endif DEBUG_NASM}
|
|
{ position in the stack (gets increased before first write) }
|
|
stackPtr := 0;
|
|
{ number of digits processed }
|
|
digits := 0;
|
|
{ did we wrap around in the stack? Necessary to know whether we should round }
|
|
overflow :=false;
|
|
{ generate a list consisting of d, d/10, d/100, ... until d < 1.0 }
|
|
while d > 1.0-roundCorr do
|
|
begin
|
|
inc(stackPtr);
|
|
inc(digits);
|
|
if stackPtr > maxDigits+1 then
|
|
begin
|
|
stackPtr := 1;
|
|
overflow := true;
|
|
end;
|
|
intPartStack[stackPtr] := d;
|
|
d := d / 10.0;
|
|
end;
|
|
{ if no integer part, exit }
|
|
if digits = 0 then
|
|
exit;
|
|
endStackPtr := stackPtr+1;
|
|
if endStackPtr > maxDigits + 1 then
|
|
endStackPtr := 1;
|
|
{ now, all digits are calculated using trunc(d*10^(-n)-int(d*10^(-n-1))*10) }
|
|
corrVal := 0.0;
|
|
{ the power of 10 with which the resulting string has to be "multiplied" }
|
|
{ if the decimal point is placed after the first significant digit }
|
|
correct := digits-1;
|
|
{$ifdef DEBUG_NASM}
|
|
writeln(stderr,'endStackPtr = ',endStackPtr);
|
|
{$endif DEBUG_NASM}
|
|
repeat
|
|
if (currprec > 0) then
|
|
begin
|
|
intPart:= trunc(intPartStack[stackPtr]-corrVal);
|
|
dec(currPrec);
|
|
inc(spos);
|
|
temp[spos] := chr(intPart+ord('0'));
|
|
{$ifdef DEBUG_NASM}
|
|
writeln(stderr,'stackptr =',stackptr,' intpart = ',intpart);
|
|
{$endif DEBUG_NASM}
|
|
if temp[spos] > '9' then
|
|
begin
|
|
temp[spos] := chr(ord(temp[spos])-10);
|
|
roundStr(temp,spos-1);
|
|
end;
|
|
end;
|
|
corrVal := int(intPartStack[stackPtr]) * 10.0;
|
|
{$ifdef DEBUG_NASM}
|
|
writeln(stderr,'trunc(corrval) = ',trunc(corrval));
|
|
{$endif DEBUG_NASM}
|
|
dec(stackPtr);
|
|
if stackPtr = 0 then
|
|
stackPtr := maxDigits+1;
|
|
until (overflow and (stackPtr = endStackPtr)) or
|
|
(not overflow and (stackPtr = maxDigits+1)) or (currPrec = 0);
|
|
{ round if we didn't use all available digits yet and if the }
|
|
{ remainder is > 5 }
|
|
if (overflow or
|
|
(stackPtr < maxDigits+1)) then
|
|
begin
|
|
{ we didn't use all available digits of the whole part -> make sure }
|
|
{ the fractional part is not used for rounding later }
|
|
currprec := -1;
|
|
{ instead, round based on the next whole digit }
|
|
if (int(intPartStack[stackPtr]-corrVal+roundcorr) >= 5.0) then
|
|
roundStr(temp,spos);
|
|
end;
|
|
{$ifdef DEBUG_NASM}
|
|
writeln(stderr,'temp at getintpart exit is = ',temp);
|
|
{$endif DEBUG_NASM}
|
|
end;
|
|
|
|
begin
|
|
case real_type of
|
|
rt_s32real :
|
|
begin
|
|
maxlen:=16;
|
|
minlen:=8;
|
|
explen:=4;
|
|
{ correction used with comparing to avoid rounding/precision errors }
|
|
roundCorr := 1.1920928955e-07;
|
|
end;
|
|
rt_s64real :
|
|
begin
|
|
maxlen := 22;
|
|
{ correction used with comparing to avoid rounding/precision errors }
|
|
roundCorr := 2.2204460493e-16;
|
|
minlen:=9;
|
|
explen:=5;
|
|
end;
|
|
rt_s80real :
|
|
begin
|
|
{ Different in TP help, but this way the output is the same (JM) }
|
|
maxlen:=25;
|
|
minlen:=10;
|
|
explen:=6;
|
|
{ correction used with comparing to avoid rounding/precision errors }
|
|
roundCorr := 1.0842021725e-19;
|
|
end;
|
|
rt_c64bit :
|
|
begin
|
|
maxlen:=23;
|
|
minlen:=10;
|
|
{ according to TP (was 5) (FK) }
|
|
explen:=6;
|
|
{ correction used with comparing to avoid rounding/precision errors }
|
|
roundCorr := 2.2204460493e-16;
|
|
end;
|
|
rt_currency :
|
|
begin
|
|
{ Different in TP help, but this way the output is the same (JM) }
|
|
maxlen:=25;
|
|
minlen:=10;
|
|
explen:=0;
|
|
{ correction used with comparing to avoid rounding/precision errors }
|
|
roundCorr := 1.0842021725e-19;
|
|
end;
|
|
rt_s128real :
|
|
begin
|
|
{ Different in TP help, but this way the output is the same (JM) }
|
|
maxlen:=25;
|
|
minlen:=10;
|
|
explen:=6;
|
|
{ correction used with comparing to avoid rounding/precision errors }
|
|
roundCorr := 1.0842021725e-19;
|
|
end;
|
|
end;
|
|
{ check parameters }
|
|
{ default value for length is -32767 }
|
|
if len=-32767 then
|
|
len:=maxlen;
|
|
{ determine sign. before precision, needs 2 less calls to abs() }
|
|
{$ifndef endian_big}
|
|
{$ifdef SUPPORT_EXTENDED}
|
|
{ extended, format (MSB): 1 Sign bit, 15 bit exponent, 64 bit mantissa }
|
|
sign := (TSplitExtended(d).w and $8000) <> 0;
|
|
expMaximal := (TSplitExtended(d).w and $7fff) = 32767;
|
|
fraczero := (TSplitExtended(d).cards[0] = 0) and
|
|
((TSplitExtended(d).cards[1] and $7fffffff) = 0);
|
|
{$else SUPPORT_EXTENDED}
|
|
{$ifdef SUPPORT_DOUBLE}
|
|
{$ifdef FPC_DOUBLE_HILO_SWAPPED}
|
|
{ double, format (MSB): 1 Sign bit, 11 bit exponent, 52 bit mantissa }
|
|
{ high and low dword are swapped when using the arm fpa }
|
|
sign := ((TSplitDouble(d).cards[0] shr 20) and $800) <> 0;
|
|
expMaximal := ((TSplitDouble(d).cards[0] shr 20) and $7ff) = 2047;
|
|
fraczero:= (TSplitDouble(d).cards[0] and $fffff = 0) and
|
|
(TSplitDouble(d).cards[1] = 0);
|
|
{$else FPC_DOUBLE_HILO_SWAPPED}
|
|
{ double, format (MSB): 1 Sign bit, 11 bit exponent, 52 bit mantissa }
|
|
sign := ((TSplitDouble(d).cards[1] shr 20) and $800) <> 0;
|
|
expMaximal := ((TSplitDouble(d).cards[1] shr 20) and $7ff) = 2047;
|
|
fraczero := (TSplitDouble(d).cards[1] and $fffff = 0) and
|
|
(TSplitDouble(d).cards[0] = 0);
|
|
{$endif FPC_DOUBLE_HILO_SWAPPED}
|
|
{$else SUPPORT_DOUBLE}
|
|
{$ifdef SUPPORT_SINGLE}
|
|
{ single, format (MSB): 1 Sign bit, 8 bit exponent, 23 bit mantissa }
|
|
sign := ((TSplitSingle(d).words[1] shr 7) and $100) <> 0;
|
|
expMaximal := ((TSplitSingle(d).words[1] shr 7) and $ff) = 255;
|
|
fraczero := (TSplitSingle(d).cards[0] and $7fffff = 0);
|
|
{$else SUPPORT_SINGLE}
|
|
{$error No little endian floating type supported yet in real2str}
|
|
{$endif SUPPORT_SINGLE}
|
|
{$endif SUPPORT_DOUBLE}
|
|
{$endif SUPPORT_EXTENDED}
|
|
{$else endian_big}
|
|
{$ifdef SUPPORT_EXTENDED}
|
|
{$error sign/NaN/Inf not yet supported for big endian CPU's in str_real}
|
|
{$else SUPPORT_EXTENDED}
|
|
{$ifdef SUPPORT_DOUBLE}
|
|
{ double, format (MSB): 1 Sign bit, 11 bit exponent, 52 bit mantissa }
|
|
sign := ((TSplitDouble(d).cards[0] shr 20) and $800) <> 0;
|
|
expMaximal := ((TSplitDouble(d).cards[0] shr 20) and $7ff) = 2047;
|
|
fraczero:= (TSplitDouble(d).cards[0] and $fffff = 0) and
|
|
(TSplitDouble(d).cards[1] = 0);
|
|
{$else SUPPORT_DOUBLE}
|
|
{$ifdef SUPPORT_SINGLE}
|
|
{ single, format (MSB): 1 Sign bit, 8 bit exponent, 23 bit mantissa }
|
|
sign := ((TSplitSingle(d).bytes[0] and $80)) <> 0;
|
|
expMaximal := ((TSplitSingle(d).words[0] shr 7) and $ff) = 255;
|
|
fraczero:= (TSplitSingle(d).cards[0] and $7fffff = 0);
|
|
{$else SUPPORT_SINGLE}
|
|
{$error No big endian floating type supported yet in real2str}
|
|
{$endif SUPPORT_SINGLE}
|
|
{$endif SUPPORT_DOUBLE}
|
|
{$endif SUPPORT_EXTENDED}
|
|
{$endif endian}
|
|
if expMaximal then
|
|
if fraczero then
|
|
if sign then
|
|
temp := '-Inf'
|
|
else temp := '+Inf'
|
|
else temp := 'Nan'
|
|
else
|
|
begin
|
|
{ d:=abs(d); this converts d to double so we loose precision }
|
|
{ for the same reason I converted d:=frac(d) to d:=d-int(d); (PM) }
|
|
if sign then
|
|
d:=-d;
|
|
{ determine precision : maximal precision is : }
|
|
currPrec := maxlen-explen-2;
|
|
{ this is also the maximal number of decimals !!}
|
|
if f>currprec then
|
|
f:=currprec;
|
|
{ when doing a fixed-point, we need less characters.}
|
|
if (f<0) {or ((d<>0) and ((d>maxexp) and (d>minexp)))} then
|
|
begin
|
|
{ determine maximal number of decimals }
|
|
if (len>=0) and (len<minlen) then
|
|
len:=minlen;
|
|
if (len>0) and (len<maxlen) then
|
|
currprec:=len-explen-2;
|
|
end;
|
|
|
|
{ leading zero, may be necessary for things like str(9.999:0:2) to }
|
|
{ be able to insert an extra character at the start of the string }
|
|
temp := ' 0';
|
|
{ position in the temporary output string }
|
|
spos := 2;
|
|
{ get the integer part }
|
|
correct := 0;
|
|
GetIntPart(d);
|
|
{ now process the fractional part }
|
|
if d > 1.0- roundCorr then
|
|
d := frac(d);
|
|
{ if we have to round earlier than the amount of available precision, }
|
|
{ only calculate digits up to that point }
|
|
if (f >= 0) and (currPrec > f) then
|
|
currPrec := f;
|
|
{ if integer part was zero, go to the first significant digit of the }
|
|
{ fractional part }
|
|
{ make sure we don't get an endless loop if d = 0 }
|
|
if (spos = 2) and (d <> 0.0) then
|
|
begin
|
|
{ take rounding errors into account }
|
|
while d < 0.1-roundCorr do
|
|
begin
|
|
d := d * 10.0;
|
|
dec(correct);
|
|
{ adjust the precision depending on how many digits we }
|
|
{ already "processed" by multiplying by 10, but only if }
|
|
{ the amount of precision is specified }
|
|
if f >= 0 then
|
|
dec(currPrec);
|
|
end;
|
|
dec(correct);
|
|
end;
|
|
{ current length of the output string in endPos }
|
|
endPos := spos;
|
|
{ always calculate at least 1 fractional digit for rounding }
|
|
if (currPrec >= 0) then
|
|
begin
|
|
corrVal := 0.5;
|
|
factor := 1;
|
|
for fracCount := 1 to currPrec do
|
|
factor := factor * 10.0;
|
|
corrval := corrval / factor;
|
|
d:=d+roundCorr;
|
|
if d >= corrVal then
|
|
d := d + corrVal;
|
|
if int(d) = 1 then
|
|
begin
|
|
roundStr(temp,spos);
|
|
d := frac(d);
|
|
if (f < 0) then
|
|
begin
|
|
dec(currprec);
|
|
if (currprec=0) then
|
|
begin
|
|
inc(spos);
|
|
temp[spos]:='0';
|
|
end;
|
|
end;
|
|
end;
|
|
{ calculate the necessary fractional digits }
|
|
for fracCount := 1 to currPrec do
|
|
begin
|
|
if d > 1.0 then
|
|
d := frac(d) * 10.0
|
|
else d := d * 10.0;
|
|
inc(spos);
|
|
temp[spos] := chr(trunc(d)+ord('0'));
|
|
if temp[spos] > '9' then
|
|
{ possible because trunc and the "*10.0" aren't exact :( }
|
|
begin
|
|
temp[spos] := chr(ord(temp[spos]) - 10);
|
|
roundStr(temp,spos-1);
|
|
end;
|
|
end;
|
|
{ new length of string }
|
|
endPos := spos;
|
|
end;
|
|
setLength(temp,endPos);
|
|
{ delete leading zero if we didn't need it while rounding at the }
|
|
{ string level }
|
|
if temp[2] = '0' then
|
|
delete(temp,2,1)
|
|
{ the rounding caused an overflow to the next power of 10 }
|
|
else inc(correct);
|
|
if sign then
|
|
temp[1] := '-';
|
|
if (f<0) or (correct>(round(ln(maxexp)/ln(10)))) then
|
|
begin
|
|
insert ('.',temp,3);
|
|
str(abs(correct),power);
|
|
if length(power)<explen-2 then
|
|
power:=copy(zero,1,explen-2-length(power))+power;
|
|
if correct<0 then
|
|
power:='-'+power
|
|
else
|
|
power:='+'+power;
|
|
temp:=temp+'E'+power;
|
|
end
|
|
else
|
|
begin
|
|
if not sign then
|
|
begin
|
|
delete(temp,1,1);
|
|
dot := 2
|
|
end
|
|
else
|
|
dot := 3;
|
|
{ set zeroes and dot }
|
|
if correct>=0 then
|
|
begin
|
|
if length(temp)<correct+dot+f-1 then
|
|
temp:=temp+copy(zero,1,correct+dot+f-length(temp));
|
|
insert ('.',temp,correct+dot);
|
|
end
|
|
else
|
|
begin
|
|
correct:=abs(correct);
|
|
insert(copy(zero,1,correct),temp,dot-1);
|
|
insert ('.',temp,dot);
|
|
end;
|
|
{ correct length to fit precision }
|
|
if f>0 then
|
|
setlength(temp,pos('.',temp)+f)
|
|
else
|
|
setLength(temp,pos('.',temp)-1);
|
|
end;
|
|
end;
|
|
if length(temp)<len then
|
|
s:=space(len-length(temp))+temp
|
|
else s:=temp;
|
|
end;
|
|
|