If these numbers were important to me I would want to see the trailing 
zeros because 0.001 may or may not be 0.0010000000.
Even 0.0010 has much more meaning than 0.001.  Trailing zeros are not 
meaning less like leading zeros. They indicate the accuracy of the number.

Doug

Bung Matrix wrote:
> The problem with FloatToStrF is, the digit parameter limit the output 
> of resulted string, e.g:
> FloatToStrF(1/1000,10,10) will result 0.001000000. That's why I made 
> the function FloatToStrNP which will convert the floating-point into 
> string using FloatToStrF and then remove all trailing zeros (1/1000 = 
> 0.001). This function of mine works just fine, but I wonder if there 
> are much more efficient way to convert float to string without 
> scientific notation and no trailing zeros, because my function 
> involves many (yes, I consider it many...) operations which can affect 
> the performance of the rapid calculation I'm going to perform with it.
>
> ---------------------
> CoolKid K-Qha
> http://kha.web.id
> ---------------------
>
> --- On *Sun, 8/10/08, Sean Roberts /<[EMAIL PROTECTED]>/* wrote:
>
>     From: Sean Roberts <[EMAIL PROTECTED]>
>     Subject: RE: [advanced_delphi] FloatToStrF without zero padding
>     To: advanced_delphi@yahoogroups.com
>     Date: Sunday, August 10, 2008, 9:37 PM
>
>     Please provide a sample number (val) that gets the padding you
>     don't want.
>      
>     I would like to use this number to see if I can get the function
>     work the way you want it to.
>
>     XFX nForce 680i LT SLI Socket 775 | Intel Core 2 Duo E6750 OC:1600
>     FSB <mailto:[EMAIL PROTECTED]> @ 3.2 MHz | 2 x OCZ SLI 1024MB PC6400
>     DDR2 800MHz| XION 600w PSU | CM Stacker 200 ATX/BTX Case | EVGA
>     GeForce 8800 GTS 512 | WD SATA 74 GB Raptor (boot/system) , WD
>     SATA 500 GB (storage) | Windows XP Pro, SP3 | Patient and
>     supportive significant other
>
>
>     ------------------------------------------------------------------------
>
>     To: advanced_delphi@ yahoogroups. com
>     From: [EMAIL PROTECTED] com
>     Date: Sun, 10 Aug 2008 00:05:27 -0700
>     Subject: [advanced_delphi] FloatToStrF without zero padding
>
>
>     Greeting,
>
>     Did anyone know how to use FloatToStrF without zero padding? I
>     need it because my code must avoid the usage of scientific
>     notation, thus I used FloatToStrF with ffFixed format. But later I
>     found out that the output of resulted string was padded with
>     zeros. For temporarily, I use these code to remove all zero
>     padding but still there must be some way which is more efficient...
>
>     ---code begin---
>     function TDMod.FloatToStrNP( val : extended) : string ;
>     var
>     a : integer ;
>     txt,retval : string ;
>     begin
>     txt := FloatToStrF( val,ffFixed, 18,18) ;
>     retval := '' ;
>     for a:=Length(txt) downto 1 do
>     begin
>     if (txt[a]='0') and (retval <> '') then retval := txt[a]+retval
>     else if txt[a]<>'0' then retval := txt[a]+retval ;
>     end ;
>     Result := retval ;
>     end;
>     ---code end---
>
>     ------------ ---------
>     CoolKid K-Qha
>     http://kha.web. id <http://kha.web.id/>
>     ------------ ---------
>
>
>  


Reply via email to