Format$ Weirdness

[Mark Newman]

Format$ doesn't seem to format fractional negative values right. In my case I'm
tying to format a number like "-.23" so that it comes out "-0.23000". So I use

Format$(-.23, "0.00000")

But the output is actually "-.23000", contrary to what the docs say about adding
leading zero to the format string.

However, this does work for positive numbers:

Format$(.23, "0.00000")

will output "0.23000"

I can work around this by using "*0.00000" for the formatting string, but I'd
like to know what's going on. Here's a short sample program:

$COMPILE EXE
$REGISTER NONE

FUNCTION PBMAIN() AS LONG
   
   DIM sTmp AS STRING
   
   ' This should display "-0.23000", but it doesn't
   sTmp = FORMAT$(-.23,"0.00000")
   
   MSGBOX sTmp, 0, "Should show '-0.23000'"
   
   
   ' Workaround
   sTmp = FORMAT$(-.23,"*0.00000")

   MSGBOX sTmp, 0, "Should show '-0.23000'"
   
   ' This should display "+0.23000", and it does
   sTmp = FORMAT$(.23,"+0.00000")

   MSGBOX sTmp, 0, "Should show '+0.23000'"

END FUNCTION

------------------
Mark Newman

[This message has been edited by Mark Newman (edited August 22, 2000).]

[Edmund Melerski]

Mark,

To achieve your goal you have to use the following format

FORMAT$(-.23,"00.00000")

This format allows for two characters in front of the dot
whilst yours makes a provision for only one character.
Hence, in the case of a negative number only minus is shown.

Edmund

------------------

[Mark Newman]

Thanks Edmund!

This isn't very clear from the docs, so perhaps a few more
examples in the manual would help.



------------------
Mark Newman

[Lance Edmonds]

As I understand it from discussions with the Documentation staff, the FORMAT$() section in the help/doc's is going to be completely revised in a future update to the compiler. Essentially, the "0" digit is also a placeholder for the negation symbol, so two leading zero's (or pound "#" signs are required to ensuree that at least one leading zero retained for negative values.

However, it creates a problem if you use zero's for digit placeholders as positive values will have leading zeros, which may not be what you want. The solution is to use pound symbols instead of zero digits, for example:

 sTmp = FORMAT$(1.23,"00.00000") ' "01.23000"
 sTmp = FORMAT$(1.23,"#0.00000") ' "01.23000"
 sTmp = FORMAT$(1.23,"##.#####") ' " 1.23000" <-- preferred
 
 sTmp = FORMAT$(-1.23,"00.00000") ' "-1.23000"
 sTmp = FORMAT$(-1.23,"#0.00000") ' "-1.23000"
 sTmp = FORMAT$(-1.23,"##.#####") ' "-1.23000" <-- preferred

------------------
Lance
PowerBASIC Support
mailto:[email protected][email protected]</A>

[Mark Newman]

sTmp = FORMAT$(-1.23,"##.#####") ' "-1.23000" <-- preferred

Using the "#" formatter in this way runs counter to the docs, which state
that the "#" will output a space if there aren't enough digits to fill the
format string on either side of the decimal.

One suggestion: Many (many!) years ago I used HP's RMB (Rocky Mountain Basic)
which had a sign placement formatter (I think it was the "+" sign), so that
a format string such as "+0.000" would always output the sign of the expression.
A handy feature; how about adding this to the wish list?

Thanks!

------------------
Mark Newman

[This message has been edited by Mark Newman (edited August 23, 2000).]

[Lance Edmonds]

I'll ask it to be put on the suggestions list. Thanks!


------------------
Lance
PowerBASIC Support
mailto:[email protected][email protected]</A>