Re: How to add thousand separators

First, this code is obsolete as written, because char is a dead data type and should not
be used for any purpose whatsoever. You should not be returning char *, and you should
DEFINITELY not be using a static variable to hold the result! None of this is required,
and it is needlessly complicated, uses obsolete and dangerous functions like sprintf,
assumes 8-bit characters, presumes the separates are ,. (in other locales they differ a
lot, and let's try to avoid discussing Indic notation which does not use groups of three).
This code is not thread-safe.

See my earlier post for a more elegant way to deal with all of this using the NLS APIs,
uses Unicode, does not use static variables anywhere, and is written for MFC, not PDP-11
C.

If you need to return values, return a CString or pass in a CString & argument. NEVER use
a static buffer.

On Sat, 03 Nov 2007 12:51:10 -0700, przemyslaw.sliwa@xxxxxxxxx wrote:

Here you go:

char* AmountComma(double field, int dec_places/*=0*/)
****
CString AmountComma(double field, int dec_places /*=0*/)
****
{
if(_isnan(field))
return "NaN";
****
return _T("NaN");
****
else if(!_finite(field))
return "Infinity";
****
return _T("INF");
Note that both of these should be stored as string resources since they might need to be
localized!
****

char buf[256];
static char out_buf[256];
****
CString buf;
CString out_buf;
****
if(fabs(field) > 1e16)
{
// probably a bad number. don't make an attempt to show all digits
sprintf(out_buf, "%18e", field);
return out_buf;
****
out_buf.Format(_T("%18e"));
retirm out_buf;
****
}

char fbuf[16];
sprintf(fbuf, "%%18.%if", dec_places);
sprintf(buf, fbuf, field);
****
Assuming it made sense to use obsolete and dangerous library calls like this, you could
have written
_tsprintf(buf, _T("%18.*f"), dec_places, field);
you don't need to construct a format string if the values of the format are themselves
variables
CString buf;
buf.Format(_T("%.*f"), dec_places, field);

Note that no fixed-size buffers are required; it is Unicode-aware, and there is no need to
format a formatting string. Also, I see no reason to construct a fixed-length string of
18 digits for any reason.
*****
int len = strlen(buf);
assert(len > 0);
len--;
int pos = 255;
out_buf[pos--] = '\0'; // populate out_buf from right to left
****
It is a lot easier to simply construct it left-to-right, and call _tcsrev to reverse it
when done. Why write such complicated code to accomplish a simple task?

First step: Reverse te string
buf.MakeReverse();

Second step: copy all digits right to left until the decimal point (whatever it is) is
reached.
int n = buf.Find(decimalsep);
if(n >= 0)
out_buf = buf.Left(n);

Third step: copy all digits right to left, putting the thousand separator in every 3,
until a sign is reached
buf = buf.Mid(n);
for(int i = 0; ; i++)
{ /* copy */
CString t = buf.Left(1);
buf = buf.Mid(1);
out_buf += t;
if(t == _T("+") || t == _T("-"))
break;
if(buf.IsEmpty())
break;
if(i > 0 && i %3 == 0)
out_buf += thousandsep;
} /* copy */

Fourth step: reverse the string
out_buf.MakeReverse();
return out_buf;

Note the above code is smaller, Unicode-aware, thread-safe, written in MFC, localizable
(except for Indic!), and much easier to understand.

Here are some guidelines to adopt in writing code:

Assume that if you ever write 'char' except in rare and exotic circumstances, you have
made a serious coding error. (Note the code shown here is neither rare nor exotic)

Assume that if you ever use an 8-bit literal except in rare and exotic circumstances, you
have made a coding error

Assume that if you EVER use a static buffer inside a function, and that declaration is not
preceded by the word 'const', you have made a VERY SERIOUS DESIGN ERROR.

Assume that if you ever write a declaration of a fixed-size character buffer, even of
TCHAR, except in certain rare and exotic circumstances (usually dealing with raw API
calls), you have made a serous design error. And even then you should avoid writing them.

Assume that if you use strcpy, wcscpy, _tcscpy, strcat, wcscat, _tcscat, sprintf,
_tsprintf, or wsprintf, you have made an EXTREMELY SERIOUS design error, so bad that you
can expect to see headlines on some anti-virus site about your program. Use only the safe
forms, such as the strsafe.h functions (VS6 through VS2005) or the _s functions in VS2005
and beyond.

Assume that if you ever write a literal string which depends on knowing how to read the
language in which the programmer coded, you have made a serious design error.

****

for(int id = 0; id<dec_places;++id)
{
out_buf[pos--] = buf[len--];
}
// and the ".":
if(dec_places)
out_buf[pos--] = buf[len--];


assert(len > 0);

int have_digits = 0;
for(int i = len; i>=0;i--)
{
char c = buf[i];
if(isdigit(c))
{
//insert a comma every 3 digits
if(have_digits == 3)
{
have_digits = 0;
out_buf[pos--] = ',';
}
have_digits++;
}
out_buf[pos--] = c;

if(isspace(c))
{ // non-white space means we got to far left (don't forget digits
and +/- etc)
break;
}
}

// start of return string is leftmost character populated
return &out_buf[pos+1];
}



On Nov 3, 6:45 pm, "Giovanni Dicanio" <giovanni.dica...@xxxxxxxxxx>
wrote:
"Charles Tam" <Charles...@xxxxxxxxxxxxxxxxxxxxxxxxx> ha scritto nel
messaggionews:0B999E85-655F-4C40-A179-7C75B3EC7008@xxxxxxxxxxxxxxxx

I've a CString containing a numeric value. How do I format it to include
the
thousand separators?

I think you may find the following CodeProject article to be interesting:

"XFormatNumber - A function to format numbers with commas"

http://www.codeproject.com/string/xformatnumber.asp

Giovanni

Joseph M. Newcomer [MVP]
email: newcomer@xxxxxxxxxxxx
Web: http://www.flounder.com
MVP Tips: http://www.flounder.com/mvp_tips.htm
.

Relevant Pages

  • Re: [C] strange bug in code
    ... >correct program output was a string of ASCII codes that might look ... > char * digits ...
    (alt.comp.lang.learn.c-cpp)
  • Re: HEXADECIMAL to STRING
    ... suppose that I have a string that is an hexadecimal number, ... void print_hex(unsigned char *bs, unsigned int n){ ... You have "len" digits. ...
    (comp.lang.c)
  • Re: what is the best way of passing floats into a string
    ... I do not null-terminate as snprintf takes care of this (according to ... But the easiest way to determine the size needed to format a number, ... int length_of_representation(double n,const char* format){ ... I get a nice result of -10.000000 in my char * string. ...
    (comp.unix.programmer)
  • Re: weird problem
    ... I already told you that the comparison between an integer and a float ... to strcmpwhich expects a pointer to a string. ... And now a question about something else: why do you use floating ... int,float, char, etc. ...
    (comp.lang.c)
  • Re: [OT] My First C# (warning - long post)
    ... cut me some slack - show me a COBOL program with less than 15 working-storage variables! ... Yes, the Trimis probably extra now, but that was my attempt to get it to quit griping at me that I had given it a "String", when it wanted a "char". ... public string IBreturn ...
    (comp.lang.cobol)