> recordset->Item[1]->Value

> I recieve a _variant_t. This contains a string which I need to convert to
a
> double. But of cause, if the string contains a space-character for
example,
> the conversion from string to double will fail (type missmatch).

> I'm using the (double) operator in the _variant_t wrapper to do the
> conversion. So before I do this conversion, I need to remove any spaces
> found in the string. How do I do this? The pseudo-code would look like:

> double dblMyVar = (double) (TRIM(recordset->Item[1]->Value)

> This would be a simple procedure in C using a standard string, but all
these
> COM/ATL datatypes/wrappers and Unicode stuff makes me confused.

> // jP

> recordset->Item[1]->Value

> I recieve a _variant_t. This contains a string which I need to convert to
a
> double. But of cause, if the string contains a space-character for
example,
> the conversion from string to double will fail (type missmatch).

> I'm using the (double) operator in the _variant_t wrapper to do the
> conversion. So before I do this conversion, I need to remove any spaces
> found in the string. How do I do this? The pseudo-code would look like:

> double dblMyVar = (double) (TRIM(recordset->Item[1]->Value)

> This would be a simple procedure in C using a standard string, but all
these
> COM/ATL datatypes/wrappers and Unicode stuff makes me confused.

> // jP

> --
> Girish Bharadwaj


> > From the following expression

> > recordset->Item[1]->Value

> > I recieve a _variant_t. This contains a string which I need to convert to
> a
> > double. But of cause, if the string contains a space-character for
> example,
> > the conversion from string to double will fail (type missmatch).

> > I'm using the (double) operator in the _variant_t wrapper to do the
> > conversion. So before I do this conversion, I need to remove any spaces
> > found in the string. How do I do this? The pseudo-code would look like:

> > double dblMyVar = (double) (TRIM(recordset->Item[1]->Value)

> > This would be a simple procedure in C using a standard string, but all
> these
> > COM/ATL datatypes/wrappers and Unicode stuff makes me confused.

> > // jP

> or

> _variant_t vnt = recordset->Item[1]->Value;
> wcscpy(wszBuffer, V_BSTR(&vnt));
> <manipulate using the C RTL wcsXXX functions>

> > _variant_t vnt = recordset->Item[1]->Value;
> > <ytour favorite string wrapper> str = V_BSTR(&vnt);
> > <trim, convert etc using your favorite class>

> > or

> > _variant_t vnt = recordset->Item[1]->Value;
> > wcscpy(wszBuffer, V_BSTR(&vnt));
> > <manipulate using the C RTL wcsXXX functions>

> But how do you manipulate a wstring (wchar_t?) string character by
> character? When using a standard single-byte char string in C++ my RTrim()
> function would look something like:

> // Removes ending spaces in string
> void RTrim(char *lpzString)
> {
>     int iOffset = strlen(lpzString);

>     // Scan for the first non-space character in string from the end and
> backwards
>     for(; *(lpzString + iOffset) != ' ' || iOffset > 0; iOffset--);
>     // Null terminate string
>     *( lpzString + iOffset + 1) = NULL;
> }

> You get the point..

> How would you do the same with a wstring (or wchar_t, multibyte-strings)?
> _bstr_t strings don't use NULL characters and in Unicode-strings every
> characters occupies two bytes and so on..

> // jP