faq swith case can be a range?

[fcvcnet]

Hi,
In swith statements, follow case is constant-expression, can it be a
range?
For example:
...
case ' 0 ' :
...
thanks.

[Gavin Deane]

fcvcnet wrote:

Hi,
In swith statements, follow case is constant-expression, can it be a
range?
For example:
...
case ' 0 ' :
...
thanks.

No. Such a range is not an integral constant expression. Maybe you need
a chain of if-else if- ... -else.

Gavin Deane

[Default User]

Gavin Deane wrote:

>
fcvcnet wrote:

Hi,
In swith statements, follow case is constant-expression, can it
be a range?
For example:
...
case ' 0 ' :
...
thanks.

No. Such a range is not an integral constant expression. Maybe you
need a chain of if-else if- ... -else.

The usual way in C or C++ is to use "fall through".
switch (c)
{
case '>':
case '0':
// do stuff
break;
}
This quickly gets messy for large ranges, so using if-else if-else
ladders are prefered by many.


Brian

[Robert Mabee]

fcvcnet wrote:

In swith statements, follow case is constant-expression, can it be a
range?

It is a simple extension; BCPL, a C ancestor, had it in 1970, as in
case 'A' to 'Z':
If you are willing to use non-standard extensions, ask your compiler
vendor or get a public-source compiler and add it yourself.

[fcvcnet]

Thanks you all.

[BobR]

fcvcnet wrote in message ...

>Thanks you all.

A litle slow on the draw here. I found this in the GCC docs:
"Case Ranges"

case 1 ... 5: // must be a space before and after the '...'

case 'A' ... 'Z':

May NOT be portable, GCC 'feature'.

The GNU crew is usually good about saying whether it is an GCC extension, ISO
C89, C90, C99. But, they didn't say in this case, so, try it in your compiler
and watch the error output. Maybe, report if it works in a compiler other
than GCC (It compiles in my MinGW 3.3).

--
Bob R
POVrookie