cross initialization of object

Consider this:

Code:

int val = 0;
switch(val) {
    case 1:
        int object = 0;
        break;
    case 2:
        break;
    default:
        break;
}

GCC 3.4.5 gives me a crosses initialization of 'int object' error. Which is fine and solvable with a pair of curly braces.

But I'm curious. I cannot find on the standard any reference to this. It is an error indeed, if I try to use object on another case label. But I'm really not doing it above.

Is this gcc specific or is indeed an error specified by the standard?

It's part of the C grammar. It's standard.
http://www.lysator.liu.se/c/ANSI-C-g...html#statement

The code within the brackets forms a compound statement. Declarations come before statements in C.

Code:

case 1:

is a statement.

...
Dumb forum won't let me post this without using code tags.

Hey, we worked HARD for that script!

Ok. That makes sense, I guess. (Am a slow grammar reader).

Always expected that to be a run-time error (and only on the case of trying to access the undefined object). Was happily surprised to see a compile-time error. But it got me curious.

Thanks.

Actually, that answer makes no sense at all. This is C++, which allows declarations anywhere.

The C++ standard does say, however, that you may not jump past a declaration. See 6.7§3.

It is possible to transfer into a block, but not in a way that bypasses declarations with initialization. A program that jumps² from a point where a local variable with automatic storage duration is not in scope to a point where it is in scope is ill-formed unless the variable has POD type and is declared without an initializer. [Example:

Code:

void f()
{
  // ...
  goto lx;    // ill-formed: jump into scope of `a'
  // ...
  ly:
  X a = 1;
  // ...
  lx:
  goto ly;    // ok, jump implies destructor
  // call for `a' followed by construction
  // again immediately following label ly
}

--end example]

² The transfer from the condition of a switch statement to a case label is considered a jump in this respect.

Aha! Thanks CornedBee. Couldn't find that reference.

case 1:
No tags necessary.
It's only the curly braces { and } which trigger the script into thinking it might be code.
Simple one-liners can be posted just fine.

> Dumb forum won't let me post this without using code tags.
Which is absolute bliss as far as I'm concerned as now 99% of code is posted properly rather than looking like trash.

> It's only the curly braces { and }
Yeah, you're a mod, you can bypass that, we know...

So the script will force all posters to enclose curly braces in code tags...