string vs. w/char*

[Tyro]

== Quote from Denis Koroskin (2korden at gmail.com)'s article
> On Tue, 01 Mar 2011 02:08:48 +0300, Tyro[a.c.edwards]
<nospam at home.com>
> wrote:
> > == Quote from Denis Koroskin (2korden at gmail.com)'s article
> >> On Mon, 28 Feb 2011 19:35:47 +0300, Tyro[a.c.edwards]
> > <nospam at home.com>
> >> wrote:
> >> > On 2/28/2011 11:08 PM, J Chapman wrote:
> >> >> == Quote from Tyro[a.c.edwards] (nospam at home.com)'s article
> >> >>> Both implementations results in error code 1812 being
> > returned from
> >> >>> GetLastError. explanation of the code reads:
> >> >>>        ERROR_RESOURCE_DATA_NOT_FOUND
> >> >>>        1812 (0x714)
> >> >>>        The specified image file did not contain a resource
> > section.
> >> >>> The code I'm porting initially consisted of a resource.h
> > file, a
> >> >>> generic.rc file and two icons. I have not tried to include
> > the icons
> >> >>> and
> >> >>> generic.rc file in the compilation because I do not know
how
> > to as yet
> >> >>> and I've only used half of the resource.h file: didn't
think
> > I need the
> >> >>> whole thing. Could this be the reason for the error? If so
> > could you
> >> >>> direct me to the explanation of how to prepare these files
> > for
> >> >>> inclusion
> >> >>> in the compilation process?
> >> >>> Thanks,
> >> >>> Andrew
> >> >>
> >> >> You need to compile the .rc file (see
> >> >> http://www.digitalmars.com/ctg/rcc.html), then add the
> > resulting .res
> >> >> file
> >> >> to dmd's command line.
> >> >
> >> > Awesome, this does the trick. However I get get a "GP
Fault"?
> > during
> >> > execution. Using windbg, I tracked it down to this piece of
> > code:
> >> >
> >> > void Create()
> >> > {
> >> >    _hwnd = CreateWindowExA(
> >> >      _exStyle,
> >> >      cast(const(char*))_wc.GetName(), // returns string
> >> >      cast(const(char*))_windowName,   // string variable
> >> >      _style,
> >> >      _x,
> >> >      _y,
> >> >      _width,
> >> >      _height,
> >> >      _hWndParent,
> >> >      _hMenu,
> >> >      _wc.GetInstance(),
> >> >      _data);
> >> >
> >> >      assert(_hwnd, "Internal error: Window Creation
Failed.");
> >> > }
> >> >
> >> > The program craps at assert() but the error is generated. It
> > just
> >> > displays a dialog box with the message: "test.exe has
stopped
> > working,
> >> > Windows is checking for a solution to the problem..."
> >> >
> >> > I'm thinking that _hwnd was never initialized and that
assert
> > is access
> >> > a null pointer but I cannot be sure. Any suggestions or
ideas?
> >> The
> >> >  cast(const(char*))_wc.GetName()
> >> line look *very* suspicious. You can't get a string and just
> > cast it to
> >> const(char)*. Most importantly, the string (most likely) is
not
> >> null-terminated.
> >> What you need to do here is the following:
> >> auto className = toStringz(_ws.GetName());
> >> auto caption = toStringz(_windowName);
> >> and pass those 2 to the function.
> >
> > Actually I've already tried that, it has no effect on the
outcome.
> > From your suggestion though, I've gone back and replace all the
> > cast(const(char*)) usage throughout the program. Final verdict:
> > the program still crashes it the same location. It actually
never
> > returns from CreateWindowExA().
> >
> >> Alternatively, you could make sure your strings are null-
> > terminated and
> >> pass the pointer directly (e.g. _windowName.ptr):
> >> string _windowName = "foo"; // null-terminated automatically
> >> string _caption = ("Hello, World" ~ "\0")[0..$-1]; // append
> > trailing zero
> >> to an existing string but exclude it from result (so that it's
> > not
> >> included in _caption.length)
> >
> This is indeed strange, but it has nothing to do with the
function itself.
> I still think the parameters you are passing might be invalid.
Try setting
> them to default values and see if that helps. Also try wrapping
the call
> with a try/catch block and output an exception you are getting
(if any).


The problem occurs at the site of the assertion. I wrapped the
function in a try/catch block and placed a call to MessageBoxA()
on either end of the the try block. Both calls to MessageBox fires
and the appropriate messages displayed. No exception is thrown:
made evident my the fact that the third call to MessageBox,
embeded in catch{}, is not fired. Nevertheless, execution haults
at the very next line following/catch and Create() never returns.