Re: FW: shell32: SHChangeNotify is really Unicode/ANSI indifferent, take2

[Francois Gouget]

On Tue, 21 Jan 2003, Rolf Kalbermatter wrote:
[...]
> @@ -289,22 +257,11 @@
>    LeaveCriticalSection(&SHELL32_ChangenotifyCS);
>
> /* if we allocated it, free it */
> - if(uFlags & SHCNF_PATHA)
> - if (uFlags & SHCNF_PATHA || uFlags & SHCNF_PATHW)
> {

Just nitpicking (since Dimitrie is on vacation ;-): it's probably not
necessary but I would be more confortable if the above test had more
parentheses:
> -
> if ((uFlags & SHCNF_PATHA) || (uFlags & SHCNF_PATHW))

FW: shell32: SHChangeNotify is really Unicode/ANSI indifferent, take2

[Dmitry Timoshkov]

"Francois Gouget"  wrote:
> On Tue, 21 Jan 2003, Rolf Kalbermatter wrote:
> [...]
> > @@ -289,22 +257,11 @@
> >  LeaveCriticalSection(&SHELL32_ChangenotifyCS);
> >
> > /* if we allocated it, free it */
> > - if(uFlags & SHCNF_PATHA)
> > - if (uFlags & SHCNF_PATHA || uFlags & SHCNF_PATHW)
> >
> > {
>
> Just nitpicking (since Dimitrie is on vacation ;-): it's probably not
> necessary but I would be more confortable if the above test had more
> parentheses:
> > -
> > if ((uFlags & SHCNF_PATHA) || (uFlags & SHCNF_PATHW))

or even better form the effectiveness point of view:

if (uFlags & (SHCNF_PATHA | SHCNF_PATHW))

FW: shell32: SHChangeNotify is really Unicode/ANSI indifferent, take2

[Rolf Kalbermatter]

> -----Original Message-----
> From: Dmitry Timoshkov [mailto:[e-mail]]
> Sent: Tuesday, January 21, 2003 7:04 AM
> To: Francois Gouget; Rolf Kalbermatter
> Cc: [e-mail]
> Subject: Re: FW: shell32: SHChangeNotify is really 
> Unicode/ANSI indifferent, take2
>
> "Francois Gouget"  wrote:
> > On Tue, 21 Jan 2003, Rolf Kalbermatter wrote:
> > [...]
> > > @@ -289,22 +257,11 @@
> > >  LeaveCriticalSection(&SHELL32_ChangenotifyCS);
> > >
> > > /* if we allocated it, free it */
> > > - if(uFlags & SHCNF_PATHA)
> > > - if (uFlags & SHCNF_PATHA || uFlags & SHCNF_PATHW)
> > >
> > > {
> >
> > Just nitpicking (since Dimitrie is on vacation ;-): it's
>
> probably not
> > necessary but I would be more confortable if the above test had more
> > parentheses:
> > > -
> > > if ((uFlags & SHCNF_PATHA) || (uFlags & SHCNF_PATHW))
>
> or even better form the effectiveness point of view:
>
> if (uFlags & (SHCNF_PATHA | SHCNF_PATHW))

Well, because of the values of the SHCNF constants

if (uFlags & (SHCNF_PATHA | SHCNF_PATHW))

really is equal to

if (uFlags & SHCNF_PATHW)

but this one although correct looks highly unintuitive ;-).

Rolf Kalbermatter