http://d.puremagic.com/issues/show_bug.cgi?id=1053

           Summary: Make 'static else if' or 'static if (...) {...} else if'
                    work
           Product: D
           Version: 1.008
          Platform: All
        OS/Version: All
            Status: NEW
          Severity: enhancement
          Priority: P2
         Component: DMD
        AssignedTo: bugzilla@digitalmars.com
        ReportedBy: wbaxter@gmail.com


It would be nice if "else static if" could be written "static else if".  Or even better just "else if".

"else static if" struck me as odd the first time I saw it, and since then I continue to mistype it as "static else if" pretty much every time I use it.


So it would be great if "static else if" just worked.

On the other hand, I'm pretty sure
  static if (foo) { }
  else if (bar) { }
is an error currently, so it would be nice if that were just taken to be a
"static else if".  There isn't a "static else {}" so why should we have to
repeat the "static" on an "else if"?  One "static" should be enough.

I guess this could be an issue?:
  static if (foo)
     if (biff)
         single_statement();
  else if (bar) { }

But I could live without being able to do that.  Oh, and besides that it's
already ambiguous with plain ifs...
  if (foo)
     if (biff)
         single_statement();
  else if (bar) { }

Exactly the same deal -- who does the else belong to?

--
As an added bonus either of these changes would make the Emacs D-mode capable of indenting static ifs with else ifs properly.  That's not a sufficient reason to make the change (emacs D-mode could be made to work somehow) But the fact that cc-mode basically sees "else static if" as an aberration was the thing that finally drove me to file an enhancement request on this.


-- 

<d-bugmail@puremagic.com> wrote in message news:bug-1053-3@http.d.puremagic.com/issues/...
> http://d.puremagic.com/issues/show_bug.cgi?id=1053
>
>           Summary: Make 'static else if' or 'static if (...) {...} else
> if'
>                    work
>           Product: D
>           Version: 1.008
>          Platform: All
>        OS/Version: All
>            Status: NEW
>          Severity: enhancement
>          Priority: P2
>         Component: DMD
>        AssignedTo: bugzilla@digitalmars.com
>        ReportedBy: wbaxter@gmail.com
>
>
> It would be nice if "else static if" could be written "static else if".
> Or
> even better just "else if".
>
> "else static if" struck me as odd the first time I saw it, and since then
> I
> continue to mistype it as "static else if" pretty much every time I use
> it.

Then stop typing it that way ;)

>
> So it would be great if "static else if" just worked.
>
> On the other hand, I'm pretty sure
>  static if (foo) { }
>  else if (bar) { }
> is an error currently, so it would be nice if that were just taken to be a
> "static else if".  There isn't a "static else {}" so why should we have to
> repeat the "static" on an "else if"?  One "static" should be enough.

It's not.  "static if" is a statement of its own, and "if" is a statement of its own.  This:

static if(foo) {}
else if(bar) {}

is legal.  It means if foo evaluates to true, compile in the static if's body.  Otherwise, compile in the "if(bar) {}" statement.

There's no "static else" because the grammar for static if is "static if(cond) statment 'else' statement'.  So you chain static ifs with "static if(){} else static if(){}" because "static if" is a statement, and it comes after that "else".

> Oh, and besides that it's
> already ambiguous with plain ifs...
>  if (foo)
>     if (biff)
>         single_statement();
>  else if (bar) { }
>
> Exactly the same deal -- who does the else belong to?

It belongs to the if(biff) statement.

> As an added bonus either of these changes would make the Emacs D-mode
> capable
> of indenting static ifs with else ifs properly.  That's not a sufficient
> reason
> to make the change (emacs D-mode could be made to work somehow) But the
> fact
> that cc-mode basically sees "else static if" as an aberration was the
> thing
> that finally drove me to file an enhancement request on this.

Unfortunately I think this would make the grammar needlessly complicated.

And don't use emacs; use vim ;)

Jarrett Billingsley wrote:
> <d-bugmail@puremagic.com> wrote in message news:bug-1053-3@http.d.puremagic.com/issues/...

> Unfortunately I think this would make the grammar needlessly complicated.

> 

Maybe perl/python/ruby/cpp/tcl were right.  Else if should be a single keyword.  Then it wouldn't be an issue.

> And don't use emacs; use vim ;)

I can't live without my M-x hanoi.


--bb

"Bill Baxter" <dnewsgroup@billbaxter.com> wrote in message news:et2qs5$2okg$1@digitalmars.com...
> Maybe perl/python/ruby/cpp/tcl were right.  Else if should be a single keyword.  Then it wouldn't be an issue.

Phh.. that's because their grammars aren't as flexible ;)

But C++ doesn't have "elseif" as a single word.  Maybe you're thinking of the "#elif" preprocessor command?

> I can't live without my M-x hanoi.

lol.

Jarrett Billingsley wrote:
> "Bill Baxter" <dnewsgroup@billbaxter.com> wrote in message news:et2qs5$2okg$1@digitalmars.com...
>> Maybe perl/python/ruby/cpp/tcl were right.  Else if should be a single keyword.  Then it wouldn't be an issue.
> 
> Phh.. that's because their grammars aren't as flexible ;)
> 
> But C++ doesn't have "elseif" as a single word.  Maybe you're thinking of the "#elif" preprocessor command?

I didn't see 'c++' mentioned, only 'cpp', which is (confusingly) also a common acronym for the C (or C++) pre-processor. So I'd guess he was indeed thinking of #elif :).

"Frits van Bommel" <fvbommel@REMwOVExCAPSs.nl> wrote in message news:et3mrr$14cu$2@digitalmars.com...

> I didn't see 'c++' mentioned, only 'cpp', which is (confusingly) also a common acronym for the C (or C++) pre-processor. So I'd guess he was indeed thinking of #elif :).

Ohh!  That's right.  You're probably right.

Jarrett Billingsley wrote:
> "Frits van Bommel" <fvbommel@REMwOVExCAPSs.nl> wrote in message news:et3mrr$14cu$2@digitalmars.com...
> 
>> I didn't see 'c++' mentioned, only 'cpp', which is (confusingly) also a common acronym for the C (or C++) pre-processor. So I'd guess he was indeed thinking of #elif :).
> 
> Ohh!  That's right.  You're probably right. 

Yep.  That's what I meant.

--bb

http://d.puremagic.com/issues/show_bug.cgi?id=1053





------- Comment #2 from wbaxter@gmail.com  2007-03-12 18:35 -------
(For the record, there was some discussion on this on the newsgroup Frits' message above was in reply to something there about which languages use an 'elseif' keyword).

Anyway, I can see the difficulties involved in the implementation of this, but I still think it would make D better for users.  In the end, it's just an enhancement request.  If Walter agrees then maybe something will be done about it, if not then he'll close it.

But here's another attempt to sell the idea: I'm pretty sure there are very few C/C++/D/Java users who really think of

if () {
}
else if () {
}
else if () {
}
else () {
}

as this:

if () {
}
else {
   if () {
   }
   else {
       if () {
       }
       else () {
       }
   }
}

which is pretty much the way you have to be thinking for 'else static if' to make sense.


-- 

http://d.puremagic.com/issues/show_bug.cgi?id=1053


deewiant@gmail.com changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Version|1.008                       |1.009




------- Comment #3 from deewiant@gmail.com  2007-03-17 12:03 -------
That _is_ the way I often think of if-else. :-P I prefer the way it currently works, as I do sometimes write:

static if
else if

And it means what I want it to mean.

Sometimes, though, I find that it's clearer to write something as:

if
else {
  if
  else
}

Instead of:

if
else if
else

Depending on what exactly is in the first else.

"static else if" would complicate the grammar a bit (written in about a minute,
not sure if the following would be correct):
  if: "if [expression] [statement] (else [statement])"
  static if: "[static if [expression] [statement] (else [statement])] | [static
if [expression] [statement] (static else [if-statement])]"

As opposed to the current:
  if or static if: "(static) if [expression] [statement] (else [statement])"


--