Help requested on C++ template syntax (for Emacs development).

Help requested on C++ template syntax (for Emacs development).

[Alan Mackenzie]

Hi, GCC list!

Apologies if this isn't the right place to ask, but it seems like a good
place to find experts on C++ and Java syntax.  If there's a better place
to ask, would somebody give me a pointer, please.

I'm the maintainer of Emacs's CC Mode, which includes modes for C, C++,
Java, etc.

Currently, C++ templates are not handled 100% accurately by C++ mode -
neither are generics in Java mode, for that matter.  I want to fix this.
My knowledge of C++ is moderate rather than comprehensive.

So, the question: is it possible to identify with 100% certainty, PURELY
SYNTACTICALLY (i.e. without access to the compiler's symbol table),
when "< ... >" is a pair of template (C++) or generic (Java) brackets?

I'm thinking of things like

    foo (a < b, c > d);

I think this is unambiguously a function call with 2 parameters, the
expressions "a < b" and "c > d".  It cannot be be one with 1 parameter
beginning with the template invocation "a < b , c >".  Or can it?

I believe that a pair of template brackets cannot enclose a semicolon,
for example.

Another related question: although there is no maximum bound on how far
apart template/generic brackets can be, I believe that in practice, they
are never that far apart (a few hundred bytes max, perhaps).  Is this, in
fact, the case?

A cc: to acm@muc.de would be most appreciated.

Thanks in advance for any and all help!

-- 
Alan Mackenzie (Nuremberg, Germany).

Re: Help requested on C++ template syntax (for Emacs development).

[Ian Lance Taylor]

Alan Mackenzie <acm@muc.de> writes:

> I'm thinking of things like
>
>     foo (a < b, c > d);
>
> I think this is unambiguously a function call with 2 parameters, the
> expressions "a < b" and "c > d".  It cannot be be one with 1 parameter
> beginning with the template invocation "a < b , c >".  Or can it?

No, it can't be, because a<b, c> is a type.  The result would be
foo(TYPE d), which can not be a function call.  On the other hand, if
there were a type before foo then this would be a function
declaration.  For example, this is valid C++ code:

template <int a1, int a2> class a;

int fn(int d, int e)
{
  const int b = 1;
  const int c = 2;
  typedef int f;
  f foo (int, int);
  f foo (a < b, c > d);
  foo (e < b, c > d);
}

The line "f foo (a < b, c > d);" uses a template, the line "foo (e <
b, c > d);" does not.

I rather doubt that you can purely syntactically, fully reliably,
determine whether <> refers to a template, but I don't know for sure.


> Another related question: although there is no maximum bound on how far
> apart template/generic brackets can be, I believe that in practice, they
> are never that far apart (a few hundred bytes max, perhaps).  Is this, in
> fact, the case?

A few hundred characters is probably a little too small, but in
practice I think one thousand characters is probably usually
sufficient.

Ian

Re: Help requested on C++ template syntax (for Emacs development).

[Ian Lance Taylor]

Ian Lance Taylor <iant@google.com> writes:

> Alan Mackenzie <acm@muc.de> writes:
>
>> I'm thinking of things like
>>
>>     foo (a < b, c > d);
>>
>> I think this is unambiguously a function call with 2 parameters, the
>> expressions "a < b" and "c > d".  It cannot be be one with 1 parameter
>> beginning with the template invocation "a < b , c >".  Or can it?
>
> No, it can't be, because a<b, c> is a type.  The result would be
> foo(TYPE d), which can not be a function call.  On the other hand, if
> there were a type before foo then this would be a function
> declaration.  For example, this is valid C++ code:
>
> template <int a1, int a2> class a;
>
> int fn(int d, int e)
> {
>   const int b = 1;
>   const int c = 2;
>   typedef int f;
>   f foo (int, int);
>   f foo (a < b, c > d);
>   foo (e < b, c > d);
> }
>
> The line "f foo (a < b, c > d);" uses a template, the line "foo (e <
> b, c > d);" does not.

Oh, wait, I forgot about constructors.  This is valid C++ code:

template <int a1, int a2> class a;

const int b = 1;
const int c = 2;
int d;
int e;

class foo
{
  foo (a < b, c > d);
};

extern void quux (int, int);

void bar()
{
  quux (e < b, c > d);
}

Inside the class foo, you've got a template, inside the function bar,
you've got a function call.

I think you're going to have a hard time with this.

Ian

Re: Help requested on C++ template syntax (for Emacs development).

[Alan Mackenzie]

Hi, Ian!

On Tue, Jun 10, 2008 at 12:05:20PM -0700, Ian Lance Taylor wrote:
> Ian Lance Taylor <iant@google.com> writes:

> > Alan Mackenzie <acm@muc.de> writes:

> >> I'm thinking of things like

> >>     foo (a < b, c > d);

[ .... ]

> Oh, wait, I forgot about constructors.  This is valid C++ code:

> template <int a1, int a2> class a;

> class foo
> {
>   foo (a < b, c > d);
> };

> extern void quux (int, int);

> void bar()
> {
>   quux (e < b, c > d);
> }

> Inside the class foo, you've got a template, inside the function bar,
> you've got a function call.

> I think you're going to have a hard time with this.

Er, yes.  ;-)  Thank you very much for all the help.  It's going to take
some while to sink in.  C++ is a difficult language.  Hopefully, in a few
months time, Emacs will be doing better than it currently is.

> Ian

-- 
Alan Mackenzie (Nuremberg, Germany).

Re: Help requested on C++ template syntax (for Emacs development).

[Tom Tromey]

>>>>> "Alan" == Alan Mackenzie <acm@muc.de> writes:

Alan> So, the question: is it possible to identify with 100% certainty, PURELY
Alan> SYNTACTICALLY (i.e. without access to the compiler's symbol table),
Alan> when "< ... >" is a pair of template (C++) or generic (Java) brackets?

In Java, yes, if the source is error-free.

Alan> I'm thinking of things like
Alan>     foo (a < b, c > d);
Alan> I think this is unambiguously a function call with 2 parameters, the
Alan> expressions "a < b" and "c > d".

In Java this is a function call.  If 'a' is a type, this is an error,
since you can't declare 'd' here.

So... if cc-mode assumes that the code it is looking at is
syntactically correct, you can always guess.  But, it is a lot nicer
for the user if errors are flagged.  That way lies the Eclipse
approach :-)

Alan> Another related question: although there is no maximum bound on how far
Alan> apart template/generic brackets can be, I believe that in practice, they
Alan> are never that far apart (a few hundred bytes max, perhaps).  Is this, in
Alan> fact, the case?

For Java, yeah, generally speaking.  There may be exceptions.  And
sometimes I think you may find surprising amounts of whitespace in
there.

For C++, I think you are just doomed.  Even if you could get the
compiler to emit perfect information, it would only emit information
about the code it actually saw -- not stuff in ignored #if groups.

Tom