public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
* [Bug c++/16234] New: Bad diagnostic
@ 2004-06-27 17:41 igodard at pacbell dot net
  2004-06-27 18:14 ` [Bug c++/16234] " pinskia at gcc dot gnu dot org
  0 siblings, 1 reply; 2+ messages in thread
From: igodard at pacbell dot net @ 2004-06-27 17:41 UTC (permalink / raw)
  To: gcc-bugs

template<typename T> struct A {
    template<typename U> void Foo(){} };
template<typename V>struct B {
    void Bar(V* v) {
        v->template Foo<int>();
        v->Foo<int>();
        }
    };
int main() {
    A<bool> a;
    B<A<bool> > b;
    b.Bar(&a);
    }

gets you:

~/ootbc/common/test/src$ g++ foo.cc
foo.cc: In member function `void B<V>::Bar(V*)':
foo.cc:6: error: expected primary-expression before "int"
foo.cc:6: error: expected `;' before "int"

The actual problem is the missing explicit "template" qualification. Surely the compiler can recognize "<int>" as a template argument list in this case for purposes of the diagnostic. There is an eroror here - the rules do require a diagnostic - but the compiler could be a lot more graceful about what it says when it doesn't get the required qualifier.


I grant that templates with data arguments cannot be as easily recognized in such a context. In:

template<typename T> struct A {
    template<int I> void Foo(){} };
template<typename V>struct B {
    void Bar(V* v) {
        v->template Foo<int>();
        v->Foo<5>();
        }
    };
int main() {
    A<bool> a;
    B<A<bool> > b;
    b.Bar(&a);
    }

ot is probably too tough to figure out what is wrong (operator <(void(*)(), int)???, although that is very unlikely and would be syntactically wrong here even if the operator existed). Just assuming that "anyTemplateFunc<" is the start of a template invocation would give the best diagnostic most likely, and the compiler can complain that the assumption is not permitted by the lanhuage. One of the fameous "ISO C++ requires that the invocation 'Foo<int>' be explicitly qualified as 'template Foo<int>'" messages, maybe?

Ivan

-- 
           Summary: Bad diagnostic
           Product: gcc
           Version: 3.4.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: c++
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: igodard at pacbell dot net
                CC: gcc-bugs at gcc dot gnu dot org


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=16234


^ permalink raw reply	[flat|nested] 2+ messages in thread

* [Bug c++/16234] Bad diagnostic
  2004-06-27 17:41 [Bug c++/16234] New: Bad diagnostic igodard at pacbell dot net
@ 2004-06-27 18:14 ` pinskia at gcc dot gnu dot org
  0 siblings, 0 replies; 2+ messages in thread
From: pinskia at gcc dot gnu dot org @ 2004-06-27 18:14 UTC (permalink / raw)
  To: gcc-bugs


------- Additional Comments From pinskia at gcc dot gnu dot org  2004-06-27 18:09 -------
I should note that the template keyword is required by the standard, this is dup of bug 16233.

Note a->b the type of b is not known untill the template is instantiated  so a->b < 5 would be valid 
until then and will cause an error when the template is instantiated.

*** This bug has been marked as a duplicate of 16233 ***

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
         Resolution|                            |DUPLICATE


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=16234


^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2004-06-27 18:09 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2004-06-27 17:41 [Bug c++/16234] New: Bad diagnostic igodard at pacbell dot net
2004-06-27 18:14 ` [Bug c++/16234] " pinskia at gcc dot gnu dot org

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).