In article <55o19v$6o5@news.service.uci.edu>, dan@cafws4.eng.uci.edu
(Dan Harkless) writes:

|> 1.  I don't understand what implication not doing lvalue-to-rvalue has.

I believe one implication is that it does not actually perform a dereference
operations when applied to an expression like '*v'.  Thus 'v' need not
contain a valid pointer in order to use 'typeid(*v)'.

|> 2.  Not doing array-to-pointer seems to indicate that this code:
|>
|>  int   a[3];
|>
|>  cout << typeid(a).name() << endl;
|>  cout << (typeid(a) == typeid(int*)) << endl;
|>
|>     will produce:
|>
|>  int[3]
|>  0

It does seem like that to me as well.

Borland looks broken to me too.

--
swf@elsegundoca.ncr.com  sarima@ix.netcom.com

The peace of God be with you.
---
[ comp.std.c++ is moderated.  To submit articles: try just posting with      ]
[ your news-reader.  If that fails, use mailto:std-c++@ncar.ucar.edu         ]
[ FAQ:      http://reality.sgi.com/employees/austern_mti/std-c++/faq.html    ]
[ Policy:   http://reality.sgi.com/employees/austern_mti/std-c++/policy.html ]
[ Comments? mailto:std-c++-request@ncar.ucar.edu                             ]

Stan Friesen <swf@elsegundoca.ncr.com> writes:
 >In article <55o19v$6o5@news.service.uci.edu>, dan@cafws4.eng.uci.edu
 >(Dan Harkless) writes:
 >
 >|> 1.  I don't understand what implication not doing lvalue-to-rvalue has.
 >
 >I believe one implication is that it does not actually perform a dereference
 >operations when applied to an expression like '*v'.  Thus 'v' need not
 >contain a valid pointer in order to use 'typeid(*v)'.

 Actually, it will dereference the v if v is a polymorphic type.  If it
isn't, it won't dereference the pointer thanks to the wording "The expression
is not evaluated", not thanks to lvalue-to-rvalue conversion not being done,
right?

Dan Harkless
dan@unitech.com
Unitech Research, Inc.
---
[ comp.std.c++ is moderated.  To submit articles: try just posting with      ]
[ your news-reader.  If that fails, use mailto:std-c++@ncar.ucar.edu         ]
[ FAQ:      http://reality.sgi.com/employees/austern_mti/std-c++/faq.html    ]
[ Policy:   http://reality.sgi.com/employees/austern_mti/std-c++/policy.html ]
[ Comments? mailto:std-c++-request@ncar.ucar.edu                             ]

[This article has been posted and then reposted by the moderator but has never
showed up at either of the (unrelated) sites at which I receive news.  I am
trying again...]

Hello.  My company is currently upgrading our compiler to incorporate RTTI
based on the latest version of the standard.  Unfortunately, 5.2.8, Type
identification, is quite vague for many cases.

The C++ Draft Standard states in 5.2.8-3: "When typeid is applied to an
expression other than an lvalue of polymorphic class type, the result refers
to a type_info object representing the static type of the expression.
Lvalue-to-rvalue (4.1), array-to-pointer (4.2), and function-to-pointer (4.3)
conversions are not applied to the expression.

1.  I don't understand what implication not doing lvalue-to-rvalue has.

2.  Not doing array-to-pointer seems to indicate that this code:

 int   a[3];

 cout << typeid(a).name() << endl;
 cout << (typeid(a) == typeid(int*)) << endl;

    will produce:

 int[3]
 0

    However, this is not what Borland C++ 5.0 and Microsoft Visual C++ 4.2 do.
    They produce:

 int*
 1

    It sounds like they are not following 5.2.8-3, right?

3.  Not doing function-to-pointer conversions leaves a real quandry for what
    to do for the following case:

 double  f(int);
 double  (*f_ptr)(int);

 cout << typeid(f(1)).name() << endl;
 cout << typeid(f).name() << endl;
 cout << typeid(f_ptr).name() << endl;
 cout << (typeid(f) == typeid(f_ptr)) << endl;

    We are guessing:

 double
 double()(int)  // or double (int)
 double(*)(int)
 0

   However, this is not what Borland C++ 5.0 and Microsoft Visual C++ 4.2 do.
   They give:

 double
     double(*)(int)
     double(*)(int)
     1

    Again, sounds like they are not following 5.2.8-3.  We realize the .name()
    member function is implementation-defined, but what would the preferred
    .name() for f above be?  "double()(int)"?  "double (int)"?


Regards,

Dan Harkless
Unitech Research, Inc.
dan@unitech.com
---
[ comp.std.c++ is moderated.  To submit articles: try just posting with      ]
[ your news-reader.  If that fails, use mailto:std-c++@ncar.ucar.edu         ]
[ FAQ:      http://reality.sgi.com/employees/austern_mti/std-c++/faq.html    ]
[ Policy:   http://reality.sgi.com/employees/austern_mti/std-c++/policy.html ]
[ Comments? mailto:std-c++-request@ncar.ucar.edu                             ]