Issue #021
March, 1997


Contents:

Pointers to Members
A New Angle on Function Pointers
Notes From ANSI/ISO - Clarifications on Exception Handling
Introduction to STL Part 8 - advance() and distance()


POINTERS TO MEMBERS

In ANSI C, function pointers are used like this:

        #include <stdio.h>

        void f(int i)
        {
                printf("%d\n", i);
        }

        typedef void (*fp)(int);

        void main()
        {
                fp p = &f;

                (*p)(37);       /* these are equivalent */

                p(37);
        }

and are employed in a variety of ways, for example to specify a
comparison function to a library function like qsort().

In C++, pointers can be similarly used, but there are a couple of
quirks to consider.  We will discuss two of them in this section, and
another one in the next section.

The first point to mention is that C++ has C-style functions in it, but
also has other types of functions, notably member functions.  For
example:

        class A {
        public:
                void f(int);
        };

In this example, A::f(int) is a member function.  That is, it
operates on object instances of class A, and the function itself has
a "this" pointer that points at the instance in question.

Because C++ is a strongly typed language, it is desirable that a
pointer to a member function be treated differently than a pointer to
a C-style function, and that a pointer to a function member of class A
be distinguished from a pointer to a member of class B.  To do this,
we can say:

        #include <iostream.h>

        class A {
        public:
                void f(int i) {cout << "value is: " << i << "\n";}
        };

        typedef void (A::*pmfA)(int);

        pmfA x = &A::f;

        void main()
        {
                A a;
                A* p = &a;

                (p->*x)(37);
        }

Note the notation for actually calling the member function.

It is not possible to intermix such a type with other pointer types,
so for example:

        void f(int) {}

        pmfA x = &f;

is invalid.

A static member function, as in:

        class A {
        public:
                static void g(int);
        };

        typedef void (*fp)(int);

        fp p = &A::g;

is treated like a C-style function.  A static function has no "this"
pointer and does not operate on actual object instances.

Pointers to members are typically implemented just like C function
pointers, but there is an issue with their implementation in cases
where inheritance is used.  In such a case, you have to worry about
computing offsets of subobjects, and so on, when calling a member
function, and for this purpose a runtime structure similar to a
virtual table used for virtual functions is used.

It's also possible to have pointers to data members of a class, with
the pointer representing an offset into a class instance.  For example:

        #include <iostream.h>

        class A {
        public:
                int x;
        };

        typedef int A::*piA;
        piA x = &A::x;

        void main()
        {
                A a;
                A* p = &a;

                a.x = 37;

                cout << "value is: " << p->*x << "\n";
        }

Note that saying "&A::x" does not take the address of an actual data
member in an instance of A, but rather computes a generic offset that
can be applied to any instance.


A NEW ANGLE ON FUNCTION POINTERS

The discussion on function pointers in this issue overlooks one key
angle that has fairly recently been introduced into the language.
This involves distinguishing between C and C++ pointers.  A C-style
pointer in C++, that is, one that does not point to a member function,
is used just like a function pointer in C.  But according to the
standard (section 7.5), such a pointer in fact has a different type.

For example, consider:

        extern "C" typedef void (*fp1)(int);

        extern "C++" typedef void (*fp2)(int);

        extern "C" void f(int);

fp1 and fp2 are not the same type, and saying:

        fp2 p = &f;

to initialize p to the f(int) declared in the 'extern "C"' will not
work.

It is possible to overload functions on this basis, so that for
example:

        extern "C" void f(void (*)(int));

        extern "C++" void f(void (*)(int));

is legal, with the appropriate f() called based on the function
pointer type passed to it.  The function pointer parameter types in
this example are not identical; the first is a pointer to a C
function, the second a pointer to a C++ one.

This feature is new and may not be implemented in your local C++
compiler.


NOTES FROM ANSI/ISO - CLARIFICATIONS ON EXCEPTION HANDLING
Jonathan Schilling, jls@sco.com

The ANSI/ISO C++ standards meeting earlier this month in Nashua, New
Hampshire, produced some clarifications of exception handling
semantics.

One interesting case is this one, which was featured in the January
1997 issue of the magazine C++ Report:

        try {
                // exception prone code here, that may do a throw
        }
        catch (...) {
                // common error code here
                
                try {
                        throw;  // re-throw to more specific handler
                }
                catch (ExceptA&) {
                        // handle ExceptA here
                }
                catch (ExceptB&) {
                        // handle ExceptB here
                }
                catch (...) {
                        // handle unknown exceptions here
                }

                throw;
        }

The idea behind the code is to factor out common error handling logic 
into the first part of the catch handler (so as not to replicate it), 
rethrow the exception to get error handling specific to the exception 
in the individual inner handlers, and then finally to rethrow the 
exception again to let functions further up the call chain do their 
handling.

The question is, does this code work as intended?  The draft standard
speaks of a throw creating a temporary object that is then deleted when
the corresponding handler exits.  Does this mean that when the inner
handlers above exit, the rethrow will be of a nonexistent temporary
object?  The standard isn't really clear on this, and some existing 
compilers have been found to do the deletion at the inner handler,
with the result that the program crashes.

The answer is that this code should indeed work as intended, and that
the existing compilers for which this does not work are wrong.
(Fortunately SCO's new C++ compiler is one of the ones that is getting
it right!).
 
Furthermore, the committee stated that the value of the standard library 
function uncaught_exception() (see C++ Newsletter #019) changes (from 
false to true) at both of the rethrows, until such time as the rethrown
exception is caught again.

Another exception handling issue that was clarified is whether base
class destructors are called when a derived class destructor throws an 
exception:
                
        class B {
        public:
                ~B() { ... }
        };

        class D : public B {
        public:
                ~D() { throw "error"; }
        };

        void f() {
                try {
                        D d;
                }
                catch (...) { }
        }

Does ~B() get called as well as ~D()?  The answer is yes.  This may 
seem almost obvious -- it is part of the general principle of C++
that constructed subobjects always get destroyed if something goes
wrong with the enclosing object -- but in fact there was some debate
on this within the committee.

Finally, one of the comments from the ANSI public review period
concerned an area of exception handling that needed no clarification
but is often misunderstood:

        try {
                throw 0;
        }
        catch (void *) {
                // does the exception get caught here?
        }

The handler should not catch the exception, but apparently in some
compilers it does.  The draft standard is clear that throw and catch
types either have to match exactly, or be related by inheritance, or
be subject to a pointer-to-pointer standard conversion.  Since 0 is
not of a pointer type, the last requirement isn't met, and no handler
is found.  Similarly note that the whole range of other standard
conversions do not apply, so that for example a handler of type long
does not catch an exception of type int.


INTRODUCTION TO STL PART 8 - ADVANCE() AND DISTANCE()

In the last issue we started discussing iterators.  They are used in
the Standard Template Library to provide access to the contents of
data structures, and to cycle across multiple data elements.

We presented two examples of iterator usage, the first involving
pointers, the second a higher-level construct.  Both of these examples
require some grasp of pointer arithmetic, a daunting subject.  There's
another way to write the example we presented before, using a couple
of STL iterator functions:

        #include <algorithm>
        #include <iterator>
        #include <vector>
        #include <iostream>

        using namespace std;

        const int N = 100;

        void main()
        {
                vector<int> iv(N);

                iv[50] = 37;
                iv[52] = 47;

                vector<int>::iterator iter = find(iv.begin(), iv.end(), 37);
                if (iter == iv.end()) {
                        cout << "not found\n";
                }
                else {
                        int d = 0;
                        distance(iv.begin(), iter, d);
                        //cout << "found at " << iter - iv.begin() << "\n";
                        cout << "found at " << d << "\n";
                }

                advance(iter, 2);
                cout << "value = " << *iter << "\n";
        }

The function distance() computes the distance between two iterator
values.  In this example, we know that we're starting at "iv.begin()",
the beginning of the integer vector.  And we've found a match at
"iter", and so we can use distance() to compute the distance between
these, and display this result.  Note that more recently distance()
has been changed to work more like a regular function, with the
beginning and ending arguments supplied and the difference returned as
the result of the function:

        d = distance(iv.begin(), iter);

A similar issue comes up with advancing an iterator.  For example,
it's possible to use "++" for this, but cumbersome when you wish to
advance the iterator a large value.  Instead of ++, advance() can be
used to advance the iterator a specified number of positions.  In the
example above, we move the iterator forward 2 positions, and then
display the value stored in the vector at that location.

These functions provide an alternative way of manipulating iterators,
that does not depend so much on pointer arithmetic.


ACKNOWLEDGEMENTS

Thanks to Nathan Myers, Eric Nagler, David Nelson, and Jonathan
Schilling for help with proofreading.


SUBSCRIPTION INFORMATION / BACK ISSUES

To subscribe to the newsletter, send mail to majordomo@world.std.com
with this line as its message body:

subscribe c_plus_plus

Back issues are available via FTP from:

        rmi.net /pub2/glenm/newslett

or on the Web at:

        http://rainbow.rmi.net/~glenm

There is also a Java newsletter.  To subscribe to it, say:

subscribe java_letter

using the same majordomo@world.std.com address.

-------------------------

Copyright (c) 1997 Glen McCluskey.  All Rights Reserved.

This newsletter may be further distributed provided that it is copied
in its entirety, including the newsletter number at the top and the
copyright and contact information at the bottom.

Glen McCluskey & Associates
Professional C++ Consulting
Internet: glenm@glenmccl.com
Phone: (800) 722-1613 or (970) 490-2462
Fax: (970) 490-2463
FTP: rmi.net /pub2/glenm/newslett (for back issues)
Web: http://rainbow.rmi.net/~glenm