By using this site, you agree to our updated Privacy Policy and our Terms of Use. Manage your Cookies Settings.
446,305 Members | 1,619 Online
Bytes IT Community
+ Ask a Question
Need help? Post your question and get tips & solutions from a community of 446,305 IT Pros & Developers. It's quick & easy.

abstract base class question

P: n/a
Is it possible to create an abstract base class without a pure virtual
function?

Thanks

Jack

Jul 9 '06 #1
Share this Question
Share on Google+
14 Replies


P: n/a
In article <11**********************@m73g2000cwd.googlegroups .com>,
ju******@gmail.com says...
Is it possible to create an abstract base class without a pure virtual
function?
It's entirely possible to create a class and _use_ it as an abstract
base class (i.e. you only ever derive from it, never instantiate it
directly) without a pure virtual function.

Using a pure virtual function enforces that intent -- i.e. the
compiler won't _let_ you instantiate a class that contains a pure
virtual function (or a derivative, unless all pure virtual functions
have been overridden).

--
Later,
Jerry.

The universe is a figment of its own imagination.
Jul 9 '06 #2

P: n/a
Thank a lot.
>
It's entirely possible to create a class and _use_ it as an abstract
base class (i.e. you only ever derive from it, never instantiate it
directly) without a pure virtual function.
So how to do it? Could you give an example?

Thanks

Jack

Jul 9 '06 #3

P: n/a
Jerry Coffin wrote:
In article <11**********************@m73g2000cwd.googlegroups .com>,
ju******@gmail.com says...
>Is it possible to create an abstract base class without a pure virtual
function?

It's entirely possible to create a class and _use_ it as an abstract
base class (i.e. you only ever derive from it, never instantiate it
directly) without a pure virtual function.

Using a pure virtual function enforces that intent -- i.e. the
compiler won't _let_ you instantiate a class that contains a pure
virtual function (or a derivative, unless all pure virtual functions
have been overridden).
I guess that OP could use a protected constructor or destructor, as well.
Jul 9 '06 #4

P: n/a
In article <11**********************@m79g2000cwm.googlegroups .com>,
ju******@gmail.com says...
Thank a lot.

It's entirely possible to create a class and _use_ it as an abstract
base class (i.e. you only ever derive from it, never instantiate it
directly) without a pure virtual function.

So how to do it? Could you give an example?
You write a class. You don't instantiate it. If you want an example
of a class that could be derived from:

class X {
virtual ~X() {}
};

If you intend something to be used as a base class, you typically
want it to have a virtual dtor. Anything else will usually be
specific to what that class is intended to do. I can't really give an
example of not instantiating it, because there's no code involved.
You write things like 'class Y : public X;', but never 'X x;' or
anything like that.

--
Later,
Jerry.

The universe is a figment of its own imagination.
Jul 9 '06 #5

P: n/a
red floyd wrote:
I guess that OP could use a protected constructor or destructor, as well.
That won't totally prevent instantiation:

class X {
protected;
X();
public:
static X* MakeX() { return new X; }
};

Jul 9 '06 #6

P: n/a
ju******@gmail.com wrote:
Is it possible to create an abstract base class without a pure virtual
function?
Well, there are way to make one that behaves similiar, but why would you
want to? You can always make the destructor pure if there is no other
member function that you could make pure.

Jul 9 '06 #7

P: n/a
Rolf Magnus wrote:
ju******@gmail.com wrote:
Is it possible to create an abstract base class without a pure virtual
function?

Well, there are way to make one that behaves similiar, but why would you
want to? You can always make the destructor pure if there is no other
member function that you could make pure.
Thats correct!
To elaborate more on Rolfs point,
A pure virtual function can have body but it can only be called from
the derived classes and not by user directly. So for example:

class A {
virtual void fun() = 0;
};

void a::fun() { _DO_SOMETHING_}

class B : public A {
void fun() {
A::fun(); // works fine
}
};

void main() {
A myA; // fails as the class is abstract..

B myB;
myB.fun(); // works fine
}

Now that said, If you dont want any function to be pure virtual it
means you intend to call them through derived class objects. In this
case you can make destructor pure virtual and provide it a body as well
because when ur class is extended (which it must since it is abstract)
it will be destroyed properly since destructor will always be called
through the derived class destructor which c++ compiler will allow.

Ramneek

Jul 9 '06 #8

P: n/a
On 9 Jul 2006 03:11:48 -0700, "rami" <on***********@gmail.comwrote
in comp.lang.c++:
Rolf Magnus wrote:
ju******@gmail.com wrote:
Is it possible to create an abstract base class without a pure virtual
function?
Well, there are way to make one that behaves similiar, but why would you
want to? You can always make the destructor pure if there is no other
member function that you could make pure.

Thats correct!
To elaborate more on Rolfs point,
A pure virtual function can have body but it can only be called from
the derived classes and not by user directly. So for example:

class A {
virtual void fun() = 0;
};

void a::fun() { _DO_SOMETHING_}

class B : public A {
void fun() {
A::fun(); // works fine
}
};

void main() {
You shouldn't post advice here until you become aware of the fact that
there is no such thing as "void main()" in C++. The standard requires
that main() be defined with a return type of int. "void main()" is
ill-formed, requiring a diagnostic. The behavior of any executable
generated is completely undefined.

--
Jack Klein
Home: http://JK-Technology.Com
FAQs for
comp.lang.c http://c-faq.com/
comp.lang.c++ http://www.parashift.com/c++-faq-lite/
alt.comp.lang.learn.c-c++
http://www.contrib.andrew.cmu.edu/~a...FAQ-acllc.html
Jul 9 '06 #9

P: n/a

Jack Klein wrote:
On 9 Jul 2006 03:11:48 -0700, "rami" <on***********@gmail.comwrote
in comp.lang.c++:
Rolf Magnus wrote:
ju******@gmail.com wrote:
>
Is it possible to create an abstract base class without a pure virtual
function?
>
Well, there are way to make one that behaves similiar, but why would you
want to? You can always make the destructor pure if there is no other
member function that you could make pure.
Thats correct!
To elaborate more on Rolfs point,
A pure virtual function can have body but it can only be called from
the derived classes and not by user directly. So for example:

class A {
virtual void fun() = 0;
};

void a::fun() { _DO_SOMETHING_}

class B : public A {
void fun() {
A::fun(); // works fine
}
};

void main() {

You shouldn't post advice here until you become aware of the fact that
there is no such thing as "void main()" in C++. The standard requires
that main() be defined with a return type of int. "void main()" is
ill-formed, requiring a diagnostic. The behavior of any executable
generated is completely undefined.

--
Jack Klein
Home: http://JK-Technology.Com
FAQs for
comp.lang.c http://c-faq.com/
comp.lang.c++ http://www.parashift.com/c++-faq-lite/
alt.comp.lang.learn.c-c++
http://www.contrib.andrew.cmu.edu/~a...FAQ-acllc.html
i think you shudnt be posting here till you spit that attitude out and
realize that this isnt my full time job and i am only trying to help
someone (It was a typo). Or is it that you dont have anything better to
post? Or is forum paying you for picking up garbage?

Jul 9 '06 #10

P: n/a
* rami:
Jack Klein wrote:
>On 9 Jul 2006 03:11:48 -0700, "rami" <on***********@gmail.comwrote
in comp.lang.c++:
>>Rolf Magnus wrote:
ju******@gmail.com wrote:

Is it possible to create an abstract base class without a pure virtual
function?
Well, there are way to make one that behaves similiar, but why would you
want to? You can always make the destructor pure if there is no other
member function that you could make pure.
Thats correct!
To elaborate more on Rolfs point,
A pure virtual function can have body but it can only be called from
the derived classes and not by user directly. So for example:

class A {
virtual void fun() = 0;
};

void a::fun() { _DO_SOMETHING_}

class B : public A {
void fun() {
A::fun(); // works fine
}
};

void main() {
You shouldn't post advice here until you become aware of the fact that
there is no such thing as "void main()" in C++. The standard requires
that main() be defined with a return type of int. "void main()" is
ill-formed, requiring a diagnostic. The behavior of any executable
generated is completely undefined.

--
Jack Klein
Home: http://JK-Technology.Com
FAQs for
comp.lang.c http://c-faq.com/
comp.lang.c++ http://www.parashift.com/c++-faq-lite/
alt.comp.lang.learn.c-c++
http://www.contrib.andrew.cmu.edu/~a...FAQ-acllc.html

i think you shudnt be posting here till you spit that attitude out and
realize that this isnt my full time job and i am only trying to help
someone (It was a typo). Or is it that you dont have anything better to
post? Or is forum paying you for picking up garbage?
Jack is right. Your posting contained a few errors and misleading text,
but formulated in such a way that many people would take it seriously
(they won't after your reply quoted above, but that's another matter).
Here are three content problems:

* You state "can only be called from the derived classes".
That is incorrect.

* You use "_DO_SOMETHING_" as if that is a valid macro name.
It isn't valid, at least not for your own macro. Names starting
with underscore followed by uppercase are reserved.

* You use "void main".
Already discussed by Jack.

So, please try to pick up a little about how this group works.

Nobody (at least, not I! ;-)) will yell at you for helping out with
facts that you know, or stating opinions, or giving advice in general;
if such postings were removed, then this group would fall from being a
high-traffic, useful group to being a dried-out very sporadic traffic
and mostly useless. It's when the advice is formulated as authoritative
but is in fact something you really don't know anything about, that it's
a problem. Because it might then mislead people and cause new "urban
legends" to be perpetuated, misleading even more people than just those
who originally read your posting and thought it was good.

Cheers,

- Alf

--
A: Because it messes up the order in which people normally read text.
Q: Why is it such a bad thing?
A: Top-posting.
Q: What is the most annoying thing on usenet and in e-mail?
Jul 9 '06 #11

P: n/a

Alf P. Steinbach wrote:
* rami:
Jack Klein wrote:
On 9 Jul 2006 03:11:48 -0700, "rami" <on***********@gmail.comwrote
in comp.lang.c++:

Rolf Magnus wrote:
ju******@gmail.com wrote:

Is it possible to create an abstract base class without a pure virtual
function?
Well, there are way to make one that behaves similiar, but why would you
want to? You can always make the destructor pure if there is no other
member function that you could make pure.
Thats correct!
To elaborate more on Rolfs point,
A pure virtual function can have body but it can only be called from
the derived classes and not by user directly. So for example:

class A {
virtual void fun() = 0;
};

void a::fun() { _DO_SOMETHING_}

class B : public A {
void fun() {
A::fun(); // works fine
}
};

void main() {
You shouldn't post advice here until you become aware of the fact that
there is no such thing as "void main()" in C++. The standard requires
that main() be defined with a return type of int. "void main()" is
ill-formed, requiring a diagnostic. The behavior of any executable
generated is completely undefined.

--
Jack Klein
Home: http://JK-Technology.Com
FAQs for
comp.lang.c http://c-faq.com/
comp.lang.c++ http://www.parashift.com/c++-faq-lite/
alt.comp.lang.learn.c-c++
http://www.contrib.andrew.cmu.edu/~a...FAQ-acllc.html
i think you shudnt be posting here till you spit that attitude out and
realize that this isnt my full time job and i am only trying to help
someone (It was a typo). Or is it that you dont have anything better to
post? Or is forum paying you for picking up garbage?

Jack is right. Your posting contained a few errors and misleading text,
but formulated in such a way that many people would take it seriously
(they won't after your reply quoted above, but that's another matter).
Here are three content problems:

* You state "can only be called from the derived classes".
That is incorrect.

* You use "_DO_SOMETHING_" as if that is a valid macro name.
It isn't valid, at least not for your own macro. Names starting
with underscore followed by uppercase are reserved.

* You use "void main".
Already discussed by Jack.

So, please try to pick up a little about how this group works.

Nobody (at least, not I! ;-)) will yell at you for helping out with
facts that you know, or stating opinions, or giving advice in general;
if such postings were removed, then this group would fall from being a
high-traffic, useful group to being a dried-out very sporadic traffic
and mostly useless. It's when the advice is formulated as authoritative
but is in fact something you really don't know anything about, that it's
a problem. Because it might then mislead people and cause new "urban
legends" to be perpetuated, misleading even more people than just those
who originally read your posting and thought it was good.

Cheers,

- Alf

--
A: Because it messes up the order in which people normally read text.
Q: Why is it such a bad thing?
A: Top-posting.
Q: What is the most annoying thing on usenet and in e-mail?
Thanks Alf,

I understand and support what you are saying and totally agree. But at
the same time i also discourage people sending such messages which
discourages other posters. If the same thing is said so that it makes
sense more than shows an attitude we would have more people helping
here.. He could have pointed out the typo's in the message and i would
have thanked him for it.. All i know is he wasted his, yours and my
energy + some google's webspace. The point he wanted to make could have
been made with a simple post after rectifying the errors.

P.S. Whats wrong here?
* You state "can only be called from the derived classes".
That is incorrect.
AFAIK, pure virtual functions with bodies can be invoked from the
derived classes only..

Ramneek

Jul 9 '06 #12

P: n/a
* rami:
>
P.S. Whats wrong here?
> * You state "can only be called from the derived classes".
That is incorrect.

AFAIK, pure virtual functions with bodies can be invoked from the
derived classes only..
Try this with your favorite compiler:

#include <iostream>
#include <ostream>

void say( char const s[] ) { std::cout << s << std::endl; }

struct Foo
{
virtual void bar() const = 0;
void g() const { Foo::bar(); }
};

void Foo::bar() const { say( "called" ); }

struct Foo2: Foo { void bar() const {} };

int main() { Foo2().g(); }
--
A: Because it messes up the order in which people normally read text.
Q: Why is it such a bad thing?
A: Top-posting.
Q: What is the most annoying thing on usenet and in e-mail?
Jul 9 '06 #13

P: n/a

Alf P. Steinbach wrote:
* rami:

P.S. Whats wrong here?
* You state "can only be called from the derived classes".
That is incorrect.
AFAIK, pure virtual functions with bodies can be invoked from the
derived classes only..

Try this with your favorite compiler:

#include <iostream>
#include <ostream>

void say( char const s[] ) { std::cout << s << std::endl; }

struct Foo
{
virtual void bar() const = 0;
void g() const { Foo::bar(); }
};

void Foo::bar() const { say( "called" ); }

struct Foo2: Foo { void bar() const {} };

int main() { Foo2().g(); }
--
A: Because it messes up the order in which people normally read text.
Q: Why is it such a bad thing?
A: Top-posting.
Q: What is the most annoying thing on usenet and in e-mail?
Thanks!

So is the point - "can only be called _through_ the derived class
objects"?

Ramneek

Jul 9 '06 #14

P: n/a
rami schrieb:
Alf P. Steinbach wrote:
>* rami:
>>AFAIK, pure virtual functions with bodies can be invoked from the
derived classes only..
Try this with your favorite compiler:

#include <iostream>
#include <ostream>

void say( char const s[] ) { std::cout << s << std::endl; }

struct Foo
{
virtual void bar() const = 0;
void g() const { Foo::bar(); }
};

void Foo::bar() const { say( "called" ); }

struct Foo2: Foo { void bar() const {} };

int main() { Foo2().g(); }
--
A: Because it messes up the order in which people normally read text.
Q: Why is it such a bad thing?
A: Top-posting.
Q: What is the most annoying thing on usenet and in e-mail?

Thanks!

So is the point - "can only be called _through_ the derived class
objects"?
Please don't quote signatures. Good newsreaders cut them automatically.

The point is that there is no restriction on how you can call pure
virtual functions.

However, you cannot instantiate a class with pure virtual functions (you
cannot have objects of this class).

--
Thomas
Jul 9 '06 #15

This discussion thread is closed

Replies have been disabled for this discussion.