rspec/rules/S5405/cfamily/rule.adoc

== Why is this an issue?

A base class and its derived class often differ in size.

 Accessing an array of a derived class through a pointer to the base class leads to wrong pointer arithmetic and can then corrupt memory.


=== Noncompliant code example

[source,cpp]
----
struct Base { /*...*/};
struct Derived : public Base { /*...*/};
void f(Base const &b);

int main() {
  const size_t size = 4;
  Derived derivedArray[size];
  Base* basePointer = derivedArray; // Noncompliant
  f(basePointer[2]); // The arithmetic will use the size of Base, not the size of Derived, and point to a random byte in the array
}
----
If you need to do a cast from derived to base in an array, do it element by element.


=== Compliant solution

[source,cpp]
----
#include<iostream>

using namespace std;

struct Base {
  int iBase = 0;
};

struct Derived : public Base {
  int iDerived = 0;
};

int main() {
  const size_t size = 4;
  Derived derivedArray[size];
  for(int i=0; i<size; ++i) {
    derivedArray[i].iBase = i;                                           // store : 0 1 2 3
    derivedArray[i].iDerived = i*1000;
  }
  for(int i=0; i<size; ++i) {
    cout<<"derivedArray["<<i<<"].iBase="<<derivedArray[i].iBase<<endl;   // display : 0 1 2 3
    cout<<"base of derivedArray["<<i<<"].iBase="<<static_cast<Base*>(derivedArray+i)->iBase<<endl;     // display : 0 1 2 3
  }
  return 0;
}
----


== Resources

* {cpp} Core Guidelines - https://github.com/isocpp/CppCoreGuidelines/blob/e49158a/CppCoreGuidelines.md#c152-never-assign-a-pointer-to-an-array-of-derived-class-objects-to-a-pointer-to-its-base[C.152: Never assign a pointer to an array of derived class objects to a pointer to its base]


ifdef::env-github,rspecator-view[]
'''
== Comments And Links
(visible only on this page)

=== relates to: S5410

=== on 25 Oct 2019, 16:27:29 Geoffray Adde wrote:
First, we try a much stronger, constraining and simpler version of the rule: no array (C-style, std::array and std::vector) of publicly derived class.

=== on 18 Dec 2019, 17:55:05 Loïc Joly wrote:
If you want to try the stronger version, why not directly document it? I think it might have value on its own, as a code smell... And I would also prevent arrays of base classes.

endif::env-github,rspecator-view[]