|Important||This document may not represent best practices for current development, links to downloads and other resources may no longer be valid. Current recommended version can be found here. ArchiveDisclaimer|
Supports arrays of CObject pointers.
These object arrays are similar to C arrays, but they can dynamically shrink and grow as necessary.
Array indexes always start at position 0. You can decide whether to fix the upper bound or allow the array to expand when you add elements past the current bound. Memory is allocated contiguously to the upper bound, even if some elements are null.
Under Win32, the size of a CObArray object is limited only to available memory.
As with a C array, the access time for a CObArray indexed element is constant and is independent of the array size.
CObArray incorporates the IMPLEMENT_SERIAL macro to support serialization and dumping of its elements. If an array of CObject pointers is stored to an archive, either with the overloaded insertion operator or with the Serialize member function, each CObject element is, in turn, serialized along with its array index.
If you need a dump of individual CObject elements in an array, you must set the depth of the CDumpContext object to 1 or greater.
When a CObArray object is deleted, or when its elements are removed, only the CObject pointers are removed, not the objects they reference.
Before using an array, use SetSize to establish its size and allocate memory for it. If you do not use SetSize, adding elements to your array causes it to be frequently reallocated and copied. Frequent reallocation and copying are inefficient and can fragment memory.
Array class derivation is similar to list derivation. For details on the derivation of a special-purpose list class, see the article Collections.
You must use the IMPLEMENT_SERIAL macro in the implementation of your derived class if you intend to serialize the array.