Hi,
I was reading the source code to figure out how iterators exactly work on
Result objects, since I knew I used them before. Apparently my projects
only use operator*().
Consider the following code:
In Result:
const Row operator[] (size_type i) const {...}
Result is derived from const_subscript_container<Result, Row, const Row>,
which handles all the iterator functionality.
Inside that, it defines a typedef for iterator like:
typedef subscript_iterator<const this_type, ReturnType, SizeType,
DiffType> iterator;
In this case, ReturnType is const Row.
subscript_iterator<> handles the heavy lifting.
In subscript_iterator, we have:
ReturnType* operator ->() const { return &((*d)[i]); }
d is a pointer to Result, and i is the current index into the Result.
Which, if I read this right, is returning a pointer to a temporary Row object
returned from Result::operator[](size_type i).
If I'm correct in this reading, I'd recommend removing (for the stable tree)
subscript_iterator::operator->() so that nobody uses it by accident.
I think this iterator mechanism is ripe for optimization based on Byrial's
earlier idea, but will probably have to wait for 2.0.
- Chris
