next iterator position without mutation

[Cagdas Ozgenc]

Greetings.

Is there a convinient way to get the next iterator position without mutating
the iterator for non-random access containers? For example,

i + 1

instead of

i++

Thanks

[Attila Feher]

Cagdas Ozgenc wrote:

Greetings.

Is there a convinient way to get the next iterator position without
mutating the iterator for non-random access containers? For example,

i + 1

instead of

i++

Nope. For input or output iterators you cannot even read the same place
twice. I might be wrong, but for forward or bidirectional iterators you may
copy the iterator and increment the copy to "take a peek".

--
Attila aka WW

[Cagdas Ozgenc]

> Nope. For input or output iterators you cannot even read the same place

twice. I might be wrong, but for forward or bidirectional iterators you may copy the iterator and increment the copy to "take a peek".

That's bad news. The following code would be quite dangerous then:

template<typena me Iterator>
Iterator operator+(Itera tor const& input,size_t offset) {
Iterator output = input;
while(offset > 0) {
++output;
offset--;
}
return output;
}

I can't find a forward_iterato r class to make the above function iterator
type-safe. What's wrong with these STL iterators?

Thanks

[Attila Feher]

Cagdas Ozgenc wrote:

Nope. For input or output iterators you cannot even read the same
place twice. I might be wrong, but for forward or bidirectional
iterators you may copy the iterator and increment the copy to "take
a peek".

That's bad news. The following code would be quite dangerous then:

template<typena me Iterator>
Iterator operator+(Itera tor const& input,size_t offset) {
Iterator output = input;
while(offset > 0) {
++output;
offset--;
}
return output;
}

I can't find a forward_iterato r class to make the above function
iterator type-safe. What's wrong with these STL iterators?

You have iterator traits, OMHO you can make a compile time assertion if the
function template is being instantiated with the wrong iterator category.

--
Attila aka WW

[Cagdas Ozgenc]

> You have iterator traits, OMHO you can make a compile time assertion if
the

function template is being instantiated with the wrong iterator category.

Can you give a small example, or give pointers on that matter.

Thanks

[Attila Feher]

Cagdas Ozgenc wrote:

You have iterator traits, OMHO you can make a compile time assertion
if the function template is being instantiated with the wrong
iterator category.

Can you give a small example, or give pointers on that matter.

10 seconds using Google:
http://www.roguewave.com/support/doc...ator-traits.ht
ml

--
Attila aka WW