Replace const std::vector& with boost::span

[vsoulgard]

close #49

Allows to accept any contiguous memory (arrays, vectors, or raw buffers) without being tied to a specific storage type.

But boost::span does not construct from initializer_list.

[biljazovic]

Hi, thanks for the PR!

Issue is that this code would stop working (heap-use-after-free)

auto op = c.async_subscribe(
    std::vector<subscribe_topic>{{ "test/mqtt-test", { qos_e::exactly_once } }}, subscribe_props {},
    asio::deferred
);
/* initiate op later */

In the current implementation, the vector is copied only if the initiation is deferred; otherwise it's forwarded by reference (async_initiate takes care of that).

I think the correct solution is to use boost::span in (un)subscribe_op like you do, and modify async_(un)subscribe like this:

template <
    typename TopicSequence,
    typename CompletionToken =
        typename asio::default_completion_token<executor_type>::type
>
decltype(auto) async_subscribe(
    const TopicSequence& topics,
    const subscribe_props& props,
    CompletionToken&& token = {}
) {
    using Signature = void (
        error_code, std::vector<reason_code>, suback_props
    );
    return asio::async_initiate<CompletionToken, Signature>(
        detail::initiate_async_subscribe(_impl), token,
        topics, props
    );
}

Regarding std::initializer_list, we could add an overload:

template <
    typename CompletionToken =
        typename asio::default_completion_token<executor_type>::type
>
decltype(auto) async_subscribe(
    std::initializer_list<subscribe_topic> topics,
    const subscribe_props& props,
    CompletionToken&& token = {}
) {
    return async_subscribe(std::vector(topics), props, std::forward<CompletionToken>(token));
}

[vsoulgard]

Thank you for the detailed review and clear guidance!

I've comitted new changes. For async_(un)subscribe, I had to add new trait and use typename = std::enable_if_t<detail::is_container_of<TopicSequence, std::string>> to properly constrain the template. I don't know if there is better way to do this.

[biljazovic]

Thanks for the update! I'd like to explore a few options for constraining the template before merging. I'll get back to you on this.