iox-#10 Create sys package

[elBoberido]

Pre-Review Checklist for the PR Author

1. Code follows the Rust coding style and is formatted with rustfmt
2. Branch follows the naming format (iox-123-this-is-a-branch)
3. Commits messages are according to this guideline
   • Commit messages have the issue ID (iox-#42 commit text)
   • Commit author matches Eclipse Contributor Agreement (and ECA is signed)
4. Update the PR title
   • Follow the same conventions as for commit messages
   • Link to the relevant issue
5. Relevant issues are linked
6. Add sensible notes for the reviewer
7. All checks have passed (except task-list-completed)
8. Assign PR to reviewer

Notes for Reviewer

This PR refactors the low level bindings to a separate iceoryx-sys crate to. This should help to fix the docs.rs build.

Checklist for the PR Reviewer

• Commits are properly organized and messages are according to the guideline
• Code according to our coding style and naming conventions
• PR title describes the changes

Post-review Checklist for the PR Author

1. All open points are addressed and tracked via issues

References

• Relates to Release on crates.io #10

[elfenpiff]

Experimental sounds always like something is most likely to explode. What do you think of a term like "Rust binding in an early stage of development" or "Rust binding where the API may still change"

I know you already merged this but when you create another branch you may adjust this as well

[elBoberido]

This will be part of #4 which will come later today

[elfenpiff]

What does Arc mean in this context?

[elBoberido]

It means Atomically Reference Counted and indicates that the lifetime of the subscriber is managed by a thread-safe reference-counting pointer

[elfenpiff]

What does Rc mean in this context?

[elBoberido]

It means Reference Counted and indicates that the lifetime of the subscriber is managed by a single-threaded reference-counting pointer

[elfenpiff]

Why do we have a strong and a weak ref subscriber?

[elBoberido]

It's quite long since I wrote this code and only God and I understood what it was for. Now only God knows.
Seriously, the Sample holds an Rc or Arc owning reference to the subscriber. The idea was to downgrade this reference to a non-owning Weak for Rc or Weak for Arc reference. But I'm not sure anymore why.

[elfenpiff]

Out of curiosity, what happens when we make the history capacity an uint16_t in cpp and forget to adjust it here?

[elBoberido]

There is a layer of indirection

let raw = cpp!([service as "const char *",
                instance as "const char *",
                event as "const char *",
                history_capacity as "uint64_t", // <- this comes from the Rust struct
                node_name as "const char *",
                offer_on_create as "bool",
                subscriber_too_slow_policy as "uint8_t"]
                -> *mut Publisher as "PublisherPortUser*"
{
    PublisherOptions options; // <- this is the C++ struct
    options.historyCapacity = history_capacity;
    options.nodeName = IdString_t(TruncateToCapacity, node_name);
    options.offerOnCreate = offer_on_create;
    options.subscriberTooSlowPolicy = static_cast<ConsumerTooSlowPolicy>(subscriber_too_slow_policy);
    auto portData = PoshRuntime::getInstance().getMiddlewarePublisher(
        {
            IdString_t(TruncateToCapacity, service),
            IdString_t(TruncateToCapacity, instance),
            IdString_t(TruncateToCapacity, event)
        },
        options
    );
    return new PublisherPortUser(portData);
});

Depending on the build flags, this will be a warning or an error.