[22321] Benchmark example updated

[Javgilavi]

Description

I remake the benchmark example to new example style maintaining the original functionality.

@Mergifyio backport 3.1.x 3.0.x

Fixes #5398

Contributor Checklist

• Commit messages follow the project guidelines.
• The code follows the style guidelines of this project.
• Tests that thoroughly check the new feature have been added/Regression tests checking the bug and its fix have been added; the added tests pass locally
• N/A Any new/modified methods have been properly documented using Doxygen.
• N/A Any new configuration API has an equivalent XML API (with the corresponding XSD extension)
• Changes are backport compatible: they do NOT break ABI nor change library core behavior.
• Changes are API compatible.
• New feature has been added to the versions.md file (if applicable).
• N/A New feature has been documented/Current behavior is correctly described in the documentation.
• Applicable backports have been included in the description.

Reviewer Checklist

• The PR has a milestone assigned.
• The title and description correctly express the PR's purpose.
• Check contributor checklist is correct.
• If this is a critical bug fix, backports to the critical-only supported branches have been requested.
• Check CI results: changes do not issue any warning.
• Check CI results: failing tests are unrelated with the changes.

[rsanchez15]

Pending example tests

[Mario-DL]

Good first iteration overall. The PR brings an interesting example.
In addition to my comments below, we would need to add this PR content Refactor benchmark example to the versions.md and replace the N/A with a tick in the checklist.

[Mario-DL]

We typically recommend not to perform blocking operations in callbacks. I think it would make more sens moving these lines to the first part of the run() method.
Additionally, I think we could change the matched_ condition to be greater of equal to 2 since we expect both of out endpoints to be matched with the publisher ones.

[Javgilavi]

The idea is to make sure the subscriber not to sent back the data until the on_publication_match happen. To solve the case of the subscriber receiving the data but not sending it properly.
I don't understand how this could be moved to run().
Also, I understand the comment of change for matched_ to be greater or equal to 2 is for the publisher and not here, right?

[Mario-DL]

The first thing in the run() method would be:

{
        // Wait for the data endpoints discovery
        std::unique_lock<std::mutex> matched_lock(mutex_);
        cv_.wait(matched_lock, [&]()
                {
                    // at least one has been discovered
                    return ((matched_ >= 2) || is_stopped());
                });
}

Then, I would remove all while (matched_ == 0){} blocks from the on_data_available(). This will make the subscriber to first wait to be matched and simplify the 'on_data_available()`. We should until greater or equal than 2 since both subscriber endpoints must be matched.

[Mario-DL]

Whats more, we could avoid waiting for any matched_ within run(), since the reliability and durability will do the late-joining this for us when specified

[Javgilavi]

As we discussed in private chat, I just delated the blocking operation on the on_data_available().

[Mario-DL]

Testing locally I found a couple of strange things to me in the publisher side:

• The first index seems to be sent and received twice
• The RECEIVED is only printed for odd sample idexes (1,3,5,...)
  Is that behavior correct ?

[Javgilavi]

The problem of repeating the first index is fix.
The RECEIVED is only printed for odd because they are sent by the subscriber and receive in the publisher, and the evens are sent by the publisher and receive by the subscriber

[Mario-DL]

Yeah, I reviewed the logic and I first thought that the subscriber was not incrementing the index but it is doing it, so it is expected.

[Mario-DL]

Please, update the commands output in this document since the index 0, index 1, index 0,... behavior was already corrected and no longer exists

[Mario-DL]

Leaving just some comments on previous suggestions

[Mario-DL]

Please, don't forget updating the versions.md and marking it in the contributor checklist

[Mario-DL]

I think we are missing this here, otherwise when the example is launched without --samples the subscriber will receive the first sample and skip.

Suggested change

	if (benchmark_.index() >= samples_)
	if (samples_ != 0 && benchmark_.index() >= samples_)