[iron] Add BagSplitInfo service call on bag close (backport #1422)

[mergify]

• Replaces Add BagSplitInfo service call on bag close #1216
• Relates "Finalize" bag files when recording #1418
• Note: The BagSplitInfo::opened_file will have an empty string to indicate that it was "bag close" and not bag split event.

  ---

  This is an automatic backport of pull request Add BagSplitInfo service call on bag close #1422 done by Mergify.

[MichaelOrlov]

Note: In this PR we are adding a new public method to the EventCallbackManager.
However, this is ABI compatible changes according to the https://acodersjourney.com/20-abi-breaking-changes/

3. Add new non-virtual methods to an existing class

[MichaelOrlov]

Gist: https://gist.githubusercontent.com/MichaelOrlov/20516443e02fa9ea05ae60f71b4f54dc/raw/1a8e502306882e59e6eb65f7ece4573d567bd913/ros2.repos
BUILD args: --packages-above-and-dependencies rosbag2_cpp rosbag2_transport ros2bag
TEST args: --packages-above rosbag2_cpp rosbag2_transport ros2bag
ROS Distro: iron
Job: ci_launcher
ci_launcher ran: https://ci.ros2.org/job/ci_launcher/13816

• Linux
• Linux-aarch64
• Windows

[MichaelOrlov]

@clalancette The Iron CI fails with bunch of violations in uncrustify and flake8 https://ci.ros2.org/job/ci_linux-aarch64/15295/#showFailuresLink.
It seems uncrustify was updated on the Iron which is not expected to be.
Should I fix all violations in rosbag2 or it is gonna be some more generic remedy by pinning uncrustify?

[clalancette]

It seems uncrustify was updated on the Iron which is not expected to be.

It was not updated; the CI was run against noble. Fixing that to jammy should fix the CI.

[MichaelOrlov]

As regards:

CI was run against noble. Fixing that to jammy should fix the CI.

Good catch. Interesting. Will need to check ros-ci-for-pr scripts if we need to update it.

[MichaelOrlov]

Re-run Linux CI jobs with "jammy" linux distro

• Linux
• Linux-aarch64