Add "Bottle Song" exercise and test suite

[rabestro]

A new exercise called "Bottle Song" has been introduced. This includes corresponding test suite for the exercise. Modification was also made to update the average run time value in the config file. Overall, this commit provides a new practice exercise for users to solve.

[rabestro]

I have an error message, so I fixed the number to 2.0

PS /Users/jegors/Projects/Exercism/awk> ./bin/configlet lint
The lint command is under development.
To check your track using the latest linting rules,
please regularly update configlet and re-run this command.

The value of `test_runner.average_run_time` is `2`, but it must be a float:
/Users/jegors/Projects/Exercism/awk/config.json

Configlet detected at least one problem.
For more information on resolving the problems, please see the documentation:
https://exercism.org/docs/building/configlet/lint

[glennj]

I have an error message, so I fixed the number to 2.0

And now CI is unhappy ("The value of test_runner.average_run_time is 2.0, but it must be an integer")

Upgrade your configlet: bin/fetch-configlet

[glennj]

There should be exactly one commented skip directive in test-bottle-song.bats

The first test should contain

    #[[ $BATS_RUN_SKIPPED == "true" ]] || skip

and every other test should contain

    [[ $BATS_RUN_SKIPPED == "true" ]] || skip

[rabestro]

[[ $BATS_RUN_SKIPPED == "true" ]] || skip

Yeah... I have to improve my prompt :)

[rabestro]

There should be exactly one commented skip directive in test-bottle-song.bats

AI generated the whole test but I provide wrong sample with deleted comments. :)

[glennj]

Based on the provided test data and the sample solution,

That's interesting. I start by translating the canonical data to create the tests (I use vim, and q recording macros takes care of a most of the repetitive grunt work), and then I write the example solution.

Thanks for making this happen @rabestro. It's good to have allies for track maintenance.