feat: customizeable recording file name

[JonasPammer]

Resolves apache/grails-core#14329

Commit Message out-take:

README.md text was copied with template from https://github.com/spockframework/spock/blob/master/docs/extensions.adoc#global-extensions

I tried to get knowledge about this ServiceLoader paradigm from grails-core code search and oracle doc

Although this whole META-INF is weird to me,
I did not find it necessary to implement something complex like spock's ConfigurationScriptLoader, which uses (try to load env var path, file system, ..) into this just for this.

PR Appendix to last paragraph: Maybe the first one (env var) seems like a viable alternative/addition to this pattern, given that all other settings can be managed through systemProperty 'grails.geb.recording.* '*'``, i did not roll with using it instead as to first try to do it as per agreed upon https://github.com/grails/geb/issues/133#issuecomment-2654122264, and to possibly get validation if i did the ServiceLoader Pattern right ❤️ 💭

[CLAassistant]

All committers have signed the CLA.

[matrei]

Awesome @JonasPammer, thank you for this PR!

I do have some concern for the tests, in that now all tests will use the custom implementation, and no tests will use the default implementation. If we create and use a TestDescriptionServiceFactory and call that instead of the ServiceLoader directly, we can modify the implementation programatically in setupSpec() and cleanupSpec().

Also, I think we should target the 4.2.x branch with this PR as we have not yet released a 4.2.0 version.

What do you think?

[JonasPammer]

Totally! I saw other 2 projects in grails/spock also use a seperate class (i think there called loader), will implement thanks to your helpful guiding comments.

[JonasPammer]

Force Push Explanation

[changed pr base branch](https://docs.github.com/en/pull-requests/collaborating-with-pull-requests/proposing-changes-to-your-work-with-pull-requests/changing-the-base-branch-of-a-pull-request)

~\IdeaProjects\geb git:[geb-133-serviceloader]
git fetch upstream 4.2.x
From https://github.com/grails/geb
 * branch            4.2.x      -> FETCH_HEAD
~\IdeaProjects\geb git:[geb-133-serviceloader]
git fetch upstream 5.0.x
From https://github.com/grails/geb
 * branch            5.0.x      -> FETCH_HEAD
~\IdeaProjects\geb git:[geb-133-serviceloader]
git rebase --onto upstream/4.2.x upstream/5.0.x
Successfully rebased and updated refs/heads/geb-133-serviceloader.
~\IdeaProjects\geb git:[geb-133-serviceloader]
git push -f origin geb-133-serviceloader
Enumerating objects: 65, done.
Counting objects: 100% (65/65), done.
Delta compression using up to 12 threads
Compressing objects: 100% (30/30), done.
Writing objects: 100% (47/47), 5.58 KiB | 815.00 KiB/s, done.
Total 47 (delta 20), reused 0 (delta 0), pack-reused 0 (from 0)
remote: Resolving deltas: 100% (20/20), completed with 7 local objects.
To https://github.com/JonasPammer/geb.git
 + b2076f6...8186a9b geb-133-serviceloader -> geb-133-serviceloader (forced update)

[JonasPammer]

This PR (and, also seeing you there ;) gets me to also maybe open a PR in grails-docs for minor fixes found along the way, its fun

i am happy to drop some commits if desired, e.g. the last one - [add ContainerGebConfiguration ability](a9045bf)

[jdaugherty]

@JonasPammer Thank you for taking the time on this. I had only minor comments / cleanup.

[matrei]

Excellent! I have some minor suggestions/questions.

[JonasPammer]

thanks for helping me improve, will implement changes sometime tomorrow

[JonasPammer]

left 2 comments open ended to either be resolved or answered with that we should do it the way written

[matrei]

Nice! Just some follow up comments/suggestions.

[JonasPammer]

👍

[JonasPammer]

last commit can be dropped, i wanted to hold a grail to the actual code of your alls which this plugin is about to not be bloated by java design patterns necessities of 4 classes

[matrei]

Thanks @JonasPammer, I think we can drop the last commit for now (it might actually be a breaking change in 4.x).
After that I'm happy with this PR and will gladly approve.

[matrei]

Talking of breaking changes. Changing ContainerGebTestDescription to an interface is actually a breaking change. @jdaugherty What do you think, should we target 5.0.x with this PR instead?

[JonasPammer]

if its just about the namespacing, i could make ContainerGebTestDescription be the current Default...., and transform the interface class name to IContainer..., if the I pattern is ok

[jdaugherty]

Talking of breaking changes. Changing ContainerGebTestDescription to an interface is actually a breaking change. @jdaugherty What do you think, should we target 5.0.x with this PR instead?

It was internal only so I think it's ok in this case.

[JonasPammer]

given i will go ahead with the Default.. Naming scheme in #146

[matrei]

@JonasPammer Should we go in the direction of a more general service loader instead, because it looks like really usable extension point mechanism that can be useful for more cases: #146 (comment)

[JonasPammer]

set do draft as per above comment linked PR design decision WIP

[JonasPammer]

as per apache/grails-core#14330 i understand when this will be closed i will reopen a new version in core

[jdaugherty]

@JonasPammer you should be safe to open this against grails core now.

grails-geb is the source location in core
grails-test-examples/geb is the test app