fix generics templates

[bendavies]

This PR fixes the templates for MicroMapperInterface and MapperInterface.
regression tests have been added.

as an aside, is the $toClass param in MapperInterface::load(...) redundant?
It seems we will only ever have concrete implementations here and will just new up a new class, which will match the class provided in the AsMapper attribute.

[bendavies]

@weaverryan i don't think these are my failures

[weaverryan]

Hmm, actually, I think they are... it's related to deprecations in PHPStan - so my guess is from the new test. But these are "Other deprecation notices"... are those direct? Indirect? I'm actually not sure - we might need to tweak our SYMFONY_DEPRECATIONS_HELPER... or figure out the underlying cause of this.

[stof]

I guess they are not classified properly because the deprecation is triggered from a class defined inside a phar, so it path has starts with phar://, which means it does not start with the absolute path of the vendor folder. I don't find it unexpected that we did not though of the case of phars in the vendor folder in that detection logic.

[weaverryan]

I wonder if changing the phpstan dependency to the stable 1.10 in composer.json would help (I can't remember why I have 1.11-dev - probably by accident).

[bendavies]

doesn't help. but i tried using phpunit directly instead of (not-so-)simple-phpunit, and there are no deprecations.

[weaverryan]

doesn't help. but i tried using phpunit directly instead of (not-so-)simple-phpunit, and there are no deprecations

Yes, but probably just because it's hiding the deprecations :p. Maybe we can just generate and point to a baseline file for these deprecations - https://symfony.com/doc/current/components/phpunit_bridge.html#baseline-deprecations

[bendavies]

done

[weaverryan]

Still not happy - not sure why - phpunit-bridge can be pesky to work with and configure. BUT, I have changed the settings so that if you push, the CI will run automatically vs waiting for my approval 🙃

[bendavies]

oh yeah, it's hard coded paths in the allowed deprations

[bendavies]

can't see the issue atm. there are deprecations in ci that i don't get locally.

[bendavies]

hallelujah

[weaverryan]

Thank you for your hard work on this Ben!