Enhancement for company names, titles and lastname extensions

[blacksenator]

Fix #38

Differentiation between companies and personal names, more detailed mappings for prefixes (multiple words/components), titles (mainly academic with multiple words/components) and lastname extensions (nobilit predicates)

[coveralls]

Coverage decreased (-5.0%) to 95.023% when pulling 92bdf0c on blacksenator:enhancement_branch into 9a54a71 on theiconic:master.

[blacksenator]

Okay, I used three commits to achieve the desired goal. But now I'm happy with the result. It's too bad that the automatic test routines have such narrow limits.
To be honest, I can't see where I can make the code even leaner so as not to exceed the limits (but I´m not a pro :/ ).
I would appreciate a thorough manual review and feedback. I would be even more pleased if my additions become part of the solution.

[rvanlaak]

Nice! Will work on Dutch.php asap to verify your implementation

I see you're also adding a PHPStan config in this PR. It would be good to add that in a separate PR, and run it via Github Actions.

You can check php-translation/symfony-storage#38 as example 👍

[blacksenator]

@rvanlaak Thank you for your fast reaction.
I'm afraid I'm not familiar with GitHub Actions. I see the advantages of all automation, but it is difficult for me to find my way around in the tools - I am probably too old for this stuff...
I added the line because I received this as a false report for my German.php.
As long as the PR is not merged, the line does not need to be added - right?

[rvanlaak]

@blacksenator I will add PHPStan in a separate PR 👍

Would be best to remove as many unrelated files from this PR, so more tests pass.

I'm also affraid that it would be best to split this PR in two. One for the titles / lastname extensions. And one for the company names.

And, to get code coverage it probably is enough to add a test case in NameTest, ParserTest and GermanParserTest.

[rvanlaak]

see #41

[blacksenator]

@wyrfel Do you agree with rvanlaak to split up this PR into two or maybe three PR´s? It is of course possible to split up the changes, but I would like to avoid the effort if it is not necessary.

[wyrfel]

Hey @blacksenator . Like i said on the issue, i think some of this functionality is out of scope. I also feel this PR is implementing some things that are already there. I'll review and get back to you, though. In the meantime, yes, i agree that splitting into separate PRs for the different aspects would be good, e.g. one for PHPStan, one for extending the german language file etc. . That way we can more easily discuss the different aspects.

[wyrfel]

@blacksenator I would like to integrate a lot of these changes into a next major release. For that to happen we will need you to remove the author, copyright and licence annotations in the code, please. We will credit you in the release notes and github provides transparency over the authorship, as well.

[blacksenator]

done

[rvanlaak]

The additional methods on LanguageInterface will be a BC break for all current implementations. Can we find a way to make this forward compatible, for example by introducing a new interface and adding deprecating notices on the old one?

[blacksenator]

Is that a question for me? If so, then I have to admit that I did not understand what your idea for a workaround should look like

[wyrfel]

@rvanlaak I'm planning to integrate this into a new major release that will have the LanguageInterface renamed to DefinitionInterface, anyway (see #46). It will also have multibyte-safe string operations. I'm afraid you will have to make the necessary adjustments when upgrading to that newer version, once ready. Do you anticipate major problems from this?

[rvanlaak]

What about releasing a minor version that adds DefinitionInterface already?

If you would keep both LanguageInterface and DefinitionInterface in place and make all related functionality BC, then Parser can check which one is used to trigger the deprecation. The LanguageInterface then is mentioned to be removed in the major version.

This will give you time to release DefinitionInterface specific functionality already and decouple that from the BC break.