Merge parsing of bracketed patterns

[k3KAW8Pnf7mkmdSMPHz27]

Fixes #6892.
Character classes (regex expressions in brackets, e.g., [a-zA-Z]) should be usable with the regex modifier, but it currently isn't for citation keys.

Bracketed patterns are parsed using several different methods in CitationKeyGenerator.java, BracketedPattern.java and AbstractCitationKeyPattern. This PR aims to merge all parsing code into BracketedPattern.java to make it easier to modify bracketed patterns without breaking the citation key generator.

Currently BracketedPattern's parser can parse expressions containing the regex modifier using character classes.

Todo

• What are AbstractCitationKeyPattern and GlobalCitationKeyPattern used for, and should they be removed?
• Why are both GlobalCitationKeyPattern and CitationKeyPatternPreferences needed to create a CitationKeyGenerator?

Checklist

• Change in CHANGELOG.md described (if applicable)
• Tests created for changes (if applicable)
• Manually tested changed features in running JabRef (always required)
• Screenshots added in PR description (for UI changes)
• Checked documentation: Is the information available and up to date? If not created an issue at https://github.com/JabRef/user-documentation/issues or, even better, submitted a pull request to the documentation repository.

[k3KAW8Pnf7mkmdSMPHz27]

I think I need an opinion on the cleaning of "unwanted characters".

In #4770 the argument were that citation patterns containing special characters ([auth]-[year]) should not remove the "-" between the two bracketed expressions/field markers as it is a wanted special character.
I'd argue based on recent changes #6300, this should no longer be the case, as it is a preference that the user can set themselves.

In particular, I believe this test case,

jabref/src/test/java/org/jabref/logic/citationkeypattern/CitationKeyGeneratorTest.java

Lines 1067 to 1074 in 4b1b7b4

	@Test
	void generateKeyWithMinusInCitationStyleOutsideAField() {
	BibEntry entry = new BibEntry()
	.withField(StandardField.AUTHOR, AUTHOR_STRING_FIRSTNAME_FULL_LASTNAME_FULL_COUNT_1)
	.withField(StandardField.YEAR, "2019");
	assertEquals("Newton-2019", generateKey(entry, "[auth]-[year]"));
	}

should fail using the default unwanted characters.

jabref/src/main/java/org/jabref/logic/citationkeypattern/CitationKeyGenerator.java

Line 28 in 4b1b7b4

public static final String DEFAULT_UNWANTED_CHARACTERS = "-`ʹ:!;?^+";

Is this reasonable? Should I open a separate PR with more information on this issue? (the answer dictates where the cleanKey method call goes and should only change 1-4 lines of code)

[Siedlerchr]

Regarding the unwanted characters, you are right. The generator should respect the user defined unwanted chars (and of course the illegal ones)

[k3KAW8Pnf7mkmdSMPHz27]

@Siedlerchr I'll change it and add some test cases

[k3KAW8Pnf7mkmdSMPHz27]

I think looking more into AbstractCitationKeyPattern and GlobalCitationKeyPattern will end up being out of scope of this PR. I have left them unchanged.

[Siedlerchr]

looks good to me, just one minor thing where I am not sure