Add multi-sentence title formatting to sentence case and title case

[tmrd993]

Fixes #6759

Added support for multiple-sentence title formatting for the "Sentence Case" and "Title Case" options by using a regular expression to first split the input into sentences, and afterwards processing those sentences individually. A RegEx might not be the best choice for natural language processing but since most titles are only 1 sentence long, the proposed RegEx should cover the vast majority of cases. Also added additional test cases for Sentence Case and Title Case formatting. The old tests are assuming one sentence titles but are still applicable.

• 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.

[koppor]

Welcome to JabRef development! I am very happy to see an ADR. Regarding the code, I just have some minor comments.

[koppor]

While you are on it, I propose to rewrite to parameterized tests:

package org.jabref.logic.formatter.casechanger;

import java.util.stream.Stream;

import org.junit.jupiter.params.ParameterizedTest;
import org.junit.jupiter.params.provider.Arguments;
import org.junit.jupiter.params.provider.MethodSource;

import static org.junit.jupiter.api.Assertions.assertEquals;

/**
 * Tests in addition to the general tests from {@link org.jabref.logic.formatter.FormatterTest}
 */
public class SentenceCaseFormatterTest {

    private SentenceCaseFormatter formatter = new SentenceCaseFormatter();

    private static Stream<Arguments> testData() {
        return Stream.of(
                Arguments.of("Upper first", "upper First"),
                Arguments.of("Upper first", "uPPER FIRST"),
                Arguments.of("Upper {NOT} first", "upper {NOT} FIRST"),
                Arguments.of("Upper {N}ot first", "upper {N}OT FIRST"),
                Arguments.of("Whose music? A sociology of musical language",
                        "Whose music? a sociology of musical language"),
                Arguments.of("Bibliographic software. A comparison.",
                        "bibliographic software. a comparison."),
                Arguments.of("England’s monitor; The history of the separation",
                        "England’s Monitor; the History of the Separation"),
                Arguments.of("Dr. schultz: a dentist turned bounty hunter.",
                        "Dr. schultz: a dentist turned bounty hunter."),
                Arguments.of("Example case. {EXCLUDED SENTENCE.}",
                        "Example case. {EXCLUDED SENTENCE.}"),
                Arguments.of("I have {Aa} dream", new SentenceCaseFormatter().getExampleInput())
        );
    }

    @ParameterizedTest
    @MethodSource("testData")
    public void test(String expected, String input) {
        assertEquals(expected, formatter.format(input));
    }
}

OK, the descriptions are lost. Maybe, the description could be added as first paramter to the test method.

[koppor]

Thank you for working on it - could you fix checkstyle? 😇

[ant:checkstyle] [ERROR] /home/runner/work/jabref/jabref/src/main/java/org/jabref/logic/formatter/casechanger/TitleCaseFormatter.java:37: There is more than 1 empty line one after another. [EmptyLineSeparator]

Even though there are many failing tests, we should at least try to keep the essential tests green. "Checkstyle" and "Unit tests" are essential. "Unit tests" go through; just checkstyle is it.

[tmrd993]

Done. How can I test for checkstyle issues?

I am using gradlew check as explained in the doc but it only does unit tests.

[Siedlerchr]

If you use intellij you can install the checkstyle plug-in and also import the jabref code style for the formatter.
Otherwise just run gradlew checkstyle.
You also see the failing tests if you click on the checkstyle CI test

[Siedlerchr]

Is there a reason why these tests are commented out?

[tmrd993]

They are redundant as I already added them as parameterized tests but I didn't know if I should delete them yet since they contain the descriptions of the tests.

[Siedlerchr]

Then just remove them. I think it's clear from the arguments what is tested

[Siedlerchr]

@tmrd993 Thanks again for your contribution!