Fixes HTML to plain text conversion

[idlira]

Description

StringCleanup#htmlToPlainText is expected to convert p and br tags to line breaks.
This works if these tags were not followed by other tags. For example:

➡️ Input: Hello<br />World
✅ Output:

Hello
World

but
➡️ Input: <strong>Hello</strong><br />World
🚫 Output: HelloWorld

This happens because StringCleanup.PATTERN_STRIP_XML will look for white-spaces leading or trailing a tag, effectively replacing </strong>\n in the example above by an empty string.

To not disrupt this pattern, and possible cause a breaking change/behavioral change, which is used in several other places, we iterate and clean each line.

Additional Notes

• This PR fixes or works on following ticket(s): OX-12298

Checklist

• Code change has been tested and works locally
• Code was formatted via IntelliJ and follows SonarLint & best practices

[Copilot]

Pull request overview

This PR fixes an issue where StringCleanup#htmlToPlainText incorrectly handled line breaks when HTML tags preceded <br /> or <p> tags. The fix processes each line individually to preserve line breaks that would otherwise be removed by the XML stripping regex pattern.

Changes:

• Added line-by-line processing in htmlToPlainText to prevent the XML stripping pattern from removing intentional line breaks
• Introduced Strings.iterateLines utility method to support line-by-line iteration with platform-independent line ending handling
• Added test case to verify the fix for HTML with inline tags followed by line breaks

Reviewed changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 1 comment.

File	Description
src/main/java/sirius/kernel/commons/StringCleanup.java	Modified htmlToPlainText to process lines individually, preserving converted line breaks
src/main/java/sirius/kernel/commons/Strings.java	Added iterateLines utility method for platform-independent line iteration
src/test/kotlin/sirius/kernel/commons/StringsTest.kt	Added test case verifying correct handling of inline tags followed by <br />

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[jakobvogel]

As mentioned in my other comment: Maybe also add special cases as tests, such as the following one.

<br />
<br />
Hello<br />
World

And I know, the test method was already like that before, but IMHO this is a case for a parameterized test. Right now, the list of unrelated tests in one method gives less reliable statistics if one of the tests fails.

[idlira]

These exist already:

        assertEquals(
            """
            first

            second
        """.trimIndent(),
            Strings.cleanup("<p>first<br><br/>second</p>", StringCleanup::htmlToPlainText, StringCleanup::trim)
        )

[jakobvogel]

That test is not equivalent to my example, as it does not contain line breaks, but it includes trimming. Therefore, it does not test the behaviour of your new approach. 🙃

I played around with this a little:

• An extension of your test case by adding a line break: <strong>bold</strong><br />\nmove gives bold\n move (with a space in front of move)
• My example <br />\n<br />\nHello<br />\nWorld gives \n Hello\n World (with a space at the front of each of the three lines!)

I am not sure whether these really are the results we want. Therefore, I would appreciate to have respective additional test cases defining the gold standard.

[idlira]

the logic in this class is so broken... will need review and posterior PR