Make FileLogger platform-independent: UTF-8 encoding and LF line separators

[elharo]

Partially fixes #206.

Problem

FileLogger was not platform-independent in two ways:

1. Charset: writing used the platform default charset. The MirrorStreamWrapper.write(byte[], int, int) decoded mirrored bytes via new String(b, off, len), and the PrintStream encoded output with the platform default. On a host whose default charset is not UTF-8, non-ASCII text (e.g. café) was encoded/decoded as ISO-8859-1 instead of UTF-8, producing mojibake like café.
2. Line separators: line endings depended on System.lineSeparator(), so on Windows the output used \r\n while on Linux/macOS it used \n, making log files differ from platform to platform.

Change

• Encode and decode UTF-8 unconditionally: the PrintStream is now created with StandardCharsets.UTF_8 (both the mirrored and non-mirrored paths), and MirrorStreamWrapper buffers bytes and decodes them with UTF-8 on flush. Neither path consults the platform default charset anymore. The write(int) byte path now also flows through the UTF-8-buffered mirror (previously it only worked for ASCII).
• Use a Unix LF (\n) line separator in consumeLine instead of System.lineSeparator(), so logs are byte-identical across Linux, macOS, and Windows.

getPrintStream() is retained as required by the ExecutionLogger API; it is now UTF-8 encoded and its output is still mirrored.

Tests

• Reworked the UTF-8 test into mirrorAndFileShouldUseUtf8, which writes non-ASCII content, asserts the mirror handler receives the correctly decoded café, and asserts the underlying file contains the exact UTF-8 bytes — verifying both the mirror and the file are UTF-8 independent of the host.
• Replaced System.lineSeparator() with \n in FileLoggerTest and TestMirrorHandler to assert the new, platform-independent LF output.

All 26 tests pass.

[elharo]

Hmm, this fails on JDK 25 Windows but in a strange way. Might be related:

[INFO] Running org.apache.maven.shared.scriptinterpreter.BeanShellScriptInterpreterTest
[INFO] Tests run: 2, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 0.122 s -- in org.apache.maven.shared.scriptinterpreter.BeanShellScriptInterpreterTest
[INFO] Running org.apache.maven.shared.scriptinterpreter.FileLoggerTest
[ERROR] Tests run: 6, Failures: 2, Errors: 0, Skipped: 0, Time elapsed: 0.108 s <<< FAILURE! -- in org.apache.maven.shared.scriptinterpreter.FileLoggerTest
[ERROR] org.apache.maven.shared.scriptinterpreter.FileLoggerTest.outputFileNoMirror(File) -- Time elapsed: 0.078 s <<< FAILURE!
org.opentest4j.AssertionFailedError: 
expected: <Test1
Test2
> but was: <Test1
Test2
>
	at org.junit.jupiter.api.AssertionFailureBuilder.build(AssertionFailureBuilder.java:151)
	at org.junit.jupiter.api.AssertionFailureBuilder.buildAndThrow(AssertionFailureBuilder.java:132)
	at org.junit.jupiter.api.AssertEquals.failNotEqual(AssertEquals.java:197)
	at org.junit.jupiter.api.AssertEquals.assertEquals(AssertEquals.java:182)
	at org.junit.jupiter.api.AssertEquals.assertEquals(AssertEquals.java:177)
	at org.junit.jupiter.api.Assertions.assertEquals(Assertions.java:1145)
	at org.apache.maven.shared.scriptinterpreter.FileLoggerTest.outputFileNoMirror(FileLoggerTest.java:93)
	at java.base/java.lang.reflect.Method.invoke(Method.java:565)
	at java.base/java.util.ArrayList.forEach(ArrayList.java:1604)
	at java.base/java.util.ArrayList.forEach(ArrayList.java:1604)

not clear why it would pass on UNix and fail on Windows

[elharo]

remove "It is always UTF-8 encoded so that
* output is independent of the platform default charset."
as it should be assumed.