TEST_BUG: removing non-ascii characters causes tests to fail

These two langtools jtreg test files contain a utf8 char:
test/tools/javac/api/6437999/Utf8.java:25: * @author Peter von der AhÃ©
test/tools/javac/api/T6306137.java:30: * @author Peter von der AhÃ©

They should contain only ASCII. The tests should generate the need files containing the utf8 chars so that the chars themselves are not in the source code in the repository.

Fixing (JDK-8008023) the above cause the tests to fail , as the tests expected these characters to exist,
I suggest we re-introduce the non-ascii characters, in a less subtle and more obvious
way such that it will not be removed inadvertently. As the previous suggestion says
these files IMO should be created on the fly.