In WordPress unit test,
wordpress-tests/wp-testdata/jacob/utf-8.txt
has UTF-8 BOM at the beginning of the file so that testcase
wordpress-tests/wp-testcase/jacob/TestFormatting.php
Test_UTF8_URI_Encode::test_percent_encodes_non_reserved_characters()
fails. Because BOM code (0xef, 0xbb, 0xbf. for UTF-8) is tried to be url encoded.
BOM should not be included in this kind of test data.
You may take out the first 3 bytes from the test file, or UTF-8 compliant editor usually has an option letting you choose whether you use BOM or not for UTF text. I think turn off the option (no BOM) is better in most case.
I confirmed after taking BOM the test case worked fine.
