The "UTF-32LE" branch is dead because "UTF-16LE" will always match it. This
patch fixes that by placing the "UTF-32LE" condition before the "UTF-16LE" one.
Also added a minor change to disambiguate UTF16LE and UTF32LE in edge case
where input consist of the UTF16LE BOM only (which would be detected as UTF32LE
otherwise)
Original issue reported on code.google.com by bruno.do.medeiros on 29 May 2013 at 1:39
Original issue reported on code.google.com by
bruno.do.medeiros
on 29 May 2013 at 1:39Attachments: