Set UTF-8 Encoding for Reading Files.

[jeschkies]

We had the following exceptions on generating the coverage: java.nio.charset.MalformedInputException: Input length = 1. It turned out that Scoverage sets an UTF-8 encoding for writing files but not for reading, e.g. in CodeGrid. I suggest Source.fromFile(mfile.source, "UTF-8").mkString because Scoverage should not rely on the JVM default encoding if it is using UTF-8 for writing.

[ckipp01]

Thanks for bringing this up, it's a solid point. It looks like this was actually added to the ScoverageHtmlWriter, but not the other writers. So while even in sbt there is a setting key to set the encoding, it will only be applied to the html reports. I plan to fix this and align the various writers to all have the same behavior where you can pass in an encoding, and if not, it will default to utf-8. Right now the behavior is actually a bit different and for backwards compatibility reason it defaulted to system, but I want to assume that most of the time, in fact almost all of the time, that's not desired.

[jeschkies]

Thanks for addressing this. I should have proposed a change back then 🙈 🙂

[ckipp01]

Closed via #390