java.nio.charset.UnmappableCharacterException: Input length = 1

[RadoBuransky]

I have some nice Slovak characters in source code. For example:

class Class2 {
  def msg_test = println("test code ŠTEFÁNIK")
}

But then if I run Scoverage it crashes with following:

Input length = 1
java.nio.charset.UnmappableCharacterException: Input length = 1
    at java.nio.charset.CoderResult.throwException(CoderResult.java:282)
    at sun.nio.cs.StreamDecoder.implRead(StreamDecoder.java:339)
    at sun.nio.cs.StreamDecoder.read(StreamDecoder.java:178)
    at java.io.InputStreamReader.read(InputStreamReader.java:184)
    at java.io.BufferedReader.read1(BufferedReader.java:210)
    at java.io.BufferedReader.read(BufferedReader.java:286)
    at java.io.Reader.read(Reader.java:140)
    at scala.io.BufferedSource.mkString(BufferedSource.scala:96)
    at scoverage.report.CodeGrid.source(CodeGrid.scala:63)
    at scoverage.report.CodeGrid.<init>(CodeGrid.scala:17)
    at scoverage.report.ScoverageHtmlWriter.filePage(ScoverageHtmlWriter.scala:78)

It is obviously related to encoding. I am able to reproduce it by explicitly setting wrong encoding in CodeGrid.scala line 63:

  private def source(mfile: MeasuredFile): String = Source.fromFile(mfile.source, "windows-1252").mkString

Current implementation uses an arbitrary default encoding which is probably not ideal. It can be different for every one of us. I would propose to hardcode UTF-8. If you like this idea, I will send a pull request soon.

[mismare]

Hi guys,

Do you have an estimate as to when the fix is going to be released?

Thanks,
Ema

[RadoBuransky]

@sksamuel Sam?

[sksamuel]

I'll do a release tonight.

[mismare]

Great, thanks!

[gslowikowski]

Is this issue fixed? Where/when?

[RadoBuransky]

@gslowikowski #135

[gslowikowski]

I missed this PR. It's late, but I must say, I don't like this solution. I would like to have possibility to parametrize source encoding. This is natural in all Maven plugins I'm using. I don't want to add warning to Scoverage Maven plugin, that UTF-8 is the only acceptable encoding since next version. It can be default, but add the possibility to change it.

[mismare]

Hi guys,

Any updates on this ?

Thanks,
Ema

[gslowikowski]

Here you have alternative approach - source encoding parametrized

[gslowikowski]

Very simple SBT plugin implementation - https://github.com/gslowikowski/sbt-scoverage/tree/source-encoding.
In SBT encoding can and should be found in scalacOptions if not specified in coverageSourceEncoding setting key. Someone with better SBT knowledge can implement this logic.

[gslowikowski]

scoverage/sbt-scoverage#86 is related to this issue and will be fixed, if my proposed solutions will be accepted.

[gslowikowski]

SBT plugin proposed implementation (https://github.com/gslowikowski/sbt-scoverage/commits/source-encoding) improved - default encoding value pulled from scalacOptions in (Compile). It's complete now.

[gslowikowski]

Nobody's interested in this functionality anymore?