Saving UTF-8 file makes non-alphabet char unrecognizable characters.

Environment

• Operating System: Windows_NT x64 10.0.16299, Chinese Home
• JDK version: 1.8.0_111
• Visual Studio Code version: 1.19.0
• Java extension version: 0.16.0

Steps To Reproduce

1. Open a UTF-8 java file having non-alphabet char (e.g., Chinese character) in in VSCode
2. Type/delete any char and save the file

[attach a sample project reproducing the error]
Only such info in server log file

!ENTRY org.eclipse.jdt.ls.core 1 0 2017-12-19 09:45:34.091
!MESSAGE >> document/hover

!ENTRY org.eclipse.jdt.ls.core 1 0 2017-12-19 09:45:34.707
!MESSAGE >> document/documentHighlight

!ENTRY org.eclipse.jdt.ls.core 1 0 2017-12-19 09:45:34.964
!MESSAGE >> document/codeAction

!ENTRY org.eclipse.jdt.ls.core 1 0 2017-12-19 09:46:54.920
!MESSAGE >> document/codeAction

!ENTRY org.eclipse.jdt.ls.core 1 0 2017-12-19 09:46:54.921
!MESSAGE >> document/documentHighlight

!ENTRY org.eclipse.jdt.ls.core 1 0 2017-12-19 09:46:55.179
!MESSAGE >> document/codeAction

All other extensions are disabled, and if this Java extension is disabled, too, VSCode saves the file correctly.