Character offsets from the client are utf-16 code units

[Marwes]

The language server protocol specifies that that character positions are in UTF-16 code units (see https://github.com/Microsoft/language-server-protocol/blob/master/protocol.md#text-documents). I encountered this after crashing https://github.com/gluon-lang/gluon_language-server due to accidentally typing a multi byte character.

While tracking down what character was supposed to mean for the protocol I also looked at what rls did to see if there was a easy way to handle this but from what I understand from the source I don't think this is handled in any way.

https://github.com/rust-lang-nursery/rls/blob/9782aa4c3c5b203732b9a1a4f3c9d7fbc912e437/src/actions/requests.rs#L1005

https://github.com/rust-lang-nursery/rls/blob/41bf9af09ccc9cbbe8a67a375f0f7fe3196dbabf/src/lsp_data.rs#L100

Feel free to just close this If I am mistaken, I figured I'd at least open an issue in case this needs to be handled.

[Xanewok]

Related discussion: microsoft/language-server-protocol#376

[Marwes]

FWIW this is the location translation I am using https://github.com/gluon-lang/gluon_language-server/blob/master/src/location_translation.rs (manually using encode_utf16 and len_utf16)

[Xanewok]

Everything old is new again: #1113

[Xanewok]

The aforementioned issue has more recent information so I'll close this as duplicate.