Language in DDI exporter

Dataverse has a metadatafield "Language". But it seems there is no way to determine the language of dataset in DDI export. It seems that the only field that has language attribute is subject keyword, but it is always English, disregarding the dataset language.

What steps does it take to reproduce the issue?
Create a dataset with metadatafield "Language" selected as non-English, for example French. Publish dataset. Export DDI through UI "Metadata->DDI" or Native API.
Field stdyDscr/stdyInfo/subject/keyword" will be always with attribute "xml:lang="en" disregarding the language of the study and the controlled vocabulary will be always in English. There is no other way to determine the language of the dataset.

• When does this issue occur?
  During DDI export.

• To whom does it occur (all users, curators, superusers)?
  All users that use DDI export.

• What did you expect to happen?
  ""xml:lang" attribute and controlled vocabulary to depend on language of the dataset.

Reproduce the issue by creating and publishing a dataset with a non-English Language value, then export DDI through Metadata→DDI and the Native API. Inspect stdyDscr/stdyInfo/subject/keyword, especially xml:lang and the controlled vocabulary, and trace the DDI export path. Done means both reflect the dataset language rather than always using English.