Ensure non-ASCII characters are handled as expected about datasets-server

[severo]

Cannot reproduce for now because neither of the reported examples is streamable, and this backend is restricted to streamable datasets.

• https://datasets-preview.huggingface.tech/rows?dataset=wmt19&config=cs-en&split=train
• https://datasets-preview.huggingface.tech/rows?dataset=russian_super_glue&config=rucos&split=train

[severo]

wmt19 does not seem to show issues (see huggingface/datasets#897 (comment) for the original issue, with the former viewer)

[severo]

For russian_super_glue, all the configs are previewable, but rucos, which is the example in the original issue: huggingface/datasets-viewer#15. But I don't think the current issue has anything to do with the encoding, and we should fix that error first before being able to check the encoding

https://huggingface.co/datasets/russian_super_glue/viewer/rucos/test

[github-actions]

This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.

Please note that issues that do not follow the contributing guidelines are likely to be ignored.

[severo]

Seems to be working as expected