[ruby] UTF-8 issues for :bytes fields

[namxam]

I have the following specification

Google::Protobuf::DescriptorPool.generated_pool.build do
  add_message "api.Mutation" do
    optional :set_json, :bytes, 1
  end
end
Mutation = Google::Protobuf::DescriptorPool.generated_pool.lookup("api.Mutation").msgclass

When I try to set a UTF-8 string like this Api::Mutation.new(set_json: 'Möp') it fails with ``initialize': U+00F6 from UTF-8 to ASCII-8BIT (Encoding::UndefinedConversionError)`.

Is this not supported or is it a bug? From what I have read, UTF-8 and ASCII Input should be fine for bytes.

[namxam]

Ok, my bad… in ruby you are supposed to call .b on a string to return a copied string whose encoding is ASCII-8BIT.

https://ruby-doc.org/core-2.3.0/String.html#method-i-b