The HTML 5 specification describes an algorithm for selecting the character encoding to be used in a multi-part form submission (e.g. UTF-8). However, it is not clear how the selected encoding should be relayed to the server so that the content can be properly decoded on the receiving end.
Often, character encodings are represented by appending a "charset" parameter to the value of the Content-Type request header. However, this parameter does not appear to be defined for the multipart/form-data MIME type:
https://www.rfc-editor.org/rfc/rfc7578#section-8
Each part in a multipart form submission may provide its own Content-Type header; however, RFC 7578 notes that "in practice, many widely deployed implementations do not supply a charset parameter in each part, but rather, they rely on the notion of a 'default charset' for a multipart/form-data instance".
RFC 7578 goes on to suggest that a hidden "_charset_" form field can be used for this purpose. However, neither Safari (9.1) nor Chrome (51) appear to populate this field, nor do they provide any per-part encoding information.
I've looked at the request headers produced by both browsers and I don't see any obvious character encoding information. Does anyone know how the browsers are conveying this information to the server?