Currently there are several matching functions not implemented, ex: doc, xls, flac, [wav](https://github.com/gabriel-vasile/mimetype/blob/master/matcher

In https://mimesniff.spec.whatwg.org/#parsing-a-mime-type the following step might be clearer if it explicitly said the trailing sequence after " was to be discarded. Might also be better if the example contains a ';' since the step mentions ';'

Current:

Collect a sequence of code points that are not U+003B (;) from input, given position.
Example: Given text/html;charset="shift_jis"iso-20