Mime type checking can't handle dash character

[pieter-edelman-nictiz]

Hi,

The most recent version of the Validator complains when it encounters the following URL in the narrative of a resource: data:application/octet-stream;base64,volmacht, with the message:

The URL is not valid because '(The mimetype portion of the data: URL is not valid (Mime type invalid) in URL: data:application/octet-stream;base64,volmacht)': data:application/octet-stream;base64,volmacht

See test.zip for the offending file.

So it claims 'application/octet-stream' is not a valid mime type, although it actually is valid. I think this is due to the checkValidMimeType method in InstanceValidator.java. Here, the following regex is used to check for validity of the mime type:

^(\w+|\*)\/(\w+|\*)((;\s*(\w+)=\s*(\S+))?)$

However, the \w character class doesn't include the normal dash, although this character is allowed and quite common in the subtype part of mime types. Changing the regex to something like:

^(\w+|\*)\/([\w-]+|\*)((;\s*(\w+)=\s*(\S+))?)$

should probably solve the problem.