Normalize content-type before extention computation in SimpleHTTPResolver

[damooo]

in SimpleHTTPResolver, when we are saving fie to cache, we are computing file-extension from request's content-type header. this extension will be then used in computing source_format at time of resolving.

But it is not normalizing content-type before dictionary lookup. Thus it doesn't support values like image/jpeg; charset=UTF-8, (archive.org responds like that.). We have to normalize it to image/jpeg before dictionary lookup.

Here is the place in code:

loris/loris/resolver.py

Line 301 in 3c641eb

extension = self.get_format(ident, constants.FORMATS_BY_MEDIA_TYPE[response.headers['content-type']])