cubicweb #1587305 Better HTTP Accept header parsing [resolved]

We currently use the same parse_accept_header() method for Accept-* HTTP headers but there are a few subtleties on the path, especially for the Accept header itself :

  • mimetypes might have additional information (such as level=1)
  • for a same score, mimetypes must be sorted from most specific ones to less specific ones (i.e. consider wildcards and additional information)

CubicWeb should provide a more robust implementation for that.

done in3.12.0
load left0.000
closed by<not specified>