A dictionary with details about the feed info.
Same as feed.info.
If this contains HTML or XHTML, it is sanitized by default.
If this contains HTML or XHTML, certain (X)HTML elements within this value may contain relative URI (Uniform Resource Identifier)`s. If so, they are :ref:`resolved according to a set of rules <advanced.base>.
The content type of the feed info.
Most likely values for
For Atom feeds, the content type is taken from the type attribute, which defaults to text/plain if not specified. For RSS feeds, the content type is auto-determined by inspecting the content, and defaults to text/html. Note that this may cause silent data loss if the value contains plain text with angle brackets. There is nothing I can do about this problem; it is a limitation of RSS.
Future enhancement: some versions of RSS clearly specify that certain values default to text/plain, and Universal Feed Parser should respect this, but it doesn’t yet.
The language of the feed info.
~feed.info_detail.language is supposed to be a language code, as
specified by :abbr:`RFC (Request For Comments) 3066
<http://www.ietf.org/rfc/rfc3066.txt>`_, but publishers have been known to
publish random values like “English” or “German”. Universal Feed
Parser does not do any parsing or normalization of language codes.
~feed.info_detail.language may come from the element’s xml:lang
attribute, or it may inherit from a parent element’s xml:lang, or the
Content-Language HTTP header. If the
feed does not specify a language,
~feed.info_detail.language will be
None, the Python null value.
The original base URI for links within the feed copyright.
~feed.info_detail.base is only useful in rare situations and can
usually be ignored. It is the original base URI for this value, as specified by the element’s xml:base attribute,
or a parent element’s xml:base, or the appropriate HTTP header, or the URI of
the feed. (See Relative Link Resolution for more details.) By the time you see
it, Universal Feed Parser has already resolved relative links in all
values where it makes sense to do so. Clients should never need to manually
resolve relative links.