avformat/id3v2: allow ID3 parsing without AVFormatContext

avformat/id3v2: allow ID3 parsing without AVFormatContext

Add ff_id3v2_read_dict() for parsing without AVFormatContext, but
instead with AVIOContext and AVDictionary.

AVFormatContext is still used for logging, if available.

Chapter parsing is the only non-logging functionality that actually
needs AVFormatContext, and AFAICS it should be modified to write the
data to ID3v2ExtraMeta first, from where it can be implanted to
AVFormatContext by a separate function (like it is done with
read_apic() and ff_id3v2_parse_apic()). That is outside the scope of
this patch, though.

Signed-off-by: Anssi Hannula <anssi.hannula@iki.fi>