Extraneous Ratings Information

Question

I was just looking at the ratings element and wondering why it has so much data in it.

For me, it looks like this (using Python, the u means its unicode):
(u'rating', {u'positive': True, u'negative': False, u'value': 4, u'neutral': False, u'unknown': False, u'type': {u'simple': True, u'star_rating': False}})

Instead of getting a string for the rating, we get all four simple ratings as booleans. Additionally, we get the star rating and both types as another dictionary.

I think this would probably be a better option, just due to neatness:
(u'rating', {u'type': u'simple', u'rating': u'positive'})
or
(u'rating', {u'type': u'stars', u'rating': u'5'})

vikhyat · Answer

Thanks for the feedback. I had left the "rating" field out of the Library Entry model documentation for exactly this reason. It will indeed be changed to {type: ("simple" | "advanced"), rating: (0.5 | 1 | 1.5 | ... | 5 | null)} shortly.

halvkant · Answer

Will we have to do the conversion between simple and advanced rating modes on the client side or will it happen via server-side methods (or at least some documentation)?

vikhyat · Answer

Format has been updated -- the rating field now returns `{"type":"advanced","value":"4.0"}`.

Conversion between rating types needs to be done client-side. The rules for conversion are:
[0, 2.4] -> negative
(2.4, 3.6) -> neutral
[3.6, 5] -> positive

I will be documenting this soon.

Hummingbird v1

6.5 / 10

65ms

100%

N/A

Extraneous Ratings Information