</div>@
I have no idea why all this info is even presented in such an easily accessible way, but it's awesome for my user script which displays artist names among other things. However, inconsistency between `data-image-tag-aliases`, `data-image-tags`, `title` and `alt` attributes stops me from displaying correct artist tags in some cases, as well as implementing more interesting features. Also pure CSS rating detection sometimes fails because of silly aliases like "solo female"="solo female explicit".
What I get now is this:
![full](https://i.imgur.com/iHp9Qoy.jpg)
I can parse title tags of course, but if I have an option of getting pure data instead of parsing formatted text for users, I'd rather go the clean way.
**Could you please add an attribute like `data-image-tag-names` with just clean tags or remove them from the `data-image-tag-aliases` attribute?**
P.S. What's the purspose of the datasets anyway? I don't see them used anywhere.
The datasets are used in the clientside query parser to evaluate complex filters
There will likely always be some inconsistency between data-image-tag-aliases and data-image-tags; the latter is the authoritative source, and the former is a cache column that includes all aliases for denormalizing searching.
I am planning to redo this properly in the future. The only thing I can say right now is that it might be a while to wait for it.
@byte[] Client-side query parser? Using currently listed posts? Where is it used? I feel like I’m missing a major feature somewhere. :D
Hmm, looks like the API thingy provides the data I need in the form I need. Am I right to assume that on most (all?) pages I can just append “.json” to the location.pathname, leaving the rest of URL intact, perform the query with credentials, and get the exact same data that is displayed on the page (unless data is changed between the requests, of course)? The official docs mention keys, but cookies seem to work just fine.