Requested feature overview description.

Provide a simpler way to get hourly and total credit limits via API. It could be at /user/credit.

The way it could work is when user sends username and api key would return a list of credit names and remaining amounts (uploads hourly, uploads total, note edits, flags,...) like

[{"tag_edits":150},{"flags":10},...]

Why would it be useful?

It would be simpler and more accurate for app makers to implement limits into their apps.

What part(s) of the site page(s) are affected?

API

I don't believe there's an API endpoint for it, but you could parse:

https://e621.net/user/upload_limit

Though it would be awkward to check it with a program, as you still need to log in to see it..

Pupslut said:
I don't believe there's an API endpoint for it, but you could parse:

https://e621.net/user/upload_limit

Though it would be awkward to check it with a program, as you still need to log in to see it..

Yes, that + more and make it an api.

There are alternative ways that don't require login but take a while.

Pupslut said:
I don't believe there's an API endpoint for it, but you could parse:

https://e621.net/user/upload_limit

This seemed like an interesting problem, so here's my solution.

You can get all the relevant elements with the xpath /html/body/div[4]/div[1]/ul/li (although it might be better to change this into a class based check rather than indexing). This is a slight modification of one produced by Firefox's Inspector.

so, Python+LXML solution:

import lxml.html as lh
# put in download code here.. for my test I just saved using the browser.
h = lh.parse(downloaded_html) 
nodes = h.xpath('/html/body/div[4]/div[1]/ul/li')
d = {}
for v in nodes:
    nodes = h.xpath('/html/body/div[4]/div[1]/ul/li')
    # filter out the 'Formula:' line
    if not val.isdigit():
        continue
    d[id.lower()] = int(val)

The result (d) looks like this:

{'approved': 1, 'deleted': 3, 'pending/flagged': 0, 'remaining hourly uploads': 30, 'remaining hourly tag edits': 150}

Login would be trivial if there was an API endpoint, but it seems like urls which are not API endpoints ignore the usual login / passwordhash POST parameters (meaning that requesting /user/upload_limit does return a legible page, but all limits are set to a value of 0).

So you'd probably have to use something like the mechanize module to pretend to be a browser. Not especially hard, but it does slow things down, since you need to wait for two requests from the server, instead of one).