this is the code that powers @tinruufu, which happens to include a small but useful danbooru api client as well as a utility for downloading all images from a danbooru search result

you need python 2.7, probably running on a unix-y system like linux or bsd or mac osx or something

there are also some python things you can install with pip install [package] to make stuff work

completely necessary:

• requests, which we use to hit the danbooru api and download files

optional:

• gevent, for if you want to download files in parallel (it'll work fine without it, but it'll be slower)
• tweepy, for tweeting

you can set twitter and danbooru credentials if you like; make a file called secrets.py in this directory and make it look something like this:

danbooru_login = 'username'
danbooru_key = 'AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA'

token_key = 'AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA'
token_secret = 'AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA'
app_key = 'AAAAAAAAAAAAAAAAAAAAAAAAA'
app_secret = 'AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA'

without danbooru credentials associated with a gold account, you won't be able to access images with censored tags

without twitter keys and a file you can read at /usr/share/dict/words, you won't be able to tweet

to download all images containing nekomimi and red hair, for example, you would run python dl.py nekomimi red_hair

files are downloaded to a downloads directory in the current working directory

nt.py is where the business happens; it includes the utilities that the rest of the stuff uses and will try to tweet when invoked as a script

all arguments are taken as tags to filter by, just like dl.py

a delayed wrapper for nt.py that takes the same arguments and does the same thing, but waits a random interval between 0 and 48 hours before doing anything.