.gitignore | ||
Pipfile | ||
Pipfile.lock | ||
pynit.py | ||
README.md | ||
requirements.txt | ||
template.sh | ||
unlicense |
pynit
an integration between saved reddit posts and pinboard.
specifically, comments and posts saved to your reddit user profile can be pulled into pinboard through this project. the following information is pulled down and used when pinning to pinboard:
- subreddit (becomes a tag)
- post title (becomes title)
- post description, if it exists (becomes description)
- post url (becomes url)
the tag "added-by-pynnit" is also added to each entry moved to pinboard in this way. This allows for easy viewing of all imported links. I found this very useful when I was writing the script in the first place, but you may not want it.
using this bullshit
installing
I use pipenv
pretty exclusively because of how it works with my editor. that's the only package manager i've actually tested with, but theoretically the typical pip install -r requirements.txt
should work just fine.
setting up the environment
You need several things to use this project:
- Reddit username
- Reddit password
- Reddit client ID
- Reddit client secret
- Pinboard api key
Your reddit un/pw you should already have. The client ID and client secret can be generated by following the instructions here: https://github.com/reddit-archive/reddit/wiki/oauth2
Your pinboard api key can be found here (assuming you are logged in): https://pinboard.in/settings/password/
You then pass those variables as arguments to the script run:
☭ python pynit.py --reddit-un '' --reddit-pw '' --reddit-cid '' --reddit-sec '' --pb-apikey ''
background and additional information
getting data from reddit
if you don't have MFA set up for your account (you should do that) then this is easy. everything works as expected. BUT if you DO have MFA set up let me tell you: this gets dumber. when you enter your reddit password you have to also include an active MFA 6 digit code, like this:
... --redditpw "thing:123123"
so good.
putting data in pinboard
This is slow. The API docs published by pinboard require that you only make a call once per 3 seconds, so if you're adding a lot of entries at once (say, during the intial move over) this can take a while, but its still fast enough for my purposes.
bonus (cursed?) information about reddit (i'm so sorry)
So, there are multiple kinds of reddit posts, and each kind of reddit post seems to have distinct names for the same things, which is REALLY fucking annoying. Its extra frustrating because there's not just a quick lookup for this, you have to just dig through Too Much json.
If you're using PRAW, a reddit /post/ has an attribute called .is_self
that's boolean. If its true, its a text only post, if its false then its a link post.
Reddit /comments/ do not have this attribute. They DO have an attribute called .is_root
, which i use to differentiate themm.
roadmap
- move reddit/pinboard script to single file.
- make the script take command line arguments (i.e. must pass
-reddit
if you want to sync saved posts) - write new function
update_tags
that will take a tag and retag to a new name (add a cli flag)