import-to-pocket.py

#!/usr/bin/env python3

'''
Import Google Reader starred items to Pocket
============================================

This script will read the list of starred items generated by exporting your
Reader data using [Google Takeout][].  Generally you will use it like this:

    python import-to-pocket.py starred.json

Where `starred.json` is the file containing your starred items.

Authentication
==============

You will need to populate the file `pocket.yml` with your Pocket OAUTH
credentials, like this:

    consumer_key: YOUR_CONSUMER_KEY
    access_token: YOUR_ACCESS_TOKEN

Getting a consumer key
----------------------

You will need to visit <http://getpocket.com/developer/apps/new> and create
a new application.

Getting an access_token
-----------------------

Get an OAUTH request token:

    curl -k -d consumer_key=CONSUMER_KEY \
            -d redirect_uri=http://localhost/ \
            https://getpocket.com/v3/oauth/request

This will return a request token:

    code=REQUEST_TOKEN

Insert that token (everything after `code=`) into the following URL and
paste it into your browser:

    https://getpocket.com/auth/authorize?request_token=REQUEST_TOKEN&redirect_uri=http://localhost/

After authorizing access to your Pocket data, convert the request token
into an access token:

    curl -k -d consumer_key=CONSUMER_KEY \
            -d code=REQUEST_TOKEN \
            https://getpocket.com/v3/oauth/authorize

This will return an access_token:

    access_token=ACCESS_TOKEN&username=larsks

Place this access token and the consumer key into `pocket.yml`.

Details of this process are at
<http://getpocket.com/developer/docs/authentication>.

API Limits
==========

Note that Pocket [rate limits][] API requests to 320/hour.  By default, this script
will load items as fast as possible, which means it will ultimately bail
out if you have more than 320 starred items.  Just wait a bit and start it
again and it will pick up where it left off.

You can pass the `-R` command line flag to have the script limit requests
to stay under the API limit.  This should allow it to run unattended until
completion, but you'll only get a new item every 10 seconds or so.

[google takeout]: https://www.google.com/takeout/
[rate limits]: http://getpocket.com/developer/docs/rate-limits
'''

import os
import argparse
import json
import requests
import time
import yaml

def parse_args():
    p = argparse.ArgumentParser()
    p.add_argument('--ratelimit', '-r', default=320,
            help='API rate limit as number of requests/hour')
    p.add_argument('--config', '-f', default='pocket.yml')
    p.add_argument('--enforce-limit', '-R', action='store_true')
    p.add_argument('input')
    return p.parse_args()

def main():
    opts = parse_args()

    with open(opts.config) as fd:
        cfg = yaml.load(fd)

    access_token = cfg['access_token']
    consumer_key = cfg['consumer_key']

    interval = (60*60) / int(opts.ratelimit)

    with open(opts.input) as fd:
        starred = json.load(fd)

    try:
        with open('restarturl') as fd:
            restart_at_url = fd.read().strip()
            print('will restart at', restart_at_url)
            os.unlink('restarturl')
    except OSError:
        restart_at_url = None

    state=0
    for item in starred['items']:
        url = item['url']  # should fail if no url

        if restart_at_url and url == restart_at_url:
            state = 1

        data = {
            'access_token': access_token,
            'consumer_key': consumer_key,
            'url': url,
            'tags': item['tags'],
        }

        print(f"{url}")
        r = requests.post('https://getpocket.com/v3/add',
                data = json.dumps(data),
                headers = {
                    'Content-type': 'application/json; charset=UTF-8',
                    'X-accept': 'application/json',
                    },
                )

        try:
            r.raise_for_status()
        except:
            open('restarturl', 'w').write(url)
            raise

        if opts.enforce_limit:
            time.sleep(interval)

if __name__ == '__main__':
    main()

reddit-link-dl.py

#! /usr/bin/env python3

"""
Download a full list of your reddit saved and upvoted items and saved into output.json.
The JSON format used is expected by ./import-to-pocket.py, so it can be imported to pocket easily.

Authentication
==============

This script uses PRAW, the official Reddit python API, for getting your upvoted and saved posts.

To use it, you'll have to create an app:

https://github.com/reddit-archive/reddit/wiki/OAuth2-Quick-Start-Example

Only need first steps with PRAW (client ID and secret); don't need an auth token.

Configuration
=============

Save the `id`, `sec`, and your `username` and `password` into reddit.yml
(using those keys). This script will read the yaml and download all posts to
output.json, which can be edited before you run ./import-to-pocket.py.
"""

import itertools
import json
import uuid

import praw
import yaml

USER_AGENT = str(uuid.uuid4())

CONFIG = 'reddit.yml'


def parse_tags(url):
    if url.startswith("/r/"):
        url = url[3:]

    return url.split('/', 1)[0]


def main():
    with open(CONFIG) as fd:
        cfg = yaml.load(fd)

    username = cfg['username']
    password = cfg['password']
    _id = cfg['id']
    sec = cfg['sec']

    reddit = praw.Reddit(
        client_id=_id,
        client_secret=sec,
        password=password,
        username=username,
        user_agent=USER_AGENT,
    )

    me = reddit.user.me()

    upvoted = [
        {"url": "https://www.reddit.com" + p.permalink, "tags": parse_tags(p.permalink) + ",reddit,upvoted"}
        for p in me.upvoted()
    ]

    saved = [
        {"url": "https://www.reddit.com" + p.permalink, "tags": parse_tags(p.permalink) + ",reddit,saved"}
        for p in me.saved()
    ]

    with open("output.json", "w") as f:
        json.dump({"items": list(itertools.chain(upvoted, saved))}, f)


main()

updated to python3, and updated the import script to be able to import reddit posts you've upvoted/saved.

meant to be used as a one-time conversion; I use IFTTT for transferring liked/saved posts on my reddit account to Pocket continuously.

Note - for anyone else using this, the script does NOT sort by timestamp of post (in fact, the reddit api will return your posts with most recent first?). So running this sync will result in your oldest reddit saved/upvoted being at the front of your pocket list, unless you sorted(reverse=True) the list somewhere before pocket API calls.