import-to-pocket.py

#!/usr/bin/python

'''
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 sys
import argparse
import json
import requests
import pprint
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']:
        title = item['title'] if 'title' in item else item['origin']['title']
        title = title.encode('utf-8')

        if not 'alternate' in item:
            continue

        for alt in item['alternate']:
            if 'href' in alt:
                url = alt['href']
                break
        else:
            print title
            continue

        if restart_at_url and url == restart_at_url:
            state = 1

        if state == 0:
            continue

        data = {
                'access_token': access_token,
                'consumer_key': consumer_key,
                'url': url,
                'title': title,
                }

        print '[', url, ']', title
        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()