Fetch recently played episodes from Overcast.fm. WIP.

overcast-recently-played.py

'''
You'll need to pip install some dependencies:

* python-dateutil
* requests

Also, populate your EMAIL and PASSWORD below.
'''

from xml.etree import ElementTree
from datetime import datetime
from dateutil.tz import UTC
from dateutil.parser import parse as parse_dt

import re
import sys
import requests
import pickle
import os.path


# configuration
EMAIL = 'EMAIL HERE'
PASSWORD = 'PASSWORD HERE'


# load stored session, or re-authenticate
if os.path.exists('session.saved'):
    print('Found saved session. Restoring!')
    session = pickle.loads(open('session.saved', 'rb').read())
else:
    print('No saved session. Authenticating!')
    session = requests.Session()
    response = session.post('https://overcast.fm/login', data={
        'email': EMAIL,
        'password': PASSWORD
    })

    if response.status_code != 200:
        print('Authentication failed')
        sys.exit(0)

    print('Authenticated successfully. Saving session.')

    with open('session.saved', 'wb') as saved_session:
        saved_session.write(pickle.dumps(session))

# fetch the latest detailed OPML export from Overcast
print('Fetching latest OPML export from Overcast')
response = session.get('https://overcast.fm/account/export_opml/extended')
if response.status_code != 200:
    print('Failed to fetch OPML. Exiting.')
    print(response.text)
    sys.exit(0)

# parse the OPML
tree = ElementTree.fromstring(response.text)

# find all podcasts and their episodes
podcasts = tree.findall(".//*[@type='rss']")

# look for recently played episodes
now = datetime.utcnow().astimezone(UTC)
for podcast in podcasts:
    pod_title = podcast.attrib['title']
    for episode in podcast.getchildren():
        # skip unplayed episodes
        played = episode.attrib.get('played', '0') == '1'
        if not played:
            continue

        # skip episodes played over one day ago
        user_activity_date = parse_dt(episode.attrib.get('userUpdatedDate'))
        recency = now - user_activity_date
        if recency.days > 1:
            continue

        # parse out the remaining details we care about
        title = episode.attrib['title']
        published = parse_dt(episode.attrib['pubDate'])
        url = episode.attrib['url']
        overcast_url = episode.attrib['overcastUrl']
        overcast_id = episode.attrib['overcastId']
        progress = episode.attrib.get('progress')

        # fetch the epside artwork
        response = session.get(overcast_url)
        results = re.findall('img class="art fullart" src="(.*)"', response.text)
        artwork_url = ''
        if len(results) == 1:
            artwork_url = results[0]

        print('Played episode of ', pod_title)
        print('    ->', title)
        print('    ->', artwork_url)
        print('    ->', url)
        print('    ->', overcast_url)
        print('    ->', user_activity_date)

Hi I'm very new to python and script running, but I'm getting "ImportError: cannot import name UTC" - I've installed the dependencies, but I don't know if that is the dateutil is the issue?

Python 2.7.10

To be honest, I'm not great with Python either. I ended up using the logic as the foundation for a Node.js version. I forked this code from here: https://gist.github.com/cleverdevil/a8215850420493c1ee06364161e281c0 I would recommend asking there.

I know this is kinda old but did you ever release the node version of this at all?

@rknightuk I built it in a private repo where I had my username and password as strings. Which I shouldn't have done.

But I just copied and pasted that into a Gist over here: https://gist.github.com/EdwardHinkle/80c1f012bdb5681961f834d1a2ada596

Thank you so much!

My pleasure! Have fun 😁