A small hack to automatically upload Podcast audios to YouTube. The code is not pretty (and the required setup neither).
Read the instructions bellow carefully (and also check the source code, it's just a few functions).
It probably won't work out-of-the-box -- sorry about that.
- Watches a Podcast feed for new episodes;
- Convert the episode to a video format using a static background image;
- Push the video to a YouTube channel.
- The script is deployed to a cloud server;
- A crontab executes the script periodically;
- First parse the podcast xml from the web, check if there is a new episode;
- If there's a new episode, download it first to the server;
- Grab a predefined image, generate a
.flvvideo file using this static image as background; - When the video file is done, push it to a YouTube channel.
If you already have a Podcast setup and wants to make your episodes available on YouTube this script will automate that task. The file conversion is necessary because YouTube doesn't let you upload audio files to their platform.
I basically made it to fit my wife's Podcast needs, where she uses Wordpress and Blubrry to generate the xml feed. I don't really understand much about podcasting, but I believe the xml feed should be more or less standarized.
The script uses Python 2.
The script might need some tweaks to make it work on your environment. In my case it's running on a Ubuntu VPS.
git clone https://github.com/vitorfs/pc2yt.git
After that, install the requirements (ideally inside a virtualenv):
pip install -r requirements.txt
Now install ffmpeg:
apt-get install ffmpeg
Within the same directory as the pc2yt.py file, create two empty folders: audios and videos.
mkdir audios videos
Then, create a .env file with the URL of your Podcast feed:
.env
FEED_URL=https://example.com/podcast/feed.xml
Save a file named background.png in the same directory as the pc2yt.py. This file will be used to generate the static background of the podcast video.
To send a file from your local machine to a server:
scp background.png root@server_ip_address:/path/to/pc2yt_directory/
Now go to the Google Cloud Platform Console, optionally create a new project or use an existing one.
Visit the APIs & Services page, then Credentials and create a new OAuth client ID. You may need to fill out the details of the OAuth consent screen before generating the credentials.
After you create your new OAuth credentials, click on the Download JSON button.
You will download a file named client_secret_9999999999-xxxxxxxxxx.apps.googleusercontent.com.json.
Rename it to just client_secret.json instead.
Now save this file to the save directory as the pc2yt.py script. If it's on a cloud server, you can send it using scp:
scp client_secret.json root@server_ip_address:/path/to/pc2yt_directory/
There's another configuration file, which is a file named .last. This file stores the reference to the last podcast converted and uploaded. The reference I use here is the id of the podcast entry. In case of podcast using Wordpress/Blubbry setup, an example of this file contents is:
.last
https://example.com/?p=596
To understand what you should put there you should first inspect your feed xml and check what is used as id. You can do that using a Python shell:
import feedparser
d = feedparser.parse('https://example.com/feed.xml')
d['entries'][0]['id']
This file is not mandatory. If you start the script without a .last file, it will download, convert and upload ALL podcasts in your feed.
So you can either use .last file to set a reference point, from what point it will start converting and uploading, or let the script figure it out.
This is important, because of the OAuth, the first time you use it is a little bit different. You also need to tell to what YouTube channel should the script upload the files.
First time you are using it, run the command below:
python pc2yt.py --noauth_local_webserver
When the code reachs the get_authenticated_service() function, you will get a notification on the console. It will give you a long URL. Access this URL on a web browser, select which Google account you want to use and what YouTube channel you want to upload the files to. Finally, get the verification code, go back to the console and paste the verification code.
After that you will see the script will create a file named youtube.dat. It will now take care of refreshing the token by itself.
This should be the last thing in the configuration. Depends on how frequently you post new episodes, you might want to tune this setting:
sudo crontab -e
An example where the script would be execute every hour, at 12:05, 13:05, 14:05, etc:
# m h dom mon dow command
5 * * * * /home/pc2yt/venv/bin/python /home/pc2yt/pc2yt.py