CLI for Readlist

I maintain the readlist in a CouchDB database. Each feed (channel) is a document in that database. I use the file's name as the primary key i.e "_id". For example, "sri-lankas-economic-crisis.json" is the key to the Sri Lankan Economic Crisis reading list. It's a single document. It has many feed items like any JSONFeed. The first few were easy to create and manage. But then I needed something simple to manage this if I was going to be serious about using it. So I borrowed the idea from BBRF client and created a simple CLI client that I can use.

Given a config file. Readlist CLI can be used to create a Channel, and add items to a channel. That's it. I have not added destructive actions to the CLI. I do it manually, as I don't do that often. 

{
    "couchdb": "https://courchdb-url.com/readlists",
    "username": "couchdb_user",
    "password": "couchdb_password",
    "channel": "current_channel.json",
    "author":
    {
        "name": "Thejesh GN",
        "url": "https://thejeshgn.com"
    },
    "home_page_url": "https://thejeshgn.com",
    "feed_base_url": "https://example.cdn.com/readlists/{0}",
    "debug": true
}

While adding the URL to an existing Channel, it will try and pull the entire content if "-f" flag is passed. Else it will add the minimum content required to form the JSONFeed item. 

Readlist Client

Usage:
  readlist get
  readlist use <channel>
  readlist new -c <channel> -t <title> -d <description>
  readlist [-c <channel>] [-f] add <url> 

Options:
  -h --help            Show this screen.
  -v --version         Version of this program
  -c <channel>         Select a channel to limit the command to. Not required when the command "use" has been run before.

I use trafilatura to scrape the articles and get the details, requests to talk to CouchDB and docopt to create the command-line interface.

it's not fully tested, and I have not written test cases yet. But I have been using it for a while now. Use it at your risk. 

It also needs a CouchDB. CLI provides the input side of the feed. On the output side, you can use a serverless function to read the document corresponding to a channel or readlist from CouchDB and serve it. Since the Document is a valid JSONFeed, you don't need to do any transformation before serving. It can be cached easily and can be served using a CDN.

You can install it from

https://github.com/thejeshgn/readlist