readme.md

Draft.dev Author Collector

This project contains the Draft.dev Author Database collection code that is run on n8n.io.

Deploying to Production

• Set up a new DO droplet with Node 12+ installed
• SSH into the new server
• Clone this repo into the web directory: git clone https://github.com/draftdev/authors-collector.git /var/www/html
• Install dependencies: npm i
• Create a .config file and copy the base config into it (see below).
• Run n8n as the nodejs user using pm2: sudo -u nodejs N8N_PORT=3000 N8N_CONFIG_FILES=.config pm2 start node_modules/.bin/n8n
• View the app and copy workflows into it: http://<DROPLET_IP_ADDRESS>

Base Config

The config adds basic auth to ensure people can't come in and mess with your workflows.

{
	"security": {
		"basicAuth": {
			"active": true,
			"user": "...",
			"password": "..."
		}
	}
}

Upgrading in Production

• Kill the node process: sudo -u nodejs pm2 kill
• Pull the latest code: git pull
• Install packages: npm i
• Start the node process again (see above).

Database Schema

Updated 10/30/20

-- Install extensions
CREATE EXTENSION IF NOT EXISTS "uuid-ossp";

-- Sources table
create table sources
(
    id uuid default uuid_generate_v1() not null
        constraint sources_pk
            primary key,
    name varchar not null,
    rss_url varchar(1024) not null,
    url varchar(1024),
    pay_rate varchar(256),
    last_checked_at timestamp
);

create unique index sources_rss_url_uindex
    on sources (rss_url);

-- Posts table
create table posts
(
    id uuid default uuid_generate_v1() not null
        constraint posts_pk
            primary key,
    title varchar(1024),
    url varchar(1024) not null,
    description text,
    source_id uuid,
    author_id uuid,
    original_author varchar(256),
    original_topics json,
    created_at timestamp default now() not null,
    published_at timestamp,
    topics_extracted_at timestamp,
    author_extracted_at timestamp
);

create unique index posts_url_uindex
    on posts (url);

-- Authors table
create table authors
(
    id uuid default uuid_generate_v1() not null
        constraint authors_pk
            primary key,
    name varchar(1024) not null,
    bio text,
    links json,
    created_at timestamp default now() not null
);

-- Topics tables
create table topics
(
    id serial
        constraint topics_pk
            primary key,
    name varchar(50) not null,
    description varchar(720)
);

create unique index topics_name_uindex
	on topics (name);

create table post_topics (
    post_id uuid REFERENCES posts (id) ON UPDATE CASCADE ON DELETE CASCADE,
    topic_id int REFERENCES topics (id) ON UPDATE CASCADE,
    CONSTRAINT post_topics_pkey PRIMARY KEY (post_id, topic_id)
);

Data Sources

• Sources include most paying community writer programs as well as select Medium blogs.
• Topics extracted from Stack Overflow.

Workflows

See the ./workflows folder.