William Perron
2b3ff1aa04
|
7 months ago | |
---|---|---|
.github/workflows | 10 months ago | |
cmd/themis-server | 7 months ago | |
correlation | 10 months ago | |
data | 11 months ago | |
migrations | 11 months ago | |
.gitignore | 1 year ago | |
Dockerfile | 10 months ago | |
LICENSE | 1 year ago | |
README.md | 10 months ago | |
absences.go | 10 months ago | |
absences_test.go | 10 months ago | |
audit_log.go | 10 months ago | |
claim_type.go | 11 months ago | |
conflicts.go | 10 months ago | |
conflicts_test.go | 10 months ago | |
errors.go | 12 months ago | |
fly.toml | 10 months ago | |
fmt.go | 10 months ago | |
fmt_test.go | 10 months ago | |
go.mod | 9 months ago | |
go.sum | 9 months ago | |
notify.go | 7 months ago | |
start.sh | 2 years ago | |
store.go | 10 months ago | |
store_test.go | 10 months ago | |
time.go | 1 year ago | |
time_test.go | 1 year ago | |
uptime_darwin.go | 11 months ago | |
uptime_linux.go | 11 months ago | |
uptime_test.go | 11 months ago |
README.md
themis
Discord App to allow EU4 players to take claims on regions and provinces.
Setup
Requirements
To develop:
- Go version 1.21 or higher installed locally
sqlite3
installed locally (already ships by default on most OSes)
To deploy:
- Register for a free account on Fly
- An AWS account
Steps
To work with the core modules that simply interact with the database, or make sure that everything compiles correctly, you can run the following commands:
go test ./...
go build -o ./bin/ ./cmd/...
Operations
This application is deployed via Fly on a single virtual machine instance. It uses an embed SQLite database as its database engine, and uses Litestream to replicate the database to an S3 bucket.
Application Configuration
The deployment configurations can be found in the fly.toml file and you can find more information on the configuration options in the official Fly documentation.
The virtual image is based off of the Dockerfile which is a multi-stage build that builds the main application binary, downloads Litestream and packages everything on top of an Ubuntu 22.04 base.
Entrypoint
The application is started using a custom entrypoint shell script that is in charge of first restoring the database file through Litestream and then starting the main application as a child process of Litestream's replication process.
It's a very simple script but is necessary since the application doesn't have a persistent volume to rely on and must rehydrate its database file after every deployment.
Environment Variables
Env Var | Defined At |
---|---|
DISCORD_TOKEN |
fly secret |
DISCORD_APP_ID |
fly.toml |
DISCORD_GUILD_ID |
fly.toml |
AWS_ACCESS_KEY_ID |
fly secret |
AWS_SECRET_ACCESS_KEY |
fly secret |
Local Development
Application Entrypoint
./cmd/themis-server
This is the main application entrypoint./cmd/themis-repl
coming soon
Core Functions
The core database functions can be developed and tested locally easily, just run
go test ./...
to test your changes locally.
You can also load a test database easily by using the init.sql
script. You can
use the cli sqlite3 client or connect using whatever SQL editor you prefer that
can open sqlite3 connections.
sqlite3 local.db < migrations/init.sql
sqlite3 local.db
# interactive SQLite session
Discord Integration
This is a work in progress, but I am currently using a dedicated Discord server
with this application already signed into it. You can contact me directly to be
invited into that server. From there you only need to set the following
environment variables before launching the ./cmd/themis-server
entrypoint:
export DISCORD_APP_ID="1014881815921705030"
export DISCORD_GUILD_ID="[test server id goes here]"
Litestream replication
Litestream is not a necessary component to run the ./cmd/themis-server
entrypoint and can be safely ignored when developing locally. Still, if you wish
to try it out, you can find the Litestream commands used in production in the
start.sh script. As per the Litestream docs, it should work fine
with Minio but I have not tested it yet nor are there any
scripts provided to run it (yet).
SQLite
Importing From CSV
This is a neat feature built-in to SQLite. Using the source file at data/eu4-provinces.csv you can import the data directly into a SQLite database using the following command:
$ sqlite3
# ...
sqlite> .mode csv
sqlite> .import data/eu4-provinces.csv provinces
sqlite> .schema provinces
CREATE TABLE provinces(
"ID" TEXT,
"Name" TEXT,
"Development" TEXT,
"BT" TEXT,
"BP" TEXT,
"BM" TEXT,
"Trade good" TEXT,
"Trade node" TEXT,
"Modifiers" TEXT,
"Type" TEXT,
"Continent" TEXT,
"Superregion" TEXT,
"Region" TEXT,
"Area" TEXT
);
sqlite> select count(1) from provinces;
3925
Creating an SQL Dump From Imported Data
The init script at migrations/init.sql was initially created after importing the CSV data and running the following commands:
# In the same SQlite session as last section
sqlite> .output dump.sql
sqlite> .dump
Note: The column names were edited manually to remove capital letters and spaces to make it easier to work with.
Claims Schema
CREATE TABLE claim_types (
claim_type TEXT PRIMARY KEY
);
CREATE TABLE claims (
id INTEGER PRIMARY KEY AUTOINCREMENT,
player TEXT,
claim_type TEXT,
val TEXT,
FOREIGN KEY(claim_type) REFERENCES claim_types(claim_type)
);