GameDB
Video game database based on IGDB.
---
## Overview
GameDB is a database of games, platforms, characters, collections, franchises, and videos sourced from
[IGDB](https://www.igdb.com) and published as a static JSON API via
[GitHub Pages](https://lizardbyte.github.io/GameDB). Video metadata is enriched with data from the YouTube API.
The data is intended to be consumed by [LizardByte](https://app.lizardbyte.dev) projects such as
[Sunshine](https://github.com/LizardByte/Sunshine).
## Data
The database is updated automatically on a schedule via the [update-db](.github/workflows/update-db.yml) workflow.
The generated JSON files are published to the `gh-pages` branch and served at
`https://app.lizardbyte.dev/GameDB/`.
| Endpoint | Description | URL |
|-------------------|-------------------------------------------------------|------------------------------------------------------------------------------|
| Buckets | Game name search index, split by first two characters | `https://app.lizardbyte.dev/GameDB/buckets/.json` |
| Localized buckets | Optional regional game title search index | `https://app.lizardbyte.dev/GameDB/buckets/localized//.json` |
| Characters | Individual character details and all characters | `https://app.lizardbyte.dev/GameDB/characters/.json` |
| Collections | Individual collection details and all collections | `https://app.lizardbyte.dev/GameDB/collections/.json` |
| Franchises | Individual franchise details and all franchises | `https://app.lizardbyte.dev/GameDB/franchises/.json` |
| Games | Individual game details (no aggregate `all.json`) | `https://app.lizardbyte.dev/GameDB/games/.json` |
| Platforms | Individual platform details and all platforms | `https://app.lizardbyte.dev/GameDB/platforms/.json` |
| Videos | Individual YouTube video metadata | `https://app.lizardbyte.dev/GameDB/videos/.json` |
| Stats | Total item counts per category | `https://app.lizardbyte.dev/GameDB/stats.json` |
`all.json` files (e.g. `characters/all.json`) contain a summary of every item in that category as a single
dictionary keyed by ID.
Buckets are used as a lightweight search index for game names. Each bucket file is named after the first two
alphanumeric characters of the game name (lowercased), e.g. `ha.json` for games starting with "Ha" such as
*Halo*. Games whose names contain a space as the second character are put into a bucket named after the first
character. Games whose names do not start with two alphanumeric characters are grouped into `@.json`. Each bucket
contains a dictionary of `{ id: { name } }` entries, keeping individual files small for fast lookups.
### Optional game title and cover localization
Individual game records retain their original `name` and may also contain IGDB's `game_localizations` array.
Each localization has a `name`, an optional `cover.url`, and a `region` with an `identifier`, `name`, and `category`.
Use the localized cover URL when present, falling back to the game's original `cover.url` otherwise. IGDB regions can
represent a locale or a continent, so the identifier should be used as supplied by IGDB rather than assumed to be
a language code. If a game has no localization for the desired region, use its original `name`.
Localized titles are indexed separately under `buckets/localized//.json`. The region path component
is the lowercased `region.identifier`. A localized bucket uses the first two Unicode alphanumeric characters of the
localized title, lowercased, ignoring punctuation and spaces; titles with no alphanumeric characters use `@`.
Bucket entries keep the familiar `{ id: { name } }` shape, with `name` set to the localized title. Only games with
a title for that region appear in its buckets. To search both translated and original titles, search the selected
region's bucket and the original bucket, then deduplicate results by game ID.
The existing game `name`, bucket URLs, and original bucket contents remain available without selecting a region.
## Development
Code contributors can use the [Developer Setup](docs/developerSetup.md) guide.
