TITLESCREEN

API

JSON over HTTPS, no key, CORS open, cached at the edge. Base: https://titlescreen.org/api/v1/. Please use the archive for bulk work rather than walking the API.

Empty means unknown: a field with no value is left out of the answer, never sent as null or zero.

Identify a copy

GET /api/v1/lookup?sha1=<hex>&size=<bytes>     the identity: SHA-1 + size of the canonical form
GET /api/v1/lookup?crc32=<hex>&size=<bytes>    a pre-check: a candidate, not a confirmation
GET /api/v1/lookup?fp1=<hex>                   the disc fingerprint: candidates
GET /api/v1/lookup?serial=SLUS-00594           by the publisher's serial
GET /api/v1/lookup?set=sf2                     an arcade set by name

The answer:

{
  "matches": [
    {
      "basis": "sha1",          // sha1, file (a list's whole-file hash), crc32, fp1, serial, set
      "dump": {"id": 123, "key": "snes/no-intro/Super Metroid (Japan, USA) (En,Ja)", "name": "...", "platform": "snes", ...},
      "releases": [{"id": "r...", "platform": "snes", "region": "JP", ...}, {"id": "r...", "region": "US", ...}],
      "work": {"id": "w...", "title": "Super Metroid"}
    }
  ],
  "identical_releases": true    // one copy, sold as each of the releases listed
}

For a whole library, ask in batches: up to 50 items a request.

GET /api/v1/lookup/batch?sha1=<hex>:<size>,<hex>:<size>,...&fp1=<hex>,...
    -> {"sha1": {"<hex>:<size>": <answer as above>, ...}, "fp1": {"<hex>": <answer>, ...}}
GET /api/v1/releases/batch?ids=r...,r...
    -> [release summaries with platform, generation, publisher, developer and genre names joined in]

How to compute the hashes: the canonical form and the disc fingerprint. Nothing about your file is sent but the hash you ask about.

Games, releases, companies

GET /api/v1/works/{w...}              a game and its releases, titles and links
GET /api/v1/releases/{r...}           a release with its identified copies, files and hashes
GET /api/v1/releases/{r...}/evidence  the statements its facts were resolved from (the same page as HTML: /release/{r...}/evidence)
GET /api/v1/companies/{c...}
GET /api/v1/search?q=&platform=&page=
GET /api/v1/platforms
GET /api/v1/genres
GET /api/v1/meta                      the build's version and inputs

A release carries support: for each published fact, how many sources state it (agree), how many state another value (against), the best join behind it (joined: id an identifier, name the title, main the main title, override a correction), and the sources on each side. How these are made: Methodology.

Ids are stable: a game, release or company id never changes meaning. Fields are only ever added within this version of the API.