Vetustas Archiva

Open, curated datasets of notable things across history and culture. movies, music, games, books, people, companies, inventions, landmarks, art and more, compiled and enriched from public sources into clean, consistent CSVs.

4,531 records across 14 datasets. Plain CSV, one file per category. No build step needed to use the data — just read the files.

The Vetustas ecosystem

This repository is the single source of truth. Everything else is built on it:

Explorer	archiva.vetustas.net	Browse, search, sort and filter every dataset, with interactive charts (value counts, histograms, choropleth maps).
Public API	api.vetustas.net	Token-less, rate-limited JSON API. Per-dataset endpoints with filters. Docs at docs.vetustas.net.
The Game	vetustas.net	A trivia / guessing game powered by these datasets.

Datasets

File	Records	What
datasets/movies.csv	440	Films, with IMDb ratings, director, genre, runtime
datasets/tv.csv	329	TV series, seasons/episodes, network, creator
datasets/anime.csv	268	Anime, studio, format, AniList scores
datasets/games.csv	272	Video games, developer, publisher, series
datasets/music.csv	309	Albums, label, tracks, certifications, popularity
datasets/literature.csv	315	Books, author, type, language
datasets/fine-art.csv	292	Paintings & sculptures, artist, movement, medium
datasets/landmarks.csv	289	Buildings & sites, architect, country
datasets/people.csv	502	Notable people, occupation, nationality
datasets/companies.csv	398	Companies, founding, industry, type
datasets/inventions.csv	296	Inventions & discoveries, inventor, category
datasets/events.csv	497	Historical events, type, era
datasets/football-clubs.csv	234	Football clubs
datasets/basketball-clubs.csv	90	Basketball clubs

Data format

• UTF-8 CSV, \r\n line endings, header row first.
• Every row has a frozen id — a UUIDv4 that is permanent and globally unique. It never changes and is never reused, even if a record is corrected. It's the stable key to reference a record over time.
• Dates use signed YYYY[-MM[-DD]], truncated to the real precision (a year-only fact is YYYY, not YYYY-01-01). BC dates are negative, -YYYY.
• Image columns (poster / cover / logo / crest / image) hold URLs to externally hosted files (this repo doesn't store the images).
• Controlled-vocabulary columns (genre, type, category, movement, era) are normalized to a small canonical set per dataset.

Using the data

It's just CSV — open it in anything (pandas, a spreadsheet, csv in any language):

import csv
with open("datasets/movies.csv") as f:
    for row in csv.DictReader(f):
        print(row["title"], row["release_date"], row["imdb_rating"])

import fs from "fs";
const movies = fs
  .readFileSync("datasets/movies.csv", "utf-8")
  .split("\n")
  .slice(1)
  .map((line) => line.split(","))
  .map(([id, title, release_date, imdb_rating]) => ({
    id,
    title,
    release_date,
    imdb_rating,
  }));

Or query it live, without cloning anything, via the public API:

curl "https://api.vetustas.net/v1/movies?genre=Sci-Fi&min_rating=8&year_min=2000"

Full endpoint reference and filters: docs.vetustas.net.

Scripts

The datasets are generated and maintained by a set of Node + Python scripts under scripts/. You only need these to add or refresh data — using the data needs nothing but the CSVs.

npm install
cp .env.example .env       # add API keys where a source needs one (e.g. TMDB)

npm run expand:movies      # one scraper per category (expand:tv, expand:games, …)
npm run backfill:id        # stamp UUIDs onto new rows
npm run validate           # schema, unique ids, date checks

📖 Full script reference — every command, what it does, and its arguments, is in scripts/README.md.

Requires Node 18+ and Python 3.8+.

License

Dual-licensed:

• Datasets (datasets/) — CC BY-SA 4.0. Free to use, share and adapt (including commercially) with attribution; derivatives stay open under the same license.
• Code (scripts/) — MIT.

Data is compiled from public sources, each under its own terms. See datasets/LICENSE for the full attribution and source notes.

Contributing

Contributions welcome and very needed. corrections, new records, new sources. See CONTRIBUTING.md for the rules and how to submit a PR.