Alaa-Taieb/map-tunisia

The most accurate open-source SVG maps of Tunisia

Country · 7 economic regions · 5 districts · 24 governorates · 279 delegations · 2,083 imadas
5 levels of detail · 8 visual modes · SVG · GeoJSON · TopoJSON · العربية · Français · English

🧭 Live explorer · ⬇️ Download · ⚡ Quick start · 📚 Docs · ✅ Accuracy report

Most Tunisia SVGs online are hand-traced, outdated, missing islands, or stop at the 24 governorates. Map Tunisia is built from OpenStreetMap, the most precise open source of Tunisian boundaries, and cross-checked against official INS and ISIE lists. The result is a complete, exactly nested set of maps that you can drop into any project.

Six partitions, exactly nested

Eight modes

Five levels of detail (Djerba, at the same zoom)

More examples: labelled maps in both scripts, dark delegations, nested sectors

governorates · medium · labeled

governorates · medium · labeled-ar

delegations · low · dark

delegations · medium · nested

1. Use a file directly. Every SVG scales to any size:

<img src="https://cdn.jsdelivr.net/gh/Alaa-Taieb/map-tunisia@main/dist/svg/governorates/tunisia-governorates-medium-light.svg"
     alt="Map of the governorates of Tunisia" width="420">

2. Inline it and style it with CSS. Every shape has an id and data-* attributes:

#TN-61 { fill: #ffb703; }                            /* Sfax, by ISO 3166-2 code */
.unit[data-region="sud-ouest"] { fill: #8ecae6; }    /* a whole economic region */
.unit:hover { fill: #e63946; }

3. Add behaviour with a few lines of JavaScript:

document.querySelectorAll('.map-tunisia .unit').forEach((el) =>
  el.addEventListener('click', () =>
    alert(`${el.dataset.nameFr} · ${el.dataset.nameAr} — ${Number(el.dataset.population2024).toLocaleString()} inhabitants`)));

4. Place your own points. All SVGs share one frame (UTM 32N, 1 unit = 1 km):

import { project } from 'https://cdn.jsdelivr.net/gh/Alaa-Taieb/map-tunisia@main/src/projection.js';
const [x, y] = project([10.7600, 34.7406]); // Sfax → SVG coordinates
svg.insertAdjacentHTML('beforeend', `<circle cx="${x}" cy="${y}" r="3" fill="crimson"/>`);

More recipes for React, D3 + TopoJSON, Leaflet, MapLibre, Python (GeoPandas, Plotly), R, QGIS, Figma, print and laser cutting are in the usage guide.

Every file lives in dist/ and follows one naming scheme:

dist/svg/<partition>/tunisia-<partition>-<lod>-<mode>.svg     140 SVG maps
dist/geojson/tunisia-<partition>-<lod>.geojson                WGS84, with all attributes
dist/topojson/tunisia-<partition>-<lod>.topojson              WGS84, shared borders
dist/data/<partition>.csv | .json                             names, codes, populations, areas, label points

Browse every combination in the file catalog or the live explorer, or grab everything with Download ZIP.

For JavaScript projects: npm install github:Alaa-Taieb/map-tunisia, then import { svgPath, project } from 'map-tunisia' (API).

Details and guidance: partitions, levels & modes · every id and attribute: data dictionary.

Map Tunisia starts from OpenStreetMap and fixes what raw OSM gets wrong. Everything is reproducible and documented in the accuracy report, regenerated with every build.

How it works: all boundaries are noded into one set of faces, conflicts are resolved by documented rules, inland lakes are split by nearest-boundary allocation, and post-2014 delegations are rebuilt from their imadas using the electoral authority's (ISIE) official lists. Then mapshaper simplifies with topology preserved. Read the methodology.

Every SVG uses the same viewBox="0 0 406 838" in WGS 84 / UTM zone 32N with 1 unit = 1 km. Files of different partitions overlay perfectly, distances can be read straight off the map, and the dependency-free project() helper (tested to < 1 mm against PROJ) places any longitude/latitude. See projection & coordinates.

npm install
npm run build        # dist/ from data/ in ~30 s
npm test             # 160+ checks: counts, ids, nesting, areas, populations, XML validity, sizes, projection

# refresh from the latest OpenStreetMap data (Python 3.10+)
pip install -r scripts/extract/requirements.txt
npm run fetch && npm run extract && npm run build

The pipeline: pyosmium + Shapely (extraction and topology) → mapshaper (dissolve, projection, simplification) → a custom SVG writer (labels with polylabel, map colouring, accessible markup). See CONTRIBUTING.md.

• 6 nested partitions, 5 levels of detail, 8 modes
• All 279 delegations of the 2024 nomenclature
• GeoJSON / TopoJSON / CSV, projection helper, live explorer
• Municipalities (baladiyat) as a separate, non-nested partition
• Optional water layer: lagoons, sebkhas, chotts and dams
• Cities and capitals as point layers
• Tamazight (Tifinagh) and English-transliteration label modes
• Published npm package and versioned release ZIPs

Ideas and votes are welcome in the issues.

Spotted a wrong border, name or code? Report a data error. It's the most valuable contribution you can make. Code, docs and new modes are welcome too: see CONTRIBUTING.md and the Code of Conduct.

• Code (scripts, src/, explorer, tests): MIT
• Maps and data (dist/, data/, illustrations): ODbL 1.0, derived from OpenStreetMap

When you use the maps, please credit:

Map data © OpenStreetMap contributors · Map Tunisia by Alaa Taieb

Using Map Tunisia in research? See CITATION.cff (GitHub's Cite this repository button).

Alaa Taieb Creator & maintainer

Map Tunisia is created and maintained by Alaa Taieb. If this project helps you, please ⭐ star the repository, share it, and credit it when you publish a map. It helps others find it.

• The thousands of OpenStreetMap contributors who mapped Tunisia's boundaries and coastline
• Institut National de la Statistique (INS) for the geocodes and census figures, and the ISIE for its public circonscription maps
• Geofabrik for the daily extracts
• mapshaper, pyosmium, Shapely, pyproj and polylabel, the open-source tools this project stands on