belgica 0.5.0__tar.gz

belgica-0.5.0/.gitignore

@@ -0,0 +1,22 @@
+node_modules/
+.DS_Store
+docs/.vitepress/dist/
+docs/.vitepress/cache/
+dist/
+coverage/
+.sources/
+csv/
+# Generated from data/courts.json and data/cantons.json by pnpm build:data-files; published, not committed
+data/courts/
+data/cantons/
+# Python package: the data and licences are copied in at build time
+python/src/belgica/data/
+python/LICENSE
+python/LICENSE-DATA
+python/tests/.cases/
+python/.venv/
+python/dist/
+__pycache__/
+.pytest_cache/
+.mypy_cache/
+.ruff_cache/

belgica-0.5.0/LICENSE

@@ -0,0 +1,24 @@
+MIT License
+
+Copyright (c) 2026 Pieter-Jan Scheir
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.
+
+This licence covers the code (src/, scripts/, test/ and the built files in
+dist/). The data in data/ is licensed under CC BY 4.0; see LICENSE-DATA.

belgica-0.5.0/LICENSE-DATA

@@ -0,0 +1,28 @@
+The data in data/ is licensed under the Creative Commons Attribution 4.0
+International licence (CC BY 4.0):
+https://creativecommons.org/licenses/by/4.0/legalcode
+
+Copyright (c) 2026 Pieter-Jan Scheir
+
+You may share and adapt the data for any purpose, including commercially, as
+long as you give appropriate credit, link to the licence, and indicate if
+changes were made. Credit "belgica (https://github.com/pieterjanscheir/belgica)"
+and pass on the source attributions below.
+
+Sources:
+
+- Municipalities (data/municipalities.json) and the REFNIS codes in
+  data/cantons.json: © Statbel (Directorate-general Statistics - Statistics
+  Belgium), CC BY 4.0, https://statbel.fgov.be/
+- Postal codes (data/postal-codes.json) and the municipality of each address
+  (address.municipality in data/courts.json, data/prisons.json and
+  data/houses-of-justice.json): Statbel's conversion table from postal codes
+  to REFNIS codes, © Statbel, CC BY 4.0, https://statbel.fgov.be/en/cc-40
+- Cantons, arrondissements and judicial areas: the annex to the Belgian
+  Judicial Code, as consolidated on Justel (https://www.ejustice.just.fgov.be/).
+  Official acts are not protected by copyright in Belgium (art. XI.172 §2 WER).
+- Interest rates: FOD Financiën and the Belgisch Staatsblad.
+- Court addresses and contacts: the courts' own websites and just.fgov.be.
+
+The data is provided as is, without warranty. It is reference data, not legal
+advice: check the official sources before relying on it.

belgica-0.5.0/PKG-INFO

@@ -0,0 +1,87 @@
+Metadata-Version: 2.5
+Name: belgica
+Version: 0.5.0
+Summary: Belgian legal reference data: courts, judicial cantons and arrondissements, municipalities and postal codes, and legal and commercial interest rates.
+Project-URL: Homepage, https://pieterjanscheir.github.io/belgica/
+Project-URL: Documentation, https://pieterjanscheir.github.io/belgica/guide/python
+Project-URL: Source, https://github.com/pieterjanscheir/belgica
+Project-URL: Issues, https://github.com/pieterjanscheir/belgica/issues
+Project-URL: Changelog, https://github.com/pieterjanscheir/belgica/blob/main/CHANGELOG.md
+Author: Pieter-Jan Scheir
+License-Expression: MIT AND CC-BY-4.0
+Keywords: arrondissements,belgian,belgium,courts,geojson,interest-rates,judicial-cantons,justice-of-the-peace,legal,municipalities,postal-codes,refnis,vredegerecht
+Classifier: Development Status :: 4 - Beta
+Classifier: Intended Audience :: Developers
+Classifier: Intended Audience :: Legal Industry
+Classifier: Natural Language :: Dutch
+Classifier: Natural Language :: French
+Classifier: Natural Language :: German
+Classifier: Programming Language :: Python :: 3
+Classifier: Topic :: Scientific/Engineering :: GIS
+Classifier: Typing :: Typed
+Requires-Python: >=3.11
+Description-Content-Type: text/markdown
+
+# belgica
+
+Belgian legal reference data for Python: every court with its registries' contacts, prisons,
+houses of justice, judicial cantons and arrondissements, municipalities and their postal codes,
+and the legal and commercial interest rates. Every record says where it comes from, and the
+scraped data is rebuilt from the official sources every week.
+
+The same data and helpers as the [`belgica` npm package](https://www.npmjs.com/package/belgica),
+with the same version number. The tests check that both give the same answers.
+
+```bash
+pip install belgica
+```
+
+Python 3.11 or later, no dependencies. Documentation:
+**https://pieterjanscheir.github.io/belgica/guide/python**
+
+It's reference data, not legal advice: check the official sources before relying on it.
+
+```python
+from belgica import (
+    find_arrondissements_at,
+    find_cantons_at,
+    find_competent_courts,
+    find_courts,
+    find_courts_for_postal_code,
+    get_cantons_for_municipality,
+    get_contact,
+    get_court,
+    get_court_location,
+    get_interest_rate,
+    get_justice_of_the_peace,
+)
+
+get_court('court-of-appeal-brussel')['name']  # {'nl': 'Hof van Beroep Brussel', 'fr': "Cour d'appel de Bruxelles"}
+get_contact(get_court('court-of-appeal-liege'), 'email', 'Greffe pénal')  # 'liege.ca.greffepenal@just.fgov.be'
+find_courts(type='enterprise-court', arrondissement='west-vlaanderen')
+
+[canton] = get_cantons_for_municipality('44088')  # merelbeke-melle
+get_justice_of_the_peace(canton['id'])['address']  # where to file
+get_cantons_for_municipality('44021', on='2025-09-15')  # Gent 1–5 on that day: a split city
+find_cantons_at(3.7268, 51.053)  # [gent-1]: the canton of a point
+find_arrondissements_at(3.7268, 51.053)  # [oost-vlaanderen]: and its arrondissement
+get_court_location('justice-of-the-peace-gent-1')['geometry']  # the court on the map, a GeoJSON Point
+find_competent_courts('23088', 'police-court')  # [police-court-vilvoorde]: which court, by REFNIS code
+find_courts_for_postal_code('1040', 'police-court')  # per municipality: Brussel and Etterbeek share the code
+
+get_interest_rate('2024-06-15', 'legal')  # InterestRateResult(rate=5.75, period_from='2024-01-01', period_to='2024-12-31')
+```
+
+The records are the dicts from the JSON files, with their keys as in the JSON (`validFrom`,
+`postalCode`), typed with the `TypedDict`s in `belgica.types`. Each data file is read on first
+use; the canton boundaries (22 MB) only by `find_cantons_at` and `get_canton_boundary`, the
+arrondissement boundaries only by `find_arrondissements_at` and `get_arrondissement_boundary`.
+
+Lookups take `on`, a day as a `datetime.date` or `'YYYY-MM-DD'`, and answer for today without it.
+A day before a dataset's coverage raises a `ValueError` rather than answer with today's data.
+
+## Licence
+
+Code MIT. Data CC BY 4.0: municipality, province, region and postal-code data © Statbel, canton and
+arrondissement boundaries and court locations © National Geographic Institute (NGI). See the
+[repository](https://github.com/pieterjanscheir/belgica) for the sources of every dataset.

belgica-0.5.0/README.md

@@ -0,0 +1,63 @@
+# belgica
+
+Belgian legal reference data for Python: every court with its registries' contacts, prisons,
+houses of justice, judicial cantons and arrondissements, municipalities and their postal codes,
+and the legal and commercial interest rates. Every record says where it comes from, and the
+scraped data is rebuilt from the official sources every week.
+
+The same data and helpers as the [`belgica` npm package](https://www.npmjs.com/package/belgica),
+with the same version number. The tests check that both give the same answers.
+
+```bash
+pip install belgica
+```
+
+Python 3.11 or later, no dependencies. Documentation:
+**https://pieterjanscheir.github.io/belgica/guide/python**
+
+It's reference data, not legal advice: check the official sources before relying on it.
+
+```python
+from belgica import (
+    find_arrondissements_at,
+    find_cantons_at,
+    find_competent_courts,
+    find_courts,
+    find_courts_for_postal_code,
+    get_cantons_for_municipality,
+    get_contact,
+    get_court,
+    get_court_location,
+    get_interest_rate,
+    get_justice_of_the_peace,
+)
+
+get_court('court-of-appeal-brussel')['name']  # {'nl': 'Hof van Beroep Brussel', 'fr': "Cour d'appel de Bruxelles"}
+get_contact(get_court('court-of-appeal-liege'), 'email', 'Greffe pénal')  # 'liege.ca.greffepenal@just.fgov.be'
+find_courts(type='enterprise-court', arrondissement='west-vlaanderen')
+
+[canton] = get_cantons_for_municipality('44088')  # merelbeke-melle
+get_justice_of_the_peace(canton['id'])['address']  # where to file
+get_cantons_for_municipality('44021', on='2025-09-15')  # Gent 1–5 on that day: a split city
+find_cantons_at(3.7268, 51.053)  # [gent-1]: the canton of a point
+find_arrondissements_at(3.7268, 51.053)  # [oost-vlaanderen]: and its arrondissement
+get_court_location('justice-of-the-peace-gent-1')['geometry']  # the court on the map, a GeoJSON Point
+find_competent_courts('23088', 'police-court')  # [police-court-vilvoorde]: which court, by REFNIS code
+find_courts_for_postal_code('1040', 'police-court')  # per municipality: Brussel and Etterbeek share the code
+
+get_interest_rate('2024-06-15', 'legal')  # InterestRateResult(rate=5.75, period_from='2024-01-01', period_to='2024-12-31')
+```
+
+The records are the dicts from the JSON files, with their keys as in the JSON (`validFrom`,
+`postalCode`), typed with the `TypedDict`s in `belgica.types`. Each data file is read on first
+use; the canton boundaries (22 MB) only by `find_cantons_at` and `get_canton_boundary`, the
+arrondissement boundaries only by `find_arrondissements_at` and `get_arrondissement_boundary`.
+
+Lookups take `on`, a day as a `datetime.date` or `'YYYY-MM-DD'`, and answer for today without it.
+A day before a dataset's coverage raises a `ValueError` rather than answer with today's data.
+
+## Licence
+
+Code MIT. Data CC BY 4.0: municipality, province, region and postal-code data © Statbel, canton and
+arrondissement boundaries and court locations © National Geographic Institute (NGI). See the
+[repository](https://github.com/pieterjanscheir/belgica) for the sources of every dataset.

belgica-0.5.0/hatch_build.py

@@ -0,0 +1,30 @@
+"""Copies the data and the licences from the repository into the package before a build.
+
+Building from the repository (python/ in a checkout) takes the files from ../; building a wheel
+from the sdist finds them already in it.
+"""
+
+import shutil
+from pathlib import Path
+from typing import Any
+
+from hatchling.builders.hooks.plugin.interface import BuildHookInterface
+
+
+def copy_from_repository(root: Path) -> None:
+    repository = root.parent
+    if not (repository / 'package.json').exists():
+        return
+    target = root / 'src' / 'belgica' / 'data'
+    target.mkdir(exist_ok=True)
+    for stale in target.glob('*.json'):
+        stale.unlink()
+    for file in (repository / 'data').glob('*.json'):
+        shutil.copyfile(file, target / file.name)
+    for name in ('LICENSE', 'LICENSE-DATA'):
+        shutil.copyfile(repository / name, root / name)
+
+
+class CustomBuildHook(BuildHookInterface):  # type: ignore[type-arg]
+    def initialize(self, version: str, build_data: dict[str, Any]) -> None:
+        copy_from_repository(Path(self.root))

belgica-0.5.0/pyproject.toml

@@ -0,0 +1,84 @@
+[project]
+name = "belgica"
+# The same version as the npm package: test_package.py checks it against ../package.json
+version = "0.5.0"
+description = "Belgian legal reference data: courts, judicial cantons and arrondissements, municipalities and postal codes, and legal and commercial interest rates."
+readme = "README.md"
+license = "MIT AND CC-BY-4.0"
+license-files = ["LICENSE", "LICENSE-DATA"]
+authors = [{ name = "Pieter-Jan Scheir" }]
+requires-python = ">=3.11"
+dependencies = []
+keywords = [
+    "belgium",
+    "belgian",
+    "legal",
+    "courts",
+    "judicial-cantons",
+    "arrondissements",
+    "municipalities",
+    "refnis",
+    "postal-codes",
+    "geojson",
+    "interest-rates",
+    "vredegerecht",
+    "justice-of-the-peace",
+]
+classifiers = [
+    "Development Status :: 4 - Beta",
+    "Intended Audience :: Developers",
+    "Intended Audience :: Legal Industry",
+    "Natural Language :: Dutch",
+    "Natural Language :: French",
+    "Natural Language :: German",
+    "Programming Language :: Python :: 3",
+    "Topic :: Scientific/Engineering :: GIS",
+    "Typing :: Typed",
+]
+
+[project.urls]
+Homepage = "https://pieterjanscheir.github.io/belgica/"
+Documentation = "https://pieterjanscheir.github.io/belgica/guide/python"
+Source = "https://github.com/pieterjanscheir/belgica"
+Issues = "https://github.com/pieterjanscheir/belgica/issues"
+Changelog = "https://github.com/pieterjanscheir/belgica/blob/main/CHANGELOG.md"
+
+[dependency-groups]
+# hatchling: the tests copy the data in with hatch_build.py
+dev = ["hatchling>=1.27", "pytest>=8", "ruff>=0.14", "mypy>=1.18"]
+
+[build-system]
+requires = ["hatchling>=1.27"]
+build-backend = "hatchling.build"
+
+[tool.hatch.build.hooks.custom]
+# hatch_build.py copies ../data, ../LICENSE and ../LICENSE-DATA in
+
+[tool.hatch.build.targets.sdist]
+# Copied in by the build hook, so not in git
+artifacts = ["src/belgica/data/*.json", "LICENSE", "LICENSE-DATA"]
+# Made by pnpm build:cases for the day they were made
+exclude = ["tests/.cases"]
+
+[tool.hatch.build.targets.wheel]
+packages = ["src/belgica"]
+artifacts = ["src/belgica/data/*.json"]
+
+[tool.pytest.ini_options]
+testpaths = ["tests"]
+
+[tool.ruff]
+line-length = 100
+target-version = "py311"
+# The examples there are laid out for reading, one call a line
+extend-exclude = ["README.md"]
+
+[tool.ruff.lint]
+select = ["E", "F", "I", "UP", "B", "SIM", "RUF"]
+
+[tool.ruff.format]
+quote-style = "single"
+
+[tool.mypy]
+strict = true
+files = ["src", "tests"]

belgica-0.5.0/src/belgica/__init__.py

@@ -0,0 +1,145 @@
+"""Belgian legal reference data: courts, judicial cantons and arrondissements, municipalities and
+postal codes, prisons, houses of justice, and legal and commercial interest rates.
+
+The records are the dicts from the JSON files, typed in `belgica.types`. Each data file is read
+on first use. Lookups take `on`, a day as a `date` or 'YYYY-MM-DD', and give today's answer
+without it. Documentation: https://pieterjanscheir.github.io/belgica/guide/python
+"""
+
+from ._arrondissement_boundaries import (
+    ARRONDISSEMENT_BOUNDARIES_COVERAGE_FROM,
+    arrondissement_boundaries,
+    find_arrondissements_at,
+    get_arrondissement_boundary,
+)
+from ._canton_boundaries import (
+    CANTON_BOUNDARIES_COVERAGE_FROM,
+    canton_boundaries,
+    find_cantons_at,
+    get_canton_boundary,
+)
+from ._competence import (
+    CantonCourt,
+    PostalCodeCourts,
+    find_competent_courts,
+    find_courts_for_postal_code,
+    find_justices_of_the_peace,
+)
+from ._contacts import format_iban, format_phone_number, get_contact
+from ._court_locations import COURT_LOCATIONS_COVERAGE_FROM, court_locations, get_court_location
+from ._courts import (
+    COURTS_COVERAGE_FROM,
+    courts,
+    find_courts,
+    format_court_name,
+    get_court,
+    get_courts_for_canton,
+    get_divisions,
+    get_justice_of_the_peace,
+)
+from ._interest_rates import (
+    InterestRateResult,
+    get_interest_rate,
+    get_latest_known_rate_date,
+    interest_rates,
+)
+from ._places import (
+    HOUSES_OF_JUSTICE_COVERAGE_FROM,
+    PRISONS_COVERAGE_FROM,
+    get_house_of_justice,
+    get_prison,
+    houses_of_justice,
+    prisons,
+)
+from ._postal_codes import (
+    POSTAL_CODES_COVERAGE_FROM,
+    get_municipalities_for_postal_code,
+    get_postal_code,
+    get_postal_codes,
+    postal_codes,
+)
+from ._territory import (
+    TERRITORY_COVERAGE_FROM,
+    arrondissements,
+    cantons,
+    get_arrondissement,
+    get_arrondissement_for_municipality,
+    get_canton,
+    get_cantons_for_municipality,
+    get_judicial_area,
+    get_municipality,
+    get_province,
+    get_province_for_municipality,
+    get_region,
+    get_region_for_municipality,
+    judicial_areas,
+    municipalities,
+    provinces,
+    regions,
+)
+from ._time import Day, applies_on, to_day
+
+__all__ = [
+    'ARRONDISSEMENT_BOUNDARIES_COVERAGE_FROM',
+    'CANTON_BOUNDARIES_COVERAGE_FROM',
+    'COURTS_COVERAGE_FROM',
+    'COURT_LOCATIONS_COVERAGE_FROM',
+    'HOUSES_OF_JUSTICE_COVERAGE_FROM',
+    'POSTAL_CODES_COVERAGE_FROM',
+    'PRISONS_COVERAGE_FROM',
+    'TERRITORY_COVERAGE_FROM',
+    'CantonCourt',
+    'Day',
+    'InterestRateResult',
+    'PostalCodeCourts',
+    'applies_on',
+    'arrondissement_boundaries',
+    'arrondissements',
+    'canton_boundaries',
+    'cantons',
+    'court_locations',
+    'courts',
+    'find_arrondissements_at',
+    'find_cantons_at',
+    'find_competent_courts',
+    'find_courts',
+    'find_courts_for_postal_code',
+    'find_justices_of_the_peace',
+    'format_court_name',
+    'format_iban',
+    'format_phone_number',
+    'get_arrondissement',
+    'get_arrondissement_boundary',
+    'get_arrondissement_for_municipality',
+    'get_canton',
+    'get_canton_boundary',
+    'get_cantons_for_municipality',
+    'get_contact',
+    'get_court',
+    'get_court_location',
+    'get_courts_for_canton',
+    'get_divisions',
+    'get_house_of_justice',
+    'get_interest_rate',
+    'get_judicial_area',
+    'get_justice_of_the_peace',
+    'get_latest_known_rate_date',
+    'get_municipalities_for_postal_code',
+    'get_municipality',
+    'get_postal_code',
+    'get_postal_codes',
+    'get_prison',
+    'get_province',
+    'get_province_for_municipality',
+    'get_region',
+    'get_region_for_municipality',
+    'houses_of_justice',
+    'interest_rates',
+    'judicial_areas',
+    'municipalities',
+    'postal_codes',
+    'prisons',
+    'provinces',
+    'regions',
+    'to_day',
+]

belgica-0.5.0/src/belgica/_arrondissement_boundaries.py

@@ -0,0 +1,69 @@
+"""The boundaries of the judicial arrondissements, and the arrondissement of a point"""
+
+from functools import cache
+
+from ._data import load
+from ._geometry import BoundingBox, bounding_box, contains, in_box, require_coordinates
+from ._territory import get_arrondissement
+from ._time import Day, applying_on, pick_version, versions_by_id
+from .types import Arrondissement, ArrondissementBoundary
+
+ARRONDISSEMENT_BOUNDARIES_COVERAGE_FROM = '2026-10-21'
+"""The boundaries describe the situation from the day the NGI's file was first read for them. A
+boundary the NGI changes replaces the previous one; an arrondissement the NGI drops keeps its
+boundary with a `validTo`. Lookups for an earlier day raise a `ValueError`."""
+
+_DATASET = 'arrondissement boundaries'
+
+
+@cache
+def arrondissement_boundaries() -> tuple[ArrondissementBoundary, ...]:
+    """Every boundary, every version, read on first use"""
+    return tuple(load('arrondissement-boundaries'))
+
+
+@cache
+def _by_id() -> dict[str, list[ArrondissementBoundary]]:
+    return versions_by_id(arrondissement_boundaries())
+
+
+def get_arrondissement_boundary(id: str, on: Day | None = None) -> ArrondissementBoundary | None:
+    """The boundary of an arrondissement, by arrondissement id: today's version, or the one of a
+    given day"""
+    return pick_version(_by_id().get(id), on, ARRONDISSEMENT_BOUNDARIES_COVERAGE_FROM, _DATASET)
+
+
+@cache
+def _index() -> list[tuple[ArrondissementBoundary, BoundingBox]]:
+    return [
+        (boundary, bounding_box(boundary['geometry'])) for boundary in arrondissement_boundaries()
+    ]
+
+
+def find_arrondissements_at(
+    longitude: float, latitude: float, on: Day | None = None
+) -> list[Arrondissement]:
+    """The arrondissements whose boundary holds a point, on a day (today by default). Longitude
+    first, then latitude, in decimal degrees (WGS 84), as GeoJSON orders them. One arrondissement
+    for a point in Belgium; none outside it; two for a point on the line between two. Raises
+    `ValueError` for coordinates that aren't finite numbers, or a day before the coverage."""
+    require_coordinates(longitude, latitude)
+    applying = {
+        id(boundary)
+        for boundary in applying_on(
+            arrondissement_boundaries(), on, ARRONDISSEMENT_BOUNDARIES_COVERAGE_FROM, _DATASET
+        )
+    }
+    found = []
+    for boundary, box in _index():
+        if (
+            id(boundary) in applying
+            and in_box(box, longitude, latitude)
+            and contains(boundary['geometry'], longitude, latitude)
+        ):
+            # Every boundary belongs to an arrondissement that applies on the same days
+            # (test/arrondissement-boundaries.test.ts)
+            arrondissement = get_arrondissement(boundary['arrondissement'], on)
+            if arrondissement:
+                found.append(arrondissement)
+    return found

belgica-0.5.0/src/belgica/_canton_boundaries.py

@@ -0,0 +1,66 @@
+"""The boundaries of the judicial cantons, and the canton of a point"""
+
+from functools import cache
+
+from ._data import load
+from ._geometry import BoundingBox, bounding_box, contains, in_box, require_coordinates
+from ._territory import get_canton
+from ._time import Day, applying_on, pick_version, versions_by_id
+from .types import Canton, CantonBoundary
+
+CANTON_BOUNDARIES_COVERAGE_FROM = '2026-10-21'
+"""The boundaries describe the situation from the day the NGI's file was first read. A boundary
+the NGI changes replaces the previous one; a canton the NGI drops keeps its boundary with a
+`validTo`. Lookups for an earlier day raise a `ValueError`."""
+
+_DATASET = 'canton boundaries'
+
+
+@cache
+def canton_boundaries() -> tuple[CantonBoundary, ...]:
+    """Every boundary, every version. About 22 MB of JSON, read on first use."""
+    return tuple(load('canton-boundaries'))
+
+
+@cache
+def _by_id() -> dict[str, list[CantonBoundary]]:
+    return versions_by_id(canton_boundaries())
+
+
+def get_canton_boundary(id: str, on: Day | None = None) -> CantonBoundary | None:
+    """The boundary of a canton, by canton id: today's version, or the one of a given day"""
+    return pick_version(_by_id().get(id), on, CANTON_BOUNDARIES_COVERAGE_FROM, _DATASET)
+
+
+@cache
+def _index() -> list[tuple[CantonBoundary, BoundingBox]]:
+    return [(boundary, bounding_box(boundary['geometry'])) for boundary in canton_boundaries()]
+
+
+def find_cantons_at(longitude: float, latitude: float, on: Day | None = None) -> list[Canton]:
+    """The cantons whose boundary holds a point, on a day (today by default). Longitude first,
+    then latitude, in decimal degrees (WGS 84), as GeoJSON orders them. One canton for a point in
+    Belgium; none outside it; two for a point on the line between two cantons. The boundary is the
+    NGI's digitisation of the annex to the Judicial Code: where a point lies close to a boundary,
+    the canton's `territory` text decides. Raises `ValueError` for coordinates that aren't finite
+    numbers, or a day before the coverage."""
+    require_coordinates(longitude, latitude)
+    applying = {
+        id(boundary)
+        for boundary in applying_on(
+            canton_boundaries(), on, CANTON_BOUNDARIES_COVERAGE_FROM, _DATASET
+        )
+    }
+    found = []
+    for boundary, box in _index():
+        if (
+            id(boundary) in applying
+            and in_box(box, longitude, latitude)
+            and contains(boundary['geometry'], longitude, latitude)
+        ):
+            # Every boundary belongs to a canton that applies on the same days
+            # (tests/test_data.py)
+            canton = get_canton(boundary['canton'], on)
+            if canton:
+                found.append(canton)
+    return found