halkyon_app: data locations as precise as a country or a state

An operator's DATA_LOCATION and BACKUP_LOCATION may name a subdivision
(ISO 3166-2, US-CA, DE-BY) besides the EU or a country. The app adds
"(EU)" to a member country and the country to a subdivision ("Germany
(EU)", "California, United States", "Bavaria, Germany (EU)"), in
English and German, and shows the most specific flag there is: a
subdivision's own where the flag package has one, else its country's.
Subdivision names come from Unicode CLDR 48.2 (tool/subdivision_names.py),
bundled with their licence, which the app registers.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: Tizian 「ティツィアーン」 <raisondetredev@pm.me>

tool/subdivision_names.py

@@ -0,0 +1,66 @@
+"""Writes halkyon_app's subdivision names from a release of Unicode CLDR.
+
+The names of states, regions and provinces (ISO 3166-2) that
+`AssetSubdivisionNames` reads, in English and German, from CLDR's
+`common/subdivisions/` at a tagged release, with CLDR's licence beside them:
+
+    python3 -I tool/subdivision_names.py [release-48-2]
+
+CLDR names a subdivision by its country and code in lower case without the
+hyphen (`usca`); the asset keys it as ISO 3166-2 does (`US-CA`). English is
+kept whole; German only where it differs, as the app falls back to English.
+Variants (`alt`) are left out. Run it again with a newer tag to update both
+files, and say the release in `assets/subdivisions/ATTRIBUTION.md`.
+"""
+
+import json
+import re
+import sys
+import urllib.request
+import xml.etree.ElementTree as ElementTree
+from pathlib import Path
+
+RELEASE = sys.argv[1] if len(sys.argv) > 1 else 'release-48-2'
+LANGUAGES = ('en', 'de')
+SOURCE = 'https://raw.githubusercontent.com/unicode-org/cldr/' + RELEASE + '/'
+ASSETS = Path(__file__).resolve().parent.parent / 'packages' / 'halkyon_app' / 'assets' / 'subdivisions'
+CODE = re.compile(r'^[A-Z]{2}-[A-Z0-9]{1,3}$')
+
+
+def fetch(path):
+    with urllib.request.urlopen(SOURCE + path, timeout=60) as response:
+        return response.read()
+
+
+def names_in(language):
+    root = ElementTree.fromstring(fetch('common/subdivisions/' + language + '.xml'))
+    names = {}
+    for element in root.iter('subdivision'):
+        if 'alt' in element.attrib or not (element.text or '').strip():
+            continue
+        cldr = element.attrib['type']
+        code = cldr[:2].upper() + '-' + cldr[2:].upper()
+        if CODE.match(code):
+            names[code] = element.text.strip()
+    return names
+
+
+def main():
+    english = names_in('en')
+    names = {'release': RELEASE, 'en': english}
+    for language in LANGUAGES[1:]:
+        names[language] = {
+            code: name
+            for code, name in names_in(language).items()
+            if english.get(code) != name
+        }
+    ASSETS.mkdir(parents=True, exist_ok=True)
+    (ASSETS / 'subdivisions.json').write_text(
+        json.dumps(names, ensure_ascii=False, sort_keys=True, separators=(',', ':')) + '\n',
+        encoding='utf-8',
+    )
+    (ASSETS / 'LICENSE-CLDR.txt').write_bytes(fetch('LICENSE'))
+    print(RELEASE, {language: len(names[language]) for language in LANGUAGES})
+
+
+main()