Add python languagedata

.github/workflows/python.yml

@@ -0,0 +1,52 @@
+name: Python build
+on:
+  [pull_request, push]
+
+jobs:
+  build:
+    runs-on: ubuntu-latest
+    strategy:
+      matrix:
+        python-version: ["3.11", "3.12", "3.13"]
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+
+      - name: Set up Python ${{ matrix.python-version }}
+        uses: actions/setup-python@v5
+        with:
+          python-version: ${{ matrix.python-version }}
+
+      - name: Install package
+        run: pip install .
+
+      - name: Run tests
+        run: |
+            pip install pytest
+            pytest tests/python
+            python -c "import languagedata; print('languagedata package loaded successfully')"
+
+  publish:
+    runs-on: ubuntu-latest
+    needs: build
+    if: startsWith(github.ref, 'refs/tags/v')
+    environment: release
+    permissions:
+      id-token: write
+      contents: read
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+
+      - name: Set up Python 3.13
+        uses: actions/setup-python@v5
+        with:
+          python-version: "3.13"
+
+      - name: Build package
+        run: |
+          pip install uv
+          uv build
+
+      - name: Publish to PyPI
+        uses: pypa/gh-action-pypi-publish@release/v1

pyproject.toml

@@ -0,0 +1,44 @@
+[build-system]
+requires = ["hatchling"]
+build-backend = "hatchling.build"
+
+[project]
+name = "languagedata"
+description = "Language data and utilities"
+version = "2.0.3"
+requires-python = ">=3.10"
+license = "GPL-2.0-or-later"
+readme = "README.md"
+authors = [{name= "Santhosh Thottingal", email= "santhosh.thottingal@gmail.com"}]
+classifiers = [
+    "Programming Language :: Python :: 3",
+    "Operating System :: OS Independent",
+]
+
+[tool.hatch.build.targets.wheel]
+packages = ["python/languagedata"]
+
+[tool.hatch.build.targets.sdist]
+include = [
+    "python/languagedata/",
+    "data/language-data.json",
+]
+exclude = [
+	"src*",
+    "docs*",
+    ".venv*",
+    ".ruff_cache*",
+	".pytest_cache*",
+]
+
+[tool.pytest.ini_options]
+testpaths = ["tests/python"]
+pythonpath = ["python"]
+
+
+[dependency-groups]
+dev = [
+    "pytest>=7.0",
+    "ruff>=0.1.0",
+]
+

python/languagedata/__init__.py

@@ -0,0 +1,153 @@
+from typing import Union, List
+from importlib.resources import files
+import functools
+import json
+
+with files('languagedata').joinpath('language-data.json').open('r') as f:
+    languageData = json.load(f)
+
+
+def isKnown(languageCode):
+    return languageCode in languageData['languages']
+
+
+def isRedirect(language):
+    if isKnown(language) and len(languageData['languages'][language]) == 1:
+        return languageData['languages'][language][0]
+    else:
+        return False
+
+
+def getLanguages():
+    return languageData['languages']
+
+
+def getScript(language):
+    target = isRedirect(language)
+    if target:
+        return getScript(target)
+    elif not isKnown(language):
+        return 'Zyyy'
+    else:
+        return languageData['languages'][language][0]
+
+
+def getRegions(language: str) -> Union[str, List[str]]:
+    target = isRedirect(language)
+    if target:
+        return getRegions(target)
+    elif isKnown(language) and languageData['languages'][language][1]:
+        return languageData['languages'][language][1]
+    else:
+        return 'UNKNOWN'
+
+
+def getAutonym(language: str) -> str:
+    target = isRedirect(language)
+    if target:
+        return getAutonym(target)
+    return languageData['languages'][language][2] if (isKnown(language) and len(languageData['languages'][language]) > 2) else language
+
+
+def getAutonyms():
+    autonymsByCode = {}
+    for language in languageData['languages']:
+        if isRedirect(language):
+            continue
+        autonymsByCode[language] = getAutonym(language)
+    return autonymsByCode
+
+
+def getLanguagesInScripts(scripts):
+    languagesInScripts = []
+    for language in languageData['languages']:
+        if isRedirect(language):
+            continue
+        for script in scripts:
+            if script == getScript(language):
+                languagesInScripts.append(language)
+                break
+    return languagesInScripts
+
+
+def getLanguagesInScript(script):
+    return getLanguagesInScripts([script])
+
+
+def getGroupOfScript(script):
+    for scriptGroup in languageData['scriptgroups']:
+        if script in languageData['scriptgroups'][scriptGroup]:
+            return scriptGroup
+    return 'Other'
+
+
+def getScriptGroupOfLanguage(language):
+    return getGroupOfScript(getScript(language))
+
+
+def getLanguagesByScriptGroup(languages):
+    languagesByScriptGroup = {}
+    for language in languages:
+        resolvedRedirect = isRedirect(language) or language
+        langScriptGroup = getScriptGroupOfLanguage(resolvedRedirect)
+        if langScriptGroup not in languagesByScriptGroup:
+            languagesByScriptGroup[langScriptGroup] = []
+        languagesByScriptGroup[langScriptGroup].append(language)
+    return languagesByScriptGroup
+
+
+def getLanguagesByScriptGroupInRegions(regions):
+    languagesByScriptGroupInRegions = {}
+    for language in languageData['languages']:
+        if isRedirect(language):
+            continue
+        for region in regions:
+            if region in getRegions(language):
+                scriptGroup = getScriptGroupOfLanguage(language)
+                if scriptGroup not in languagesByScriptGroupInRegions:
+                    languagesByScriptGroupInRegions[scriptGroup] = []
+                languagesByScriptGroupInRegions[scriptGroup].append(language)
+                break
+    return languagesByScriptGroupInRegions
+
+
+def getLanguagesByScriptGroupInRegion(region):
+    return getLanguagesByScriptGroupInRegions([region])
+
+
+def sortByScriptGroup(languages):
+    groupedLanguages = getLanguagesByScriptGroup(languages)
+    scriptGroups = sorted(groupedLanguages.keys())
+    allLanguages = []
+    for scriptGroup in scriptGroups:
+        allLanguages.extend(groupedLanguages[scriptGroup])
+    return allLanguages
+
+
+def _cmpByAutonym(a, b):
+    autonymA = getAutonym(a) or a
+    autonymB = getAutonym(b) or b
+    return -1 if autonymA.lower() < autonymB.lower() else 1
+
+
+def sortByAutonym(languages):
+    return sorted(languages, key=functools.cmp_to_key(_cmpByAutonym))
+
+
+def isRtl(language):
+    return getScript(language) in languageData['rtlscripts']
+
+
+def getDir(language):
+    return 'rtl' if isRtl(language) else 'ltr'
+
+
+def getLanguagesInTerritory(territory):
+    return languageData.get('territories', {}).get(territory, [])
+
+
+def addLanguage(code, options):
+    if options.get('target'):
+        languageData['languages'][code] = [options['target']]
+    else:
+        languageData['languages'][code] = [options.get('script'), options.get('regions'), options.get('autonym')]

python/languagedata/language-data.json

@@ -0,0 +1 @@
+../../data/language-data.json