PIVOT-13757: Fix locale bundling for JDK23+

pyproject.toml

@@ -28,7 +28,7 @@ keywords = ["jdk", "java", "jvm", "jre"]
 name = "jdk4py"
 readme = "README.md"
 requires-python = ">=3.10"
-version = "25.0.2.0"
+version = "25.0.2.1"
 
 [project.urls]
 Repository = "https://github.com/activeviam/jdk4py"
@@ -47,6 +47,7 @@ ignore = [
     "D203",
     "D211",
     "D212",
+    "E501",
     "EM102",
     "ISC001",
     "S101",

src/jdk4py/_included_locales.py

@@ -1,17 +1,29 @@
+from collections.abc import Collection
+from typing import Literal, TypeAlias, get_args
+
+_RegionSpecificLocale: TypeAlias = Literal[
+    "bn-IN",
+    "da-DK",
+    "de-DE",
+    "en-GB",
+    "en-US",
+    "es-ES",
+    "es-MX",
+    "fr-FR",
+    "it-IT",
+    "ja-JP",
+    "pt-BR",
+    "ru-RU",
+    "zh-CN",
+]
+_REGION_SPECIFIC_LOCALES: Collection[str] = get_args(_RegionSpecificLocale)
+
 INCLUDED_LOCALES = frozenset(
-    [
-        "bn-IN",
-        "da-DK",
-        "de-DE",
-        "en-US",
-        "en-GB",
-        "es-ES",
-        "es-MX",
-        "fr-FR",
-        "it-IT",
-        "ja-JP",
-        "pt-BR",
-        "ru-RU",
-        "zh-CN",
-    ],
+    {
+        locale
+        for tag in _REGION_SPECIFIC_LOCALES
+        # Expand each region-specific locale to also include its language-only locale so that `jlink --include-locales` retains the language-only locale bundles (e.g. `FormatData_fr.class`).
+        # See https://github.com/openjdk/jdk/blob/jdk-25%2B35/src/jdk.jlink/share/classes/jdk/tools/jlink/internal/plugins/IncludeLocalesPlugin.java.
+        for locale in (tag, tag.split("-", maxsplit=1)[0])
+    },
 )

tests/resources/PrintLocaleNumberFormats.java

@@ -0,0 +1,19 @@
+package resources;
+
+import java.text.DecimalFormatSymbols;
+import java.util.Locale;
+
+public class PrintLocaleNumberFormats {
+
+    public static void main(String[] args) {
+        for (final Locale locale: DecimalFormatSymbols.getAvailableLocales()) {
+            final DecimalFormatSymbols dfs = DecimalFormatSymbols.getInstance(locale);
+            System.out.println(
+                locale.toLanguageTag()
+                    + "\t" + ((int) dfs.getDecimalSeparator())
+                    + "\t" + ((int) dfs.getGroupingSeparator())
+            );
+        }
+    }
+
+}

tests/test_jdk4py.py

@@ -1,9 +1,11 @@
 import re
+from dataclasses import dataclass
 from pathlib import Path
 from subprocess import run
+from typing import get_args
 
 from jdk4py import JAVA, JAVA_VERSION
-from jdk4py._included_locales import INCLUDED_LOCALES
+from jdk4py._included_locales import INCLUDED_LOCALES, _RegionSpecificLocale
 
 _TEST_RESOURCES_DIRECTORY = Path(__file__).parent / "resources"
 
@@ -43,4 +45,75 @@ def test_included_locales() -> None:
         locale.replace("_", "-")
         for locale in completed_process.stdout.strip().splitlines()
     }
-    assert locales.issuperset(INCLUDED_LOCALES)
+    assert locales >= INCLUDED_LOCALES
+
+
+@dataclass(frozen=True)
+class _NumberFormattingSeparators:
+    decimal_separator: str
+    grouping_separator: str
+
+
+_COMMA = ","
+_DOT = "."
+_NARROW_NO_BREAK_SPACE = chr(0x202F)
+_NO_BREAK_SPACE = chr(0x00A0)
+
+
+def _get_number_formatting_separators(
+    locale: _RegionSpecificLocale, /
+) -> _NumberFormattingSeparators:
+    match locale:
+        case "bn-IN" | "en-GB" | "en-US" | "es-MX" | "ja-JP" | "zh-CN":
+            return _NumberFormattingSeparators(
+                decimal_separator=_DOT, grouping_separator=_COMMA
+            )
+        case "da-DK" | "de-DE" | "es-ES" | "it-IT" | "pt-BR":
+            return _NumberFormattingSeparators(
+                decimal_separator=_COMMA, grouping_separator=_DOT
+            )
+        case "fr-FR":
+            return _NumberFormattingSeparators(
+                decimal_separator=_COMMA,
+                grouping_separator=_NARROW_NO_BREAK_SPACE,
+            )
+        case "ru-RU":
+            return _NumberFormattingSeparators(
+                decimal_separator=_COMMA, grouping_separator=_NO_BREAK_SPACE
+            )
+
+
+def test_locale_data_inclusion() -> None:
+    """
+    Check that each included locale uses its expected number formatting separators.
+
+    ``jlink --include-locales=fr-FR`` does not retain ``FormatData_fr.class``,
+    so ``DecimalFormatSymbols.getInstance(Locale.forLanguageTag("fr-FR"))``
+    silently returns ROOT symbols.
+
+    ``INCLUDED_LOCALES`` must therefore list each region-specific locale alongside its language-only parent.
+    """
+    expected = {
+        locale: _get_number_formatting_separators(locale)
+        for locale in get_args(_RegionSpecificLocale)
+    }
+
+    completed_process = run(  # noqa: S603
+        [JAVA, "-jar", _TEST_RESOURCES_DIRECTORY / "PrintLocaleNumberFormats.jar"],
+        capture_output=True,
+        check=True,
+        text=True,
+    )
+    actual = {
+        locale: _NumberFormattingSeparators(
+            decimal_separator=chr(int(decimal_cp)),
+            grouping_separator=chr(int(grouping_cp)),
+        )
+        for line in completed_process.stdout.strip().splitlines()
+        for locale, decimal_cp, grouping_cp in [line.split("\t")]
+        if locale in expected
+    }
+
+    assert actual == expected, (
+        "CLDR data not loaded for some locales (likely a missing language-only bundle)"
+    )