gh-64243: Implement locale.getlocale fall back in gettext.find

Doc/library/gettext.rst

@@ -130,9 +130,10 @@ install themselves in the built-in namespace as the function :func:`!_`.
    strings, where each string is a language code.
 
    If *localedir* is not given, then the default system locale directory is used.
-   [#]_  If *languages* is not given, then the following environment variables are
-   searched: :envvar:`LANGUAGE`, :envvar:`LC_ALL`, :envvar:`LC_MESSAGES`, and
-   :envvar:`LANG`.  The first one returning a non-empty value is used for the
+   [#]_  If *languages* is not given, then the environment variable :envvar:`LANGUAGE`
+   is searched, it falls back to the current locale or to the environment
+   variables :envvar:`LC_ALL`, :envvar:`LC_MESSAGES`, and
+   :envvar:`LANG` where the first one returning a non-empty value is used for the
    *languages* variable. The environment variables should contain a colon separated
    list of languages, which will be split on the colon to produce the expected list
    of language code strings.
@@ -147,6 +148,9 @@ install themselves in the built-in namespace as the function :func:`!_`.
    of all file names, in the order in which they appear in the languages list or
    the environment variables.
 
+   .. versionchanged:: next
+      :func:`locale.setlocale` is used to generate *languages* if *languages* is
+      not provided.
 
 .. function:: translation(domain, localedir=None, languages=None, class_=None, fallback=False)
 

Lib/gettext.py

@@ -46,6 +46,7 @@
 import operator
 import os
 import sys
+import locale
 
 
 __all__ = ['NullTranslations', 'GNUTranslations', 'Catalog',
@@ -229,7 +230,6 @@ def func(n):
 
 
 def _expand_lang(loc):
-    import locale
     loc = locale.normalize(loc)
     COMPONENT_CODESET   = 1 << 0
     COMPONENT_TERRITORY = 1 << 1
@@ -491,11 +491,16 @@ def find(domain, localedir=None, languages=None, all=False):
         localedir = _default_localedir
     if languages is None:
         languages = []
-        for envar in ('LANGUAGE', 'LC_ALL', 'LC_MESSAGES', 'LANG'):
-            val = os.environ.get(envar)
-            if val:
-                languages = val.split(':')
-                break
+        if val := os.environ.get('LANGUAGE'):
+            languages = val.split(':')
+        elif (loc := locale.getlocale()) != (None, None):
+            languages = [".".join(filter(None, loc))]
+        else:
+            for envar in ('LC_ALL', 'LC_MESSAGES', 'LANG'):
+                val = os.environ.get(envar)
+                if val:
+                    languages = val.split(':')
+                    break
         if 'C' not in languages:
             languages.append('C')
     # now normalize and expand the languages

Lib/test/test_gettext.py

@@ -1,3 +1,4 @@
+import locale
 import os
 import base64
 import gettext
@@ -875,17 +876,65 @@ def create_mo_file(self, lang):
             f.write(GNU_MO_DATA)
         return mo_file
 
-    def test_find_with_env_vars(self):
-        # test that find correctly finds the environment variables
-        # when languages are not supplied
-        mo_file = self.create_mo_file("ga_IE")
+    def _for_all_vars(self, mo_file, locale, expected=True):
         for var in ('LANGUAGE', 'LC_ALL', 'LC_MESSAGES', 'LANG'):
-            self.env.set(var, 'ga_IE')
+            self.env.set(var, locale)
             result = gettext.find("mofile",
                                   localedir=os.path.join(self.tempdir, "locale"))
-            self.assertEqual(result, mo_file)
+            if expected:
+                self.assertEqual(mo_file, result)
+            else:
+                self.assertIsNone(result)
             self.env.unset(var)
 
+    @unittest.mock.patch("locale.getlocale", return_value=(None, None))
+    def test_find_with_env_vars(self, patch_getlocale):
+        # test that find correctly finds the environment variables
+        # when languages are not supplied
+        mo_file = self.create_mo_file("ca_ES")
+        self._for_all_vars(mo_file, "ca_ES")
+        self._for_all_vars(mo_file, "ca_ES.UTF-8")
+        self._for_all_vars(mo_file, "ca_ES.UTF-8.mo")
+        self._for_all_vars(mo_file, "es_ES:ca_ES:fr_FR")
+        self._for_all_vars(mo_file, "ca_ES@euro")
+        self._for_all_vars(mo_file, "ca_ES.UTF-8@euro")
+        self._for_all_vars(mo_file, "ca_ES@valencia")
+        self._for_all_vars(mo_file, "C", expected=False)
+        self._for_all_vars(mo_file, "C.UTF-8", expected=False)
+
+    @unittest.mock.patch('gettext._expand_lang')
+    def test_encoding_not_ignored(self, patch_expand_lang):
+        self.env.set('LANGUAGE', 'ga_IE.UTF-8')
+        gettext.find("mofile")
+        patch_expand_lang.assert_any_call('ga_IE.UTF-8')
+        self.env.unset('LANGUAGE')
+
+    def test_find_LANGUAGE_priority(self):
+        self.env.set('LANGUAGE', 'ga_IE')
+        self.env.set('LC_ALL', 'C')
+        orig = locale.setlocale(locale.LC_ALL)
+        self.addCleanup(lambda: locale.setlocale(locale.LC_ALL, orig))
+        locale.setlocale(locale.LC_ALL, 'C')
+        mo_file = self.create_mo_file("ga_IE")
+        result = gettext.find("mofile", localedir=os.path.join(self.tempdir, "locale"))
+        self.assertEqual(result, mo_file)
+
+    def test_process_vars_override(self):
+        orig = locale.setlocale(locale.LC_ALL)
+        self.addCleanup(lambda: locale.setlocale(locale.LC_ALL, orig))
+        mo_file = self.create_mo_file("ca_ES")
+        for loc in ("ca_ES", "ca_ES.UTF-8", "ca_ES@euro", "ca_ES@valencia"):
+            try:
+                locale.setlocale(locale.LC_ALL, loc)
+            except locale.Error:
+                self.skipTest('platform does not support locale')
+            result = gettext.find("mofile", localedir=os.path.join(self.tempdir, "locale"))
+            self.assertEqual(mo_file, result)
+        for loc in ("C", "C.UTF-8"):
+            locale.setlocale(locale.LC_ALL, loc)
+            result = gettext.find("mofile", localedir=os.path.join(self.tempdir, "locale"))
+            self.assertIsNone(result)
+
     def test_find_with_languages(self):
         # test that passed languages are used
         self.env.set('LANGUAGE', 'pt_BR')
@@ -934,14 +983,14 @@ def test__all__(self):
 
     @cpython_only
     def test_lazy_import(self):
-        ensure_lazy_imports("gettext", {"re", "warnings", "locale"})
+        ensure_lazy_imports("gettext", {"re", "warnings"})
 
 
 if __name__ == '__main__':
     unittest.main()
 
 
-# For reference, here's the .po file used to created the GNU_MO_DATA above.
+# For reference, here's the .po file used to create the GNU_MO_DATA above.
 #
 # The original version was automatically generated from the sources with
 # pygettext. Later it was manually modified to add plural forms support.

Misc/NEWS.d/next/Library/2025-03-19-17-30-00.gh-issue-64243.fuheq3.rst

@@ -0,0 +1,2 @@
+Implement a fall back to :func:`locale.getlocale` in :func:`gettext.find` if
+*languages* is not provided and :envvar:`LANGUAGE` is not set.