[Docutils-checkins] SF.net SVN: docutils:[9074] trunk/docutils

[<mi...@us...>]

Revision: 9074
          http://sourceforge.net/p/docutils/code/9074
Author:   milde
Date:     2022-06-15 15:28:35 +0000 (Wed, 15 Jun 2022)
Log Message:
-----------
Ensure locale_encoding is lower case.

Use `locale.getpreferredencoding(False)` since we dropped support
for older Python versions with side-effects.

Modified Paths:
--------------
    trunk/docutils/docutils/io.py
    trunk/docutils/docutils/utils/error_reporting.py
    trunk/docutils/test/test_io.py
    trunk/docutils/test/test_parsers/test_rst/test_directives/test_date.py

Modified: trunk/docutils/docutils/io.py
===================================================================
--- trunk/docutils/docutils/io.py	2022-06-15 15:28:26 UTC (rev 9073)
+++ trunk/docutils/docutils/io.py	2022-06-15 15:28:35 UTC (rev 9074)
@@ -19,16 +19,20 @@
 from docutils import TransformSpec
 
 
-# Guess the locale's encoding.
+# Guess the locale's preferred encoding.
 # If no valid guess can be made, locale_encoding is set to `None`:
+#
+# TODO: check whether this is set correctly with every OS and Python version
+#       or whether front-end tools need to call `locale.setlocale()`
+#       before importing this module
 try:
-    locale_encoding = locale.getlocale()[1] or locale.getdefaultlocale()[1]
+    locale_encoding = locale.getpreferredencoding(do_setlocale=False).lower()
 except ValueError as error:  # OS X may set UTF-8 without language code
     # See https://bugs.python.org/issue18378 fixed in 3.8
     # and https://sourceforge.net/p/docutils/bugs/298/.
     # Drop the special case after requiring Python >= 3.8
     if "unknown locale: UTF-8" in error.args:
-        locale_encoding = "UTF-8"
+        locale_encoding = "utf-8"
     else:
         locale_encoding = None
 except:  # noqa  any other problems determining the locale -> use None

Modified: trunk/docutils/docutils/utils/error_reporting.py
===================================================================
--- trunk/docutils/docutils/utils/error_reporting.py	2022-06-15 15:28:26 UTC (rev 9073)
+++ trunk/docutils/docutils/utils/error_reporting.py	2022-06-15 15:28:35 UTC (rev 9074)
@@ -45,41 +45,17 @@
 common exceptions.
 """
 
-import codecs
 import sys
 import warnings
 
+from docutils.io import locale_encoding
+
 warnings.warn('The `docutils.utils.error_reporting` module is deprecated '
               'and will be removed in Docutils 0.21 or later.\n'
               'Details with help("docutils.utils.error_reporting").',
               DeprecationWarning, stacklevel=2)
 
-# Guess the locale's encoding.
-# If no valid guess can be made, locale_encoding is set to `None`:
-try:
-    import locale # module missing in Jython
-except ImportError:
-    locale_encoding = None
-else:
-    try:
-        locale_encoding = locale.getlocale()[1] or locale.getdefaultlocale()[1]
-        # locale.getpreferredencoding([do_setlocale=True|False])
-        # has side-effects | might return a wrong guess.
-    except ValueError as error: # OS X may set UTF-8 without language code
-        # see http://bugs.python.org/issue18378
-        # and https://sourceforge.net/p/docutils/bugs/298/
-        if "unknown locale: UTF-8" in error.args:
-            locale_encoding = "UTF-8"
-        else:
-            locale_encoding = None
-    except: # noqa  any other problems determining the locale -> use None
-        locale_encoding = None
-    try:
-        codecs.lookup(locale_encoding or '') # None -> ''
-    except LookupError:
-        locale_encoding = None
 
-
 if sys.version_info >= (3, 0):
     unicode = str  # noqa
 

Modified: trunk/docutils/test/test_io.py
===================================================================
--- trunk/docutils/test/test_io.py	2022-06-15 15:28:26 UTC (rev 9073)
+++ trunk/docutils/test/test_io.py	2022-06-15 15:28:35 UTC (rev 9074)
@@ -132,7 +132,7 @@
     def test_heuristics_no_utf8(self):
         # if no encoding is given and decoding with utf-8 fails,
         # use either the locale encoding (if specified) or latin-1:
-        if io.locale_encoding.lower() not in ('utf-8', 'utf8'):
+        if io.locale_encoding not in ('utf-8', 'utf8'):
             # in Py3k, the locale encoding is used without --input-encoding
             # skipping the heuristic unless decoding fails.
             return

Modified: trunk/docutils/test/test_parsers/test_rst/test_directives/test_date.py
===================================================================
--- trunk/docutils/test/test_parsers/test_rst/test_directives/test_date.py	2022-06-15 15:28:26 UTC (rev 9073)
+++ trunk/docutils/test/test_parsers/test_rst/test_directives/test_date.py	2022-06-15 15:28:35 UTC (rev 9074)
@@ -63,7 +63,7 @@
 
 # some locales return non-ASCII characters for names of days or months
 # ensure the directive handles them correctly
-if locale_encoding.lower() in ('utf8', 'utf-8', 'latin-1', 'iso8859-1'):
+if locale_encoding in ('utf-8', 'utf8', 'latin-1', 'iso-8859-1'):
     totest['decode date'] = [
     ["""\
 .. |date| date:: täglich

This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.