Handle language codes according to BCP 47

(Best Current Praxis: Tags for Identifying Languages).

Normalize language tags: downcase, - to _, optional subtags.

E.g. Given "de-AT-1901", look for language modules
     "de_at_1901", "de_at", "de_1901", and "de"
     
TODO: warn (instead of fail), if no module found.

git-svn-id: https://svn.code.sf.net/p/docutils/code/trunk/docutils@6423 929543f6-e4f2-0310-98a6-ba3bd3dd1d04

1 files changed, 14 insertions, 5 deletions

diff --git a/docutils/languages/__init__.py b/docutils/languages/__init__.py
index 7d6521fc1..c61871b90 100644
--- a/docutils/languages/__init__.py
+++ b/docutils/languages/__init__.py
@@ -11,11 +11,20 @@ This package contains modules for language-dependent features of Docutils.
 
 __docformat__ = 'reStructuredText'
 
+from docutils.utils import normalize_language_tag
+
 _languages = {}
 
 def get_language(language_code):
-    if language_code in _languages:
-        return _languages[language_code]
-    module = __import__(language_code, globals(), locals())
-    _languages[language_code] = module
-    return module
+    for tag in normalize_language_tag(language_code):
+        if tag in _languages:
+            return _languages[tag]
+        try:
+            module = __import__(tag, globals(), locals())
+        except ImportError:
+            continue
+        _languages[tag] = module
+        return module
+    # TODO: use Docutils reporter, warn instead of fail
+    # warn ('language "%s" not supported' % tag)
+    raise