fix(autofixes): use NO-BREAK SPACE with colons in French and Breton (#…

…12549)

* fix(autofixes): use NO-BREAK SPACE with colons in French and Breton

The French style guide recommends a non-breaking space before a colon
character, not a narrow one.

Fixes #11237

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

docs/changes.rst

@@ -15,6 +15,7 @@ Not yet released.
 * :ref:`mt-deepl` now supports specifying translation context.
 * :ref:`mt-aws` now supports :ref:`glossary-mt`.
 * :ref:`autofix` for Devanagari danda now better handles latin script.
+* :ref:`autofix` for French and Breton now uses a non-breaking space before colons instead of a narrow one.
 
 **Bug fixes**
 

weblate/checks/chars.py

@@ -19,12 +19,22 @@
 
     from weblate.trans.models import Unit
 
-FRENCH_PUNCTUATION = {";", ":", "?", "!"}
+FRENCH_PUNCTUATION_NBSP = {":"}
+FRENCH_PUNCTUATION_NNBSP = {";", "?", "!"}
+FRENCH_PUNCTUATION = FRENCH_PUNCTUATION_NBSP.union(FRENCH_PUNCTUATION_NNBSP)
 FRENCH_PUNCTUATION_SPACING = {"Zs", "Ps", "Pe"}
-FRENCH_PUNCTUATION_FIXUP_RE = "([ \u00a0\u2009])([{}])".format(
-    "".join(FRENCH_PUNCTUATION)
+FRENCH_PUNCTUATION_FIXUP_RE_NBSP = "([ \u2009\u202f])([{}])".format(
+    "".join(FRENCH_PUNCTUATION_NBSP)
+)
+FRENCH_PUNCTUATION_FIXUP_RE_NNBSP = "([ \u00a0\u2009])([{}])".format(
+    "".join(FRENCH_PUNCTUATION_NNBSP)
+)
+FRENCH_PUNCTUATION_MISSING_RE_NBSP = "([^\u00a0])([{}])".format(
+    "".join(FRENCH_PUNCTUATION_NBSP)
+)
+FRENCH_PUNCTUATION_MISSING_RE_NNBSP = "([^\u202f])([{}])".format(
+    "".join(FRENCH_PUNCTUATION_NNBSP)
 )
-FRENCH_PUNCTUATION_MISSING_RE = "([^\u202f])([{}])".format("".join(FRENCH_PUNCTUATION))
 MY_QUESTION_MARK = "\u1038\u104b"
 INTERROBANGS = ("?!", "!?", "？！", "！？", "⁈", "⁉")
 
@@ -487,13 +497,23 @@ def get_fixup(self, unit: Unit) -> Iterable[tuple[str, str, str]] | None:
         return [
             # First fix possibly wrong whitespace
             (
-                FRENCH_PUNCTUATION_FIXUP_RE,
+                FRENCH_PUNCTUATION_FIXUP_RE_NBSP,
+                "\u00a0$2",
+                "gu",
+            ),
+            (
+                FRENCH_PUNCTUATION_FIXUP_RE_NNBSP,
                 "\u202f$2",
                 "gu",
             ),
             # Then add missing ones
             (
-                FRENCH_PUNCTUATION_MISSING_RE,
+                FRENCH_PUNCTUATION_MISSING_RE_NBSP,
+                "$1\u00a0$2",
+                "gu",
+            ),
+            (
+                FRENCH_PUNCTUATION_MISSING_RE_NNBSP,
                 "$1\u202f$2",
                 "gu",
             ),

weblate/trans/autofixes/chars.py

@@ -10,7 +10,8 @@
 from django.utils.translation import gettext_lazy
 
 from weblate.checks.chars import (
-    FRENCH_PUNCTUATION_FIXUP_RE,
+    FRENCH_PUNCTUATION_FIXUP_RE_NBSP,
+    FRENCH_PUNCTUATION_FIXUP_RE_NNBSP,
     EndEllipsisCheck,
     PunctuationSpacingCheck,
     ZeroWidthSpaceCheck,
@@ -111,7 +112,10 @@ def fix_single_target(
             and "ignore-punctuation-spacing" not in unit.all_flags
         ):
             # Fix existing
-            new_target = re.sub(FRENCH_PUNCTUATION_FIXUP_RE, "\u202f\\2", target)
+            new_target = re.sub(FRENCH_PUNCTUATION_FIXUP_RE_NBSP, "\u00a0\\2", target)
+            new_target = re.sub(
+                FRENCH_PUNCTUATION_FIXUP_RE_NNBSP, "\u202f\\2", new_target
+            )
             # Do not add missing as that is likely to trigger issues with other content
             # such as URLs or Markdown syntax.
             return new_target, new_target != target

weblate/trans/tests/test_autofix.py

@@ -176,9 +176,9 @@ def test_punctuation_spacing(self) -> None:
         cs_unit = MockUnit(source="Foo:", code="cs")
         self.assertEqual(fix.fix_target(["Bar:"], non_unit), (["Bar:"], False))
         self.assertEqual(
-            fix.fix_target(["Bar\u202f:"], fr_unit), (["Bar\u202f:"], False)
+            fix.fix_target(["Bar\u00a0:"], fr_unit), (["Bar\u00a0:"], False)
         )
-        self.assertEqual(fix.fix_target(["Bar :"], fr_unit), (["Bar\u202f:"], True))
+        self.assertEqual(fix.fix_target(["Bar :"], fr_unit), (["Bar\u00a0:"], True))
         self.assertEqual(fix.fix_target(["Bar:"], fr_unit), (["Bar:"], False))
         self.assertEqual(fix.fix_target(["Bar:"], fr_ca_unit), (["Bar:"], False))
         self.assertEqual(fix.fix_target(["Bar:"], cs_unit), (["Bar:"], False))