Add check for zero-width space (issue #146)

WeblateOrg · Nov 13, 2012 · 9c16b4b · 9c16b4b
1 parent afc7f5c
commit 9c16b4b
Show file tree

Hide file tree

Showing 3 changed files with 21 additions and 0 deletions.
diff --git a/docs/usage.rst b/docs/usage.rst
@@ -311,6 +311,14 @@ Mismatched BBcode
 BBcode in translation does not match source. The method for detecting BBcode is
 currently quite simple.
 
+.. _check-zero-width-space:
+
+Zero-width space
+~~~~~~~~~~~~~~~~
+
+Translation contains extra zero-width space (<U+200B>) character. This
+character is usually inserted by mistake.
+
 .. _check-optional-plural:
 
 Source checks

diff --git a/weblate/settings_example.py b/weblate/settings_example.py
@@ -338,6 +338,7 @@
     'weblate.trans.checks.DirectionCheck',
     'weblate.trans.checks.NewlineCountingCheck',
     'weblate.trans.checks.BBCodeCheck',
+    'weblate.trans.checks.ZeroWidthSpaceCheck',
     'weblate.trans.checks.OptionalPluralCheck',
     'weblate.trans.checks.EllipsisCheck',
 )

diff --git a/weblate/trans/checks.py b/weblate/trans/checks.py
@@ -154,6 +154,7 @@
     'weblate.trans.checks.DirectionCheck',
     'weblate.trans.checks.NewlineCountingCheck',
     'weblate.trans.checks.BBCodeCheck',
+    'weblate.trans.checks.ZeroWidthSpaceCheck',
     'weblate.trans.checks.OptionalPluralCheck',
     'weblate.trans.checks.EllipsisCheck',
 )
@@ -601,6 +602,17 @@ def check_single(self, source, target, flags, language, unit):
         tgt_tags = set([x[0] for x in tgt_match])
         return (src_tags != tgt_tags)
 
+class ZeroWidthSpaceCheck(TargetCheck):
+    '''
+    Check for zero width space char (<U+200B>).
+    '''
+    check_id = 'zero-width-space'
+    name = _('Zero-width space')
+    description = _('Translation contains extra zero-width space character')
+
+    def check_single(self, source, target, flags, language, unit):
+        return u'\u200b' in target and not u'\u200b' in source
+
 class OptionalPluralCheck(SourceCheck):
     '''
     Check for not used plural form.