Parameterize should accept malformed utf8 characters [#4323 state:res…

…olved] Signed-off-by: Jeremy Kemper <jeremy@bitsweat.net>
rails · Apr 4, 2010 · 642d5d2 · 642d5d2
1 parent e617af1
commit 642d5d2
Show file tree

Hide file tree

Showing 2 changed files with 13 additions and 5 deletions.
diff --git a/activesupport/lib/active_support/inflector.rb b/activesupport/lib/active_support/inflector.rb
@@ -1,6 +1,7 @@
 # encoding: utf-8
 require 'singleton'
 require 'iconv'
+require 'kconv'
 
 module ActiveSupport
   # The Inflector transforms words from singular to plural, class names to table names, modularized class names to ones without,
@@ -257,6 +258,8 @@ def demodulize(class_name_in_module)
     #   <%= link_to(@person.name, person_path(@person)) %>
     #   # => <a href="/person/1-donald-e-knuth">Donald E. Knuth</a>
     def parameterize(string, sep = '-')
+      # remove malformed utf8 characters
+      string = string.toutf8 unless string.is_utf8?
       # replace accented chars with ther ascii equivalents
       parameterized_string = transliterate(string)
       # Turn unwanted chars into the seperator

diff --git a/activesupport/test/inflector_test_cases.rb b/activesupport/test/inflector_test_cases.rb
@@ -152,7 +152,8 @@ module InflectorTestCases
     "Trailing bad characters!@#"          => "trailing-bad-characters",
     "!@#Leading bad characters"           => "leading-bad-characters",
     "Squeeze   separators"                => "squeeze-separators",
-    "Test with + sign"                    => "test-with-sign"
+    "Test with + sign"                    => "test-with-sign",
+    "Test with malformed utf8 \251"       => "test-with-malformed-utf8"
   }
 
   StringToParameterizeWithNoSeparator = {
@@ -161,7 +162,8 @@ module InflectorTestCases
     "Trailing bad characters!@#"          => "trailingbadcharacters",
     "!@#Leading bad characters"           => "leadingbadcharacters",
     "Squeeze   separators"                => "squeezeseparators",
-    "Test with + sign"                    => "testwithsign"
+    "Test with + sign"                    => "testwithsign",
+    "Test with malformed utf8 \251"       => "testwithmalformedutf8"
   }
 
   StringToParameterizeWithUnderscore = {
@@ -170,19 +172,22 @@ module InflectorTestCases
     "Trailing bad characters!@#"          => "trailing_bad_characters",
     "!@#Leading bad characters"           => "leading_bad_characters",
     "Squeeze   separators"                => "squeeze_separators",
-    "Test with + sign"                    => "test_with_sign"
+    "Test with + sign"                    => "test_with_sign",
+    "Test with malformed utf8 \251"       => "test_with_malformed_utf8"
   }
 
   # Ruby 1.9 doesn't do Unicode normalization yet.
   if RUBY_VERSION >= '1.9'
     StringToParameterizedAndNormalized = {
       "Malmö"                               => "malm",
-      "Garçons"                             => "gar-ons"
+      "Garçons"                             => "gar-ons",
+      "Ops \251"                            => "ops"
     }
   else
     StringToParameterizedAndNormalized = {
       "Malmö"                               => "malmo",
-      "Garçons"                             => "garcons"
+      "Garçons"                             => "garcons",
+      "Ops \251"                            => "ops"
     }
   end