* string.c: Raise ArgumentError when invalid string is detected in

case mapping methods.
* enc/unicode.c: Check for invalid string and signal with negative
  length value.
* test/ruby/enc/test_case_mapping.rb: Add tests for above.
* test/ruby/test_m17n_comb.rb: Add a message to clarify test failure.


git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55253 b2dd03c8-39d4-4d8f-98ff-823fe69b080e

ChangeLog

@@ -1,3 +1,15 @@
+Thu Jun  2 10:24:48 2016  Martin Duerst  <duerst@it.aoyama.ac.jp>
+
+	* string.c: Raise ArgumentError when invalid string is detected in
+	  case mapping methods.
+
+	* enc/unicode.c: Check for invalid string and signal with negative
+	  length value.
+
+	* test/ruby/enc/test_case_mapping.rb: Add tests for above.
+
+	* test/ruby/test_m17n_comb.rb: Add a message to clarify test failure.
+
 Wed Jun  1 21:41:05 2016  Kazuki Yamaguchi  <k@rhe.jp>
 
 	* ext/openssl/extconf.rb: Check existence of ASN1_TIME_adj(). The old

enc/unicode.c

@@ -669,14 +669,20 @@ onigenc_unicode_case_map(OnigCaseFoldType* flagP,
     OnigCodePoint code;
     OnigUChar *to_start = to;
     OnigCaseFoldType flags = *flagP;
+    int codepoint_length;
+
     to_end -= CASE_MAPPING_SLACK;
     /* copy flags ONIGENC_CASE_UPCASE     and ONIGENC_CASE_DOWNCASE over to
      *            ONIGENC_CASE_UP_SPECIAL and ONIGENC_CASE_DOWN_SPECIAL */
     flags |= (flags&(ONIGENC_CASE_UPCASE|ONIGENC_CASE_DOWNCASE))<<ONIGENC_CASE_SPECIAL_OFFSET;
 
     while (*pp<end && to<=to_end) {
+	codepoint_length = ONIGENC_PRECISE_MBC_ENC_LEN(enc, *pp, end);
+	if (codepoint_length < 0)
+	    return codepoint_length; /* encoding invalid */
 	code = ONIGENC_MBC_TO_CODE(enc, *pp, end);
-	*pp += enclen(enc, *pp, end);
+	*pp += codepoint_length;
+
 	if (code<='z') { /* ASCII comes first */
 	    if (code>='a' && code<='z') {
 	        if (flags&ONIGENC_CASE_UPCASE) {

string.c

@@ -5785,6 +5785,7 @@ rb_str_casemap(VALUE source, OnigCaseFoldType *flags, rb_encoding *enc)
     mapping_buffer pre_buffer, /* only next pointer used */
 		  *current_buffer = &pre_buffer;
     int buffer_count = 0;
+    int buffer_length_or_invalid;
 
     if (RSTRING_LEN(source) == 0) return rb_str_dup(source);
 
@@ -5799,12 +5800,23 @@ rb_str_casemap(VALUE source, OnigCaseFoldType *flags, rb_encoding *enc)
 	current_buffer = current_buffer->next;
 	current_buffer->next = NULL;
 	current_buffer->capa = capa;
-	target_length  += current_buffer->used
-			= onigenc_unicode_case_map(flags,
-				(const OnigUChar**)&source_current, source_end,
-				current_buffer->space,
-				current_buffer->space+current_buffer->capa,
-				enc);
+	buffer_length_or_invalid = onigenc_unicode_case_map(flags,
+				   (const OnigUChar**)&source_current, source_end,
+				   current_buffer->space,
+				   current_buffer->space+current_buffer->capa,
+				   enc);
+	if (buffer_length_or_invalid < 0) {
+	    mapping_buffer *previous_buffer;
+
+	    current_buffer = pre_buffer.next;
+	    while (current_buffer) {
+		previous_buffer = current_buffer;
+		current_buffer  = current_buffer->next;
+		xfree(previous_buffer);
+	    }
+	    rb_raise(rb_eArgError, "input string invalid");
+	}
+	target_length  += current_buffer->used = buffer_length_or_invalid;
     }
 /* fprintf(stderr, "Buffer count is %d\n", buffer_count); *//* for tuning */
 
@@ -5819,7 +5831,7 @@ rb_str_casemap(VALUE source, OnigCaseFoldType *flags, rb_encoding *enc)
 	    memcpy(target_current, current_buffer->space, current_buffer->used);
 	    target_current += current_buffer->used;
 	    previous_buffer = current_buffer;
-	    current_buffer=current_buffer->next;
+	    current_buffer  = current_buffer->next;
 	    xfree(previous_buffer);
 	}
     }

test/ruby/enc/test_case_mapping.rb

@@ -60,6 +60,13 @@ class TestCaseMappingPreliminary < Test::Unit::TestCase
     check_swapcase_properties   'yUKIHIRO matsumoto (MAtz)', 'Yukihiro MATSUMOTO (maTZ)', :lithuanian
   end
 
+  def test_invalid
+    assert_raise(ArgumentError, "Should not be possible to upcase invalid string.") { "\xEB".force_encoding('UTF-8').upcase }
+    assert_raise(ArgumentError, "Should not be possible to downcase invalid string.") { "\xEB".force_encoding('UTF-8').downcase }
+    assert_raise(ArgumentError, "Should not be possible to capitalize invalid string.") { "\xEB".force_encoding('UTF-8').capitalize }
+    assert_raise(ArgumentError, "Should not be possible to swapcase invalid string.") { "\xEB".force_encoding('UTF-8').swapcase }
+  end
+
   def test_general
     check_downcase_properties   'résumé dürst ĭñŧėřŋãţijňőńæłĩżàťïōņ', 'RÉSUMÉ DÜRST ĬÑŦĖŘŊÃŢIJŇŐŃÆŁĨŻÀŤÏŌŅ', :lithuanian
     check_upcase_properties     'RÉSUMÉ DÜRST ĬÑŦĖŘŊÃŢIJŇŐŃÆŁĨŻÀŤÏŌŅ', 'résumé dürst ĭñŧėřŋãţijňőńæłĩżàťïōņ', :lithuanian

test/ruby/test_m17n_comb.rb

@@ -808,7 +808,7 @@ class TestM17NComb < Test::Unit::TestCase
   def test_str_downcase
     STRINGS.each {|s|
       if !s.valid_encoding?
-        assert_raise(ArgumentError) { s.downcase }
+        assert_raise(ArgumentError, "Offending string: #{s.inspect}, encoding: #{s.encoding}") { s.downcase }
         next
       end
       t = s.downcase