nobu	2015-02-23 15:26:42 +0900 (Mon, 23 Feb 2015)

  New Revision: 49695

  http://svn.ruby-lang.org/cgi-bin/viewvc.cgi?view=revision&revision=49695

  Log:
    string.c: proper exception
    
    * string.c (rb_str_split_m): raise ArgumentError at broken string
      not RegexpError, as Regexp is not involved in.
      [ruby-core:68229] [Bug #10886]

  Modified files:
    trunk/ChangeLog
    trunk/string.c
    trunk/test/ruby/test_string.rb
Index: ChangeLog
===================================================================
--- ChangeLog	(revision 49694)
+++ ChangeLog	(revision 49695)
@@ -1,3 +1,9 @@ https://github.com/ruby/ruby/blob/trunk/ChangeLog#L1
+Mon Feb 23 15:26:39 2015  Nobuyoshi Nakada  <nobu@r...>
+
+	* string.c (rb_str_split_m): raise ArgumentError at broken string
+	  not RegexpError, as Regexp is not involved in.
+	  [ruby-core:68229] [Bug #10886]
+
 Mon Feb 23 07:25:29 2015  Benoit Daloze  <eregontp@g...>
 
 	* time.c: Zone encoding should be US-ASCII if all 7-bits. Fix r46907.
Index: string.c
===================================================================
--- string.c	(revision 49694)
+++ string.c	(revision 49695)
@@ -6387,10 +6387,11 @@ rb_str_split_m(int argc, VALUE *argv, VA https://github.com/ruby/ruby/blob/trunk/string.c#L6387
     }
     else {
       fs_set:
-	spat = get_pat_quoted(spat, 1);
+	spat = get_pat_quoted(spat, 0);
 	if (BUILTIN_TYPE(spat) == T_STRING) {
 	    rb_encoding *enc2 = STR_ENC_GET(spat);
 
+	    mustnot_broken(spat);
 	    split_type = string;
 	    if (RSTRING_LEN(spat) == 0) {
 		/* Special case - split into chars */
@@ -6485,7 +6486,6 @@ rb_str_split_m(int argc, VALUE *argv, VA https://github.com/ruby/ruby/blob/trunk/string.c#L6486
 	long slen = RSTRING_LEN(spat);
 
 	mustnot_broken(str);
-	mustnot_broken(spat);
 	enc = rb_enc_check(str, spat);
 	while (ptr < eptr &&
 	       (end = rb_memsearch(sptr, slen, ptr, eptr - ptr, enc)) >= 0) {
Index: test/ruby/test_string.rb
===================================================================
--- test/ruby/test_string.rb	(revision 49694)
+++ test/ruby/test_string.rb	(revision 49695)
@@ -1345,14 +1345,18 @@ class TestString < Test::Unit::TestCase https://github.com/ruby/ruby/blob/trunk/test/ruby/test_string.rb#L1345
     assert_equal([], "".split(//, 1))
 
     assert_equal("[2, 3]", [1,2,3].slice!(1,10000).inspect, "moved from btest/knownbug")
+  end
 
+  def test_split_encoding
     bug6206 = '[ruby-dev:45441]'
     Encoding.list.each do |enc|
       next unless enc.ascii_compatible?
       s = S("a:".force_encoding(enc))
       assert_equal([enc]*2, s.split(":", 2).map(&:encoding), bug6206)
     end
+  end
 
+  def test_split_wchar
     bug8642 = '[ruby-core:56036] [Bug #8642]'
     [
      Encoding::UTF_16BE, Encoding::UTF_16LE,
@@ -1365,6 +1369,20 @@ class TestString < Test::Unit::TestCase https://github.com/ruby/ruby/blob/trunk/test/ruby/test_string.rb#L1369
     end
   end
 
+  def test_split_invalid_sequence
+    bug10886 = '[ruby-core:68229] [Bug #10886]'
+    broken = S("\xa1".force_encoding("utf-8"))
+    assert_raise(ArgumentError, bug10886) {
+      S("a,b").split(broken)
+    }
+  end
+
+  def test_split_invalid_argument
+    assert_raise(TypeError) {
+      S("a,b").split(BasicObject.new)
+    }
+  end
+
   def test_squeeze
     assert_equal(S("abc"), S("aaabbbbccc").squeeze)
     assert_equal(S("aa bb cc"), S("aa   bb      cc").squeeze(S(" ")))

--
ML: ruby-changes@q...
Info: http://www.atdot.net/~ko1/quickml/