merge revision(s) 49096,51353,53168,53169: [Backport #11834]

	test_m17n.rb: split test_scrub

	* test/ruby/test_m17n.rb (TestM17N#test_scrub): split into some
  tests.
	* include/ruby/ruby.h: add raw FL macros, which assume always the
	  argument object is not a special constant.

	* internal.h (STR_EMBED_P, STR_SHARED_P): valid only for T_STRING.

	* string.c: deal with taint flags directly across String instances.

	* transcode.c (rb_econv_substr_append, econv_primitive_convert):
	  the result should be infected by the original string.

	* string.c (rb_str_scrub): the result should be infected by the
	  original string.


git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_1@54260 b2dd03c8-39d4-4d8f-98ff-823fe69b080e

2 files changed, 32 insertions, 3 deletions

diff --git a/test/ruby/test_econv.rb b/test/ruby/test_econv.rb
index bbdb14a50b..b2425ee7ea 100644
--- a/test/ruby/test_econv.rb
+++ b/test/ruby/test_econv.rb
@@ -684,6 +684,7 @@ class TestEncodingConverter < Test::Unit::TestCase
     ec = Encoding::Converter.new("utf-8", "euc-jp")
     assert_raise(Encoding::InvalidByteSequenceError) { ec.convert("a\x80") }
     assert_raise(Encoding::UndefinedConversionError) { ec.convert("\ufffd") }
+    assert_predicate(ec.convert("abc".taint), :tainted?)
     ret = ec.primitive_convert(nil, "", nil, nil)
     assert_equal(:finished, ret)
     assert_raise(ArgumentError) { ec.convert("a") }
diff --git a/test/ruby/test_m17n.rb b/test/ruby/test_m17n.rb
index d520b1c9d9..a6e2d3c92a 100644
--- a/test/ruby/test_m17n.rb
+++ b/test/ruby/test_m17n.rb
@@ -1519,20 +1519,32 @@ class TestM17N < Test::Unit::TestCase
     assert_equal(a("\xE3\x81\x82"), s.b)
     assert_equal(Encoding::ASCII_8BIT, s.b.encoding)
     s.taint
-    assert_equal(true, s.b.tainted?)
+    assert_predicate(s.b, :tainted?)
     s = "abc".b
-    assert_equal(true, s.b.ascii_only?)
+    assert_predicate(s.b, :ascii_only?)
   end
 
-  def test_scrub
+  def test_scrub_valid_string
+    str = "foo"
+    assert_equal(str, str.scrub)
+    assert_not_same(str, str.scrub)
+    assert_predicate(str.dup.taint.scrub, :tainted?)
     str = "\u3042\u3044"
+    assert_equal(str, str.scrub)
     assert_not_same(str, str.scrub)
+    assert_predicate(str.dup.taint.scrub, :tainted?)
     str.force_encoding(Encoding::ISO_2022_JP) # dummy encoding
+    assert_equal(str, str.scrub)
     assert_not_same(str, str.scrub)
     assert_nothing_raised(ArgumentError) {str.scrub(nil)}
+    assert_predicate(str.dup.taint.scrub, :tainted?)
+  end
 
+  def test_scrub_replace_default
     assert_equal("\uFFFD\uFFFD\uFFFD", u("\x80\x80\x80").scrub)
     assert_equal("\uFFFDA", u("\xF4\x80\x80A").scrub)
+    assert_predicate(u("\x80\x80\x80").taint.scrub, :tainted?)
+    assert_predicate(u("\xF4\x80\x80A").taint.scrub, :tainted?)
 
     # examples in Unicode 6.1.0 D93b
     assert_equal("\x41\uFFFD\uFFFD\x41\uFFFD\x41",
@@ -1543,14 +1555,28 @@ class TestM17N < Test::Unit::TestCase
                  u("\x61\xF1\x80\x80\xE1\x80\xC2\x62\x80\x63\x80\xBF\x64").scrub)
     assert_equal("abcdefghijklmnopqrstuvwxyz\u0061\uFFFD\uFFFD\uFFFD\u0062\uFFFD\u0063\uFFFD\uFFFD\u0064",
                  u("abcdefghijklmnopqrstuvwxyz\x61\xF1\x80\x80\xE1\x80\xC2\x62\x80\x63\x80\xBF\x64").scrub)
+  end
 
+  def test_scrub_replace_argument
+    assert_equal("foo", u("foo").scrub("\u3013"))
+    assert_predicate(u("foo").taint.scrub("\u3013"), :tainted?)
+    assert_not_predicate(u("foo").scrub("\u3013".taint), :tainted?)
+    assert_equal("\u3042\u3044", u("\xE3\x81\x82\xE3\x81\x84").scrub("\u3013"))
+    assert_predicate(u("\xE3\x81\x82\xE3\x81\x84").taint.scrub("\u3013"), :tainted?)
+    assert_not_predicate(u("\xE3\x81\x82\xE3\x81\x84").scrub("\u3013".taint), :tainted?)
     assert_equal("\u3042\u3013", u("\xE3\x81\x82\xE3\x81").scrub("\u3013"))
+    assert_predicate(u("\xE3\x81\x82\xE3\x81").taint.scrub("\u3013"), :tainted?)
+    assert_predicate(u("\xE3\x81\x82\xE3\x81").scrub("\u3013".taint), :tainted?)
     assert_raise(Encoding::CompatibilityError){ u("\xE3\x81\x82\xE3\x81").scrub(e("\xA4\xA2")) }
     assert_raise(TypeError){ u("\xE3\x81\x82\xE3\x81").scrub(1) }
     assert_raise(ArgumentError){ u("\xE3\x81\x82\xE3\x81\x82\xE3\x81").scrub(u("\x81")) }
     assert_equal(e("\xA4\xA2\xA2\xAE"), e("\xA4\xA2\xA4").scrub(e("\xA2\xAE")))
+  end
 
+  def test_scrub_replace_block
     assert_equal("\u3042<e381>", u("\xE3\x81\x82\xE3\x81").scrub{|x|'<'+x.unpack('H*')[0]+'>'})
+    assert_predicate(u("\xE3\x81\x82\xE3\x81").taint.scrub{|x|'<'+x.unpack('H*')[0]+'>'}, :tainted?)
+    assert_predicate(u("\xE3\x81\x82\xE3\x81").scrub{|x|('<'+x.unpack('H*')[0]+'>').taint}, :tainted?)
     assert_raise(Encoding::CompatibilityError){ u("\xE3\x81\x82\xE3\x81").scrub{e("\xA4\xA2")} }
     assert_raise(TypeError){ u("\xE3\x81\x82\xE3\x81").scrub{1} }
     assert_raise(ArgumentError){ u("\xE3\x81\x82\xE3\x81\x82\xE3\x81").scrub{u("\x81")} }
@@ -1558,7 +1584,9 @@ class TestM17N < Test::Unit::TestCase
 
     assert_equal(u("\x81"), u("a\x81").scrub {|c| break c})
     assert_raise(ArgumentError) {u("a\x81").scrub {|c| c}}
+  end
 
+  def test_scrub_widechar
     assert_equal("\uFFFD\u3042".encode("UTF-16BE"),
                  "\xD8\x00\x30\x42".force_encoding(Encoding::UTF_16BE).
                  scrub)