Fresh start

10 files changed, 426 insertions, 0 deletions

diff --git a/jni/ruby/test/-ext-/string/test_coderange.rb b/jni/ruby/test/-ext-/string/test_coderange.rb
new file mode 100644
index 0000000..83cebf1
--- /dev/null
+++ b/jni/ruby/test/-ext-/string/test_coderange.rb
@@ -0,0 +1,59 @@
+# coding: ascii-8bit
+require 'test/unit'
+require "-test-/string/string"
+require "rbconfig/sizeof"
+
+class Test_StringCoderange < Test::Unit::TestCase
+  def setup
+    @sizeof_voidp = RbConfig::SIZEOF["void*"]
+    @a8 = Encoding::ASCII_8BIT
+    @a7 = Encoding::US_ASCII
+    @u8 = Encoding::UTF_8
+  end
+
+  def test_ascii8bit
+    enc = @a8
+    str = "a"
+    str.force_encoding(enc)
+    assert_equal :"7bit", Bug::String.new(str).coderange_scan
+
+    str = "a\xBE".force_encoding(enc)
+    assert_equal :valid, Bug::String.new(str).coderange_scan
+  end
+
+  def test_usascii
+    enc = @a7
+    str = "a"
+    str.force_encoding(enc)
+    assert_equal :"7bit", Bug::String.new(str).coderange_scan
+
+    str = "a" * (@sizeof_voidp * 2)
+    str << "\xBE"
+    str.force_encoding(enc)
+    assert_equal :broken, Bug::String.new(str).coderange_scan
+  end
+
+  def test_utf8
+    enc = @u8
+    str = "a"
+    str.force_encoding(enc)
+    assert_equal :"7bit", Bug::String.new(str).coderange_scan
+
+    str = "a" * (@sizeof_voidp * 3)
+    str << "aa\xC2\x80"
+    str.force_encoding(enc)
+    assert_equal :valid, Bug::String.new(str).coderange_scan
+
+    str = "a" * (@sizeof_voidp * 2)
+    str << "\xC2\x80"
+    str << "a" * (@sizeof_voidp * 2)
+    str.force_encoding(enc)
+    assert_equal :valid, Bug::String.new(str).coderange_scan
+
+    str = "a" * (@sizeof_voidp * 2)
+    str << "\xC1\x80"
+    str << "a" * (@sizeof_voidp * 2)
+    str.force_encoding(enc)
+    assert_equal :broken, Bug::String.new(str).coderange_scan
+  end
+end
diff --git a/jni/ruby/test/-ext-/string/test_cstr.rb b/jni/ruby/test/-ext-/string/test_cstr.rb
new file mode 100644
index 0000000..272e090
--- /dev/null
+++ b/jni/ruby/test/-ext-/string/test_cstr.rb
@@ -0,0 +1,119 @@
+require 'test/unit'
+require "-test-/string/string"
+
+class Test_StringCStr < Test::Unit::TestCase
+  Bug4319 = '[ruby-dev:43094]'
+
+  def test_embed
+    s = Bug::String.new("abcdef")
+    s.set_len(3)
+    assert_equal(0, s.cstr_term, Bug4319)
+  end
+
+  def test_long
+    s = Bug::String.new("abcdef")*100000
+    assert_equal(0, s.cstr_term, Bug4319)
+  end
+
+  WCHARS = [Encoding::UTF_16BE, Encoding::UTF_16LE, Encoding::UTF_32BE, Encoding::UTF_32LE]
+
+  def test_wchar_embed
+    WCHARS.each do |enc|
+      s = Bug::String.new("\u{4022}a".encode(enc))
+      assert_nothing_raised(ArgumentError) {s.cstr_term}
+      s.set_len(s.bytesize / 2)
+      assert_equal(1, s.size)
+      assert_equal(0, s.cstr_term)
+    end
+  end
+
+  def test_wchar_long
+    str = "\u{4022}abcdef"
+    n = 100
+    len = str.size * n
+    WCHARS.each do |enc|
+      s = Bug::String.new(str.encode(enc))*n
+      assert_nothing_raised(ArgumentError, enc.name) {s.cstr_term}
+      s.set_len(s.bytesize / 2)
+      assert_equal(len / 2, s.size, enc.name)
+      assert_equal(0, s.cstr_term, enc.name)
+    end
+  end
+
+  def test_wchar_lstrip!
+    assert_wchars_term_char(" a") {|s| s.lstrip!}
+  end
+
+  def test_wchar_rstrip!
+    assert_wchars_term_char("a ") {|s| s.rstrip!}
+  end
+
+  def test_wchar_chop!
+    assert_wchars_term_char("a\n") {|s| s.chop!}
+  end
+
+  def test_wchar_chomp!
+    assert_wchars_term_char("a\n") {|s| s.chomp!}
+  end
+
+  def test_wchar_aset
+    assert_wchars_term_char("a"*30) {|s| s[29,1] = ""}
+  end
+
+  def test_wchar_sub!
+    assert_wchars_term_char("foobar") {|s| s.sub!(/#{"foo".encode(s.encoding)}/, "")}
+  end
+
+  def test_wchar_delete!
+    assert_wchars_term_char("foobar") {|s| s.delete!("ao".encode(s.encoding))}
+  end
+
+  def test_wchar_squeeze!
+    assert_wchars_term_char("foo!") {|s| s.squeeze!}
+  end
+
+  def test_wchar_tr!
+    assert_wchars_term_char("\u{3042}foobar") {|s|
+      enc = s.encoding
+      s.tr!("\u{3042}".encode(enc), "c".encode(enc))
+    }
+  end
+
+  def test_wchar_tr_s!
+    assert_wchars_term_char("\u{3042}foobar") {|s|
+      enc = s.encoding
+      s.tr_s!("\u{3042}".encode(enc), "c".encode(enc))
+    }
+  end
+
+  def test_embedded_from_heap
+    gh821 = "[GH-821]"
+    embedded_string = "abcdefghi"
+    string = embedded_string.gsub("efg", "123")
+    {}[string] = 1
+    non_terminated = "#{string}#{nil}"
+    assert_nil(Bug::String.cstr_term_char(non_terminated), gh821)
+
+    result = {}
+    WCHARS.map do |enc|
+      embedded_string = "ab".encode(enc)
+      string = embedded_string.gsub("b".encode(enc), "1".encode(enc))
+      {}[string] = 1
+      non_terminated = "#{string}#{nil}"
+      c = Bug::String.cstr_term_char(non_terminated)
+      result[enc] = c if c
+    end
+    assert_empty(result, gh821)
+  end
+
+  def assert_wchars_term_char(str)
+    result = {}
+    WCHARS.map do |enc|
+      s = Bug::String.new(str.encode(enc))
+      yield s
+      c = s.cstr_term_char
+      result[enc] = c if c
+    end
+    assert_empty(result)
+  end
+end
diff --git a/jni/ruby/test/-ext-/string/test_ellipsize.rb b/jni/ruby/test/-ext-/string/test_ellipsize.rb
new file mode 100644
index 0000000..2c14c0c
--- /dev/null
+++ b/jni/ruby/test/-ext-/string/test_ellipsize.rb
@@ -0,0 +1,46 @@
+require 'test/unit'
+require "-test-/string/string"
+
+class Test_StringEllipsize < Test::Unit::TestCase
+  def setup
+    @foobar = Bug::String.new("foobar")
+  end
+
+  def assert_equal_with_class(expected, result, *rest)
+    assert_equal(expected.encoding, result.encoding, *rest)
+    assert_equal(expected, result, result.encoding.name)
+    assert_instance_of(Bug::String, result, *rest)
+  end
+
+  def test_longer
+    assert_equal_with_class("", @foobar.ellipsize(0))
+    assert_equal_with_class(".", @foobar.ellipsize(1))
+    assert_equal_with_class("..", @foobar.ellipsize(2))
+    assert_equal_with_class("...", @foobar.ellipsize(3))
+    assert_equal_with_class("f...", @foobar.ellipsize(4))
+    assert_equal_with_class("fo...", @foobar.ellipsize(5))
+  end
+
+  def test_shorter
+    assert_same(@foobar, @foobar.ellipsize(6))
+    assert_same(@foobar, @foobar.ellipsize(7))
+  end
+
+  def test_negative_length
+    assert_raise(IndexError) {@foobar.ellipsize(-1)}
+  end
+
+  def test_nonascii
+    a = "\u3042"
+    Encoding.list.each do |enc|
+      next if enc.dummy?
+      begin
+        s = a.encode(enc)
+        e = "...".encode(enc)
+      rescue
+      else
+        assert_equal_with_class(s*12+e, Bug::String.new(s*20).ellipsize(15))
+      end
+    end
+  end
+end
diff --git a/jni/ruby/test/-ext-/string/test_enc_associate.rb b/jni/ruby/test/-ext-/string/test_enc_associate.rb
new file mode 100644
index 0000000..edddfb4
--- /dev/null
+++ b/jni/ruby/test/-ext-/string/test_enc_associate.rb
@@ -0,0 +1,12 @@
+require 'test/unit'
+require "-test-/string/string"
+
+class Test_StrEncAssociate < Test::Unit::TestCase
+  def test_frozen
+    s = Bug::String.new("abc")
+    s.force_encoding(Encoding::US_ASCII)
+    s.freeze
+    assert_raise(RuntimeError) {s.associate_encoding!(Encoding::US_ASCII)}
+    assert_raise(RuntimeError) {s.associate_encoding!(Encoding::UTF_8)}
+  end
+end
diff --git a/jni/ruby/test/-ext-/string/test_enc_str_buf_cat.rb b/jni/ruby/test/-ext-/string/test_enc_str_buf_cat.rb
new file mode 100644
index 0000000..9582fe2
--- /dev/null
+++ b/jni/ruby/test/-ext-/string/test_enc_str_buf_cat.rb
@@ -0,0 +1,15 @@
+require 'test/unit'
+require "-test-/string/string"
+
+class Test_StringEncStrBufCat < Test::Unit::TestCase
+  Bug6509 = '[ruby-dev:45688]'
+
+  def test_unknown
+    a8_str = "a\xBE".force_encoding(Encoding::ASCII_8BIT)
+    cr_unknown_str = [0x62].pack('C*')
+    assert_equal(true, a8_str.valid_encoding?, "an assertion for following tests")
+    assert_equal(:valid, Bug::String.new(a8_str).coderange, "an assertion for following tests")
+    assert_equal(:unknown, Bug::String.new(cr_unknown_str).coderange, "an assertion for following tests")
+    assert_equal(:valid, Bug::String.new(a8_str).enc_str_buf_cat(cr_unknown_str).coderange, Bug6509)
+  end
+end
diff --git a/jni/ruby/test/-ext-/string/test_modify_expand.rb b/jni/ruby/test/-ext-/string/test_modify_expand.rb
new file mode 100644
index 0000000..34b7be7
--- /dev/null
+++ b/jni/ruby/test/-ext-/string/test_modify_expand.rb
@@ -0,0 +1,15 @@
+require 'test/unit'
+require "-test-/string/string"
+
+class Test_StringModifyExpand < Test::Unit::TestCase
+  def test_modify_expand_memory_leak
+    assert_no_memory_leak(["-r-test-/string/string"],
+                          <<-PRE, <<-CMD, "rb_str_modify_expand()", limit: 2.5)
+      s=Bug::String.new
+    PRE
+      size = $initial_size
+      10.times{s.modify_expand!(size)}
+      s.replace("")
+    CMD
+  end
+end
diff --git a/jni/ruby/test/-ext-/string/test_nofree.rb b/jni/ruby/test/-ext-/string/test_nofree.rb
new file mode 100644
index 0000000..234c84d
--- /dev/null
+++ b/jni/ruby/test/-ext-/string/test_nofree.rb
@@ -0,0 +1,10 @@
+require 'test/unit'
+
+class Test_StringNoFree < Test::Unit::TestCase
+  def test_no_memory_leak
+    bug10942 = '[ruby-core:68436] [Bug #10942] no leak on nofree string'
+    assert_no_memory_leak(%w(-r-test-/string/string), '',
+                          '1000000.times {Bug::String.nofree << "a" * 100}',
+                          bug10942, rss: true, limit: 2.0)
+  end
+end
diff --git a/jni/ruby/test/-ext-/string/test_normalize.rb b/jni/ruby/test/-ext-/string/test_normalize.rb
new file mode 100644
index 0000000..283ca93
--- /dev/null
+++ b/jni/ruby/test/-ext-/string/test_normalize.rb
@@ -0,0 +1,106 @@
+require 'test/unit'
+require "-test-/string/string"
+require "tempfile"
+
+class Test_StringNormalize < Test::Unit::TestCase
+=begin
+  def test_normalize_all
+    exclude = [
+      #0x340, 0x341, 0x343, 0x344
+    ]
+    (0x0080..0xFFFD).each do |n|
+      next if 0xD800 <= n && n <= 0xDFFF
+      next if exclude.include? n
+      code = n.to_s(16)
+      Tempfile.create("#{code}-#{n.chr(Encoding::UTF_8)}-") do |tempfile|
+        ary = Dir.glob(File.expand_path("../#{code}-*", tempfile.path))
+        assert_equal 1, ary.size
+        result = ary[0]
+        rn = result[/\/\h+-(.+?)-/, 1]
+        #assert_equal tempfile.path, result, "#{rn.dump} is not U+#{n.to_s(16)}"
+        r2 = Bug::String.new(result ).normalize_ospath
+        rn2 = r2[/\/\h+-(.+?)-/, 1]
+        if tempfile.path == result
+          if tempfile.path == r2
+          else
+            puts "U+#{n.to_s(16)} shouldn't be r2#{rn2.dump}"
+          end
+        else
+          if tempfile.path == r2
+            # puts "U+#{n.to_s(16)} shouldn't be r#{rn.dump}"
+          elsif result == r2
+            puts "U+#{n.to_s(16)} shouldn't be #{rn.dump}"
+          else
+            puts "U+#{n.to_s(16)} shouldn't be r#{rn.dump} r2#{rn2.dump}"
+          end
+        end
+      end
+    end
+  end
+=end
+
+  def test_normalize
+    %[
+      \u304C \u304B\u3099
+      \u3077 \u3075\u309A
+      \u308F\u3099 \u308F\u3099
+      \u30F4 \u30A6\u3099
+      \u30DD \u30DB\u309A
+      \u30AB\u303A \u30AB\u303A
+      \u00C1 A\u0301
+      B\u030A B\u030A
+      \u0386 \u0391\u0301
+      \u03D3 \u03D2\u0301
+      \u0401 \u0415\u0308
+      \u2260 =\u0338
+      \u{c548} \u{110b}\u{1161}\u{11ab}
+    ].scan(/(\S+)\s+(\S+)/) do |expected, src|
+      result = Bug::String.new(src).normalize_ospath
+      assert_equal expected, result,
+        "#{expected.dump} is expected but #{src.dump}"
+    end
+  rescue NotImplementedError
+  end
+
+  def test_not_normalize_kc
+    %[
+      \u2460
+      \u2162
+      \u3349
+      \u33A1
+      \u337B
+      \u2116
+      \u33CD
+      \u2121
+      \u32A4
+      \u3231
+    ].split.each do |src|
+      result = Bug::String.new(src).normalize_ospath
+      assert_equal src, result,
+        "#{src.dump} is expected not to be normalized, but #{result.dump}"
+    end
+  rescue NotImplementedError
+  end
+
+  def test_dont_normalize_hfsplus
+    %[
+      \u2190\u0338
+      \u219A
+      \u212B
+      \uF90A
+      \uF9F4
+      \uF961 \uF9DB
+      \uF96F \uF3AA
+      \uF915 \uF95C \uF9BF
+      \uFA0C
+      \uFA10
+      \uFA19
+      \uFA26
+    ].split.each do |src|
+      result = Bug::String.new(src).normalize_ospath
+      assert_equal src, result,
+        "#{src.dump} is expected not to be normalized, but #{result.dump}"
+    end
+  rescue NotImplementedError
+  end
+end
diff --git a/jni/ruby/test/-ext-/string/test_qsort.rb b/jni/ruby/test/-ext-/string/test_qsort.rb
new file mode 100644
index 0000000..3a58523
--- /dev/null
+++ b/jni/ruby/test/-ext-/string/test_qsort.rb
@@ -0,0 +1,19 @@
+require 'test/unit'
+require "-test-/string/string"
+
+class Test_StringQSort < Test::Unit::TestCase
+  def test_qsort
+    s = Bug::String.new("xxozfxx")
+    s.qsort!
+    assert_equal("foxxxxz", s)
+  end
+
+  def test_qsort_slice
+    s = Bug::String.new("xxofzx1")
+    s.qsort!(nil, nil, 3)
+    assert_equal("fzxxxo1", s)
+    s = Bug::String.new("xxofzx231")
+    s.qsort!(nil, nil, 3)
+    assert_equal("231fzxxxo", s)
+  end
+end
diff --git a/jni/ruby/test/-ext-/string/test_set_len.rb b/jni/ruby/test/-ext-/string/test_set_len.rb
new file mode 100644
index 0000000..e257cac
--- /dev/null
+++ b/jni/ruby/test/-ext-/string/test_set_len.rb
@@ -0,0 +1,25 @@
+require 'test/unit'
+require "-test-/string/string"
+
+class Test_StrSetLen < Test::Unit::TestCase
+  def setup
+    @s0 = [*"a".."z"].join("").freeze
+    @s1 = Bug::String.new(@s0)
+  end
+
+  def teardown
+    orig = [*"a".."z"].join("")
+    assert_equal(orig, @s0)
+  end
+
+  def test_non_shared
+    @s1.modify!
+    assert_equal("abc", @s1.set_len(3))
+  end
+
+  def test_shared
+    assert_raise(RuntimeError) {
+      assert_equal("abc", @s1.set_len(3))
+    }
+  end
+end