diff options
Diffstat (limited to 'test/ruby/test_regexp.rb')
| -rw-r--r-- | test/ruby/test_regexp.rb | 1033 |
1 files changed, 992 insertions, 41 deletions
diff --git a/test/ruby/test_regexp.rb b/test/ruby/test_regexp.rb index 2bf4649f14..805c57b472 100644 --- a/test/ruby/test_regexp.rb +++ b/test/ruby/test_regexp.rb @@ -40,19 +40,6 @@ class TestRegexp < Test::Unit::TestCase assert_equal("a".gsub(/a\Z/, ""), "") end - def test_yoshidam_net_20041111_1 - s = "[\xC2\xA0-\xC3\xBE]" - r = assert_deprecated_warning(/ignored/) {Regexp.new(s, nil, "u")} - assert_match(r, "\xC3\xBE") - end - - def test_yoshidam_net_20041111_2 - assert_raise(RegexpError) do - s = "[\xFF-\xFF]".force_encoding("utf-8") - assert_warning(/ignored/) {Regexp.new(s, nil, "u")} - end - end - def test_ruby_dev_31309 assert_equal('Ruby', 'Ruby'.sub(/[^a-z]/i, '-')) end @@ -85,12 +72,141 @@ class TestRegexp < Test::Unit::TestCase end end + def test_to_s_under_gc_compact_stress + omit "compaction doesn't work well on s390x" if RUBY_PLATFORM =~ /s390x/ # https://github.com/ruby/ruby/pull/5077 + EnvUtil.under_gc_compact_stress do + str = "abcd\u3042" + [:UTF_16BE, :UTF_16LE, :UTF_32BE, :UTF_32LE].each do |es| + enc = Encoding.const_get(es) + rs = Regexp.new(str.encode(enc)).to_s + assert_equal("(?-mix:abcd\u3042)".encode(enc), rs) + assert_equal(enc, rs.encoding) + end + end + end + def test_to_s_extended_subexp re = /#\g#{"\n"}/x re = /#{re}/ assert_warn('', '[ruby-core:82328] [Bug #13798]') {re.to_s} end + def test_extended_comment_invalid_escape_bug_18294 + assert_separately([], <<-RUBY) + re = / C:\\\\[a-z]{5} # e.g. C:\\users /x + assert_match(re, 'C:\\users') + assert_not_match(re, 'C:\\user') + + re = / + foo # \\M-ca + bar + /x + assert_match(re, 'foobar') + assert_not_match(re, 'foobaz') + + re = / + f[#o]o # \\M-ca + bar + /x + assert_match(re, 'foobar') + assert_not_match(re, 'foobaz') + + re = / + f[[:alnum:]#]o # \\M-ca + bar + /x + assert_match(re, 'foobar') + assert_not_match(re, 'foobaz') + + re = / + f(?# \\M-ca)oo # \\M-ca + bar + /x + assert_match(re, 'foobar') + assert_not_match(re, 'foobaz') + + re = /f(?# \\M-ca)oobar/ + assert_match(re, 'foobar') + assert_not_match(re, 'foobaz') + + re = /[-(?# fca)]oobar/ + assert_match(re, 'foobar') + assert_not_match(re, 'foobaz') + + re = /f(?# ca\0\\M-ca)oobar/ + assert_match(re, 'foobar') + assert_not_match(re, 'foobaz') + RUBY + + assert_raise(SyntaxError) {eval "/\\users/x"} + assert_raise(SyntaxError) {eval "/[\\users]/x"} + assert_raise(SyntaxError) {eval "/(?<\\users)/x"} + assert_raise(SyntaxError) {eval "/# \\users/"} + end + + def test_nonextended_section_of_extended_regexp_bug_19379 + assert_separately([], <<-'RUBY') + re = /(?-x:#)/x + assert_match(re, '#') + assert_not_match(re, '-') + + re = /(?xi:# + y)/ + assert_match(re, 'Y') + assert_not_match(re, '-') + + re = /(?mix:# + y)/ + assert_match(re, 'Y') + assert_not_match(re, '-') + + re = /(?x-im:# + y)/i + assert_match(re, 'y') + assert_not_match(re, 'Y') + + re = /(?-imx:(?xim:# + y))/x + assert_match(re, 'y') + assert_not_match(re, '-') + + re = /(?x)# + y/ + assert_match(re, 'y') + assert_not_match(re, 'Y') + + re = /(?mx-i)# + y/i + assert_match(re, 'y') + assert_not_match(re, 'Y') + + re = /(?-imx:(?xim:# + (?-x)y#))/x + assert_match(re, 'Y#') + assert_not_match(re, '-#') + + re = /(?imx:# + (?-xim:#(?im)#(?x)# + )# + (?x)# + y)/ + assert_match(re, '###Y') + assert_not_match(re, '###-') + + re = %r{#c-\w+/comment/[\w-]+} + re = %r{https?://[^/]+#{re}}x + assert_match(re, 'http://foo#c-x/comment/bar') + assert_not_match(re, 'http://foo#cx/comment/bar') + RUBY + end + + def test_utf8_comment_in_usascii_extended_regexp_bug_19455 + assert_separately([], <<-RUBY) + assert_equal(Encoding::UTF_8, /(?#\u1000)/x.encoding) + assert_equal(Encoding::UTF_8, /#\u1000/x.encoding) + RUBY + end + def test_union assert_equal :ok, begin Regexp.union( @@ -208,6 +324,9 @@ class TestRegexp < Test::Unit::TestCase assert_equal({'a' => '1', 'b' => '2', 'c' => '3'}, /^(?<a>.)(?<b>.)(?<c>.)?/.match('123').named_captures) assert_equal({'a' => '1', 'b' => '2', 'c' => ''}, /^(?<a>.)(?<b>.)(?<c>.?)/.match('12').named_captures) + assert_equal({a: '1', b: '2', c: ''}, /^(?<a>.)(?<b>.)(?<c>.?)/.match('12').named_captures(symbolize_names: true)) + assert_equal({'a' => '1', 'b' => '2', 'c' => ''}, /^(?<a>.)(?<b>.)(?<c>.?)/.match('12').named_captures(symbolize_names: false)) + assert_equal({'a' => 'x'}, /(?<a>x)|(?<a>y)/.match('x').named_captures) assert_equal({'a' => 'y'}, /(?<a>x)|(?<a>y)/.match('y').named_captures) @@ -351,6 +470,13 @@ class TestRegexp < Test::Unit::TestCase assert_equal('/\/\xF1\xF2\xF3/i', /\/#{s}/i.inspect) end + def test_inspect_under_gc_compact_stress + omit "compaction doesn't work well on s390x" if RUBY_PLATFORM =~ /s390x/ # https://github.com/ruby/ruby/pull/5077 + EnvUtil.under_gc_compact_stress do + assert_equal('/(?-mix:\\/)|/', Regexp.union(/\//, "").inspect) + end + end + def test_char_to_option assert_equal("BAR", "FOOBARBAZ"[/b../i]) assert_equal("bar", "foobarbaz"[/ b . . /x]) @@ -424,6 +550,37 @@ class TestRegexp < Test::Unit::TestCase assert_equal([2, 3], m.offset(3)) end + def test_match_byteoffset_begin_end + m = /(?<x>b..)/.match("foobarbaz") + assert_equal([3, 6], m.byteoffset("x")) + assert_equal(3, m.begin("x")) + assert_equal(6, m.end("x")) + assert_raise(IndexError) { m.byteoffset("y") } + assert_raise(IndexError) { m.byteoffset(2) } + assert_raise(IndexError) { m.begin(2) } + assert_raise(IndexError) { m.end(2) } + assert_raise(IndexError) { m.bytebegin(2) } + assert_raise(IndexError) { m.byteend(2) } + + m = /(?<x>q..)?/.match("foobarbaz") + assert_equal([nil, nil], m.byteoffset("x")) + assert_equal(nil, m.begin("x")) + assert_equal(nil, m.end("x")) + assert_equal(nil, m.bytebegin("x")) + assert_equal(nil, m.byteend("x")) + + m = /\A\u3042(.)(.)?(.)\z/.match("\u3042\u3043\u3044") + assert_equal([3, 6], m.byteoffset(1)) + assert_equal(3, m.bytebegin(1)) + assert_equal(6, m.byteend(1)) + assert_equal([nil, nil], m.byteoffset(2)) + assert_equal(nil, m.bytebegin(2)) + assert_equal(nil, m.byteend(2)) + assert_equal([6, 9], m.byteoffset(3)) + assert_equal(6, m.bytebegin(3)) + assert_equal(9, m.byteend(3)) + end + def test_match_to_s m = /(?<x>b..)/.match("foobarbaz") assert_equal("bar", m.to_s) @@ -467,6 +624,7 @@ class TestRegexp < Test::Unit::TestCase assert_nil(m[5]) assert_raise(IndexError) { m[:foo] } assert_raise(TypeError) { m[nil] } + assert_equal(["baz", nil], m[-2, 3]) end def test_match_values_at @@ -533,17 +691,73 @@ class TestRegexp < Test::Unit::TestCase assert_equal('#<MatchData "foobarbaz" 1:"foo" 2:"bar" 3:"baz" 4:nil>', m.inspect) end + def test_match_data_deconstruct + m = /foo.+/.match("foobarbaz") + assert_equal([], m.deconstruct) + + m = /(foo).+(baz)/.match("foobarbaz") + assert_equal(["foo", "baz"], m.deconstruct) + + m = /(...)(...)(...)(...)?/.match("foobarbaz") + assert_equal(["foo", "bar", "baz", nil], m.deconstruct) + end + + def test_match_data_deconstruct_keys + m = /foo.+/.match("foobarbaz") + assert_equal({}, m.deconstruct_keys([:a])) + + m = /(?<a>foo).+(?<b>baz)/.match("foobarbaz") + assert_equal({a: "foo", b: "baz"}, m.deconstruct_keys(nil)) + assert_equal({a: "foo", b: "baz"}, m.deconstruct_keys([:a, :b])) + assert_equal({b: "baz"}, m.deconstruct_keys([:b])) + assert_equal({}, m.deconstruct_keys([:c, :a])) + assert_equal({a: "foo"}, m.deconstruct_keys([:a, :c])) + assert_equal({}, m.deconstruct_keys([:a, :b, :c])) + + assert_raise(TypeError) { + m.deconstruct_keys(0) + } + + assert_raise(TypeError) { + m.deconstruct_keys(["a", "b"]) + } + end + + def test_match_no_match_no_matchdata + EnvUtil.without_gc do + h = {} + ObjectSpace.count_objects(h) + prev_matches = h[:T_MATCH] || 0 + _md = /[A-Z]/.match('1') # no match + ObjectSpace.count_objects(h) + new_matches = h[:T_MATCH] || 0 + assert_equal prev_matches, new_matches, "Bug [#20104]" + end + end + def test_initialize assert_raise(ArgumentError) { Regexp.new } assert_equal(/foo/, assert_warning(/ignored/) {Regexp.new(/foo/, Regexp::IGNORECASE)}) + assert_equal(/foo/, assert_no_warning(/ignored/) {Regexp.new(/foo/)}) + assert_equal(/foo/, assert_no_warning(/ignored/) {Regexp.new(/foo/, timeout: nil)}) + + arg_encoding_none = //n.options # ARG_ENCODING_NONE is implementation defined value - assert_equal(Encoding.find("US-ASCII"), Regexp.new("b..", nil, "n").encoding) - assert_equal("bar", "foobarbaz"[Regexp.new("b..", nil, "n")]) - assert_equal(//n, Regexp.new("", nil, "n")) + assert_deprecated_warning('') do + assert_equal(Encoding.find("US-ASCII"), Regexp.new("b..", Regexp::NOENCODING).encoding) + assert_equal("bar", "foobarbaz"[Regexp.new("b..", Regexp::NOENCODING)]) + assert_equal(//, Regexp.new("")) + assert_equal(//, Regexp.new("", timeout: 1)) + assert_equal(//n, Regexp.new("", Regexp::NOENCODING)) + assert_equal(//n, Regexp.new("", Regexp::NOENCODING, timeout: 1)) - arg_encoding_none = 32 # ARG_ENCODING_NONE is implementation defined value - assert_equal(arg_encoding_none, Regexp.new("", nil, "n").options) - assert_equal(arg_encoding_none, Regexp.new("", nil, "N").options) + assert_equal(arg_encoding_none, Regexp.new("", Regexp::NOENCODING).options) + + assert_nil(Regexp.new("").timeout) + assert_equal(1.0, Regexp.new("", timeout: 1.0).timeout) + assert_nil(Regexp.compile("").timeout) + assert_equal(1.0, Regexp.compile("", timeout: 1.0).timeout) + end assert_raise(RegexpError) { Regexp.new(")(") } assert_raise(RegexpError) { Regexp.new('[\\40000000000') } @@ -551,6 +765,35 @@ class TestRegexp < Test::Unit::TestCase assert_raise(RegexpError) { Regexp.new("((?<v>))\\g<0>") } end + def test_initialize_from_regex_memory_corruption + assert_ruby_status([], <<-'end;') + 10_000.times { Regexp.new(Regexp.new("(?<name>)")) } + end; + end + + def test_initialize_bool_warning + assert_warning(/expected true or false as ignorecase/) do + Regexp.new("foo", :i) + end + end + + def test_initialize_option + assert_equal(//i, Regexp.new("", "i")) + assert_equal(//m, Regexp.new("", "m")) + assert_equal(//x, Regexp.new("", "x")) + assert_equal(//imx, Regexp.new("", "imx")) + assert_equal(//, Regexp.new("", "")) + assert_equal(//imx, Regexp.new("", "mimix")) + + assert_raise(ArgumentError) { Regexp.new("", "e") } + assert_raise(ArgumentError) { Regexp.new("", "n") } + assert_raise(ArgumentError) { Regexp.new("", "s") } + assert_raise(ArgumentError) { Regexp.new("", "u") } + assert_raise(ArgumentError) { Regexp.new("", "o") } + assert_raise(ArgumentError) { Regexp.new("", "j") } + assert_raise(ArgumentError) { Regexp.new("", "xmen") } + end + def test_match_control_meta_escape assert_equal(0, /\c\xFF/ =~ "\c\xFF") assert_equal(0, /\c\M-\xFF/ =~ "\c\M-\xFF") @@ -659,6 +902,14 @@ class TestRegexp < Test::Unit::TestCase $_ = nil; assert_nil(~/./) end + def test_match_under_gc_compact_stress + omit "compaction doesn't work well on s390x" if RUBY_PLATFORM =~ /s390x/ # https://github.com/ruby/ruby/pull/5077 + EnvUtil.under_gc_compact_stress do + m = /(?<foo>.)(?<n>[^aeiou])?(?<bar>.+)/.match("hoge\u3042") + assert_equal("h", m.match(:foo)) + end + end + def test_match_p /backref/ =~ 'backref' # must match here, but not in a separate method, e.g., assert_send, @@ -724,7 +975,7 @@ class TestRegexp < Test::Unit::TestCase def test_dup assert_equal(//, //.dup) - assert_raise(TypeError) { //.dup.instance_eval { initialize_copy(nil) } } + assert_raise(FrozenError) { //.dup.instance_eval { initialize_copy(/a/) } } end def test_regsub @@ -748,6 +999,30 @@ class TestRegexp < Test::Unit::TestCase assert_equal('foobazquux/foobazquux', result, bug8856) end + def test_regsub_no_memory_leak + assert_no_memory_leak([], "#{<<~"begin;"}", "#{<<~"end;"}", rss: true) + code = proc do + "aaaaaaaaaaa".gsub(/a/, "") + end + + 1_000.times(&code) + begin; + 100_000.times(&code) + end; + end + + def test_regsub_no_memory_leak_many_captures + assert_no_memory_leak([], "#{<<~"begin;"}", "#{<<~"end;"}", rss: true) + code = proc do + "aaaaaaaaaaa".gsub(/(a)(b)?(c)?(d)?(e)?(f)?(g)?(h)?/, "") + end + + 1_000.times(&code) + begin; + 100_000.times(&code) + end; + end + def test_ignorecase v = assert_deprecated_warning(/variable \$= is no longer effective/) { $= } assert_equal(false, v) @@ -773,10 +1048,12 @@ class TestRegexp < Test::Unit::TestCase [Encoding::UTF_8, Encoding::Shift_JIS, Encoding::EUC_JP].each do |enc| idx = key.encode(enc) pat = /#{idx}/ - test.call {|m| assert_raise_with_message(IndexError, pat, bug10877) {m[idx]} } - test.call {|m| assert_raise_with_message(IndexError, pat, bug18160) {m.offset(idx)} } - test.call {|m| assert_raise_with_message(IndexError, pat, bug18160) {m.begin(idx)} } - test.call {|m| assert_raise_with_message(IndexError, pat, bug18160) {m.end(idx)} } + EnvUtil.with_default_internal(enc) do + test.call {|m| assert_raise_with_message(IndexError, pat, bug10877) {m[idx]} } + test.call {|m| assert_raise_with_message(IndexError, pat, bug18160) {m.offset(idx)} } + test.call {|m| assert_raise_with_message(IndexError, pat, bug18160) {m.begin(idx)} } + test.call {|m| assert_raise_with_message(IndexError, pat, bug18160) {m.end(idx)} } + end end test.call {|m| assert_equal(/a/, m.regexp) } test.call {|m| assert_equal("abc", m.string) } @@ -1045,6 +1322,9 @@ class TestRegexp < Test::Unit::TestCase assert_match(/\A[[:space:]]+\z/, "\r\n\v\f\r\s\u0085") assert_match(/\A[[:ascii:]]+\z/, "\x00\x7F") assert_no_match(/[[:ascii:]]/, "\x80\xFF") + + assert_match(/[[:word:]]/, "\u{200C}") + assert_match(/[[:word:]]/, "\u{200D}") end def test_cclass_R @@ -1165,25 +1445,223 @@ class TestRegexp < Test::Unit::TestCase end def test_unicode_age - assert_match(/^\p{Age=6.0}$/u, "\u261c") - assert_match(/^\p{Age=1.1}$/u, "\u261c") - assert_no_match(/^\P{age=6.0}$/u, "\u261c") - - assert_match(/^\p{age=6.0}$/u, "\u31f6") - assert_match(/^\p{age=3.2}$/u, "\u31f6") - assert_no_match(/^\p{age=3.1}$/u, "\u31f6") - assert_no_match(/^\p{age=3.0}$/u, "\u31f6") - assert_no_match(/^\p{age=1.1}$/u, "\u31f6") + assert_unicode_age("\u261c", matches: %w"6.0 1.1", unmatches: []) + + assert_unicode_age("\u31f6", matches: %w"6.0 3.2", unmatches: %w"3.1 3.0 1.1") + assert_unicode_age("\u2754", matches: %w"6.0", unmatches: %w"5.0 4.0 3.0 2.0 1.1") + + assert_unicode_age("\u32FF", matches: %w"12.1", unmatches: %w"12.0") + end + + def test_unicode_age_14_0 + @matches = %w"14.0" + @unmatches = %w"13.0" + + assert_unicode_age("\u{10570}") + assert_unicode_age("\u9FFF") + assert_unicode_age("\u{2A6DF}") + assert_unicode_age("\u{2B738}") + end + + def test_unicode_age_15_0 + @matches = %w"15.0" + @unmatches = %w"14.0" + + assert_unicode_age("\u{0CF3}", + "KANNADA SIGN COMBINING ANUSVARA ABOVE RIGHT") + assert_unicode_age("\u{0ECE}", "LAO YAMAKKAN") + assert_unicode_age("\u{10EFD}".."\u{10EFF}", + "ARABIC SMALL LOW WORD SAKTA..ARABIC SMALL LOW WORD MADDA") + assert_unicode_age("\u{1123F}".."\u{11241}", + "KHOJKI LETTER QA..KHOJKI VOWEL SIGN VOCALIC R") + assert_unicode_age("\u{11B00}".."\u{11B09}", + "DEVANAGARI HEAD MARK..DEVANAGARI SIGN MINDU") + assert_unicode_age("\u{11F00}".."\u{11F10}", + "KAWI SIGN CANDRABINDU..KAWI LETTER O") + assert_unicode_age("\u{11F12}".."\u{11F3A}", + "KAWI LETTER KA..KAWI VOWEL SIGN VOCALIC R") + assert_unicode_age("\u{11F3E}".."\u{11F59}", + "KAWI VOWEL SIGN E..KAWI DIGIT NINE") + assert_unicode_age("\u{1342F}", + "EGYPTIAN HIEROGLYPH V011D") + assert_unicode_age("\u{13439}".."\u{1343F}", + "EGYPTIAN HIEROGLYPH INSERT AT MIDDLE..EGYPTIAN HIEROGLYPH END WALLED ENCLOSURE") + assert_unicode_age("\u{13440}".."\u{13455}", + "EGYPTIAN HIEROGLYPH MIRROR HORIZONTALLY..EGYPTIAN HIEROGLYPH MODIFIER DAMAGED") + assert_unicode_age("\u{1B132}", "HIRAGANA LETTER SMALL KO") + assert_unicode_age("\u{1B155}", "KATAKANA LETTER SMALL KO") + assert_unicode_age("\u{1D2C0}".."\u{1D2D3}", + "KAKTOVIK NUMERAL ZERO..KAKTOVIK NUMERAL NINETEEN") + assert_unicode_age("\u{1DF25}".."\u{1DF2A}", + "LATIN SMALL LETTER D WITH MID-HEIGHT LEFT HOOK..LATIN SMALL LETTER T WITH MID-HEIGHT LEFT HOOK") + assert_unicode_age("\u{1E030}".."\u{1E06D}", + "MODIFIER LETTER CYRILLIC SMALL A..MODIFIER LETTER CYRILLIC SMALL STRAIGHT U WITH STROKE") + assert_unicode_age("\u{1E08F}", + "COMBINING CYRILLIC SMALL LETTER BYELORUSSIAN-UKRAINIAN I") + assert_unicode_age("\u{1E4D0}".."\u{1E4F9}", + "NAG MUNDARI LETTER O..NAG MUNDARI DIGIT NINE") + assert_unicode_age("\u{1F6DC}", "WIRELESS") + assert_unicode_age("\u{1F774}".."\u{1F776}", + "LOT OF FORTUNE..LUNAR ECLIPSE") + assert_unicode_age("\u{1F77B}".."\u{1F77F}", + "HAUMEA..ORCUS") + assert_unicode_age("\u{1F7D9}", "NINE POINTED WHITE STAR") + assert_unicode_age("\u{1FA75}".."\u{1FA77}", + "LIGHT BLUE HEART..PINK HEART") + assert_unicode_age("\u{1FA87}".."\u{1FA88}", + "MARACAS..FLUTE") + assert_unicode_age("\u{1FAAD}".."\u{1FAAF}", + "FOLDING HAND FAN..KHANDA") + assert_unicode_age("\u{1FABB}".."\u{1FABD}", + "HYACINTH..WING") + assert_unicode_age("\u{1FABF}", "GOOSE") + assert_unicode_age("\u{1FACE}".."\u{1FACF}", + "MOOSE..DONKEY") + assert_unicode_age("\u{1FADA}".."\u{1FADB}", + "GINGER ROOT..PEA POD") + assert_unicode_age("\u{1FAE8}", "SHAKING FACE") + assert_unicode_age("\u{1FAF7}".."\u{1FAF8}", + "LEFTWARDS PUSHING HAND..RIGHTWARDS PUSHING HAND") + assert_unicode_age("\u{2B739}", + "CJK UNIFIED IDEOGRAPH-2B739") + assert_unicode_age("\u{31350}".."\u{323AF}", + "CJK UNIFIED IDEOGRAPH-31350..CJK UNIFIED IDEOGRAPH-323AF") + end + + def test_unicode_age_15_1 + @matches = %w"15.1" + @unmatches = %w"15.0" + + # https://www.unicode.org/Public/15.1.0/ucd/DerivedAge.txt + assert_unicode_age("\u{2FFC}".."\u{2FFF}", + "IDEOGRAPHIC DESCRIPTION CHARACTER SURROUND FROM RIGHT..IDEOGRAPHIC DESCRIPTION CHARACTER ROTATION") + assert_unicode_age("\u{31EF}", + "IDEOGRAPHIC DESCRIPTION CHARACTER SUBTRACTION") + assert_unicode_age("\u{2EBF0}".."\u{2EE5D}", + "CJK UNIFIED IDEOGRAPH-2EBF0..CJK UNIFIED IDEOGRAPH-2EE5D") + end + + def test_unicode_age_16_0 + @matches = %w"16.0" + @unmatches = %w"15.1" + + # https://www.unicode.org/Public/16.0.0/ucd/DerivedAge.txt + assert_unicode_age("\u{0897}", + "ARABIC PEPET") + assert_unicode_age("\u{1B4E}".."\u{1B4F}", + "BALINESE INVERTED CARIK SIKI..BALINESE INVERTED CARIK PAREREN") + assert_unicode_age("\u{1B7F}", + "BALINESE PANTI BAWAK") + assert_unicode_age("\u{1C89}".."\u{1C8A}", + "CYRILLIC CAPITAL LETTER TJE..CYRILLIC SMALL LETTER TJE") + assert_unicode_age("\u{2427}".."\u{2429}", + "SYMBOL FOR DELETE SQUARE CHECKER BOARD FORM..SYMBOL FOR DELETE MEDIUM SHADE FORM") + assert_unicode_age("\u{31E4}".."\u{31E5}", + "CJK STROKE HXG..CJK STROKE SZP") + assert_unicode_age("\u{A7CB}".."\u{A7CD}", + "LATIN CAPITAL LETTER RAMS HORN..LATIN SMALL LETTER S WITH DIAGONAL STROKE") + assert_unicode_age("\u{A7DA}".."\u{A7DC}", + "LATIN CAPITAL LETTER LAMBDA..LATIN CAPITAL LETTER LAMBDA WITH STROKE") + assert_unicode_age("\u{105C0}".."\u{105F3}", + "TODHRI LETTER A..TODHRI LETTER OO") + assert_unicode_age("\u{10D40}".."\u{10D65}", + "GARAY DIGIT ZERO..GARAY CAPITAL LETTER OLD NA") + assert_unicode_age("\u{10D69}".."\u{10D85}", + "GARAY VOWEL SIGN E..GARAY SMALL LETTER OLD NA") + assert_unicode_age("\u{10D8E}".."\u{10D8F}", + "GARAY PLUS SIGN..GARAY MINUS SIGN") + assert_unicode_age("\u{10EC2}".."\u{10EC4}", + "ARABIC LETTER DAL WITH TWO DOTS VERTICALLY BELOW..ARABIC LETTER KAF WITH TWO DOTS VERTICALLY BELOW") + assert_unicode_age("\u{10EFC}", + "ARABIC COMBINING ALEF OVERLAY") + assert_unicode_age("\u{11380}".."\u{11389}", + "TULU-TIGALARI LETTER A..TULU-TIGALARI LETTER VOCALIC LL") + assert_unicode_age("\u{1138B}", + "TULU-TIGALARI LETTER EE") + assert_unicode_age("\u{1138E}", + "TULU-TIGALARI LETTER AI") + assert_unicode_age("\u{11390}".."\u{113B5}", + "TULU-TIGALARI LETTER OO..TULU-TIGALARI LETTER LLLA") + assert_unicode_age("\u{113B7}".."\u{113C0}", + "TULU-TIGALARI SIGN AVAGRAHA..TULU-TIGALARI VOWEL SIGN VOCALIC LL") + assert_unicode_age("\u{113C2}", + "TULU-TIGALARI VOWEL SIGN EE") + assert_unicode_age("\u{113C5}", + "TULU-TIGALARI VOWEL SIGN AI") + assert_unicode_age("\u{113C7}".."\u{113CA}", + "TULU-TIGALARI VOWEL SIGN OO..TULU-TIGALARI SIGN CANDRA ANUNASIKA") + assert_unicode_age("\u{113CC}".."\u{113D5}", + "TULU-TIGALARI SIGN ANUSVARA..TULU-TIGALARI DOUBLE DANDA") + assert_unicode_age("\u{113D7}".."\u{113D8}", + "TULU-TIGALARI SIGN OM PUSHPIKA..TULU-TIGALARI SIGN SHRII PUSHPIKA") + assert_unicode_age("\u{113E1}".."\u{113E2}", + "TULU-TIGALARI VEDIC TONE SVARITA..TULU-TIGALARI VEDIC TONE ANUDATTA") + assert_unicode_age("\u{116D0}".."\u{116E3}", + "MYANMAR PAO DIGIT ZERO..MYANMAR EASTERN PWO KAREN DIGIT NINE") + assert_unicode_age("\u{11BC0}".."\u{11BE1}", + "SUNUWAR LETTER DEVI..SUNUWAR SIGN PVO") + assert_unicode_age("\u{11BF0}".."\u{11BF9}", + "SUNUWAR DIGIT ZERO..SUNUWAR DIGIT NINE") + assert_unicode_age("\u{11F5A}", + "KAWI SIGN NUKTA") + assert_unicode_age("\u{13460}".."\u{143FA}", + "EGYPTIAN HIEROGLYPH-13460..EGYPTIAN HIEROGLYPH-143FA") + assert_unicode_age("\u{16100}".."\u{16139}", + "GURUNG KHEMA LETTER A..GURUNG KHEMA DIGIT NINE") + assert_unicode_age("\u{16D40}".."\u{16D79}", + "KIRAT RAI SIGN ANUSVARA..KIRAT RAI DIGIT NINE") + assert_unicode_age("\u{18CFF}", + "KHITAN SMALL SCRIPT CHARACTER-18CFF") + assert_unicode_age("\u{1CC00}".."\u{1CCF9}", + "UP-POINTING GO-KART..OUTLINED DIGIT NINE") + assert_unicode_age("\u{1CD00}".."\u{1CEB3}", + "BLOCK OCTANT-3..BLACK RIGHT TRIANGLE CARET") + assert_unicode_age("\u{1E5D0}".."\u{1E5FA}", + "OL ONAL LETTER O..OL ONAL DIGIT NINE") + assert_unicode_age("\u{1E5FF}", + "OL ONAL ABBREVIATION SIGN") + assert_unicode_age("\u{1F8B2}".."\u{1F8BB}", + "RIGHTWARDS ARROW WITH LOWER HOOK..SOUTH WEST ARROW FROM BAR") + assert_unicode_age("\u{1F8C0}".."\u{1F8C1}", + "LEFTWARDS ARROW FROM DOWNWARDS ARROW..RIGHTWARDS ARROW FROM DOWNWARDS ARROW") + assert_unicode_age("\u{1FA89}", + "HARP") + assert_unicode_age("\u{1FA8F}", + "SHOVEL") + assert_unicode_age("\u{1FABE}", + "LEAFLESS TREE") + assert_unicode_age("\u{1FAC6}", + "FINGERPRINT") + assert_unicode_age("\u{1FADC}", + "ROOT VEGETABLE") + assert_unicode_age("\u{1FADF}", + "SPLATTER") + assert_unicode_age("\u{1FAE9}", + "FACE WITH BAGS UNDER EYES") + assert_unicode_age("\u{1FBCB}".."\u{1FBEF}", + "WHITE CROSS MARK..TOP LEFT JUSTIFIED LOWER RIGHT QUARTER BLACK CIRCLE") + end + + UnicodeAgeRegexps = Hash.new do |h, age| + h[age] = [/\A\p{age=#{age}}+\z/u, /\A\P{age=#{age}}+\z/u].freeze + end + + def assert_unicode_age(char, mesg = nil, matches: @matches, unmatches: @unmatches) + if Range === char + char = char.to_a.join("") + end - assert_match(/^\p{age=6.0}$/u, "\u2754") - assert_no_match(/^\p{age=5.0}$/u, "\u2754") - assert_no_match(/^\p{age=4.0}$/u, "\u2754") - assert_no_match(/^\p{age=3.0}$/u, "\u2754") - assert_no_match(/^\p{age=2.0}$/u, "\u2754") - assert_no_match(/^\p{age=1.1}$/u, "\u2754") + matches.each do |age| + pos, neg = UnicodeAgeRegexps[age] + assert_match(pos, char, mesg) + assert_not_match(neg, char, mesg) + end - assert_no_match(/^\p{age=12.0}$/u, "\u32FF") - assert_match(/^\p{age=12.1}$/u, "\u32FF") + unmatches.each do |age| + pos, neg = UnicodeAgeRegexps[age] + assert_not_match(pos, char, mesg) + assert_match(neg, char, mesg) + end end MatchData_A = eval("class MatchData_\u{3042} < MatchData; self; end") @@ -1203,6 +1681,65 @@ class TestRegexp < Test::Unit::TestCase assert_equal("hoge fuga", h["body"]) end + def test_matchdata_large_capture_groups_stack + env = {"RUBY_THREAD_MACHINE_STACK_SIZE" => (256 * 1024).to_s} + assert_separately([env], <<~'RUBY') + n = 20000 + require "rbconfig/sizeof" + stack = RubyVM::DEFAULT_PARAMS[:thread_machine_stack_size] + size = RbConfig::SIZEOF["long"] + required = (n + 1) * 4 * size + if !stack || stack == 0 || stack >= required + omit "thread machine stack size not reduced (#{stack}:#{required})" + end + + inspect = Thread.new do + str = "\u{3042}" * n + m = Regexp.new("(.)" * n).match(str) + assert_not_nil(m) + assert_equal([n - 1, n], m.offset(n)) + m.inspect + end.value + + assert_include(inspect, "MatchData") + RUBY + end + + def test_match_integer_at + m = /(\d{4})(\d{2})(\d{2})/.match("20260308") + assert_equal(20260308, m.integer_at(0)) + assert_equal(2026, m.integer_at(1)) + assert_equal(3, m.integer_at(2)) + assert_equal(8, m.integer_at(3)) + assert_equal(nil, m.integer_at(4)) + assert_equal(8, m.integer_at(-1)) + assert_equal(3, m.integer_at(-2)) + assert_equal(2026, m.integer_at(-3)) + assert_equal(nil, m.integer_at(-4)) + + re = /[a-z]+|(\d+)/ + assert_equal(123, re.match("123").integer_at(1)) + assert_equal(nil, re.match("abc").integer_at(1)) + end + + def test_match_integer_at_name + m = /(?<y>\d{4})(?<m>\d{2})(?<d>\d{2})/.match("20260308") + assert_equal(2026, m.integer_at("y")) + assert_equal(3, m.integer_at("m")) + assert_equal(8, m.integer_at("d")) + end + + def test_match_integer_at_base + assert_equal(91, /\w+/.match("111").integer_at(0, 9)) + assert_equal(10_0000, /\w+/.match("10_0000").integer_at(0)) + assert_equal(0d1_0000, /\w+/.match("01_0000").integer_at(0)) + assert_equal(0o1_0000, /\w+/.match("01_0000").integer_at(0, 0)) + assert_equal(0b1_0000, /\w+/.match("0b1_0000").integer_at(0, 0)) + assert_equal(0o1_0000, /\w+/.match("0o1_0000").integer_at(0, 0)) + assert_equal(0d1_0000, /\w+/.match("0d1_0000").integer_at(0, 0)) + assert_equal(0x1_0000, /\w+/.match("0x1_0000").integer_at(0, 0)) + end + def test_regexp_popped EnvUtil.suppress_warning do assert_nothing_raised { eval("a = 1; /\#{ a }/; a") } @@ -1277,6 +1814,33 @@ class TestRegexp < Test::Unit::TestCase assert_raise(RegexpError, bug12418){ Regexp.new('(0?0|(?(5)||)|(?(5)||))?') } end + def test_quick_search + assert_match_at('(?i) *TOOKY', 'Mozilla/5.0 (Linux; Android 4.0.3; TOOKY', [[34, 40]]) # Issue #120 + end + + def test_ss_in_look_behind + assert_match_at("(?i:ss)", "ss", [[0, 2]]) + assert_match_at("(?i:ss)", "Ss", [[0, 2]]) + assert_match_at("(?i:ss)", "SS", [[0, 2]]) + assert_match_at("(?i:ss)", "\u017fS", [[0, 2]]) # LATIN SMALL LETTER LONG S + assert_match_at("(?i:ss)", "s\u017f", [[0, 2]]) + assert_match_at("(?i:ss)", "\u00df", [[0, 1]]) # LATIN SMALL LETTER SHARP S + assert_match_at("(?i:ss)", "\u1e9e", [[0, 1]]) # LATIN CAPITAL LETTER SHARP S + assert_match_at("(?i:xssy)", "xssy", [[0, 4]]) + assert_match_at("(?i:xssy)", "xSsy", [[0, 4]]) + assert_match_at("(?i:xssy)", "xSSy", [[0, 4]]) + assert_match_at("(?i:xssy)", "x\u017fSy", [[0, 4]]) + assert_match_at("(?i:xssy)", "xs\u017fy", [[0, 4]]) + assert_match_at("(?i:xssy)", "x\u00dfy", [[0, 3]]) + assert_match_at("(?i:xssy)", "x\u1e9ey", [[0, 3]]) + assert_match_at("(?i:\u00df)", "ss", [[0, 2]]) + assert_match_at("(?i:\u00df)", "SS", [[0, 2]]) + assert_match_at("(?i:[\u00df])", "ss", [[0, 2]]) + assert_match_at("(?i:[\u00df])", "SS", [[0, 2]]) + assert_match_at("(?i)(?<!ss)\u2728", "qq\u2728", [[2, 3]]) # Issue #92 + assert_match_at("(?i)(?<!xss)\u2728", "qq\u2728", [[2, 3]]) + end + def test_options_in_look_behind assert_nothing_raised { assert_match_at("(?<=(?i)ab)cd", "ABcd", [[2,4]]) @@ -1388,6 +1952,9 @@ class TestRegexp < Test::Unit::TestCase assert_equal(0, /(?~(a)c)/ =~ "abb") assert_nil($1) + + assert_equal(0, /(?~(a))/ =~ "") + assert_nil($1) end def test_backref_overrun @@ -1396,6 +1963,27 @@ class TestRegexp < Test::Unit::TestCase end end + def test_bug18631 + assert_kind_of MatchData, /(?<x>a)(?<x>aa)\k<x>/.match("aaaaa") + assert_kind_of MatchData, /(?<x>a)(?<x>aa)\k<x>/.match("aaaa") + assert_kind_of MatchData, /(?<x>a)(?<x>aa)\k<x>/.match("aaaab") + end + + def test_invalid_group + assert_separately([], "#{<<-"begin;"}\n#{<<-'end;'}") + begin; + assert_raise_with_message(RegexpError, /invalid conditional pattern/) do + Regexp.new("((?(1)x|x|)x)+") + end + end; + end + + def test_too_big_number_for_repeat_range + assert_raise_with_message(SyntaxError, /too big number for repeat range/) do + eval(%[/|{1000000}/]) + end + end + # This assertion is for porting x2() tests in testpy.py of Onigmo. def assert_match_at(re, str, positions, msg = nil) re = Regexp.new(re) unless re.is_a?(Regexp) @@ -1425,4 +2013,367 @@ class TestRegexp < Test::Unit::TestCase } assert_empty(errs, msg) end + + def test_s_timeout + assert_separately([], "#{<<-"begin;"}\n#{<<-'end;'}") + timeout = #{ EnvUtil.apply_timeout_scale(0.2).inspect } + begin; + Regexp.timeout = timeout + assert_in_delta(timeout, Regexp.timeout, timeout * 2 * Float::EPSILON) + + t = Time.now + assert_raise_with_message(Regexp::TimeoutError, "regexp match timeout") do + # A typical ReDoS case + /^(a*)*\1$/ =~ "a" * 1000000 + "x" + end + t = Time.now - t + + assert_operator(timeout, :<=, [timeout * 1.5, 1].max) + end; + end + + def test_s_timeout_corner_cases + assert_separately([], "#{<<-"begin;"}\n#{<<-'end;'}") + begin; + assert_nil(Regexp.timeout) + + # This is just an implementation detail that users should not depend on: + # If Regexp.timeout is set to a value greater than the value that can be + # represented in the internal representation of timeout, it uses the + # maximum value that can be represented. + Regexp.timeout = Float::INFINITY + assert_equal(((1<<64)-1) / 1000000000.0, Regexp.timeout) + + Regexp.timeout = 1e300 + assert_equal(((1<<64)-1) / 1000000000.0, Regexp.timeout) + + assert_raise(ArgumentError) { Regexp.timeout = Float::NAN } + assert_raise(ArgumentError) { Regexp.timeout = 0 } + assert_raise(ArgumentError) { Regexp.timeout = -1 } + + Regexp.timeout = nil + assert_nil(Regexp.timeout) + end; + end + + def test_s_timeout_memory_leak + assert_no_memory_leak([], "#{<<~"begin;"}", "#{<<~"end;"}", "[Bug #20228]", rss: true) + Regexp.timeout = 0.001 + regex = /^(a*)*$/ + str = "a" * 1000000 + "x" + + code = proc do + regex =~ str + rescue + end + + 10.times(&code) + begin; + 1_000.times(&code) + end; + end + + def test_bug_20453 + re = Regexp.new("^(a*)x$", timeout: 0.001) + + assert_raise(Regexp::TimeoutError) do + re =~ "a" * 1000000 + "x" + end + end + + def test_bug_20886 + re = Regexp.new("d()*+|a*a*bc", timeout: 0.02) + assert_raise(Regexp::TimeoutError) do + re === "b" + "a" * 1000 + end + end + + def per_instance_redos_test(global_timeout, per_instance_timeout, expected_timeout) + assert_separately([], "#{<<-"begin;"}\n#{<<-'end;'}", timeout: 60) + global_timeout = #{ EnvUtil.apply_timeout_scale(global_timeout).inspect } + per_instance_timeout = #{ (per_instance_timeout ? EnvUtil.apply_timeout_scale(per_instance_timeout) : nil).inspect } + expected_timeout = #{ EnvUtil.apply_timeout_scale(expected_timeout).inspect } + begin; + Regexp.timeout = global_timeout + + re = Regexp.new("^(a*)\\1b?a*$", timeout: per_instance_timeout) + if per_instance_timeout + assert_in_delta(per_instance_timeout, re.timeout, per_instance_timeout * 2 * Float::EPSILON) + else + assert_nil(re.timeout) + end + + t = Time.now + assert_raise_with_message(Regexp::TimeoutError, "regexp match timeout") do + re =~ "a" * 1000000 + "x" + end + t = Time.now - t + + assert_in_delta(expected_timeout, t, expected_timeout * 3 / 4) + end; + end + + def test_timeout_shorter_than_global + omit "timeout test is too unstable on s390x" if RUBY_PLATFORM =~ /s390x/ + per_instance_redos_test(10, 0.5, 0.5) + end + + def test_timeout_longer_than_global + omit "timeout test is too unstable on s390x" if RUBY_PLATFORM =~ /s390x/ + per_instance_redos_test(0.01, 0.5, 0.5) + end + + def test_timeout_nil + per_instance_redos_test(0.5, nil, 0.5) + end + + def test_timeout_corner_cases + assert_separately([], "#{<<-"begin;"}\n#{<<-'end;'}") + begin; + assert_nil(//.timeout) + + # This is just an implementation detail that users should not depend on: + # If Regexp.timeout is set to a value greater than the value that can be + # represented in the internal representation of timeout, it uses the + # maximum value that can be represented. + assert_equal(((1<<64)-1) / 1000000000.0, Regexp.new("foo", timeout: Float::INFINITY).timeout) + + assert_equal(((1<<64)-1) / 1000000000.0, Regexp.new("foo", timeout: 1e300).timeout) + + assert_raise(ArgumentError) { Regexp.new("foo", timeout: Float::NAN) } + assert_raise(ArgumentError) { Regexp.new("foo", timeout: 0) } + assert_raise(ArgumentError) { Regexp.new("foo", timeout: -1) } + end; + end + + def test_timeout_memory_leak + assert_no_memory_leak([], "#{<<~"begin;"}", "#{<<~'end;'}", "[Bug #20650]", timeout: 100, rss: true) + regex = Regexp.new("^#{"(a*)" * 10_000}x$", timeout: 0.000001) + str = "a" * 1_000_000 + "x" + + code = proc do + regex =~ str + rescue + end + + 10.times(&code) + begin; + 1_000.times(&code) + end; + end + + def test_match_cache_exponential + assert_separately([], "#{<<-"begin;"}\n#{<<-'end;'}") + timeout = #{ EnvUtil.apply_timeout_scale(10).inspect } + begin; + Regexp.timeout = timeout + assert_nil(/^(a*)*$/ =~ "a" * 1000000 + "x") + end; + end + + def test_match_cache_square + assert_separately([], "#{<<-"begin;"}\n#{<<-'end;'}") + timeout = #{ EnvUtil.apply_timeout_scale(10).inspect } + begin; + Regexp.timeout = timeout + assert_nil(/^a*b?a*$/ =~ "a" * 1000000 + "x") + end; + end + + def test_match_cache_atomic + assert_separately([], "#{<<-"begin;"}\n#{<<-'end;'}") + timeout = #{ EnvUtil.apply_timeout_scale(10).inspect } + begin; + Regexp.timeout = timeout + assert_nil(/^a*?(?>a*a*)$/ =~ "a" * 1000000 + "x") + end; + end + + def test_match_cache_atomic_complex + assert_separately([], "#{<<-"begin;"}\n#{<<-'end;'}") + timeout = #{ EnvUtil.apply_timeout_scale(10).inspect } + begin; + Regexp.timeout = timeout + assert_nil(/a*(?>a*)ab/ =~ "a" * 1000000 + "b") + end; + end + + def test_match_cache_positive_look_ahead + assert_separately([], "#{<<-"begin;"}\n#{<<-'end;'}", timeout: 30) + timeout = #{ EnvUtil.apply_timeout_scale(10).inspect } + begin; + Regexp.timeout = timeout + assert_nil(/^a*?(?=a*a*)$/ =~ "a" * 1000000 + "x") + end; + end + + def test_match_cache_positive_look_ahead_complex + assert_separately([], "#{<<-"begin;"}\n#{<<-'end;'}", timeout: 30) + timeout = #{ EnvUtil.apply_timeout_scale(10).inspect } + begin; + Regexp.timeout = timeout + assert_equal(/(?:(?=a*)a)*/ =~ "a" * 1000000, 0) + end; + end + + def test_match_cache_negative_look_ahead + assert_separately([], "#{<<-"begin;"}\n#{<<-'end;'}") + timeout = #{ EnvUtil.apply_timeout_scale(10).inspect } + begin; + Regexp.timeout = timeout + assert_nil(/^a*?(?!a*a*)$/ =~ "a" * 1000000 + "x") + end; + end + + def test_match_cache_positive_look_behind + assert_separately([], "#{<<-"begin;"}\n#{<<-'end;'}") + timeout = #{ EnvUtil.apply_timeout_scale(10).inspect } + begin; + Regexp.timeout = timeout + assert_nil(/(?<=abc|def)(a|a)*$/ =~ "abc" + "a" * 1000000 + "x") + end; + end + + def test_match_cache_negative_look_behind + assert_separately([], "#{<<-"begin;"}\n#{<<-'end;'}") + timeout = #{ EnvUtil.apply_timeout_scale(10).inspect } + begin; + Regexp.timeout = timeout + assert_nil(/(?<!x)(a|a)*$/ =~ "a" * 1000000 + "x") + end; + end + + def test_match_cache_with_peek_optimization + assert_separately([], "#{<<-"begin;"}\n#{<<-'end;'}") + timeout = #{ EnvUtil.apply_timeout_scale(10).inspect } + begin; + Regexp.timeout = timeout + assert_nil(/a+z/ =~ "a" * 1000000 + "xz") + end; + end + + def test_cache_opcodes_initialize + str = 'test1-test2-test3-test4-test_5' + re = '^([0-9a-zA-Z\-/]*){1,256}$' + 100.times do + assert !Regexp.new(re).match?(str) + end + end + + def test_bug_19273 # [Bug #19273] + pattern = /(?:(?:-?b)|(?:-?(?:1_?(?:0_?)*)?0))(?::(?:(?:-?b)|(?:-?(?:1_?(?:0_?)*)?0))){0,3}/ + assert_equal("10:0:0".match(pattern)[0], "10:0:0") + end + + def test_bug_19467 # [Bug #19467] + assert_separately([], "#{<<-"begin;"}\n#{<<-'end;'}") + timeout = #{ EnvUtil.apply_timeout_scale(10).inspect } + begin; + Regexp.timeout = timeout + + assert_nil(/\A.*a.*z\z/ =~ "a" * 1000000 + "y") + end; + end + + def test_bug_19476 # [Bug #19476] + assert_equal("123456789".match(/(?:x?\dx?){2,10}/)[0], "123456789") + assert_equal("123456789".match(/(?:x?\dx?){2,}/)[0], "123456789") + end + + def test_encoding_flags_are_preserved_when_initialized_with_another_regexp + re = Regexp.new("\u2018hello\u2019".encode("UTF-8")) + str = "".encode("US-ASCII") + + assert_nothing_raised do + str.match?(re) + str.match?(Regexp.new(re)) + end + end + + def test_bug_19537 # [Bug #19537] + str = 'aac' + re = '^([ab]{1,3})(a?)*$' + 100.times do + assert !Regexp.new(re).match?(str) + end + end + + def test_bug_20083 # [Bug #20083] + re = /([\s]*ABC)$/i + (1..100).each do |n| + text = "#{"0" * n}ABC" + assert text.match?(re) + end + end + + def test_bug_20098 # [Bug #20098] + assert(/a((.|.)|bc){,4}z/.match? 'abcbcbcbcz') + assert(/a(b+?c*){4,5}z/.match? 'abbbccbbbccbcbcz') + assert(/a(b+?(.|.)){2,3}z/.match? 'abbbcbbbcbbbcz') + assert(/a(b*?(.|.)[bc]){2,5}z/.match? 'abcbbbcbcccbcz') + assert(/^(?:.+){2,4}?b|b/.match? "aaaabaa") + end + + def test_bug_20207 # [Bug #20207] + assert(!'clan'.match?(/(?=.*a)(?!.*n)/)) + end + + def test_bug_20212 # [Bug #20212] + regex = Regexp.new( + /\A((?=.*?[a-z])(?!.*--)[a-z\d]+[a-z\d-]*[a-z\d]+).((?=.*?[a-z])(?!.*--)[a-z\d]+[a-z\d-]*[a-z\d]+).((?=.*?[a-z])(?!.*--)[a-z]+[a-z-]*[a-z]+).((?=.*?[a-z])(?!.*--)[a-z]+[a-z-]*[a-z]+)\Z/x + ) + string = "www.google.com" + 100.times.each { assert(regex.match?(string)) } + end + + def test_bug_20246 # [Bug #20246] + assert_equal '1.2.3', '1.2.3'[/(\d+)(\.\g<1>){2}/] + assert_equal '1.2.3', '1.2.3'[/((?:\d|foo|bar)+)(\.\g<1>){2}/] + end + + def test_linear_time_p + assert_send [Regexp, :linear_time?, /a/] + assert_send [Regexp, :linear_time?, 'a'] + assert_send [Regexp, :linear_time?, 'a', Regexp::IGNORECASE] + assert_not_send [Regexp, :linear_time?, /(a)\1/] + assert_not_send [Regexp, :linear_time?, "(a)\\1"] + + assert_not_send [Regexp, :linear_time?, /(?=(a))/] + assert_not_send [Regexp, :linear_time?, /(?!(a))/] + + assert_raise(TypeError) {Regexp.linear_time?(nil)} + assert_raise(TypeError) {Regexp.linear_time?(Regexp.allocate)} + end + + def test_linear_performance + pre = ->(n) {[Regexp.new("a?" * n + "a" * n), "a" * n]} + assert_linear_performance([10, 29], pre: pre) do |re, s| + re =~ s + end + end + + def test_bug_16145_and_bug_21176_caseinsensitive_small # [Bug#16145] [Bug#21176] + encodings = [Encoding::UTF_8, Encoding::ISO_8859_1] + encodings.each do |enc| + o_acute_lower = "\u00F3".encode(enc) + o_acute_upper = "\u00D3".encode(enc) + assert_match(/[x#{o_acute_lower}]/i, "abc#{o_acute_upper}", "should match o acute case insensitive") + + e_acute_lower = "\u00E9".encode(enc) + e_acute_upper = "\u00C9".encode(enc) + assert_match(/[x#{e_acute_lower}]/i, "CAF#{e_acute_upper}", "should match e acute case insensitive") + end + end + + def test_too_many_range_repeat + source = '(?:foobar){0,100}' * 100000 + assert_raise(RegexpError) { Regexp.new(source) } + assert_raise(SyntaxError) { eval("/#{source}/") } + end + + def test_too_many_null_check + source = '(?:(?:foo)?|(?:bar)?)*' * 100000 + assert_raise(RegexpError) { Regexp.new(source) } + assert_raise(SyntaxError) { eval("/#{source}/") } + end end |