diff options
Diffstat (limited to 'activesupport/test/multibyte_chars_test.rb')
| -rw-r--r-- | activesupport/test/multibyte_chars_test.rb | 88 |
1 files changed, 81 insertions, 7 deletions
diff --git a/activesupport/test/multibyte_chars_test.rb b/activesupport/test/multibyte_chars_test.rb index 680936ded5..0e489c10e1 100644 --- a/activesupport/test/multibyte_chars_test.rb +++ b/activesupport/test/multibyte_chars_test.rb @@ -100,9 +100,14 @@ class MultibyteCharsUTF8BehaviourTest < Test::Unit::TestCase def setup @chars = UNICODE_STRING.dup.mb_chars - # NEWLINE, SPACE, EM SPACE - @whitespace = "\n#{[32, 8195].pack('U*')}" - @whitespace.force_encoding(Encoding::UTF_8) if @whitespace.respond_to?(:force_encoding) + if RUBY_VERSION < '1.9' + # Multibyte support all kinds of whitespace (ie. NEWLINE, SPACE, EM SPACE) + @whitespace = "\n\t#{[32, 8195].pack('U*')}" + else + # Ruby 1.9 only supports basic whitespace + @whitespace = "\n\t ".force_encoding(Encoding::UTF_8) + end + @byte_order_mark = [65279].pack('U') end @@ -163,6 +168,7 @@ class MultibyteCharsUTF8BehaviourTest < Test::Unit::TestCase assert chars('').strip.kind_of?(ActiveSupport::Multibyte.proxy_class) assert chars('').reverse.kind_of?(ActiveSupport::Multibyte.proxy_class) assert chars(' ').slice(0).kind_of?(ActiveSupport::Multibyte.proxy_class) + assert chars('').limit(0).kind_of?(ActiveSupport::Multibyte.proxy_class) assert chars('').upcase.kind_of?(ActiveSupport::Multibyte.proxy_class) assert chars('').downcase.kind_of?(ActiveSupport::Multibyte.proxy_class) assert chars('').capitalize.kind_of?(ActiveSupport::Multibyte.proxy_class) @@ -190,7 +196,9 @@ class MultibyteCharsUTF8BehaviourTest < Test::Unit::TestCase def test_should_return_character_offset_for_regexp_matches assert_nil(@chars =~ /wrong/u) assert_equal 0, (@chars =~ /こ/u) + assert_equal 0, (@chars =~ /こに/u) assert_equal 1, (@chars =~ /に/u) + assert_equal 2, (@chars =~ /ち/u) assert_equal 3, (@chars =~ /わ/u) end @@ -220,10 +228,10 @@ class MultibyteCharsUTF8BehaviourTest < Test::Unit::TestCase assert !@chars.include?('a') end - def test_include_raises_type_error_when_nil_is_passed - assert_raise(TypeError) do - @chars.include?(nil) - end + def test_include_raises_when_nil_is_passed + @chars.include?(nil) + flunk "Expected chars.include?(nil) to raise TypeError or NoMethodError" + rescue Exception => e end def test_index_should_return_character_offset @@ -378,6 +386,17 @@ class MultibyteCharsUTF8BehaviourTest < Test::Unit::TestCase assert_equal 'わちにこ', @chars.reverse end + def test_reverse_should_work_with_normalized_strings + str = 'bös' + reversed_str = 'söb' + assert_equal chars(reversed_str).normalize(:kc), chars(str).normalize(:kc).reverse + assert_equal chars(reversed_str).normalize(:c), chars(str).normalize(:c).reverse + assert_equal chars(reversed_str).normalize(:d), chars(str).normalize(:d).reverse + assert_equal chars(reversed_str).normalize(:kd), chars(str).normalize(:kd).reverse + assert_equal chars(reversed_str).decompose, chars(str).decompose.reverse + assert_equal chars(reversed_str).compose, chars(str).compose.reverse + end + def test_slice_should_take_character_offsets assert_equal nil, ''.mb_chars.slice(0) assert_equal 'こ', @chars.slice(0) @@ -476,6 +495,43 @@ class MultibyteCharsExtrasTest < Test::Unit::TestCase end end + def test_limit_should_not_break_on_blank_strings + example = chars('') + assert_equal example, example.limit(0) + assert_equal example, example.limit(1) + end + + def test_limit_should_work_on_a_multibyte_string + example = chars(UNICODE_STRING) + bytesize = UNICODE_STRING.respond_to?(:bytesize) ? UNICODE_STRING.bytesize : UNICODE_STRING.size + + assert_equal UNICODE_STRING, example.limit(bytesize) + assert_equal '', example.limit(0) + assert_equal '', example.limit(1) + assert_equal 'こ', example.limit(3) + assert_equal 'こに', example.limit(6) + assert_equal 'こに', example.limit(8) + assert_equal 'こにち', example.limit(9) + assert_equal 'こにちわ', example.limit(50) + end + + def test_limit_should_work_on_an_ascii_string + ascii = chars(ASCII_STRING) + assert_equal ASCII_STRING, ascii.limit(ASCII_STRING.length) + assert_equal '', ascii.limit(0) + assert_equal 'o', ascii.limit(1) + assert_equal 'oh', ascii.limit(2) + assert_equal 'ohay', ascii.limit(4) + assert_equal 'ohayo', ascii.limit(50) + end + + def test_limit_should_keep_under_the_specified_byte_limit + example = chars(UNICODE_STRING) + (1..UNICODE_STRING.length).each do |limit| + assert example.limit(limit).to_s.length <= limit + end + end + def test_composition_exclusion_is_set_up_properly # Normalization of DEVANAGARI LETTER QA breaks when composition exclusion isn't used correctly qa = [0x915, 0x93c].pack('U*') @@ -586,3 +642,21 @@ class MultibyteCharsExtrasTest < Test::Unit::TestCase end.pack('U*') end end + +class MultibyteInternalsTest < ActiveSupport::TestCase + include MultibyteTestHelpers + + test "Chars translates a character offset to a byte offset" do + example = chars("Puisque c'était son erreur, il m'a aidé") + [ + [0, 0], + [3, 3], + [12, 11], + [14, 13], + [41, 39] + ].each do |byte_offset, character_offset| + assert_equal character_offset, example.send(:translate_offset, byte_offset), + "Expected byte offset #{byte_offset} to translate to #{character_offset}" + end + end +end |