RubyGems - nokogiri - Versions diffs - 1.8.5 → 1.15.3 - Mend

nokogiri 1.8.5 → 1.15.3

Potentially problematic release.

This version of nokogiri might be problematic. Click here for more details.

Files changed (358) hide show

checksums.yaml +4 -4
data/Gemfile +40 -18
data/LICENSE-DEPENDENCIES.md +1636 -1024
data/LICENSE.md +5 -28
data/README.md +203 -90
data/bin/nokogiri +63 -50
data/dependencies.yml +33 -61
data/ext/nokogiri/depend +38 -358
data/ext/nokogiri/extconf.rb +867 -417
data/ext/nokogiri/gumbo.c +594 -0
data/ext/nokogiri/html4_document.c +165 -0
data/ext/nokogiri/html4_element_description.c +299 -0
data/ext/nokogiri/html4_entity_lookup.c +37 -0
data/ext/nokogiri/html4_sax_parser_context.c +108 -0
data/ext/nokogiri/html4_sax_push_parser.c +95 -0
data/ext/nokogiri/libxml2_backwards_compat.c +121 -0
data/ext/nokogiri/nokogiri.c +251 -105
data/ext/nokogiri/nokogiri.h +215 -90
data/ext/nokogiri/test_global_handlers.c +40 -0
data/ext/nokogiri/xml_attr.c +42 -37
data/ext/nokogiri/xml_attribute_decl.c +22 -22
data/ext/nokogiri/xml_cdata.c +40 -31
data/ext/nokogiri/xml_comment.c +20 -27
data/ext/nokogiri/xml_document.c +401 -237
data/ext/nokogiri/xml_document_fragment.c +13 -17
data/ext/nokogiri/xml_dtd.c +64 -58
data/ext/nokogiri/xml_element_content.c +63 -55
data/ext/nokogiri/xml_element_decl.c +31 -31
data/ext/nokogiri/xml_encoding_handler.c +54 -21
data/ext/nokogiri/xml_entity_decl.c +37 -35
data/ext/nokogiri/xml_entity_reference.c +17 -19
data/ext/nokogiri/xml_namespace.c +136 -62
data/ext/nokogiri/xml_node.c +1387 -678
data/ext/nokogiri/xml_node_set.c +246 -216
data/ext/nokogiri/xml_processing_instruction.c +18 -20
data/ext/nokogiri/xml_reader.c +347 -212
data/ext/nokogiri/xml_relax_ng.c +86 -77
data/ext/nokogiri/xml_sax_parser.c +149 -124
data/ext/nokogiri/xml_sax_parser_context.c +145 -103
data/ext/nokogiri/xml_sax_push_parser.c +64 -36
data/ext/nokogiri/xml_schema.c +138 -81
data/ext/nokogiri/xml_syntax_error.c +42 -21
data/ext/nokogiri/xml_text.c +36 -26
data/ext/nokogiri/xml_xpath_context.c +366 -178
data/ext/nokogiri/xslt_stylesheet.c +335 -189
data/gumbo-parser/CHANGES.md +63 -0
data/gumbo-parser/Makefile +111 -0
data/gumbo-parser/THANKS +27 -0
data/gumbo-parser/src/Makefile +34 -0
data/gumbo-parser/src/README.md +41 -0
data/gumbo-parser/src/ascii.c +75 -0
data/gumbo-parser/src/ascii.h +115 -0
data/gumbo-parser/src/attribute.c +42 -0
data/gumbo-parser/src/attribute.h +17 -0
data/gumbo-parser/src/char_ref.c +22225 -0
data/gumbo-parser/src/char_ref.h +29 -0
data/gumbo-parser/src/char_ref.rl +2154 -0
data/gumbo-parser/src/error.c +630 -0
data/gumbo-parser/src/error.h +148 -0
data/gumbo-parser/src/foreign_attrs.c +103 -0
data/gumbo-parser/src/foreign_attrs.gperf +27 -0
data/gumbo-parser/src/insertion_mode.h +33 -0
data/gumbo-parser/src/macros.h +91 -0
data/gumbo-parser/src/nokogiri_gumbo.h +944 -0
data/gumbo-parser/src/parser.c +4891 -0
data/gumbo-parser/src/parser.h +41 -0
data/gumbo-parser/src/replacement.h +33 -0
data/gumbo-parser/src/string_buffer.c +103 -0
data/gumbo-parser/src/string_buffer.h +68 -0
data/gumbo-parser/src/string_piece.c +48 -0
data/gumbo-parser/src/svg_attrs.c +174 -0
data/gumbo-parser/src/svg_attrs.gperf +77 -0
data/gumbo-parser/src/svg_tags.c +137 -0
data/gumbo-parser/src/svg_tags.gperf +55 -0
data/gumbo-parser/src/tag.c +223 -0
data/gumbo-parser/src/tag_lookup.c +382 -0
data/gumbo-parser/src/tag_lookup.gperf +170 -0
data/gumbo-parser/src/tag_lookup.h +13 -0
data/gumbo-parser/src/token_buffer.c +79 -0
data/gumbo-parser/src/token_buffer.h +71 -0
data/gumbo-parser/src/token_type.h +17 -0
data/gumbo-parser/src/tokenizer.c +3463 -0
data/gumbo-parser/src/tokenizer.h +112 -0
data/gumbo-parser/src/tokenizer_states.h +339 -0
data/gumbo-parser/src/utf8.c +245 -0
data/gumbo-parser/src/utf8.h +164 -0
data/gumbo-parser/src/util.c +66 -0
data/gumbo-parser/src/util.h +34 -0
data/gumbo-parser/src/vector.c +111 -0
data/gumbo-parser/src/vector.h +45 -0
data/lib/nokogiri/class_resolver.rb +67 -0
data/lib/nokogiri/css/node.rb +10 -8
data/lib/nokogiri/css/parser.rb +397 -377
data/lib/nokogiri/css/parser.y +250 -245
data/lib/nokogiri/css/parser_extras.rb +54 -49
data/lib/nokogiri/css/syntax_error.rb +3 -1
data/lib/nokogiri/css/tokenizer.rb +107 -104
data/lib/nokogiri/css/tokenizer.rex +3 -2
data/lib/nokogiri/css/xpath_visitor.rb +224 -95
data/lib/nokogiri/css.rb +56 -17
data/lib/nokogiri/decorators/slop.rb +9 -7
data/lib/nokogiri/encoding_handler.rb +57 -0
data/lib/nokogiri/extension.rb +32 -0
data/lib/nokogiri/gumbo.rb +15 -0
data/lib/nokogiri/html.rb +38 -27
data/lib/nokogiri/{html → html4}/builder.rb +4 -2
data/lib/nokogiri/html4/document.rb +214 -0
data/lib/nokogiri/html4/document_fragment.rb +54 -0
data/lib/nokogiri/{html → html4}/element_description.rb +3 -1
data/lib/nokogiri/html4/element_description_defaults.rb +2040 -0
data/lib/nokogiri/html4/encoding_reader.rb +121 -0
data/lib/nokogiri/{html → html4}/entity_lookup.rb +4 -2
data/lib/nokogiri/{html → html4}/sax/parser.rb +17 -16
data/lib/nokogiri/html4/sax/parser_context.rb +20 -0
data/lib/nokogiri/{html → html4}/sax/push_parser.rb +12 -11
data/lib/nokogiri/html4.rb +47 -0
data/lib/nokogiri/html5/document.rb +168 -0
data/lib/nokogiri/html5/document_fragment.rb +90 -0
data/lib/nokogiri/html5/node.rb +103 -0
data/lib/nokogiri/html5.rb +392 -0
data/lib/nokogiri/jruby/dependencies.rb +3 -0
data/lib/nokogiri/jruby/nokogiri_jars.rb +43 -0
data/lib/nokogiri/syntax_error.rb +2 -0
data/lib/nokogiri/version/constant.rb +6 -0
data/lib/nokogiri/version/info.rb +223 -0
data/lib/nokogiri/version.rb +3 -108
data/lib/nokogiri/xml/attr.rb +55 -3
data/lib/nokogiri/xml/attribute_decl.rb +6 -2
data/lib/nokogiri/xml/builder.rb +98 -54
data/lib/nokogiri/xml/cdata.rb +3 -1
data/lib/nokogiri/xml/character_data.rb +2 -0
data/lib/nokogiri/xml/document.rb +312 -126
data/lib/nokogiri/xml/document_fragment.rb +104 -48
data/lib/nokogiri/xml/dtd.rb +4 -2
data/lib/nokogiri/xml/element_content.rb +12 -2
data/lib/nokogiri/xml/element_decl.rb +6 -2
data/lib/nokogiri/xml/entity_decl.rb +7 -3
data/lib/nokogiri/xml/entity_reference.rb +2 -0
data/lib/nokogiri/xml/namespace.rb +45 -0
data/lib/nokogiri/xml/node/save_options.rb +23 -8
data/lib/nokogiri/xml/node.rb +1093 -411
data/lib/nokogiri/xml/node_set.rb +173 -67
data/lib/nokogiri/xml/notation.rb +13 -0
data/lib/nokogiri/xml/parse_options.rb +145 -52
data/lib/nokogiri/xml/pp/character_data.rb +9 -6
data/lib/nokogiri/xml/pp/node.rb +42 -30
data/lib/nokogiri/xml/pp.rb +4 -2
data/lib/nokogiri/xml/processing_instruction.rb +4 -1
data/lib/nokogiri/xml/reader.rb +21 -28
data/lib/nokogiri/xml/relax_ng.rb +8 -2
data/lib/nokogiri/xml/sax/document.rb +45 -49
data/lib/nokogiri/xml/sax/parser.rb +39 -36
data/lib/nokogiri/xml/sax/parser_context.rb +8 -3
data/lib/nokogiri/xml/sax/push_parser.rb +6 -5
data/lib/nokogiri/xml/sax.rb +6 -4
data/lib/nokogiri/xml/schema.rb +19 -9
data/lib/nokogiri/xml/searchable.rb +120 -72
data/lib/nokogiri/xml/syntax_error.rb +6 -4
data/lib/nokogiri/xml/text.rb +2 -0
data/lib/nokogiri/xml/xpath/syntax_error.rb +4 -2
data/lib/nokogiri/xml/xpath.rb +15 -4
data/lib/nokogiri/xml/xpath_context.rb +3 -3
data/lib/nokogiri/xml.rb +38 -37
data/lib/nokogiri/xslt/stylesheet.rb +3 -1
data/lib/nokogiri/xslt.rb +101 -22
data/lib/nokogiri.rb +59 -75
data/lib/xsd/xmlparser/nokogiri.rb +29 -25
data/patches/libxml2/0001-Remove-script-macro-support.patch +40 -0
data/patches/libxml2/0002-Update-entities-to-remove-handling-of-ssi.patch +44 -0
data/patches/libxml2/0003-libxml2.la-is-in-top_builddir.patch +25 -0
data/patches/libxml2/0009-allow-wildcard-namespaces.patch +77 -0
data/patches/libxml2/0010-update-config.guess-and-config.sub-for-libxml2.patch +224 -0
data/patches/libxml2/0011-rip-out-libxml2-s-libc_single_threaded-support.patch +30 -0
data/patches/libxslt/0001-update-config.guess-and-config.sub-for-libxslt.patch +224 -0
data/ports/archives/libxml2-2.11.4.tar.xz +0 -0
data/ports/archives/libxslt-1.1.38.tar.xz +0 -0
metadata +126 -399
data/.autotest +0 -22
data/.cross_rubies +0 -8
data/.editorconfig +0 -17
data/.gemtest +0 -0
data/.travis.yml +0 -63
data/CHANGELOG.md +0 -1368
data/CONTRIBUTING.md +0 -42
data/C_CODING_STYLE.rdoc +0 -33
data/Gemfile-libxml-ruby +0 -3
data/Manifest.txt +0 -370
data/ROADMAP.md +0 -111
data/Rakefile +0 -348
data/SECURITY.md +0 -19
data/STANDARD_RESPONSES.md +0 -47
data/Y_U_NO_GEMSPEC.md +0 -155
data/appveyor.yml +0 -29
data/build_all +0 -44
data/ext/nokogiri/html_document.c +0 -170
data/ext/nokogiri/html_document.h +0 -10
data/ext/nokogiri/html_element_description.c +0 -279
data/ext/nokogiri/html_element_description.h +0 -10
data/ext/nokogiri/html_entity_lookup.c +0 -32
data/ext/nokogiri/html_entity_lookup.h +0 -8
data/ext/nokogiri/html_sax_parser_context.c +0 -116
data/ext/nokogiri/html_sax_parser_context.h +0 -11
data/ext/nokogiri/html_sax_push_parser.c +0 -87
data/ext/nokogiri/html_sax_push_parser.h +0 -9
data/ext/nokogiri/xml_attr.h +0 -9
data/ext/nokogiri/xml_attribute_decl.h +0 -9
data/ext/nokogiri/xml_cdata.h +0 -9
data/ext/nokogiri/xml_comment.h +0 -9
data/ext/nokogiri/xml_document.h +0 -23
data/ext/nokogiri/xml_document_fragment.h +0 -10
data/ext/nokogiri/xml_dtd.h +0 -10
data/ext/nokogiri/xml_element_content.h +0 -10
data/ext/nokogiri/xml_element_decl.h +0 -9
data/ext/nokogiri/xml_encoding_handler.h +0 -8
data/ext/nokogiri/xml_entity_decl.h +0 -10
data/ext/nokogiri/xml_entity_reference.h +0 -9
data/ext/nokogiri/xml_io.c +0 -61
data/ext/nokogiri/xml_io.h +0 -11
data/ext/nokogiri/xml_libxml2_hacks.c +0 -112
data/ext/nokogiri/xml_libxml2_hacks.h +0 -12
data/ext/nokogiri/xml_namespace.h +0 -15
data/ext/nokogiri/xml_node.h +0 -13
data/ext/nokogiri/xml_node_set.h +0 -12
data/ext/nokogiri/xml_processing_instruction.h +0 -9
data/ext/nokogiri/xml_reader.h +0 -10
data/ext/nokogiri/xml_relax_ng.h +0 -9
data/ext/nokogiri/xml_sax_parser.h +0 -39
data/ext/nokogiri/xml_sax_parser_context.h +0 -10
data/ext/nokogiri/xml_sax_push_parser.h +0 -9
data/ext/nokogiri/xml_schema.h +0 -9
data/ext/nokogiri/xml_syntax_error.h +0 -13
data/ext/nokogiri/xml_text.h +0 -9
data/ext/nokogiri/xml_xpath_context.h +0 -10
data/ext/nokogiri/xslt_stylesheet.h +0 -14
data/lib/nokogiri/html/document.rb +0 -335
data/lib/nokogiri/html/document_fragment.rb +0 -49
data/lib/nokogiri/html/element_description_defaults.rb +0 -671
data/lib/nokogiri/html/sax/parser_context.rb +0 -16
data/patches/libxml2/0001-Revert-Do-not-URI-escape-in-server-side-includes.patch +0 -78
data/patches/libxml2/0002-Fix-nullptr-deref-with-XPath-logic-ops.patch +0 -54
data/patches/libxml2/0003-Fix-infinite-loop-in-LZMA-decompression.patch +0 -50
data/patches/sort-patches-by-date +0 -25
data/ports/archives/libxml2-2.9.8.tar.gz +0 -0
data/ports/archives/libxslt-1.1.32.tar.gz +0 -0
data/suppressions/README.txt +0 -1
data/suppressions/nokogiri_ruby-2.supp +0 -10
data/tasks/test.rb +0 -100
data/test/css/test_nthiness.rb +0 -226
data/test/css/test_parser.rb +0 -386
data/test/css/test_tokenizer.rb +0 -215
data/test/css/test_xpath_visitor.rb +0 -96
data/test/decorators/test_slop.rb +0 -23
data/test/files/2ch.html +0 -108
data/test/files/GH_1042.html +0 -18
data/test/files/address_book.rlx +0 -12
data/test/files/address_book.xml +0 -10
data/test/files/atom.xml +0 -344
data/test/files/bar/bar.xsd +0 -4
data/test/files/bogus.xml +0 -0
data/test/files/dont_hurt_em_why.xml +0 -422
data/test/files/encoding.html +0 -82
data/test/files/encoding.xhtml +0 -84
data/test/files/exslt.xml +0 -8
data/test/files/exslt.xslt +0 -35
data/test/files/foo/foo.xsd +0 -4
data/test/files/metacharset.html +0 -10
data/test/files/namespace_pressure_test.xml +0 -1684
data/test/files/noencoding.html +0 -47
data/test/files/po.xml +0 -32
data/test/files/po.xsd +0 -66
data/test/files/saml/saml20assertion_schema.xsd +0 -283
data/test/files/saml/saml20protocol_schema.xsd +0 -302
data/test/files/saml/xenc_schema.xsd +0 -146
data/test/files/saml/xmldsig_schema.xsd +0 -318
data/test/files/shift_jis.html +0 -10
data/test/files/shift_jis.xml +0 -5
data/test/files/shift_jis_no_charset.html +0 -9
data/test/files/slow-xpath.xml +0 -25509
data/test/files/snuggles.xml +0 -3
data/test/files/staff.dtd +0 -10
data/test/files/staff.xml +0 -59
data/test/files/staff.xslt +0 -32
data/test/files/test_document_url/bar.xml +0 -2
data/test/files/test_document_url/document.dtd +0 -4
data/test/files/test_document_url/document.xml +0 -6
data/test/files/tlm.html +0 -851
data/test/files/to_be_xincluded.xml +0 -2
data/test/files/valid_bar.xml +0 -2
data/test/files/xinclude.xml +0 -4
data/test/helper.rb +0 -271
data/test/html/sax/test_parser.rb +0 -168
data/test/html/sax/test_parser_context.rb +0 -46
data/test/html/sax/test_parser_text.rb +0 -163
data/test/html/sax/test_push_parser.rb +0 -87
data/test/html/test_attributes.rb +0 -85
data/test/html/test_builder.rb +0 -164
data/test/html/test_document.rb +0 -712
data/test/html/test_document_encoding.rb +0 -143
data/test/html/test_document_fragment.rb +0 -310
data/test/html/test_element_description.rb +0 -105
data/test/html/test_named_characters.rb +0 -14
data/test/html/test_node.rb +0 -212
data/test/html/test_node_encoding.rb +0 -91
data/test/namespaces/test_additional_namespaces_in_builder_doc.rb +0 -14
data/test/namespaces/test_namespaces_aliased_default.rb +0 -24
data/test/namespaces/test_namespaces_in_builder_doc.rb +0 -75
data/test/namespaces/test_namespaces_in_cloned_doc.rb +0 -31
data/test/namespaces/test_namespaces_in_created_doc.rb +0 -75
data/test/namespaces/test_namespaces_in_parsed_doc.rb +0 -80
data/test/namespaces/test_namespaces_preservation.rb +0 -31
data/test/test_convert_xpath.rb +0 -135
data/test/test_css_cache.rb +0 -47
data/test/test_encoding_handler.rb +0 -48
data/test/test_memory_leak.rb +0 -156
data/test/test_nokogiri.rb +0 -138
data/test/test_soap4r_sax.rb +0 -52
data/test/test_xslt_transforms.rb +0 -314
data/test/xml/node/test_save_options.rb +0 -28
data/test/xml/node/test_subclass.rb +0 -44
data/test/xml/sax/test_parser.rb +0 -402
data/test/xml/sax/test_parser_context.rb +0 -115
data/test/xml/sax/test_parser_text.rb +0 -202
data/test/xml/sax/test_push_parser.rb +0 -265
data/test/xml/test_attr.rb +0 -74
data/test/xml/test_attribute_decl.rb +0 -86
data/test/xml/test_builder.rb +0 -341
data/test/xml/test_c14n.rb +0 -180
data/test/xml/test_cdata.rb +0 -54
data/test/xml/test_comment.rb +0 -40
data/test/xml/test_document.rb +0 -982
data/test/xml/test_document_encoding.rb +0 -31
data/test/xml/test_document_fragment.rb +0 -298
data/test/xml/test_dtd.rb +0 -187
data/test/xml/test_dtd_encoding.rb +0 -31
data/test/xml/test_element_content.rb +0 -56
data/test/xml/test_element_decl.rb +0 -73
data/test/xml/test_entity_decl.rb +0 -122
data/test/xml/test_entity_reference.rb +0 -262
data/test/xml/test_namespace.rb +0 -96
data/test/xml/test_node.rb +0 -1325
data/test/xml/test_node_attributes.rb +0 -115
data/test/xml/test_node_encoding.rb +0 -75
data/test/xml/test_node_inheritance.rb +0 -32
data/test/xml/test_node_reparenting.rb +0 -592
data/test/xml/test_node_set.rb +0 -809
data/test/xml/test_parse_options.rb +0 -64
data/test/xml/test_processing_instruction.rb +0 -30
data/test/xml/test_reader.rb +0 -620
data/test/xml/test_reader_encoding.rb +0 -134
data/test/xml/test_relax_ng.rb +0 -60
data/test/xml/test_schema.rb +0 -142
data/test/xml/test_syntax_error.rb +0 -36
data/test/xml/test_text.rb +0 -60
data/test/xml/test_unparented_node.rb +0 -483
data/test/xml/test_xinclude.rb +0 -83
data/test/xml/test_xpath.rb +0 -470
data/test/xslt/test_custom_functions.rb +0 -133
data/test/xslt/test_exception_handling.rb +0 -37

data/lib/nokogiri/html/document.rb DELETED Viewed

@@ -1,335 +0,0 @@
-module Nokogiri
-  module HTML
-    class Document < Nokogiri::XML::Document
-      ###
-      # Get the meta tag encoding for this document.  If there is no meta tag,
-      # then nil is returned.
-      def meta_encoding
-        case
-        when meta = at('//meta[@charset]')
-          meta[:charset]
-        when meta = meta_content_type
-          meta['content'][/charset\s*=\s*([\w-]+)/i, 1]
-        end
-      end
-      ###
-      # Set the meta tag encoding for this document.
-      #
-      # If an meta encoding tag is already present, its content is
-      # replaced with the given text.
-      #
-      # Otherwise, this method tries to create one at an appropriate
-      # place supplying head and/or html elements as necessary, which
-      # is inside a head element if any, and before any text node or
-      # content element (typically <body>) if any.
-      #
-      # The result when trying to set an encoding that is different
-      # from the document encoding is undefined.
-      #
-      # Beware in CRuby, that libxml2 automatically inserts a meta tag
-      # into a head element.
-      def meta_encoding= encoding
-        case
-        when meta = meta_content_type
-          meta['content'] = 'text/html; charset=%s' % encoding
-          encoding
-        when meta = at('//meta[@charset]')
-          meta['charset'] = encoding
-        else
-          meta = XML::Node.new('meta', self)
-          if dtd = internal_subset and dtd.html5_dtd?
-            meta['charset'] = encoding
-          else
-            meta['http-equiv'] = 'Content-Type'
-            meta['content'] = 'text/html; charset=%s' % encoding
-          end
-          case
-          when head = at('//head')
-            head.prepend_child(meta)
-          else
-            set_metadata_element(meta)
-          end
-          encoding
-        end
-      end
-      def meta_content_type
-        xpath('//meta[@http-equiv and boolean(@content)]').find { |node|
-          node['http-equiv'] =~ /\AContent-Type\z/i
-        }
-      end
-      private :meta_content_type
-      ###
-      # Get the title string of this document.  Return nil if there is
-      # no title tag.
-      def title
-        title = at('//title') and title.inner_text
-      end
-      ###
-      # Set the title string of this document.
-      #
-      # If a title element is already present, its content is replaced
-      # with the given text.
-      #
-      # Otherwise, this method tries to create one at an appropriate
-      # place supplying head and/or html elements as necessary, which
-      # is inside a head element if any, right after a meta
-      # encoding/charset tag if any, and before any text node or
-      # content element (typically <body>) if any.
-      def title=(text)
-        tnode = XML::Text.new(text, self)
-        if title = at('//title')
-          title.children = tnode
-          return text
-        end
-        title = XML::Node.new('title', self) << tnode
-        case
-        when head = at('//head')
-          head << title
-        when meta = at('//meta[@charset]') || meta_content_type
-          # better put after charset declaration
-          meta.add_next_sibling(title)
-        else
-          set_metadata_element(title)
-        end
-        text
-      end
-      def set_metadata_element(element)
-        case
-        when head = at('//head')
-          head << element
-        when html = at('//html')
-          head = html.prepend_child(XML::Node.new('head', self))
-          head.prepend_child(element)
-        when first = children.find { |node|
-            case node
-            when XML::Element, XML::Text
-              true
-            end
-          }
-          # We reach here only if the underlying document model
-          # allows <html>/<head> elements to be omitted and does not
-          # automatically supply them.
-          first.add_previous_sibling(element)
-        else
-          html = add_child(XML::Node.new('html', self))
-          head = html.add_child(XML::Node.new('head', self))
-          head.prepend_child(element)
-        end
-      end
-      private :set_metadata_element
-      ####
-      # Serialize Node using +options+.  Save options can also be set using a
-      # block. See SaveOptions.
-      #
-      # These two statements are equivalent:
-      #
-      #  node.serialize(:encoding => 'UTF-8', :save_with => FORMAT | AS_XML)
-      #
-      # or
-      #
-      #   node.serialize(:encoding => 'UTF-8') do |config|
-      #     config.format.as_xml
-      #   end
-      #
-      def serialize options = {}
-        options[:save_with] ||= XML::Node::SaveOptions::DEFAULT_HTML
-        super
-      end
-      ####
-      # Create a Nokogiri::XML::DocumentFragment from +tags+
-      def fragment tags = nil
-        DocumentFragment.new(self, tags, self.root)
-      end
-      class << self
-        ###
-        # Parse HTML.  +string_or_io+ may be a String, or any object that
-        # responds to _read_ and _close_ such as an IO, or StringIO.
-        # +url+ is resource where this document is located.  +encoding+ is the
-        # encoding that should be used when processing the document. +options+
-        # is a number that sets options in the parser, such as
-        # Nokogiri::XML::ParseOptions::RECOVER.  See the constants in
-        # Nokogiri::XML::ParseOptions.
-        def parse string_or_io, url = nil, encoding = nil, options = XML::ParseOptions::DEFAULT_HTML
-          options = Nokogiri::XML::ParseOptions.new(options) if Integer === options
-          # Give the options to the user
-          yield options if block_given?
-          if string_or_io.respond_to?(:encoding)
-            unless string_or_io.encoding.name == "ASCII-8BIT"
-              encoding ||= string_or_io.encoding.name
-            end
-          end
-          if string_or_io.respond_to?(:read)
-            url ||= string_or_io.respond_to?(:path) ? string_or_io.path : nil
-            unless encoding
-              # Libxml2's parser has poor support for encoding
-              # detection.  First, it does not recognize the HTML5
-              # style meta charset declaration.  Secondly, even if it
-              # successfully detects an encoding hint, it does not
-              # re-decode or re-parse the preceding part which may be
-              # garbled.
-              #
-              # EncodingReader aims to perform advanced encoding
-              # detection beyond what Libxml2 does, and to emulate
-              # rewinding of a stream and make Libxml2 redo parsing
-              # from the start when an encoding hint is found.
-              string_or_io = EncodingReader.new(string_or_io)
-              begin
-                return read_io(string_or_io, url, encoding, options.to_i)
-              rescue EncodingFound => e
-                encoding = e.found_encoding
-              end
-            end
-            return read_io(string_or_io, url, encoding, options.to_i)
-          end
-          # read_memory pukes on empty docs
-          if string_or_io.nil? or string_or_io.empty?
-            return encoding ? new.tap { |i| i.encoding = encoding } : new
-          end
-          encoding ||= EncodingReader.detect_encoding(string_or_io)
-          read_memory(string_or_io, url, encoding, options.to_i)
-        end
-      end
-      class EncodingFound < StandardError # :nodoc:
-        attr_reader :found_encoding
-        def initialize(encoding)
-          @found_encoding = encoding
-          super("encoding found: %s" % encoding)
-        end
-      end
-      class EncodingReader # :nodoc:
-        class SAXHandler < Nokogiri::XML::SAX::Document # :nodoc:
-          attr_reader :encoding
-          def initialize
-            @encoding = nil
-            super()
-          end
-          def start_element(name, attrs = [])
-            return unless name == 'meta'
-            attr = Hash[attrs]
-            charset = attr['charset'] and
-              @encoding = charset
-            http_equiv = attr['http-equiv'] and
-              http_equiv.match(/\AContent-Type\z/i) and
-              content = attr['content'] and
-              m = content.match(/;\s*charset\s*=\s*([\w-]+)/) and
-              @encoding = m[1]
-          end
-        end
-        class JumpSAXHandler < SAXHandler
-          def initialize(jumptag)
-            @jumptag = jumptag
-            super()
-          end
-          def start_element(name, attrs = [])
-            super
-            throw @jumptag, @encoding if @encoding
-            throw @jumptag, nil if name =~ /\A(?:div|h1|img|p|br)\z/
-          end
-        end
-        def self.detect_encoding(chunk)
-          if Nokogiri.jruby? && EncodingReader.is_jruby_without_fix?
-            return EncodingReader.detect_encoding_for_jruby_without_fix(chunk)
-          end
-          m = chunk.match(/\A(<\?xml[ \t\r\n]+[^>]*>)/) and
-            return Nokogiri.XML(m[1]).encoding
-          if Nokogiri.jruby?
-            m = chunk.match(/(<meta\s)(.*)(charset\s*=\s*([\w-]+))(.*)/i) and
-              return m[4]
-            catch(:encoding_found) {
-              Nokogiri::HTML::SAX::Parser.new(JumpSAXHandler.new(:encoding_found)).parse(chunk)
-              nil
-            }
-          else
-            handler = SAXHandler.new
-            parser = Nokogiri::HTML::SAX::PushParser.new(handler)
-            parser << chunk rescue Nokogiri::SyntaxError
-            handler.encoding
-          end
-        end
-        def self.is_jruby_without_fix?
-          JRUBY_VERSION.split('.').join.to_i < 165
-        end
-        def self.detect_encoding_for_jruby_without_fix(chunk)
-          m = chunk.match(/\A(<\?xml[ \t\r\n]+[^>]*>)/) and
-            return Nokogiri.XML(m[1]).encoding
-          m = chunk.match(/(<meta\s)(.*)(charset\s*=\s*([\w-]+))(.*)/i) and
-            return m[4]
-          catch(:encoding_found) {
-            Nokogiri::HTML::SAX::Parser.new(JumpSAXHandler.new(:encoding_found.to_s)).parse(chunk)
-            nil
-          }
-        rescue Nokogiri::SyntaxError, RuntimeError
-          # Ignore parser errors that nokogiri may raise
-          nil
-        end
-        def initialize(io)
-          @io = io
-          @firstchunk = nil
-          @encoding_found = nil
-        end
-        # This method is used by the C extension so that
-        # Nokogiri::HTML::Document#read_io() does not leak memory when
-        # EncodingFound is raised.
-        attr_reader :encoding_found
-        def read(len)
-          # no support for a call without len
-          if !@firstchunk
-            @firstchunk = @io.read(len) or return nil
-            # This implementation expects that the first call from
-            # htmlReadIO() is made with a length long enough (~1KB) to
-            # achieve advanced encoding detection.
-            if encoding = EncodingReader.detect_encoding(@firstchunk)
-              # The first chunk is stored for the next read in retry.
-              raise @encoding_found = EncodingFound.new(encoding)
-            end
-          end
-          @encoding_found = nil
-          ret = @firstchunk.slice!(0, len)
-          if (len -= ret.length) > 0
-            rest = @io.read(len) and ret << rest
-          end
-          if ret.empty?
-            nil
-          else
-            ret
-          end
-        end
-      end
-    end
-  end
-end

data/lib/nokogiri/html/document_fragment.rb DELETED Viewed

@@ -1,49 +0,0 @@
-module Nokogiri
-  module HTML
-    class DocumentFragment < Nokogiri::XML::DocumentFragment
-      ####
-      # Create a Nokogiri::XML::DocumentFragment from +tags+, using +encoding+
-      def self.parse tags, encoding = nil
-        doc = HTML::Document.new
-        encoding ||= if tags.respond_to?(:encoding)
-                       encoding = tags.encoding
-                       if encoding == ::Encoding::ASCII_8BIT
-                         'UTF-8'
-                       else
-                         encoding.name
-                       end
-                     else
-                       'UTF-8'
-                     end
-        doc.encoding = encoding
-        new(doc, tags)
-      end
-      def initialize document, tags = nil, ctx = nil
-        return self unless tags
-        if ctx
-          preexisting_errors = document.errors.dup
-          node_set = ctx.parse("<div>#{tags}</div>")
-          node_set.first.children.each { |child| child.parent = self } unless node_set.empty?
-          self.errors = document.errors - preexisting_errors
-        else
-          # This is a horrible hack, but I don't care
-          if tags.strip =~ /^<body/i
-            path = "/html/body"
-          else
-            path = "/html/body/node()"
-          end
-          temp_doc = HTML::Document.parse "<html><body>#{tags}", nil, document.encoding
-          temp_doc.xpath(path).each { |child| child.parent = self }
-          self.errors = temp_doc.errors
-        end
-        children
-      end
-    end
-  end
-end