Module: RDF::RDFa::Reader::Nokogiri
- Defined in:
- vendor/bundler/ruby/3.3.0/bundler/gems/rdf-rdfa-ea6265716853/lib/rdf/rdfa/reader/nokogiri.rb
Overview
Nokogiri implementation of an XML parser.
Defined Under Namespace
Classes: NodeProxy, NodeSetProxy
Class Method Summary collapse
-
.library ⇒ Symbol
Returns the name of the underlying XML library.
Instance Method Summary collapse
-
#detect_host_language_version(input, **options) ⇒ Object
Determine the host language and/or version from options and the input document.
-
#doc_base(base) ⇒ String
Find value of document base.
-
#doc_errors ⇒ Object
Document errors.
-
#initialize_xml(input, **options)
Initializes the underlying XML library.
-
#root ⇒ Object
Return proxy for document root.
Class Method Details
.library ⇒ Symbol
Returns the name of the underlying XML library.
12 13 14 |
# File 'vendor/bundler/ruby/3.3.0/bundler/gems/rdf-rdfa-ea6265716853/lib/rdf/rdfa/reader/nokogiri.rb', line 12 def self.library :nokogiri end |
Instance Method Details
#detect_host_language_version(input, **options) ⇒ Object
Determine the host language and/or version from options and the input document
191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 |
# File 'vendor/bundler/ruby/3.3.0/bundler/gems/rdf-rdfa-ea6265716853/lib/rdf/rdfa/reader/nokogiri.rb', line 191 def detect_host_language_version(input, **) @host_language = [:host_language] ? [:host_language].to_sym : nil @version = [:version] ? [:version].to_sym : nil return if @host_language && @version # Snif version based on input case input when ::Nokogiri::XML::Document, ::Nokogiri::HTML::Document doc_type_string = input.children.detect {|c| c.is_a?(::Nokogiri::XML::DTD)} version_attr = input.root && input.root.attribute("version").to_s root_element = input.root.name.downcase content_type = case when root_element == "html" && input.is_a?(::Nokogiri::HTML::Document) "text/html" when root_element == "html" && input.is_a?(::Nokogiri::XML::Document) "application/xhtml+html" end else content_type = input.content_type if input.respond_to?(:content_type) # Determine from head of document head = if input.respond_to?(:read) input.rewind string = input.read(1000) input.rewind string.to_s else input.to_s[0..1000] end doc_type_string = head.match(%r(<!DOCTYPE[^>]*>)m).to_s root = head.match(%r(<[^!\?>]*>)m).to_s root_element = root.match(%r(^<(\S+)[ >])) ? $1 : "" version_attr = root.match(/version\s*=\s*"([^"]+)"/m) ? $1 : "" head_element = head.match(%r(<head.*<\/head>)mi) head_doc = ::Nokogiri::HTML.parse(head_element.to_s) # May determine content-type and/or charset from meta # Easist way is to parse head into a document and iterate # of CSS matches head_doc.css("meta").each do |e| if e.attr("http-equiv").to_s.downcase == 'content-type' content_type, e = e.attr("content").to_s.downcase.split(";") [:encoding] = $1.downcase if e.to_s =~ /charset=([^\s]*)$/i elsif e.attr("charset") [:encoding] = e.attr("charset").to_s.downcase end end end # Already using XML parser, determine from DOCTYPE and/or root element @version ||= :"rdfa1.0" if doc_type_string =~ /RDFa 1\.0/ @version ||= :"rdfa1.0" if version_attr =~ /RDFa 1\.0/ @version ||= :"rdfa1.1" if version_attr =~ /RDFa 1\.1/ @version ||= :"rdfa1.1" @host_language ||= :xhtml1 if @version == :"rdfa1.0" && doc_type_string =~ /html/i @host_language ||= case content_type when "application/xml" then :xml when "image/svg+xml" then :svg when "text/html" case doc_type_string when /html 4/i then :html4 when /xhtml/i then :xhtml1 when /html/i then :html5 else :html5 end when "application/xhtml+xml" case doc_type_string when /html 4/i then :html4 when /xhtml/i then :xhtml1 when /html/i then :xhtml5 else :xhtml5 end else case root_element when /svg/i then :svg else :html5 end end end |
#doc_base(base) ⇒ String
Find value of document base
299 300 301 302 303 304 305 306 307 308 309 310 311 |
# File 'vendor/bundler/ruby/3.3.0/bundler/gems/rdf-rdfa-ea6265716853/lib/rdf/rdfa/reader/nokogiri.rb', line 299 def doc_base(base) # find if the document has a base element case @host_language when :xhtml1, :xhtml5, :html4, :html5 base_el = @doc.at_css("html>head>base") base = base.join(base_el.attribute("href").to_s.split("#").first) if base_el else xml_base = root.attribute_with_ns("base", "http://www.w3.org/XML/1998/namespace") || root.attribute('xml:base') if root base = base.join(xml_base) if xml_base end base end |
#doc_errors ⇒ Object
Document errors
284 285 286 287 288 289 290 291 292 |
# File 'vendor/bundler/ruby/3.3.0/bundler/gems/rdf-rdfa-ea6265716853/lib/rdf/rdfa/reader/nokogiri.rb', line 284 def doc_errors if @host_language == :html5 @doc.errors.reject do |e| e.to_s =~ %r{(The doctype must be the first token in the document)|(Expected a doctype token)|(Unexpected '\?' where start tag name is expected)} end else @doc.errors.reject {|e| e.to_s =~ /(?:Tag \w+ invalid)|(?:Missing attribute name)/} end end |
#initialize_xml(input, **options)
This method returns an undefined value.
Initializes the underlying XML library.
161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 |
# File 'vendor/bundler/ruby/3.3.0/bundler/gems/rdf-rdfa-ea6265716853/lib/rdf/rdfa/reader/nokogiri.rb', line 161 def initialize_xml(input, **) require 'nokogiri' unless defined?(::Nokogiri) @doc = case input when ::Nokogiri::HTML::Document, ::Nokogiri::XML::Document input else # Try to detect charset from input [:encoding] ||= input.charset if input.respond_to?(:charset) # Otherwise, default is utf-8 [:encoding] ||= 'utf-8' [:encoding] = [:encoding].to_s if [:encoding] case @host_language when :html4 ::Nokogiri::HTML.parse(input, base_uri.to_s, [:encoding]) when :html5 begin input = input.read if input.respond_to?(:read) ::Nokogiri::HTML5(input.force_encoding([:encoding]), max_parse_errors: 1000) rescue LoadError, NoMethodError ::Nokogiri::HTML.parse(input, base_uri.to_s, [:encoding]) end else ::Nokogiri::XML.parse(input, base_uri.to_s, [:encoding]) end end end |