Class: Moxml::Adapter::Oga

Inherits:
Base
  • Object
show all
Defined in:
lib/moxml/adapter/oga.rb

Constant Summary collapse

STANDARD_XML_ENTITIES =

Standard XML entities handled natively by parsers

%w[amp lt gt quot apos].freeze
ENTITY_PATTERN =

Shared entity name pattern (W3C: 2-31 chars, starts with alpha)

"([a-zA-Z][a-zA-Z0-9]{1,30})"
ENTITY_MARKER =

Marker character for entity preservation through Oga’s parser. U+0001 is preserved literally by Oga through parse/serialize cycle.

"\x01"
ENTITY_MARKER_REGEX =

Regular expression for entity marker post-processing

/#{ENTITY_MARKER}#{ENTITY_PATTERN};/
ENTITY_REF_REGEX =

Simple entity-only regex with no nested quantifiers

/&#{ENTITY_PATTERN};/

Class Method Summary collapse

Methods inherited from Base

create_cdata, create_comment, create_declaration, create_doctype, create_element, create_entity_reference, create_namespace, create_processing_instruction, create_text, duplicate_node, patch_node, prepare_for_new_document, sax_supported?, set_attribute_name, set_attribute_value

Methods included from XmlUtils

#encode_entities, #normalize_xml_value, #validate_comment_content, #validate_declaration_encoding, #validate_declaration_standalone, #validate_declaration_version, #validate_element_name, #validate_entity_reference_name, #validate_pi_target, #validate_prefix, #validate_uri

Class Method Details

.add_child(element, child_or_text) ⇒ Object 



268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
 
# File 'lib/moxml/adapter/oga.rb', line 268

def add_child(element, child_or_text)
  child =
    if child_or_text.is_a?(String)
      create_native_text(child_or_text)
    else
      child_or_text
    end

  # Special handling for declarations on Oga documents
  if element.is_a?(::Oga::XML::Document) &&
      child.is_a?(::Oga::XML::XmlDeclaration)
    # Set as document's xml_declaration
    element.instance_variable_set(:@xml_declaration, child)
  end

  element.children << child
end

.add_next_sibling(node, sibling) ⇒ Object 



297
298
299
300
301
302
303
304
305
306
 
# File 'lib/moxml/adapter/oga.rb', line 297

def add_next_sibling(node, sibling)
  if node.parent == sibling.parent
    # Oga doesn't manipulate children of the same parent
    dup_sibling = node.node_set.delete(sibling)
    index = node.node_set.index(node) + 1
    node.node_set.insert(index, dup_sibling)
  else
    node.after(sibling)
  end
end

.add_previous_sibling(node, sibling) ⇒ Object 



286
287
288
289
290
291
292
293
294
295
 
# File 'lib/moxml/adapter/oga.rb', line 286

def add_previous_sibling(node, sibling)
  if node.parent == sibling.parent
    # Oga doesn't manipulate children of the same parent
    dup_sibling = node.node_set.delete(sibling)
    index = node.node_set.index(node)
    node.node_set.insert(index, dup_sibling)
  else
    node.before(sibling)
  end
end

.at_xpath(node, expression, namespaces = nil) ⇒ Object 



426
427
428
429
430
431
432
433
434
435
 
# File 'lib/moxml/adapter/oga.rb', line 426

def at_xpath(node, expression, namespaces = nil)
  node.at_xpath(expression, namespaces: namespaces)
rescue ::Oga::XPath::Error => e
  raise Moxml::XPathError.new(
    e.message,
    expression: expression,
    adapter: "Oga",
    node: node,
  )
end

.attribute_element(attr) ⇒ Object 



227
228
229
 
# File 'lib/moxml/adapter/oga.rb', line 227

def attribute_element(attr)
  attr.element
end

.attributes(element) ⇒ Object 



231
232
233
234
235
236
237
238
 
# File 'lib/moxml/adapter/oga.rb', line 231

def attributes(element)
  return [] unless element.is_a?(::Oga::XML::Element)

  # remove attributes-namespaces
  element.attributes.reject do |attr|
    attr.name == ::Oga::XML::Element::XMLNS_PREFIX || attr.namespace_name == ::Oga::XML::Element::XMLNS_PREFIX
  end
end

.cdata_content(node) ⇒ Object 



350
351
352
 
# File 'lib/moxml/adapter/oga.rb', line 350

def cdata_content(node)
  node.text
end

.children(node) ⇒ Object 



187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
 
# File 'lib/moxml/adapter/oga.rb', line 187

def children(node)
  all_children = []

  if node.is_a?(::Oga::XML::Document)
    all_children += [node.xml_declaration,
                     node.doctype].compact
  end

  return all_children unless node.is_a?(::Oga::XML::Node) || node.is_a?(::Oga::XML::Document)

  all_children + node.children.reject do |child|
    child.is_a?(::Oga::XML::Text) &&
      child.text.strip.empty? &&
      !(child.previous.nil? && child.next.nil?)
  end
end

.comment_content(node) ⇒ Object 



358
359
360
 
# File 'lib/moxml/adapter/oga.rb', line 358

def comment_content(node)
  node.text
end

.create_document(_native_doc = nil) ⇒ Object 



62
63
64
 
# File 'lib/moxml/adapter/oga.rb', line 62

def create_document(_native_doc = nil)
  ::Oga::XML::Document.new
end

.create_native_cdata(content, _owner_doc = nil) ⇒ Object 



85
86
87
 
# File 'lib/moxml/adapter/oga.rb', line 85

def create_native_cdata(content, _owner_doc = nil)
  ::Oga::XML::Cdata.new(text: content)
end

.create_native_comment(content, _owner_doc = nil) ⇒ Object 



89
90
91
 
# File 'lib/moxml/adapter/oga.rb', line 89

def create_native_comment(content, _owner_doc = nil)
  ::Oga::XML::Comment.new(text: content)
end

.create_native_declaration(version, encoding, standalone) ⇒ Object 



104
105
106
107
108
109
110
111
 
# File 'lib/moxml/adapter/oga.rb', line 104

def create_native_declaration(version, encoding, standalone)
  attrs = {
    version: version,
    encoding: encoding,
    standalone: standalone,
  }.compact
  ::Moxml::Adapter::CustomizedOga::XmlDeclaration.new(attrs)
end

.create_native_doctype(name, external_id, system_id) ⇒ Object 



93
94
95
96
97
98
 
# File 'lib/moxml/adapter/oga.rb', line 93

def create_native_doctype(name, external_id, system_id)
  ::Oga::XML::Doctype.new(
    name: name, public_id: external_id, system_id: system_id,
    type: external_id ? "PUBLIC" : "SYSTEM"
  )
end

.create_native_element(name, _owner_doc = nil) ⇒ Object 



66
67
68
 
# File 'lib/moxml/adapter/oga.rb', line 66

def create_native_element(name, _owner_doc = nil)
  ::Oga::XML::Element.new(name: name)
end

.create_native_entity_reference(name) ⇒ Object 



74
75
76
77
78
79
 
# File 'lib/moxml/adapter/oga.rb', line 74

def create_native_entity_reference(name)
  text = ::Oga::XML::Text.new
  text.text = "#{ENTITY_MARKER}#{name};"
  text.instance_variable_set(:@moxml_entity_name, name)
  text
end

.create_native_namespace(element, prefix, uri) ⇒ Object 



129
130
131
132
133
134
135
136
137
138
 
# File 'lib/moxml/adapter/oga.rb', line 129

def create_native_namespace(element, prefix, uri)
  ns = element.available_namespaces[prefix]
  return ns unless ns.nil?

  # Oga creates an attribute and registers a namespace
  set_attribute(element,
                [::Oga::XML::Element::XMLNS_PREFIX, prefix].compact.join(":"), uri)
  element.register_namespace(prefix, uri)
  ::Oga::XML::Namespace.new(name: prefix, uri: uri)
end

.create_native_processing_instruction(target, content) ⇒ Object 



100
101
102
 
# File 'lib/moxml/adapter/oga.rb', line 100

def create_native_processing_instruction(target, content)
  ::Oga::XML::ProcessingInstruction.new(name: target, text: content)
end

.create_native_text(content, _owner_doc = nil) ⇒ Object 



70
71
72
 
# File 'lib/moxml/adapter/oga.rb', line 70

def create_native_text(content, _owner_doc = nil)
  ::Oga::XML::Text.new(text: encode_entity_markers(content))
end

.declaration_attribute(declaration, attr_name) ⇒ Object 



113
114
115
116
117
118
119
 
# File 'lib/moxml/adapter/oga.rb', line 113

def declaration_attribute(declaration, attr_name)
  unless ::Moxml::Declaration::ALLOWED_ATTRIBUTES.include?(attr_name.to_s)
    return
  end

  declaration.public_send(attr_name)
end

.doctype_external_id(native) ⇒ Object 



398
399
400
401
402
403
404
 
# File 'lib/moxml/adapter/oga.rb', line 398

def doctype_external_id(native)
  if native.type == "SYSTEM"
    nil
  else
    native.public_id
  end
end

.doctype_name(native) ⇒ Object

Doctype accessor methods Note: Oga stores SYSTEM identifier in public_id for SYSTEM doctypes. See: Oga::XML::Doctype puts SYSTEM dtd in public_id, system_id is nil.



394
395
396
 
# File 'lib/moxml/adapter/oga.rb', line 394

def doctype_name(native)
  native.name
end

.doctype_system_id(native) ⇒ Object 



406
407
408
409
410
411
412
 
# File 'lib/moxml/adapter/oga.rb', line 406

def doctype_system_id(native)
  if native.type == "SYSTEM"
    native.public_id
  else
    native.system_id
  end
end

.document(node) ⇒ Object 



216
217
218
219
220
221
 
# File 'lib/moxml/adapter/oga.rb', line 216

def document(node)
  current = node
  current = current.parent while parent(current)

  current
end

.entity_reference_name(node) ⇒ Object 



81
82
83
 
# File 'lib/moxml/adapter/oga.rb', line 81

def entity_reference_name(node)
  node.instance_variable_get(:@moxml_entity_name)
end

.get_attribute(element, name) ⇒ Object 



255
256
257
 
# File 'lib/moxml/adapter/oga.rb', line 255

def get_attribute(element, name)
  element.attribute(name.to_s)
end

.get_attribute_value(element, name) ⇒ Object 



259
260
261
 
# File 'lib/moxml/adapter/oga.rb', line 259

def get_attribute_value(element, name)
  restore_entity_markers(element[name.to_s])
end

.inner_text(node) ⇒ Object 



332
333
334
335
336
337
338
339
 
# File 'lib/moxml/adapter/oga.rb', line 332

def inner_text(node)
  text = if node.is_a?(::Oga::XML::Element)
           node.inner_text
         else
           node.text
         end
  restore_entity_markers(text)
end

.namespace(element) ⇒ Object 



144
145
146
147
148
149
150
151
152
153
 
# File 'lib/moxml/adapter/oga.rb', line 144

def namespace(element)
  case element
  when ::Oga::XML::Element, ::Oga::XML::Attribute
    element.namespace
  end
rescue NoMethodError
  # Oga attributes fail with NoMethodError:
  # undefined method `available_namespaces' for nil:NilClass
  nil
end

.namespace_definitions(node) ⇒ Object 



385
386
387
388
389
 
# File 'lib/moxml/adapter/oga.rb', line 385

def namespace_definitions(node)
  return [] unless node.is_a?(::Oga::XML::Element)

  node.namespaces.values
end

.namespace_prefix(namespace) ⇒ Object 



374
375
376
377
378
379
 
# File 'lib/moxml/adapter/oga.rb', line 374

def namespace_prefix(namespace)
  # nil for the default namespace
  return if namespace.name == ::Oga::XML::Element::XMLNS_PREFIX

  namespace.name
end

.namespace_uri(namespace) ⇒ Object 



381
382
383
 
# File 'lib/moxml/adapter/oga.rb', line 381

def namespace_uri(namespace)
  namespace.uri
end

.next_sibling(node) ⇒ Object 



208
209
210
 
# File 'lib/moxml/adapter/oga.rb', line 208

def next_sibling(node)
  node.next
end

.node_name(node) ⇒ Object 



179
180
181
 
# File 'lib/moxml/adapter/oga.rb', line 179

def node_name(node)
  node.name
end

.node_type(node) ⇒ Object 



159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
 
# File 'lib/moxml/adapter/oga.rb', line 159

def node_type(node)
  case node
  when ::Oga::XML::Element then :element
  when ::Oga::XML::Text
    if node.instance_variable_get(:@moxml_entity_name)
      :entity_reference
    else
      :text
    end
  when ::Oga::XML::Cdata then :cdata
  when ::Oga::XML::Comment then :comment
  when ::Oga::XML::Attribute then :attribute
  when ::Oga::XML::Namespace then :namespace
  when ::Oga::XML::ProcessingInstruction then :processing_instruction
  when ::Oga::XML::Document then :document
  when ::Oga::XML::Doctype then :doctype
  else :unknown
  end
end

.parent(node) ⇒ Object 



204
205
206
 
# File 'lib/moxml/adapter/oga.rb', line 204

def parent(node)
  node.parent if node.is_a?(::Oga::XML::Node)
end

.parse(xml, options = {}, _context = nil) ⇒ Object 



22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
 
# File 'lib/moxml/adapter/oga.rb', line 22

def parse(xml, options = {}, _context = nil)
  # Pre-process XML to convert named entities to marker form (\x01name;).
  # Oga drops named entity references like &nbsp; during parsing.
  processed_xml = preprocess_named_entities(xml)

  native_doc = begin
    ::Oga.parse_xml(processed_xml, strict: options[:strict])
  rescue LL::ParserError => e
    raise Moxml::ParseError.new(
      e.message,
      source: xml.is_a?(String) ? xml[0..100] : nil,
    )
  end

  ctx = _context || Context.new(:oga)
  DocumentBuilder.new(ctx).build(native_doc)
end

.previous_sibling(node) ⇒ Object 



212
213
214
 
# File 'lib/moxml/adapter/oga.rb', line 212

def previous_sibling(node)
  node.previous
end

.processing_instruction_content(node) ⇒ Object 



366
367
368
 
# File 'lib/moxml/adapter/oga.rb', line 366

def processing_instruction_content(node)
  node.text
end

.processing_instruction_target(node) ⇒ Object 



155
156
157
 
# File 'lib/moxml/adapter/oga.rb', line 155

def processing_instruction_target(node)
  node.name
end

.remove(node) ⇒ Object 



308
309
310
311
312
313
314
315
316
317
 
# File 'lib/moxml/adapter/oga.rb', line 308

def remove(node)
  # Special handling for declarations on Oga documents
  if node.is_a?(::Oga::XML::XmlDeclaration) &&
      node.parent.is_a?(::Oga::XML::Document)
    # Clear document's xml_declaration when removing declaration
    node.parent.instance_variable_set(:@xml_declaration, nil)
  end

  node.remove
end

.remove_attribute(element, name) ⇒ Object 



263
264
265
266
 
# File 'lib/moxml/adapter/oga.rb', line 263

def remove_attribute(element, name)
  attr = element.attribute(name.to_s)
  element.attributes.delete(attr) if attr
end

.replace(node, new_node) ⇒ Object 



319
320
321
 
# File 'lib/moxml/adapter/oga.rb', line 319

def replace(node, new_node)
  node.replace(new_node)
end

.replace_children(node, new_children) ⇒ Object 



323
324
325
326
 
# File 'lib/moxml/adapter/oga.rb', line 323

def replace_children(node, new_children)
  node.children = []
  new_children.each { |child| add_child(node, child) }
end

.root(document) ⇒ Object 



223
224
225
 
# File 'lib/moxml/adapter/oga.rb', line 223

def root(document)
  document.children.find { |node| node.is_a?(::Oga::XML::Element) }
end

.sax_parse(xml, handler) ⇒ void

This method returns an undefined value.

SAX parsing implementation for Oga

Parameters:



45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
 
# File 'lib/moxml/adapter/oga.rb', line 45

def sax_parse(xml, handler)
  bridge = OgaSAXBridge.new(handler)

  xml_string = xml.is_a?(IO) || xml.is_a?(StringIO) ? xml.read : xml.to_s

  # Manually call start_document (Oga doesn't)
  handler.on_start_document

  ::Oga.sax_parse_xml(bridge, xml_string)

  # Manually call end_document (Oga doesn't)
  handler.on_end_document
rescue StandardError => e
  error = Moxml::ParseError.new(e.message)
  handler.on_error(error)
end

.serialize(node, options = {}) ⇒ Object 



437
438
439
440
441
 
# File 'lib/moxml/adapter/oga.rb', line 437

def serialize(node, options = {})
  output = serialize_without_entity_processing(node, options)
  # Post-process: convert entity markers back to entity references
  output.gsub(ENTITY_MARKER_REGEX, '&\1;')
end

.set_attribute(element, name, value) ⇒ Object 



240
241
242
243
244
245
246
247
248
249
250
251
252
253
 
# File 'lib/moxml/adapter/oga.rb', line 240

def set_attribute(element, name, value)
  namespace_name = nil
  if name.to_s.include?(":")
    namespace_name, name = name.to_s.split(":",
                                           2)
  end

  attr = ::Oga::XML::Attribute.new(
    name: name.to_s,
    namespace_name: namespace_name,
    value: encode_entity_markers(value.to_s),
  )
  element.add_attribute(attr)
end

.set_cdata_content(node, content) ⇒ Object 



354
355
356
 
# File 'lib/moxml/adapter/oga.rb', line 354

def set_cdata_content(node, content)
  node.text = content
end

.set_comment_content(node, content) ⇒ Object 



362
363
364
 
# File 'lib/moxml/adapter/oga.rb', line 362

def set_comment_content(node, content)
  node.text = content
end

.set_declaration_attribute(declaration, attr_name, value) ⇒ Object 



121
122
123
124
125
126
127
 
# File 'lib/moxml/adapter/oga.rb', line 121

def set_declaration_attribute(declaration, attr_name, value)
  unless ::Moxml::Declaration::ALLOWED_ATTRIBUTES.include?(attr_name.to_s)
    return
  end

  declaration.public_send("#{attr_name}=", value)
end

.set_namespace(element, ns_or_string) ⇒ Object 



140
141
142
 
# File 'lib/moxml/adapter/oga.rb', line 140

def set_namespace(element, ns_or_string)
  element.namespace_name = ns_or_string.to_s
end

.set_node_name(node, name) ⇒ Object 



183
184
185
 
# File 'lib/moxml/adapter/oga.rb', line 183

def set_node_name(node, name)
  node.name = name
end

.set_processing_instruction_content(node, content) ⇒ Object 



370
371
372
 
# File 'lib/moxml/adapter/oga.rb', line 370

def set_processing_instruction_content(node, content)
  node.text = content
end

.set_root(doc, element) ⇒ Object 



14
15
16
17
18
19
20
 
# File 'lib/moxml/adapter/oga.rb', line 14

def set_root(doc, element)
  # Clear existing root element if any - Oga's NodeSet needs special handling
  # We need to manually remove elements since NodeSet doesn't support clear or delete_if
  elements_to_remove = doc.children.grep(::Oga::XML::Element)
  elements_to_remove.each { |elem| doc.children.delete(elem) }
  doc.children << element
end

.set_text_content(node, content) ⇒ Object 



341
342
343
344
345
346
347
348
 
# File 'lib/moxml/adapter/oga.rb', line 341

def set_text_content(node, content)
  encoded = encode_entity_markers(content)
  if node.is_a?(::Oga::XML::Element)
    node.inner_text = encoded
  else
    node.text = encoded
  end
end

.text_content(node) ⇒ Object 



328
329
330
 
# File 'lib/moxml/adapter/oga.rb', line 328

def text_content(node)
  restore_entity_markers(node.text)
end

.xpath(node, expression, namespaces = nil) ⇒ Object 



414
415
416
417
418
419
420
421
422
423
424
 
# File 'lib/moxml/adapter/oga.rb', line 414

def xpath(node, expression, namespaces = nil)
  node.xpath(expression, {},
             namespaces: namespaces&.transform_keys(&:to_s)).to_a
rescue ::LL::ParserError => e
  raise Moxml::XPathError.new(
    e.message,
    expression: expression,
    adapter: "Oga",
    node: node,
  )
end