Class: Moxml::Adapter::Oga

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

Defined Under Namespace

Classes: OgaSAXBridge

Class Method Summary collapse

Methods inherited from Base

actual_native, attribute_name, bulk_materialize?, create_cdata, create_comment, create_declaration, create_doctype, create_element, create_entity_reference, create_namespace, create_processing_instruction, create_text, decode_entities, entity_bearing?, in_scope_namespaces, line_number, patch_node, patches_children?, preprocess_entities, restore_entities, sax_supported?, set_attribute_name, set_attribute_value, wrappers_recyclable?

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 



360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
 
# File 'lib/moxml/adapter/oga.rb', line 360

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

  if element.is_a?(::Oga::XML::Document) &&
      child.is_a?(::Oga::XML::XmlDeclaration)
    attachments.set(element, :xml_declaration, child)
    return
  end

  # Insert doctype before root element in document
  if element.is_a?(::Oga::XML::Document) && child.is_a?(::Oga::XML::Doctype)
    root_idx = nil
    element.children.each_with_index do |n, i|
      if n.is_a?(::Oga::XML::Element)
        root_idx = i
        break
      end
    end
    if root_idx
      element.children.insert(root_idx, child)
      return
    end
  end

  element.children << child
end

.add_next_sibling(node, sibling) ⇒ Object 



403
404
405
406
407
408
409
410
411
412
 
# File 'lib/moxml/adapter/oga.rb', line 403

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 



392
393
394
395
396
397
398
399
400
401
 
# File 'lib/moxml/adapter/oga.rb', line 392

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

.adjacent_to_entity_reference?(node) ⇒ Boolean

Returns:

  • (Boolean) 


244
245
246
 
# File 'lib/moxml/adapter/oga.rb', line 244

def adjacent_to_entity_reference?(node)
  entity_ref?(node.previous) || entity_ref?(node.next)
end

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



533
534
535
536
537
538
539
540
541
542
 
# File 'lib/moxml/adapter/oga.rb', line 533

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

.attachments ⇒ Object 



19
20
21
 
# File 'lib/moxml/adapter/oga.rb', line 19

def attachments
  @attachments ||= Moxml::NativeAttachment.new
end

.attribute_element(attr) ⇒ Object 



308
309
310
 
# File 'lib/moxml/adapter/oga.rb', line 308

def attribute_element(attr)
  attr.element
end

.attributes(element) ⇒ Object 



312
313
314
315
316
317
318
319
 
# File 'lib/moxml/adapter/oga.rb', line 312

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 



455
456
457
 
# File 'lib/moxml/adapter/oga.rb', line 455

def cdata_content(node)
  node.text
end

.children(node) ⇒ Object 



220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
 
# File 'lib/moxml/adapter/oga.rb', line 220

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)

  child_nodes = node.children.to_a
  # Filter out whitespace-only text nodes at document level only.
  # Document-level whitespace (between <?xml?> and <root>) is
  # formatting, not content, and differs across adapters.
  # Whitespace inside elements (e.g. "FigureA.1" spacing) is
  # meaningful and must be preserved.
  if node.is_a?(::Oga::XML::Document)
    child_nodes = child_nodes.reject do |child|
      child.is_a?(::Oga::XML::Text) && child.text.strip.empty?
    end
  end
  all_children + child_nodes
end

.comment_content(node) ⇒ Object 



463
464
465
 
# File 'lib/moxml/adapter/oga.rb', line 463

def comment_content(node)
  node.text
end

.create_document(_native_doc = nil) ⇒ Object 



87
88
89
 
# File 'lib/moxml/adapter/oga.rb', line 87

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

.create_native_cdata(content, _owner_doc = nil) ⇒ Object 



117
118
119
 
# File 'lib/moxml/adapter/oga.rb', line 117

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

.create_native_comment(content, _owner_doc = nil) ⇒ Object 



121
122
123
 
# File 'lib/moxml/adapter/oga.rb', line 121

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

.create_native_declaration(version, encoding, standalone) ⇒ Object 



136
137
138
139
140
141
142
143
 
# File 'lib/moxml/adapter/oga.rb', line 136

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 



125
126
127
128
129
130
 
# File 'lib/moxml/adapter/oga.rb', line 125

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 



91
92
93
 
# File 'lib/moxml/adapter/oga.rb', line 91

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

.create_native_entity_reference(name) ⇒ Object 



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

def create_native_entity_reference(name)
  text = ::Oga::XML::Text.new
  text.text = "#{Entity::MARKER}#{name};"
  attachments.set(text, :entity_name, name)
  text
end

.create_native_namespace(element, prefix, uri) ⇒ Object 



161
162
163
164
165
166
167
168
169
170
 
# File 'lib/moxml/adapter/oga.rb', line 161

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 



132
133
134
 
# File 'lib/moxml/adapter/oga.rb', line 132

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

.create_native_text(content, _owner_doc = nil) ⇒ Object 



95
96
97
98
99
100
101
102
103
104
 
# File 'lib/moxml/adapter/oga.rb', line 95

def create_native_text(content, _owner_doc = nil)
  processed = preprocess_entities(content)
  text = ::Oga::XML::Text.new(text: processed)
  # Oga::XML::Text.new stores the value via ivar, bypassing the
  # RawValueOverride setter. Set the sidecar explicitly so reads
  # return the user-supplied value verbatim rather than running
  # it through the lazy decoder.
  attachments.set(text, :raw_text, processed)
  text
end

.declaration_attribute(declaration, attr_name) ⇒ Object 



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

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 



503
504
505
506
507
508
509
 
# File 'lib/moxml/adapter/oga.rb', line 503

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.



499
500
501
 
# File 'lib/moxml/adapter/oga.rb', line 499

def doctype_name(native)
  native.name
end

.doctype_system_id(native) ⇒ Object 



511
512
513
514
515
516
517
 
# File 'lib/moxml/adapter/oga.rb', line 511

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

.document(node) ⇒ Object 



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

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

  current
end

.duplicate_node(node) ⇒ Object

Oga's Node#dup shallow-copies the ivars, leaving attributes and children shared with the original; the public dup contract needs a full deep copy.



260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
 
# File 'lib/moxml/adapter/oga.rb', line 260

def duplicate_node(node)
  case node
  when ::Oga::XML::Element
    copy = ::Oga::XML::Element.new(name: node.name)
    copy.namespace_name = node.namespace_name if node.namespace_name
    node.attributes.each do |attr|
      copy.add_attribute(::Oga::XML::Attribute.new(
                           name: attr.name,
                           namespace_name: attr.namespace_name,
                           value: attr.value,
                         ))
    end
    node.children.each do |child|
      copy.children << duplicate_node(child)
    end
    copy
  when ::Oga::XML::Text
    ::Oga::XML::Text.new(text: node.text)
  when ::Oga::XML::Cdata
    ::Oga::XML::Cdata.new(text: node.text)
  when ::Oga::XML::Comment
    ::Oga::XML::Comment.new(text: node.text)
  when ::Oga::XML::ProcessingInstruction
    ::Oga::XML::ProcessingInstruction.new(name: node.name, text: node.text)
  else
    node.dup
  end
end

.entity_ref?(node) ⇒ Boolean

Returns:

  • (Boolean) 


248
249
250
251
 
# File 'lib/moxml/adapter/oga.rb', line 248

def entity_ref?(node)
  node.is_a?(::Oga::XML::Text) &&
    attachments.get(node, :entity_name)
end

.entity_reference_name(node) ⇒ Object 



113
114
115
 
# File 'lib/moxml/adapter/oga.rb', line 113

def entity_reference_name(node)
  attachments.get(node, :entity_name)
end

.get_attribute(element, name) ⇒ Object 



340
341
342
 
# File 'lib/moxml/adapter/oga.rb', line 340

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

.get_attribute_value(element, name) ⇒ Object 



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

def get_attribute_value(element, name)
  attr = element.attribute(name.to_s)
  attr&.value
end

.has_declaration?(native_doc, _wrapper) ⇒ Boolean

Returns:

  • (Boolean) 


548
549
550
551
552
553
554
555
 
# File 'lib/moxml/adapter/oga.rb', line 548

def has_declaration?(native_doc, _wrapper)
  decl = attachments.get(native_doc, :xml_declaration)
  if decl.nil? && !attachments.key?(native_doc, :xml_declaration)
    native_doc.is_a?(::Oga::XML::Document) && !native_doc.xml_declaration.nil?
  else
    !decl.nil?
  end
end

.inner_text(node) ⇒ Object 



438
439
440
441
442
443
444
 
# File 'lib/moxml/adapter/oga.rb', line 438

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

.namespace(element) ⇒ Object 



177
178
179
180
181
182
183
184
185
186
 
# File 'lib/moxml/adapter/oga.rb', line 177

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 



490
491
492
493
494
 
# File 'lib/moxml/adapter/oga.rb', line 490

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

  node.namespaces.values
end

.namespace_prefix(namespace) ⇒ Object 



479
480
481
482
483
484
 
# File 'lib/moxml/adapter/oga.rb', line 479

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 



486
487
488
 
# File 'lib/moxml/adapter/oga.rb', line 486

def namespace_uri(namespace)
  namespace.uri
end

.next_sibling(node) ⇒ Object 



289
290
291
 
# File 'lib/moxml/adapter/oga.rb', line 289

def next_sibling(node)
  node.next
end

.node_name(node) ⇒ Object 



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

def node_name(node)
  node.name
end

.node_type(node) ⇒ Object 



192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
 
# File 'lib/moxml/adapter/oga.rb', line 192

def node_type(node)
  case node
  when ::Oga::XML::Element then :element
  when ::Oga::XML::Text
    if attachments.key?(node, :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 



253
254
255
 
# File 'lib/moxml/adapter/oga.rb', line 253

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

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



31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
 
# File 'lib/moxml/adapter/oga.rb', line 31

def parse(xml, options = {}, _context = nil)
  processed_xml = preprocess_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

  # Oga parses multiple root elements leniently; the moxml
  # contract requires a single root. to_a because the vendored
  # fork's NodeSet#count ignores its block.
  roots = native_doc.children.to_a.count do |child|
    child.instance_of?(::Oga::XML::Element)
  end
  if roots > 1
    raise Moxml::ParseError.new(
      "multiple root elements",
      source: xml.is_a?(String) ? xml[0..100] : nil,
    )
  end

  ctx = _context || Context.new(:oga)
  doc = Document.new(native_doc, ctx)
  Entity::Restorer.new(doc).run if ctx.config.restore_entities
  doc
end

.previous_sibling(node) ⇒ Object 



293
294
295
 
# File 'lib/moxml/adapter/oga.rb', line 293

def previous_sibling(node)
  node.previous
end

.processing_instruction_content(node) ⇒ Object 



471
472
473
 
# File 'lib/moxml/adapter/oga.rb', line 471

def processing_instruction_content(node)
  node.text
end

.processing_instruction_target(node) ⇒ Object 



188
189
190
 
# File 'lib/moxml/adapter/oga.rb', line 188

def processing_instruction_target(node)
  node.name
end

.remove(node) ⇒ Object 



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

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 declaration state in attachment map
    attachments.set(node.parent, :xml_declaration, nil)
  end

  node.remove
end

.remove_attribute(element, name) ⇒ Object 



349
350
351
352
 
# File 'lib/moxml/adapter/oga.rb', line 349

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

.remove_attribute_native(attr) ⇒ Object 



354
355
356
357
358
 
# File 'lib/moxml/adapter/oga.rb', line 354

def remove_attribute_native(attr)
  element = attr.element
  element&.attributes&.delete(attr)
  attr
end

.remove_declaration(native_doc) ⇒ Object 



557
558
559
 
# File 'lib/moxml/adapter/oga.rb', line 557

def remove_declaration(native_doc)
  attachments.set(native_doc, :xml_declaration, nil)
end

.replace(node, new_node) ⇒ Object 



425
426
427
 
# File 'lib/moxml/adapter/oga.rb', line 425

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

.replace_children(node, new_children) ⇒ Object 



429
430
431
432
 
# File 'lib/moxml/adapter/oga.rb', line 429

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

.root(document) ⇒ Object 



304
305
306
 
# File 'lib/moxml/adapter/oga.rb', line 304

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.

Driven off Oga.parse_xml plus a DOM walk because Oga 3.4's native SAX parser decodes on_text content before delivery and exposes no hook to override. The prepended EntityDecoderOverride ensures node.text invoked during the walk produces the correct single-pass decode, so no separate entity pass is needed here. Trade-off: peak memory scales with document size (Oga 3.4 is the final release of an unmaintained gem; the trade is accepted).

Parameters:



75
76
77
78
79
80
81
82
83
84
85
 
# File 'lib/moxml/adapter/oga.rb', line 75

def sax_parse(xml, handler)
  xml_string = xml.is_a?(IO) || xml.is_a?(StringIO) ? xml.read : xml.to_s
  native_doc = ::Oga.parse_xml(xml_string)

  bridge = OgaSAXBridge.new(handler)
  bridge.on_start_document
  native_doc.children.each { |child| bridge.emit(child) }
  bridge.on_end_document
rescue StandardError => e
  handler.on_error(Moxml::ParseError.new(e.message))
end

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



544
545
546
 
# File 'lib/moxml/adapter/oga.rb', line 544

def serialize(node, options = {})
  serialize_without_entity_processing(node, options)
end

.set_attribute(element, name, value) ⇒ Object 



321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
 
# File 'lib/moxml/adapter/oga.rb', line 321

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

  processed = preprocess_entities(value.to_s)
  attr = ::Oga::XML::Attribute.new(
    name: name.to_s,
    namespace_name: namespace_name,
    value: processed,
  )
  # See create_native_text: Attribute.new bypasses the override
  # setter, so populate the sidecar here.
  attachments.set(attr, :raw_value, processed)
  element.add_attribute(attr)
end

.set_cdata_content(node, content) ⇒ Object 



459
460
461
 
# File 'lib/moxml/adapter/oga.rb', line 459

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

.set_comment_content(node, content) ⇒ Object 



467
468
469
 
# File 'lib/moxml/adapter/oga.rb', line 467

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

.set_declaration_attribute(declaration, attr_name, value) ⇒ Object 



153
154
155
156
157
158
159
 
# File 'lib/moxml/adapter/oga.rb', line 153

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 



172
173
174
175
 
# File 'lib/moxml/adapter/oga.rb', line 172

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

.set_node_name(node, name) ⇒ Object 



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

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

.set_processing_instruction_content(node, content) ⇒ Object 



475
476
477
 
# File 'lib/moxml/adapter/oga.rb', line 475

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

.set_root(doc, element) ⇒ Object 



23
24
25
26
27
28
29
 
# File 'lib/moxml/adapter/oga.rb', line 23

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 



446
447
448
449
450
451
452
453
 
# File 'lib/moxml/adapter/oga.rb', line 446

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

.text_content(node) ⇒ Object 



434
435
436
 
# File 'lib/moxml/adapter/oga.rb', line 434

def text_content(node)
  node.text
end

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



519
520
521
522
523
524
525
526
527
528
529
530
531
 
# File 'lib/moxml/adapter/oga.rb', line 519

def xpath(node, expression, namespaces = nil)
  result = node.xpath(expression, {},
                      namespaces: namespaces&.transform_keys(&:to_s))
  # Adapter contract: Array<native> | scalar
  result.is_a?(Enumerable) ? result.to_a : result
rescue ::LL::ParserError => e
  raise Moxml::XPathError.new(
    e.message,
    expression: expression,
    adapter: "Oga",
    node: node,
  )
end