Class: Nokogiri::XML::Range

Inherits:
Object
  • Object
show all
Defined in:
lib/nokogiri/xml/range/version.rb,
lib/nokogiri/xml/range/refinement.rb,
lib/nokogiri/xml/range.rb

Defined Under Namespace

Modules: Refinement Classes: HierarchyRequestError, IndexSizeError, InvalidNodeTypeError, InvalidStateError, NotFoundError, NotSupportedError, WrongDocumentError

Constant Summary collapse

VERSION =

nokogiri-xml-range version

"0.1.1"
START_TO_START =
0
START_TO_END =
1
END_TO_END =
2
END_TO_START =
3

Instance Attribute Summary collapse

Class Method Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(start_container, start_offset, end_container, end_offset) ⇒ Range

Returns a new instance of Range.


55
56
57
58
# File 'lib/nokogiri/xml/range.rb', line 55

def initialize(start_container, start_offset, end_container, end_offset)
  @start_container, @start_offset, @end_container, @end_offset =
    start_container, start_offset, end_container, end_offset
end

Instance Attribute Details

#end_containerObject (readonly) Also known as: end_node

Returns the value of attribute end_container


51
52
53
# File 'lib/nokogiri/xml/range.rb', line 51

def end_container
  @end_container
end

#end_offsetObject (readonly)

Returns the value of attribute end_offset


51
52
53
# File 'lib/nokogiri/xml/range.rb', line 51

def end_offset
  @end_offset
end

#start_containerObject (readonly) Also known as: start_node

Returns the value of attribute start_container


51
52
53
# File 'lib/nokogiri/xml/range.rb', line 51

def start_container
  @start_container
end

#start_offsetObject (readonly)

Returns the value of attribute start_offset


51
52
53
# File 'lib/nokogiri/xml/range.rb', line 51

def start_offset
  @start_offset
end

Class Method Details

.compare_points(node1, offset1, node2, offset2) ⇒ Object


24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
# File 'lib/nokogiri/xml/range.rb', line 24

def compare_points(node1, offset1, node2, offset2)
  return unless node1.document == node2.document

  case node1 <=> node2
  when 0
    offset1 <=> offset2
  when 1
    compare_points(node2, offset2, node1, offset1) * -1
  else
    ancestors = node2.ancestors_to(node1) # nil or [node2, parent of node2, ..., child of node1, node1]
    if ancestors
      child = nil
      ancestors.reverse_each do |anc|
        child = anc if anc.parent == node1
      end
      if node1.children.index(child) < offset1
        1
      else
        -1
      end
    else
      -1
    end
  end
end

Instance Method Details

#clone_contentsObject


289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
# File 'lib/nokogiri/xml/range.rb', line 289

def clone_contents
  fragment = DocumentFragment.new(document)
  return fragment if collapsed?

  if @start_container == @end_container and @start_container.replacable?
    cloned = @start_container.clone(0)
    cloned.content = @start_container.substring_data(@start_offset, @end_offset - @start_offset)
    fragment << cloned
    return fragment
  end

  common_ancestor = common_ancestor_container
  first_partially_contained_child = nil
  @end_node_ancestors = [@end_container] + @end_container.ancestors
  unless @end_node_ancestors.include?(@start_container)
    first_partially_contained_child = common_ancestor.children.find {|child|
      partially_contain_node? child
    }
  end
  last_partially_contained_child = nil
  unless ([@start_container] + @start_container.ancestors).include? @end_container
    last_partially_contained_child = common_ancestor.children.reverse_each.find {|child|
      partially_contain_node? child
    }
  end

  contained_children = common_ancestor.children.select {|child|
    contain_node? child
  }

  raise HierarchyRequestError if contained_children.any? {|child|
    child.type == Node::DOCUMENT_TYPE_NODE
  }

  if first_partially_contained_child && first_partially_contained_child.replacable?

    cloned = @start_container.clone(0)
    cloned.content = @start_container.substring_data(@start_offset, @start_container.length - @start_offset)
    fragment << cloned
  elsif first_partially_contained_child
    cloned =first_partially_contained_child.clone(0)
    fragment << cloned
    subrange = self.class.new(@start_container, @start_offset, first_partially_contained_child, first_partially_contained_child.length)
    subfragment = subrange.clone_contents
    cloned << subfragment
  end

  contained_children.each do |contained_child|
    cloned = contained_child.clone(1)
    fragment << cloned
  end

  if last_partially_contained_child && last_partially_contained_child.replacable?
    cloned = @end_container.clone(0)
    cloned.content = @end_container.substring_data(0, @end_offset)
    fragment << cloned
  elsif last_partially_contained_child
    cloned = last_partially_contained_child.clone(0)
    fragment << cloned
    subrange = self.class.new(last_partially_contained_child, 0, @end_container, @end_offset)
    subfragment = subrange.clone_contents
    cloned << subfragment
  end

  fragment
end

#collapse(to_start = false) ⇒ Object Also known as: collapse!


125
126
127
# File 'lib/nokogiri/xml/range.rb', line 125

def collapse(to_start=false)
  to_start ? set_end(*start_point) : set_start(*end_point)
end

#collapsed?Boolean

Returns:

  • (Boolean)

120
121
122
123
# File 'lib/nokogiri/xml/range.rb', line 120

def collapsed?
  @start_offset == @end_offset and
    @start_container == @end_container
end

#common_ancestor_containerObject


130
131
132
133
134
135
136
137
# File 'lib/nokogiri/xml/range.rb', line 130

def common_ancestor_container
  container = @start_container
  ancestors_of_end = [@end_container] + @end_container.ancestors
  until ancestors_of_end.include?(container)
    container = container.parent
  end
  container
end

#compare_boundary_points(how, source_range) ⇒ Object

Raises:


153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
# File 'lib/nokogiri/xml/range.rb', line 153

def compare_boundary_points(how, source_range)
  raise WrongDocumentError, 'different document' unless source_range.document == document
  this_point, other_point =
    case how
    when START_TO_START
      [start_point, source_range.start_point]
    when START_TO_END
      [end_point, source_range.start_point]
    when END_TO_END
      [end_point, source_range.end_point]
    when END_TO_START
      [start_point, source_range.end_point]
    else
      raise NotSupportedError, 'unsupported way to compare'
    end
  self.class.compare_points(this_point[0], this_point[1], other_point[0], other_point[1])
end

#compare_point(node, offset) ⇒ Object

Raises:


517
518
519
520
521
522
523
524
# File 'lib/nokogiri/xml/range.rb', line 517

def compare_point(node, offset)
  raise WrongDocumentError unless node.ancestors.last == @start_container.ancestors.last
  raise InvalidNodeTypeError if node.type == Node::DOCUMENT_TYPE_NODE
  raise IndexSizeError if offset > node.length
  return -1 if self.class.compare_points(node, offset, @start_container, @start_offset) == -1
  return 1 if self.class.compare_points(node, offset, @end_container, @end_offset) == 1
  0
end

#contain_node?(node) ⇒ Boolean Also known as: include_node?, cover_node?

Returns:

  • (Boolean)

477
478
479
480
481
# File 'lib/nokogiri/xml/range.rb', line 477

def contain_node?(node)
  document == node.document and
    self.class.compare_points(node, 0, @start_container, @start_offset) == 1 and
    self.class.compare_points(node, node.length, @end_container, @end_offset) == -1
end

#containing_nodesObject


485
486
487
488
489
490
# File 'lib/nokogiri/xml/range.rb', line 485

def containing_nodes
  nodes = NodeSet.new(document)
  select_containing_nodes common_ancestor_container, nodes

  nodes
end

#delete_contentsObject


171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
# File 'lib/nokogiri/xml/range.rb', line 171

def delete_contents
  return if collapsed?

  if @start_container == @end_container and @start_container.replacable?
    @start_container.replace_data @start_offset, @end_offset - @start_offset, ''
    return
  end

  nodes_to_remove = NodeSet.new(document)
  common_ancestor = common_ancestor_container
  select_containing_children common_ancestor, nodes_to_remove

  if @end_container.ancestors_to @start_container
    new_node, new_offset = @start_container, @start_offset
  else
    reference_node = @start_container
    parent = reference_node.parent
    while parent and !@end_container.ancestors_to(parent)
      reference_node = parent
      parent = reference_node.parent
    end
    new_node = parent
    new_offset = parent.children.index(reference_node) + 1
  end

  if @start_container.replacable?
    @start_container.replace_data @start_offset, @start_container.length - @start_offset, ''
  end

  nodes_to_remove.each &:remove

  if @end_container.replacable?
    @end_container.replace_data 0, @end_offset, ''
  end

  @start_container = @end_container = new_node
  @start_offset = @end_offset = new_offset
end

#documentObject


68
69
70
# File 'lib/nokogiri/xml/range.rb', line 68

def document
  @start_container.document
end

#end_pointObject


64
65
66
# File 'lib/nokogiri/xml/range.rb', line 64

def end_point
  [@end_container, @end_offset]
end

#extract_contentsObject


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
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
# File 'lib/nokogiri/xml/range.rb', line 210

def extract_contents
  fragment = DocumentFragment.new(document)
  return fragment if collapsed?

  if @start_container == @end_container and @start_container.replacable?
    cloned = @start_container.clone(0)
    cloned.content = @start_container.substring_data(@start_offset, @end_offset - @start_offset)
    fragment << cloned
    @start_container.replace_data @start_offset, @end_offset - @start_offset, ''
    return fragment
  end
  common_ancestor = common_ancestor_container
  end_node_ancestors = [@end_container] + @end_container.ancestors
  first_partially_contained_child =nil
  unless end_node_ancestors.include? @start_container
    first_partially_contained_child = common_ancestor.children.find {|child|
      partially_contain_node? child
    }
  end
  last_partially_contained_child = nil
  unless @start_container.ancestors_to @end_container
    last_partially_contained_child = common_ancestor.children.reverse_each.find {|child|
      partially_contain_node? child
    }
  end
  contained_children = common_ancestor.children.select {|child|
    contain_node? child
  }
  raise HierarchyRequestError if contained_children.any? {|child|
    child.type == Node::DOCUMENT_TYPE_NODE
  }

  if end_node_ancestors.include? @start_container
    new_node, new_offset = @start_container, @start_offset
  else
    reference_node = @start_container
    parent = reference_node.parent
    while parent and !end_node_ancestors.include?(parent)
      reference_node = reference_node.parent
      parent = reference_node.parent
    end
    new_node = parent
    new_offset = parent.children.index(reference_node) + 1
  end

  if first_partially_contained_child && first_partially_contained_child.replacable?

    cloned = @start_container.clone(0)
    cloned.content = @start_container.substring_data(@start_offset, @start_container.length - @start_offset)
    fragment << cloned
    @start_container.replace_data @start_offset, @start_container.length - @start_offset, ''
  elsif first_partially_contained_child
    cloned = first_partially_contained_child.clone(0)
    fragment << cloned
    subrange = Range.new(@start_container, @start_offset, first_partially_contained_child, first_partially_contained_child.length)
    subfragment = subrange.extract_contents
    cloned << subfragment
  end
  contained_children.each do |contained_child|
    fragment << contained_child
  end
  if last_partially_contained_child && last_partially_contained_child.replacable?

    cloned = @end_container.clone(0)
    cloned.content = @end_container.substring_data(0, @end_offset)
    fragment << cloned
    @end_container.replace_data 0, @end_offset, ''
  elsif last_partially_contained_child
    cloned = last_partially_contained_child.clone(0)
    fragment << cloned
    subrange = Range.new(last_partially_contained_child, 0, @end_container, @end_offset)
    subfragment = subrange.extract_contents
    cloned << subfragment
  end
  @start_container = @end_container = new_node
  @start_offset = @end_offset = new_offset
  fragment
end

#insert_node(node) ⇒ Object


356
357
358
359
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
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
# File 'lib/nokogiri/xml/range.rb', line 356

def insert_node(node)
  if [Node::PI_NODE, Node::COMMENT_NODE].include?(@start_container.type) or
    @start_container.text? && @start_container.parent.nil?
    raise HierarchyRequestError
  end
  reference_node = nil
  if @start_container.text?
    reference_node = @start_container
  else
    reference_node = @start_container.children[@start_offset]
  end
  if reference_node
    parent = reference_node.parent
  else
    parent = @start_container
  end
  node.validate_pre_insertion parent, reference_node
  if @start_container.text?
    #7 reference_node = @start_container.split()
  end
  # Nokogiri doesn't support serial text node,
  # so we need to handle it ourselves
  split_node = nil
  if @start_container.text?
    split_node = self.class.new(@start_container, @start_offset, @start_container, @start_container.length).extract_contents
    reference_node = split_node
  end
  if node == reference_node
    reference_node = node.next_sibling
  end
  if node.parent
    node.remove
  end
  if reference_node
    if split_node
      @start_container.parent.children.index(@start_container) + 1
    else
      new_offset = reference_node.parent.children.index(reference_node)
    end
  else
    new_offset = parent.length
  end

  # pre-insert
  if split_node
    # pre-insert validation node parent reference_node(@start_container or split_node)
    unless [Node::DOCUMENT_NODE, Node::DOCUMENT_FRAG_NODE, Node::ELEMENT_NODE].include? parent.type
      raise HierarchyRequestError
    end
    raise hierarchyrequesterror if parent.host_including_inclusive_ancestor? node
    raise Hierarchyrequesterror if reference_node and @start_container.parent != parent
    unless [Node::DOCUMENT_FRAG_NODE, Node::DOCUMENT_TYPE_NODE, Node::ELEMENT_NODE, Node::TEXT_NODE, Node::PI_NODE, Node::COMMENT_NODE].include? node.type
      raise Hierarchyrequesterror
    end
    raise HierarchyRequestError if node.text? && parent.document?
    raise HierarchyRequestError if node.type == Node::DOCUMENT_TYPE_NODE and !parent.document?
    if parent.document?
      case node.type
      when Node::DOCUMENT_FRAG_NODE
        child_element_count = 0
        node.children.each do |n|
          raise HierarchyRequestError if n.text?
          child_element_count += 1 if n.element?
          raise HierarchyRequestError if child_element_count > 1
        end
        if child_element_count == 1
          raise HierarchyRequestError if parent.children.any?(&:element?)
          if reference_node
            raise HierarchyRequestError if reference_node.type == Node::DOCUMENT_TYPE_NODE
            raise HierarchyRequestError if @start_container.following_node.type == Node::DOCUMENT_TYPE_NODE
          end
        end
      when Node::ELEMENT_NODE
        raise HierarchyRequestError if parent.children.any?(&:element?)
        if reference_node
          raise HierarchyRequestError if reference_node.child == Node::DOCUMENT_TYPE_NODE
          raise HierarchyRequestError if @start_container.following_node.type == Node::DOCUMENT_TYPE_NODE
        end
      when Node::DOCUMENT_TYPE_NODE
        raise HierarchyRequestError if parent.children.any? {|n|
          n.type == Node::DOCUMENT_TYPE_NODE
        }
        if reference_node
          raise HierarchyRequestError if @start_container.preceding_node.element?
          raise HierarchyRequestError if parent.children.any?(&:element?)
        end
      end
    end

    reference_child = @start_container
    if reference_child == parent
      reference_child = split_node
    end
    parent.document.adopt node
    @start_container.after node
    node.after split_node
  else
    node.validate_pre_insertion parent, reference_node
    reference_child = reference_node
    if reference_child == parent
      reference_child = parent.next_sibling
    end
    parent.document.adopt node
    reference_child.before node
  end

  if collapsed?
    @end_container, @end_offset = parent, new_offset
  end
end

#intersect_node?(node) ⇒ Boolean

Returns:

  • (Boolean)

526
527
528
529
530
531
532
533
534
# File 'lib/nokogiri/xml/range.rb', line 526

def intersect_node?(node)
  return false unless node.ancestors.last == @start_container.ancestors.last
  return true unless node.respond_to?(:parent)
  parent = node.parent
  return true unless parent
  offset = parent.children.index(node)
  (self.class.compare_points(parent, offset, @end_container, @end_offset) == -1) and
    (self.class.compare_points(parent, offset + 1, @start_container, @start_offset) == 1)
end

#partially_contain_node?(node) ⇒ Boolean Also known as: partially_include_node?, partially_cover_node?

Returns:

  • (Boolean)

492
493
494
495
496
497
# File 'lib/nokogiri/xml/range.rb', line 492

def partially_contain_node?(node)
  path_to_start = @start_container.ancestors_to(node)
  path_to_end = @end_container.ancestors_to(node)
  !path_to_start.nil? && path_to_end.nil? or
    path_to_start.nil? && !path_to_end.nil?
end

#partially_containing_nodesObject


501
502
503
504
505
506
# File 'lib/nokogiri/xml/range.rb', line 501

def partially_containing_nodes
  inclusive_ancestors_of_start = @start_container.inclusive_ancestors
  inclusive_ancestors_of_end = @end_container.inclusive_ancestors
  (inclusive_ancestors_of_start | inclusive_ancestors_of_end) -
    (inclusive_ancestors_of_start & inclusive_ancestors_of_end)
end

#point_in_range?(node, offset) ⇒ Boolean

Returns:

  • (Boolean)

Raises:


508
509
510
511
512
513
514
515
# File 'lib/nokogiri/xml/range.rb', line 508

def point_in_range?(node, offset)
  return false unless node.ancestors.last == @start_container.ancestors.last
  raise InvalidNodeTypeError if node.type == Node::DOCUMENT_TYPE_NODE
  raise IndexSizeError if offset > node.length
  return false if self.class.compare_points(node, offset, @start_container, @start_offset) == -1
  return false if self.class.compare_points(node, offset, @end_container, @end_offset) == 1
  true
end

#rootObject


72
73
74
# File 'lib/nokogiri/xml/range.rb', line 72

def root
  document.root
end

#select_node(node) ⇒ Object


139
140
141
142
143
144
145
# File 'lib/nokogiri/xml/range.rb', line 139

def select_node(node)
  parent = node.parent
  raise InvalidNodeTypeError, 'parent node is empty' unless parent
  index = parent.children.index(node)
  set_start parent, index
  set_end parent, index + 1
end

#select_node_contents(node) ⇒ Object


147
148
149
150
151
# File 'lib/nokogiri/xml/range.rb', line 147

def select_node_contents(node)
  raise InvalidNodeTypeError, 'document type declaration is passed' if node.type == Node::DOCUMENT_TYPE_NODE
  set_start node, 0
  set_end node, node.length
end

#set_end(node, offset) ⇒ Object Also known as: end=


86
87
88
89
90
91
92
93
# File 'lib/nokogiri/xml/range.rb', line 86

def set_end(node, offset)
  validate_boundary_point node, offset
  if document != node.document or
    self.class.compare_points(node, offset, @start_container, @start_offset) == -1
    set_start node, offset
  end
  @end_container, @end_offset = node, offset
end

#set_end_after(node) ⇒ Object


114
115
116
117
118
# File 'lib/nokogiri/xml/range.rb', line 114

def set_end_after(node)
  parent = node.parent
  raise InvalidNodeTypeError, 'parent node is empty' unless parent
  set_end(parent, parent.children.index(node) + 1)
end

#set_end_before(node) ⇒ Object


108
109
110
111
112
# File 'lib/nokogiri/xml/range.rb', line 108

def set_end_before(node)
  parent = node.parent
  raise InvalidNodeTypeError, 'parent node is empty' unless parent
  set_end(parent, parent.children.index(node))
end

#set_start(node, offset) ⇒ Object Also known as: start=


76
77
78
79
80
81
82
83
# File 'lib/nokogiri/xml/range.rb', line 76

def set_start(node, offset)
  validate_boundary_point node, offset
  if document != node.document or
    self.class.compare_points(node, offset, @end_container, @end_offset) == 1
    set_end node, offset
  end
  @start_container, @start_offset = node, offset
end

#set_start_after(node) ⇒ Object


102
103
104
105
106
# File 'lib/nokogiri/xml/range.rb', line 102

def set_start_after(node)
  parent = node.parent
  raise InvalidNodeTypeError, 'parent node is empty' unless parent
  set_start(parent, parent.children.index(node) + 1)
end

#set_start_before(node) ⇒ Object


96
97
98
99
100
# File 'lib/nokogiri/xml/range.rb', line 96

def set_start_before(node)
  parent = node.parent
  raise InvalidNodeTypeError, 'parent node is empty' unless parent
  set_start(parent, parent.children.index(node))
end

#start_pointObject


60
61
62
# File 'lib/nokogiri/xml/range.rb', line 60

def start_point
  [@start_container, @start_offset]
end

#surround_contents(new_parent) ⇒ Object

Raises:


467
468
469
470
471
472
473
474
475
# File 'lib/nokogiri/xml/range.rb', line 467

def surround_contents(new_parent)
  raise InvalidStateError unless partially_containing_nodes.all?(&:text?)
  raise InvalidNodeTypeError if [Node::DOCUMENT_NODE, Node::DOCUMENT_TYPE_NODE, Node::DOCUMENT_FRAG_NODE].include? new_parent.type
  fragment = extract_contents
  new_parent.replace_all_with nil if new_parent.child
  insert_node new_parent
  new_parent << fragment
  select_node new_parent
end

#to_sObject


536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
# File 'lib/nokogiri/xml/range.rb', line 536

def to_s
  s = ''
  if @start_container == @end_container and @start_container.text?
    return @start_container.substring_data(@start_offset, @end_offset - @start_offset)
  end
  if @start_container.text?
    s << @start_container.substring_data(@start_offset, @start_container.length)
  end
  containing_nodes.reduce s do |concatenated, node|
    concatenated << node.content if node.text?
    concatenated
  end
  if @end_container.text?
    s << @end_container.substring_data(0, @end_offset)
  end
  s
end