Class: Nokolexbor::NodeSet

Inherits:
Node
  • Object
show all
Includes:
Enumerable
Defined in:
lib/nokolexbor/node_set.rb,
ext/nokolexbor/nl_node_set.c

Constant Summary

Constants inherited from Node

Nokolexbor::Node::ATTRIBUTE_NODE, Nokolexbor::Node::CDATA_SECTION_NODE, Nokolexbor::Node::COMMENT_NODE, Nokolexbor::Node::DOCUMENT_FRAG_NODE, Nokolexbor::Node::DOCUMENT_NODE, Nokolexbor::Node::DOCUMENT_TYPE_NODE, Nokolexbor::Node::ELEMENT_NODE, Nokolexbor::Node::ENTITY_NODE, Nokolexbor::Node::ENTITY_REF_NODE, Nokolexbor::Node::LOOKS_LIKE_XPATH, Nokolexbor::Node::NOTATION_NODE, Nokolexbor::Node::PI_NODE, Nokolexbor::Node::TEXT_NODE

Instance Attribute Summary

Attributes inherited from Node

#document

Class Method Summary collapse

Instance Method Summary collapse

Methods inherited from Node

#[]=, #add_child, #add_next_sibling, #add_previous_sibling, #add_sibling, #ancestors, #at, #at_css_impl, #at_xpath, #attribute_nodes, #attributes, #attrs, #cdata?, #child, #children=, #classes, #clone, #comment?, #content=, #css_impl, #css_path, #document?, #element?, #element_children, #first_element_child, #fragment, #fragment?, #key?, #keys, #kwattr_add, #kwattr_append, #kwattr_remove, #kwattr_values, #last_element_child, #matches?, #name, #next, #next_element, #node_type, #nokogiri_at_css, #parent, #parent=, #parse, #path, #pointer_id, #prepend_child, #previous, #previous_element, #processing_instruction?, #replace, #search, #source_location, #swap, #text?, #traverse, #value?, #values, #write_to

Class Method Details

.new(document, list = []) {|Document| ... } ⇒ Document

Create a NodeSet with document defaulting to list.

Yields:

Returns:



12
13
14
15
16
17
18
19
20
21
22
23
24
 
# File 'lib/nokolexbor/node_set.rb', line 12

def self.new(document, list = [])
  obj = allocate
  if document.is_a?(Document) || document.nil?
    obj.instance_variable_set(:@document, document)
  elsif document.is_a?(Node)
    obj.instance_variable_set(:@document, document.document)
  else
    raise ArgumentError, "Expected a Document or Node, got #{document.class}"
  end
  list.each { |x| obj << x }
  yield obj if block_given?
  obj
end

Instance Method Details

#&(other) ⇒ NodeSet

Returns A new NodeSet with the common nodes only.

Returns:

  • (NodeSet) —

    A new NodeSet with the common nodes only.



306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
 
# File 'ext/nokolexbor/nl_node_set.c', line 306

static VALUE
nl_node_set_intersection(VALUE self, VALUE other)
{
  if (!rb_obj_is_kind_of(other, cNokolexborNodeSet)) {
    rb_raise(rb_eArgError, "Parameter must be a Nokolexbor::NodeSet");
  }

  lexbor_array_t *self_array = nl_rb_node_set_unwrap(self);
  lexbor_array_t *other_array = nl_rb_node_set_unwrap(other);

  lexbor_array_t *new_array = lexbor_array_create();

  for (size_t i = 0; i < self_array->length; i++) {
    for (size_t j = 0; j < other_array->length; j++) {
      if (self_array->list[i] == other_array->list[j]) {
        lexbor_array_push(new_array, self_array->list[i]);
        break;
      }
    }
  }

  return nl_rb_node_set_create_with_data(new_array, nl_rb_document_get(self));
}

#-(other) ⇒ NodeSet

Returns A new NodeSet with the nodes in this NodeSet that aren't in other.

Returns:

  • (NodeSet) —

    A new NodeSet with the nodes in this NodeSet that aren't in other



333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
 
# File 'ext/nokolexbor/nl_node_set.c', line 333

static VALUE
nl_node_set_difference(VALUE self, VALUE other)
{
  if (!rb_obj_is_kind_of(other, cNokolexborNodeSet)) {
    rb_raise(rb_eArgError, "Parameter must be a Nokolexbor::NodeSet");
  }

  lexbor_array_t *self_array = nl_rb_node_set_unwrap(self);
  lexbor_array_t *other_array = nl_rb_node_set_unwrap(other);

  lexbor_array_t *new_array = lexbor_array_create();

  for (size_t i = 0; i < self_array->length; i++) {
    bool found = false;
    for (size_t j = 0; j < other_array->length; j++) {
      if (self_array->list[i] == other_array->list[j]) {
        found = true;
        break;
      }
    }
    if (!found) {
      lexbor_array_push(new_array, self_array->list[i]);
    }
  }

  return nl_rb_node_set_create_with_data(new_array, nl_rb_document_get(self));
}

#==(other) ⇒ Boolean

Returns true if two NodeSets contain the same number of elements and each element is equal to the corresponding element in the other NodeSet.

Returns:

  • (Boolean) —

    true if two NodeSets contain the same number of elements and each element is equal to the corresponding element in the other NodeSet.



148
149
150
151
152
153
154
155
156
 
# File 'lib/nokolexbor/node_set.rb', line 148

def ==(other)
  return false unless other.is_a?(NodeSet)
  return false unless length == other.length

  each_with_index do |node, i|
    return false unless node == other[i]
  end
  true
end

#[](index) ⇒ Node? #[](start, length) ⇒ NodeSet? #[](range) ⇒ NodeSet? Also known as: slice

The Nokolexbor::Node at index, or returns a Nokolexbor::NodeSet containing nodes starting at start and continuing for length elements, or returns a Nokolexbor::NodeSet containing nodes specified by range. Negative indices count backward from the end of the node_set (-1 is the last node). Returns nil if the index (or start) are out of range.

Overloads:

Returns:

  • (Node, NodeSet, nil) —

    the Nokolexbor::Node at index, or returns a Nokolexbor::NodeSet containing nodes starting at start and continuing for length elements, or returns a Nokolexbor::NodeSet containing nodes specified by range. Negative indices count backward from the end of the node_set (-1 is the last node). Returns nil if the index (or start) are out of range.



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
 
# File 'ext/nokolexbor/nl_node_set.c', line 212

static VALUE
nl_node_set_slice(int argc, VALUE *argv, VALUE self)
{
  VALUE arg;
  long beg, len;

  lexbor_array_t *array = nl_rb_node_set_unwrap(self);

  if (argc == 2) {
    beg = NUM2LONG(argv[0]);
    len = NUM2LONG(argv[1]);
    if (beg < 0) {
      beg += array->length;
    }
    return nl_node_set_subseq(self, beg, len);
  }

  if (argc != 1) {
    rb_scan_args(argc, argv, "11", NULL, NULL);
  }
  arg = argv[0];

  if (FIXNUM_P(arg)) {
    return nl_node_set_index_at(self, FIX2LONG(arg));
  }

  /* if arg is Range */
  switch (rb_range_beg_len(arg, &beg, &len, array->length, 0)) {
  case Qfalse:
    break;
  case Qnil:
    return Qnil;
  default:
    return nl_node_set_subseq(self, beg, len);
  }

  return nl_node_set_index_at(self, NUM2LONG(arg));
}

#add_class(name) ⇒ Object

Add the class attribute name to all containing nodes.

See Also:



191
192
193
194
195
196
 
# File 'lib/nokolexbor/node_set.rb', line 191

def add_class(name)
  each do |el|
    el.add_class(name)
  end
  self
end

#after(node) ⇒ Object

Insert node after the last Node in this NodeSet



69
70
71
 
# File 'lib/nokolexbor/node_set.rb', line 69

def after(node)
  last.after(node)
end

#append_class(name) ⇒ Object

Append the class attribute name to all containing nodes.

See Also:



201
202
203
204
205
206
 
# File 'lib/nokolexbor/node_set.rb', line 201

def append_class(name)
  each do |el|
    el.append_class(name)
  end
  self
end

#at_css(selector) ⇒ Node?

Like #css, but returns the first match.

This method uses Lexbor as the selector engine. Its performance is much higher than Nokolexbor::Node#at_xpath or Nokolexbor::Node#nokogiri_at_css.

Returns:

  • (Node, nil) —

    The first matched Node.

See Also:



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
 
# File 'ext/nokolexbor/nl_node_set.c', line 414

static VALUE
nl_node_set_at_css(VALUE self, VALUE selector)
{
  lexbor_array_t *array = lexbor_array_create();
  lxb_dom_document_t *doc = nl_rb_document_unwrap(nl_rb_document_get(self));

  lxb_status_t status = nl_node_set_find(self, selector, nl_node_at_css_callback, array);

  if (status != LXB_STATUS_OK) {
    lexbor_array_destroy(array, true);
    nl_raise_lexbor_error(status);
  }

  if (array->length == 0) {
    lexbor_array_destroy(array, true);
    return Qnil;
  }

  nl_sort_nodes_if_necessary(selector, doc, array);

  VALUE ret = nl_rb_node_create(array->list[0], nl_rb_document_get(self));

  lexbor_array_destroy(array, true);

  return ret;
}

#attr(key, value = nil, &block) ⇒ NodeSet Also known as: set, attribute

Set attributes on each Node in the NodeSet, or get an attribute from the first Node in the NodeSet.

Examples:

Get an attribute from the first Node in a NodeSet.

node_set.attr("href")

Set attributes on each node.

node_set.attr("href" => "http://example.com", "class" => "a")
node_set.attr("href", "http://example.com")
node_set.attr("href") { |node| "http://example.com" }

Returns:

  • (NodeSet) —

    self, to support chaining of calls.



239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
 
# File 'lib/nokolexbor/node_set.rb', line 239

def attr(key, value = nil, &block)
  unless key.is_a?(Hash) || (key && (value || block))
    return first&.attribute(key)
  end

  hash = key.is_a?(Hash) ? key : { key => value }

  hash.each do |k, v|
    each do |node|
      node[k] = v || yield(node)
    end
  end

  self
end

#before(node) ⇒ Object

Insert node before the first Node in this NodeSet



64
65
66
 
# File 'lib/nokolexbor/node_set.rb', line 64

def before(node)
  first.before(node)
end

#children ⇒ NodeSet

Returns A new NodeSet containing all the children of all the nodes in the NodeSet.

Returns:

  • (NodeSet) —

    A new NodeSet containing all the children of all the nodes in the NodeSet.



160
161
162
163
164
165
166
 
# File 'lib/nokolexbor/node_set.rb', line 160

def children
  node_set = NodeSet.new(@document)
  each do |node|
    node.children.each { |n| node_set.push(n) }
  end
  node_set
end

#content ⇒ String Also known as: text, inner_text, to_str

Get the content of all contained Nodes.

Returns:

  • (String) 


86
87
88
 
# File 'lib/nokolexbor/node_set.rb', line 86

def content
  self.map(&:content).join
end

#css(selector) ⇒ NodeSet

Search this object for CSS rules. rules must be one or more CSS selectors.

This method uses Lexbor as the selector engine. Its performance is much higher than #xpath or #nokogiri_css.

Examples:

node.css('title')
node.css('body h1.bold')
node.css('div + p.green', 'div#one')

Returns:

  • (NodeSet) —

    The matched set of Nodes.

See Also:



444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
 
# File 'ext/nokolexbor/nl_node_set.c', line 444

static VALUE
nl_node_set_css(VALUE self, VALUE selector)
{
  lexbor_array_t *array = lexbor_array_create();
  lxb_dom_document_t *doc = nl_rb_document_unwrap(nl_rb_document_get(self));

  lxb_status_t status = nl_node_set_find(self, selector, nl_node_css_callback, array);
  if (status != LXB_STATUS_OK) {
    lexbor_array_destroy(array, true);
    nl_raise_lexbor_error(status);
  }

  nl_sort_nodes_if_necessary(selector, doc, array);

  return nl_rb_node_set_create_with_data(array, nl_rb_document_get(self));
}

#delete(node) ⇒ Node?

Delete node from the NodeSet.

Parameters:

Returns:

  • (Node, nil) —

    The deleted node if found, otherwise returns nil.



111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
 
# File 'ext/nokolexbor/nl_node_set.c', line 111

static VALUE
nl_node_set_delete(VALUE self, VALUE rb_node)
{
  lexbor_array_t *array = nl_rb_node_set_unwrap(self);
  lxb_dom_node_t *node = nl_rb_node_unwrap(rb_node);

  size_t i;
  for (i = 0; i < array->length; i++)
    if (array->list[i] == node) {
      break;
    }

  if (i >= array->length) {
    // not found
    return Qnil;
  }
  lexbor_array_delete(array, i, 1);
  return rb_node;
}

#destroy ⇒ Object

Destroy all nodes in the NodeSet.

See Also:



125
126
127
 
# File 'lib/nokolexbor/node_set.rb', line 125

def destroy
  self.each(&:destroy)
end

#each {|Node| ... } ⇒ Object

Iterate over each node.

Yields:



29
30
31
32
33
34
35
36
 
# File 'lib/nokolexbor/node_set.rb', line 29

def each
  return to_enum unless block_given?

  0.upto(length - 1) do |x|
    yield self[x]
  end
  self
end

#empty? ⇒ Boolean

Returns true if this NodeSet is empty.

Returns:

  • (Boolean) —

    true if this NodeSet is empty.



59
60
61
 
# File 'lib/nokolexbor/node_set.rb', line 59

def empty?
  length == 0
end

#first(n = nil) ⇒ Node+

Get the first n elements of the NodeSet.

Parameters:

  • n (Numeric, nil) (defaults to: nil)

Returns:



43
44
45
46
47
48
49
 
# File 'lib/nokolexbor/node_set.rb', line 43

def first(n = nil)
  return self[0] unless n

  list = []
  [n, length].min.times { |i| list << self[i] }
  list
end

#include?(node) ⇒ Boolean

Returns true if any member of this NodeSet equals node.

Returns:

  • (Boolean) —

    true if any member of this NodeSet equals node.



137
138
139
140
141
142
143
144
145
146
147
148
149
 
# File 'ext/nokolexbor/nl_node_set.c', line 137

static VALUE
nl_node_set_is_include(VALUE self, VALUE rb_node)
{
  lexbor_array_t *array = nl_rb_node_set_unwrap(self);
  lxb_dom_node_t *node = nl_rb_node_unwrap(rb_node);

  for (size_t i = 0; i < array->length; i++)
    if (array->list[i] == node) {
      return Qtrue;
    }

  return Qfalse;
}

#index(node = nil) ⇒ Integer

Returns The index of the first node in this NodeSet that is equal to node or meets the given block. Returns nil if no match is found.

Returns:

  • (Integer) —

    The index of the first node in this NodeSet that is equal to node or meets the given block. Returns nil if no match is found.



74
75
76
77
78
79
80
81
 
# File 'lib/nokolexbor/node_set.rb', line 74

def index(node = nil)
  if node
    each_with_index { |member, j| return j if member == node }
  elsif block_given?
    each_with_index { |member, j| return j if yield(member) }
  end
  nil
end

#inner_html(*args) ⇒ String

Get the inner html of all contained Nodes.

Returns:

  • (String) 


97
98
99
 
# File 'lib/nokolexbor/node_set.rb', line 97

def inner_html(*args)
  self.map { |n| n.inner_html(*args) }.join
end

#inspect ⇒ Object 



284
285
286
 
# File 'lib/nokolexbor/node_set.rb', line 284

def inspect
  "[#{map(&:inspect).join(', ')}]"
end

#last ⇒ Node?

Get the last element of the NodeSet.

Returns:



54
55
56
 
# File 'lib/nokolexbor/node_set.rb', line 54

def last
  self[-1]
end

#length ⇒ Integer Also known as: size

Get the length of this NodeSet.

Returns:

  • (Integer) 


71
72
73
74
75
 
# File 'ext/nokolexbor/nl_node_set.c', line 71

static VALUE
nl_node_set_length(VALUE self)
{
  return INT2NUM(nl_rb_node_set_unwrap(self)->length);
}

#nokogiri_css(*args) ⇒ NodeSet

Search this object for CSS rules. rules must be one or more CSS selectors. It supports a mixed syntax of CSS selectors and XPath.

This method uses libxml2 as the selector engine. It works the same way as Nokogiri::Node#css.

Returns:

  • (NodeSet) —

    The matched set of Nodes.

See Also:



271
272
273
274
275
276
277
278
279
280
281
282
 
# File 'lib/nokolexbor/node_set.rb', line 271

def nokogiri_css(*args)
  rules, handler, ns, _ = extract_params(args)
  paths = css_rules_to_xpath(rules, ns)

  NodeSet.new(@document) do |set|
    each do |node|
      node.send(:xpath_internal, node, paths, handler, ns, nil).each do |inner_node|
        set << inner_node
      end
    end
  end
end

#outer_html(*args) ⇒ String Also known as: to_s, to_html, serialize

Convert this NodeSet to HTML.

Returns:

  • (String) 


104
105
106
 
# File 'lib/nokolexbor/node_set.rb', line 104

def outer_html(*args)
  self.map { |n| n.outer_html(*args) }.join
end

#pop ⇒ Node?

Returns The last element of this NodeSet and removes it. Returns nil if the set is empty.

Returns:

  • (Node, nil) —

    The last element of this NodeSet and removes it. Returns nil if the set is empty.



131
132
133
134
135
 
# File 'lib/nokolexbor/node_set.rb', line 131

def pop
  return nil if length == 0

  delete(last)
end

#push(node) ⇒ NodeSet Also known as: <<

Append node to the NodeSet.

Parameters:

Returns:

  • (NodeSet) —

    self, to support chaining of calls.



87
88
89
90
91
92
93
94
95
96
97
98
99
 
# File 'ext/nokolexbor/nl_node_set.c', line 87

static VALUE
nl_node_set_push(VALUE self, VALUE rb_node)
{
  lexbor_array_t *array = nl_rb_node_set_unwrap(self);
  lxb_dom_node_t *node = nl_rb_node_unwrap(rb_node);

  lxb_status_t status = lexbor_array_push_unique(array, node);
  if (status != LXB_STATUS_OK && status != LXB_STATUS_STOPPED) {
    nl_raise_lexbor_error(status);
  }

  return self;
}

#remove ⇒ Object Also known as: unlink

Remove all nodes in this NodeSet.

See Also:



115
116
117
 
# File 'lib/nokolexbor/node_set.rb', line 115

def remove
  self.each(&:remove)
end

#remove_attr(name) ⇒ Object Also known as: remove_attribute

Remove the attributed named name from all containing nodes.

See Also:



221
222
223
224
 
# File 'lib/nokolexbor/node_set.rb', line 221

def remove_attr(name)
  each { |el| el.delete(name) }
  self
end

#remove_class(name = nil) ⇒ Object

Remove the class attribute name from all containing nodes.

See Also:



211
212
213
214
215
216
 
# File 'lib/nokolexbor/node_set.rb', line 211

def remove_class(name = nil)
  each do |el|
    el.remove_class(name)
  end
  self
end

#reverse ⇒ NodeSet

Returns A new NodeSet containing all the nodes in the NodeSet in reverse order.

Returns:

  • (NodeSet) —

    A new NodeSet containing all the nodes in the NodeSet in reverse order.



170
171
172
173
174
175
176
 
# File 'lib/nokolexbor/node_set.rb', line 170

def reverse
  node_set = NodeSet.new(@document)
  (length - 1).downto(0) do |x|
    node_set.push(self[x])
  end
  node_set
end

#shift ⇒ Node?

Returns The first element of this NodeSet and removes it. Returns nil if the set is empty.

Returns:

  • (Node, nil) —

    The first element of this NodeSet and removes it. Returns nil if the set is empty.



139
140
141
142
143
 
# File 'lib/nokolexbor/node_set.rb', line 139

def shift
  return nil if length == 0

  delete(first)
end

#to_a ⇒ Array<Node> Also known as: to_ary

Returns This list as an Array.

Returns:

  • (Array<Node>) —

    This list as an Array



254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
 
# File 'ext/nokolexbor/nl_node_set.c', line 254

static VALUE
nl_node_set_to_array(VALUE self)
{
  lexbor_array_t *array = nl_rb_node_set_unwrap(self);

  VALUE list = rb_ary_new2(array->length);
  VALUE doc = nl_rb_document_get(self);
  for (size_t i = 0; i < array->length; i++) {
    lxb_dom_node_t *node = (lxb_dom_node_t *)array->list[i];
    VALUE rb_node = nl_rb_node_create(node, doc);
    rb_ary_push(list, rb_node);
  }

  return list;
}

#wrap(node_or_tags) ⇒ NodeSet

Wrap all nodes of this NodeSet with node_or_tags.

Returns:

  • (NodeSet) —

    self, to support chaining.

See Also:



183
184
185
186
 
# File 'lib/nokolexbor/node_set.rb', line 183

def wrap(node_or_tags)
  map { |node| node.wrap(node_or_tags) }
  self
end

#xpath(*args) ⇒ NodeSet

Search this node for XPath paths. paths must be one or more XPath queries.

It works the same way as Nokogiri::Node#xpath.

Examples:

node.xpath('.//title')

Returns:

  • (NodeSet) —

    The matched set of Nodes.



258
259
260
261
262
263
264
265
266
267
268
 
# File 'lib/nokolexbor/node_set.rb', line 258

def xpath(*args)
  paths, handler, ns, binds = extract_params(args)

  NodeSet.new(@document) do |set|
    each do |node|
      node.send(:xpath_internal, node, paths, handler, ns, binds).each do |inner_node|
        set << inner_node
      end
    end
  end
end

#|(other) ⇒ NodeSet Also known as: +

Returns A new set built by merging the other set, excluding duplicates.

Returns:

  • (NodeSet) —

    A new set built by merging the other set, excluding duplicates.



273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
 
# File 'ext/nokolexbor/nl_node_set.c', line 273

static VALUE
nl_node_set_union(VALUE self, VALUE other)
{
  if (!rb_obj_is_kind_of(other, cNokolexborNodeSet)) {
    rb_raise(rb_eArgError, "Parameter must be a Nokolexbor::NodeSet");
  }

  lexbor_array_t *self_array = nl_rb_node_set_unwrap(self);
  lexbor_array_t *other_array = nl_rb_node_set_unwrap(other);

  if (self_array->length + other_array->length == 0) {
    return nl_rb_node_set_create_with_data(NULL, nl_rb_document_get(self));
  }

  lexbor_array_t *new_array = lexbor_array_create();
  lxb_status_t status = lexbor_array_init(new_array, self_array->length + other_array->length);
  if (status != LXB_STATUS_OK) {
    nl_raise_lexbor_error(status);
  }

  memcpy(new_array->list, self_array->list, sizeof(lxb_dom_node_t *) * self_array->length);
  new_array->length = self_array->length;

  for (size_t i = 0; i < other_array->length; i++) {
    lexbor_array_push_unique(new_array, other_array->list[i]);
  }

  return nl_rb_node_set_create_with_data(new_array, nl_rb_document_get(self));
}