nokolexbor 0.7.0-aarch64-linux

data/lib/nokolexbor/node_set.rb

@@ -0,0 +1,293 @@
+# frozen_string_literal: true
+
+module Nokolexbor
+  class NodeSet < Nokolexbor::Node
+    include Enumerable
+
+    # Create a NodeSet with +document+ defaulting to +list+.
+    #
+    # @yield [Document]
+    #
+    # @return [Document]
+    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
+
+    # Iterate over each node.
+    #
+    # @yield [Node]
+    def each
+      return to_enum unless block_given?
+
+      0.upto(length - 1) do |x|
+        yield self[x]
+      end
+      self
+    end
+
+    # Get the first +n+ elements of the NodeSet.
+    #
+    # @param n [Numeric,nil]
+    #
+    # @return [Node,Array<Node>] {Node} if +n+ is nil, otherwise {Array<Node>}
+    def first(n = nil)
+      return self[0] unless n
+
+      list = []
+      [n, length].min.times { |i| list << self[i] }
+      list
+    end
+
+    # Get the last element of the NodeSet.
+    #
+    # @return [Node,nil]
+    def last
+      self[-1]
+    end
+
+    # @return [Boolean] true if this NodeSet is empty.
+    def empty?
+      length == 0
+    end
+
+    # Insert +node+ before the first Node in this NodeSet
+    def before(node)
+      first.before(node)
+    end
+
+    # Insert +node+ after the last Node in this NodeSet
+    def after(node)
+      last.after(node)
+    end
+
+    # @return [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.
+    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
+
+    # Get the content of all contained Nodes.
+    #
+    # @return [String]
+    def content
+      self.map(&:content).join
+    end
+
+    alias_method :text, :content
+    alias_method :inner_text, :content
+    alias_method :to_str, :content
+
+    # Get the inner html of all contained Nodes.
+    #
+    # @return [String]
+    def inner_html(*args)
+      self.map { |n| n.inner_html(*args) }.join
+    end
+
+    # Convert this NodeSet to HTML.
+    #
+    # @return [String]
+    def outer_html(*args)
+      self.map { |n| n.outer_html(*args) }.join
+    end
+
+    alias_method :to_s, :outer_html
+    alias_method :to_html, :outer_html
+    alias_method :serialize, :outer_html
+
+    # Remove all nodes in this NodeSet.
+    #
+    # @see Node#remove
+    def remove
+      self.each(&:remove)
+    end
+
+    alias_method :unlink, :remove
+    alias_method :to_ary, :to_a
+
+    # Destroy all nodes in the NodeSet.
+    #
+    # @see Node#destroy
+    def destroy
+      self.each(&:destroy)
+    end
+
+    # @return [Node,nil] The last element of this NodeSet and removes it. Returns
+    #   +nil+ if the set is empty.
+    def pop
+      return nil if length == 0
+
+      delete(last)
+    end
+
+    # @return [Node,nil] The first element of this NodeSet and removes it. Returns
+    #   +nil+ if the set is empty.
+    def shift
+      return nil if length == 0
+
+      delete(first)
+    end
+
+    # @return [Boolean] true if two NodeSets contain the same number
+    #   of elements and each element is equal to the corresponding
+    #   element in the other NodeSet.
+    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
+
+    # @return [NodeSet] A new NodeSet containing all the children of all the nodes in
+    #   the NodeSet.
+    def children
+      node_set = NodeSet.new(@document)
+      each do |node|
+        node.children.each { |n| node_set.push(n) }
+      end
+      node_set
+    end
+
+    # @return [NodeSet] A new NodeSet containing all the nodes in the NodeSet
+    #   in reverse order.
+    def reverse
+      node_set = NodeSet.new(@document)
+      (length - 1).downto(0) do |x|
+        node_set.push(self[x])
+      end
+      node_set
+    end
+
+    # Wrap all nodes of this NodeSet with +node_or_tags+.
+    #
+    # @see Node#wrap
+    #
+    # @return [NodeSet] +self+, to support chaining.
+    def wrap(node_or_tags)
+      map { |node| node.wrap(node_or_tags) }
+      self
+    end
+
+    # Add the class attribute +name+ to all containing nodes.
+    #
+    # @see Node#add_class
+    def add_class(name)
+      each do |el|
+        el.add_class(name)
+      end
+      self
+    end
+
+    # Append the class attribute +name+ to all containing nodes.
+    #
+    # @see Node#append_class
+    def append_class(name)
+      each do |el|
+        el.append_class(name)
+      end
+      self
+    end
+
+    # Remove the class attribute +name+ from all containing nodes.
+    #
+    # @see Node#remove_class
+    def remove_class(name = nil)
+      each do |el|
+        el.remove_class(name)
+      end
+      self
+    end
+
+    # Remove the attributed named +name+ from all containing nodes.
+    #
+    # @see Node#remove_attr
+    def remove_attr(name)
+      each { |el| el.delete(name) }
+      self
+    end
+    alias_method :remove_attribute, :remove_attr
+
+    # Set attributes on each Node in the NodeSet, or get an
+    # attribute from the first Node in the NodeSet.
+    #
+    # @example Get an attribute from the first Node in a NodeSet.
+    #   node_set.attr("href")
+    #
+    # @example 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" }
+    #
+    # @return [NodeSet] +self+, to support chaining of calls.
+    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
+    alias_method :set, :attr
+    alias_method :attribute, :attr
+
+    # (see Node#xpath)
+    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
+
+    # (see Node#nokogiri_css)
+    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
+
+    def inspect
+      "[#{map(&:inspect).join(', ')}]"
+    end
+
+    private
+
+    IMPLIED_XPATH_CONTEXTS = [".//", "self::"].freeze # :nodoc:
+      
+  end
+end

data/lib/nokolexbor/version.rb

@@ -0,0 +1,5 @@
+# frozen_string_literal: true
+
+module Nokolexbor
+  VERSION = '0.7.0'
+end

data/lib/nokolexbor/xpath.rb

@@ -0,0 +1,69 @@
+# frozen_string_literal: true
+
+module Nokolexbor
+  module XPath
+    class SyntaxError < StandardError
+      attr_reader :domain
+      attr_reader :code
+      attr_reader :level
+      attr_reader :file
+      attr_reader :line
+      attr_reader :str1
+      attr_reader :str2
+      attr_reader :str3
+      attr_reader :int1
+      attr_reader :column
+
+      ###
+      # return true if this is a non error
+      def none?
+        level == 0
+      end
+
+      ###
+      # return true if this is a warning
+      def warning?
+        level == 1
+      end
+
+      ###
+      # return true if this is an error
+      def error?
+        level == 2
+      end
+
+      ###
+      # return true if this error is fatal
+      def fatal?
+        level == 3
+      end
+
+      def to_s
+        message = super.chomp
+        [location_to_s, level_to_s, message]
+          .compact.join(": ")
+          .force_encoding(message.encoding)
+      end
+
+      private
+
+      def level_to_s
+        case level
+        when 3 then "FATAL"
+        when 2 then "ERROR"
+        when 1 then "WARNING"
+        end
+      end
+
+      def nil_or_zero?(attribute)
+        attribute.nil? || attribute.zero?
+      end
+
+      def location_to_s
+        return nil if nil_or_zero?(line) && nil_or_zero?(column)
+
+        "#{line}:#{column}"
+      end
+    end
+  end
+end

data/lib/nokolexbor/xpath_context.rb

@@ -0,0 +1,14 @@
+# frozen_string_literal: true
+
+module Nokolexbor
+  class XPathContext
+    ###
+    # Register namespaces in +namespaces+
+    def register_namespaces(namespaces)
+      namespaces.each do |k, v|
+        k = k.to_s.gsub(/.*:/, "") # strip off 'xmlns:' or 'xml:'
+        register_ns(k, v)
+      end
+    end
+  end
+end

data/lib/nokolexbor.rb

@@ -0,0 +1,50 @@
+# frozen_string_literal: true
+
+begin
+  # pre-compiled extension by rake-compiler is located inside lib/nokolexbor/<ruby_version>/
+  RUBY_VERSION =~ /(\d+\.\d+)/
+  require "nokolexbor/#{Regexp.last_match(1)}/nokolexbor"
+rescue LoadError => e
+  if /GLIBC/.match?(e.message)
+    warn(<<~EOM)
+      ERROR: It looks like you're trying to use Nokolexbor as a precompiled native gem on a system
+             with an unsupported version of glibc.
+        #{e.message}
+        If that's the case, then please install Nokolexbor via the `ruby` platform gem:
+            gem install nokolexbor --platform=ruby
+        or:
+            bundle config set force_ruby_platform true
+    EOM
+    raise e
+  end
+
+  require 'nokolexbor/nokolexbor'
+end
+
+require 'nokolexbor/version'
+require 'nokolexbor/node'
+require 'nokolexbor/document'
+require 'nokolexbor/node_set'
+require 'nokolexbor/document_fragment'
+require 'nokolexbor/xpath'
+require 'nokolexbor/xpath_context'
+require 'nokolexbor/builder'
+
+module Nokolexbor
+  class << self
+    def parse(*args)
+      Document.parse(*args)
+    end
+
+    alias_method :HTML, :parse
+  end
+end
+
+# Parse an HTML document, or build one with the DSL when a block is given.
+def Nokolexbor(string_or_io = nil, &block)
+  if block
+    Nokolexbor::Builder.new(&block).parent
+  else
+    Nokolexbor.parse(string_or_io)
+  end
+end

metadata

@@ -0,0 +1,85 @@
+--- !ruby/object:Gem::Specification
+name: nokolexbor
+version: !ruby/object:Gem::Version
+  version: 0.7.0
+platform: aarch64-linux
+authors:
+- Yicheng Zhou
+bindir: bin
+cert_chain: []
+date: 2026-03-25 00:00:00.000000000 Z
+dependencies:
+- !ruby/object:Gem::Dependency
+  name: rake-compiler
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '1.0'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '1.0'
+- !ruby/object:Gem::Dependency
+  name: minitest
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '5.0'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '5.0'
+description: Nokolexbor is a high-performance HTML5 parser, with support for both
+  CSS selectors and XPath. It's API is designed to be compatible with Nokogiri.
+email: zyc9012@gmail.com
+executables: []
+extensions: []
+extra_rdoc_files: []
+files:
+- lib/nokolexbor.rb
+- lib/nokolexbor/3.1/nokolexbor.so
+- lib/nokolexbor/3.2/nokolexbor.so
+- lib/nokolexbor/3.3/nokolexbor.so
+- lib/nokolexbor/3.4/nokolexbor.so
+- lib/nokolexbor/4.0/nokolexbor.so
+- lib/nokolexbor/builder.rb
+- lib/nokolexbor/document.rb
+- lib/nokolexbor/document_fragment.rb
+- lib/nokolexbor/node.rb
+- lib/nokolexbor/node_set.rb
+- lib/nokolexbor/version.rb
+- lib/nokolexbor/xpath.rb
+- lib/nokolexbor/xpath_context.rb
+homepage: https://github.com/serpapi/nokolexbor
+licenses:
+- MIT
+metadata: {}
+rdoc_options: []
+require_paths:
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '3.1'
+  - - "<"
+    - !ruby/object:Gem::Version
+      version: 4.1.dev
+required_rubygems_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '0'
+requirements: []
+rubygems_version: 4.0.3
+specification_version: 4
+summary: High-performance HTML5 parser, with support for both CSS selectors and XPath.
+test_files: []