Module: JsRegex::SecondPass

Defined in:
lib/js_regex/second_pass.rb

Overview

After conversion of a full Regexp::Expression tree, this class checks for Node instances that need further processing.

E.g. subexpression calls (such as g<1>) can be look-ahead, so the full Regexp must have been processed first, and only then can they be substituted with the conversion result of their targeted group.

Class Method Summary collapse

Class Method Details

.alternate_conditional_permutations(tree) ⇒ Object 



36
37
38
39
40
41
42
43
44
 
# File 'lib/js_regex/second_pass.rb', line 36

def alternate_conditional_permutations(tree)
  permutations = conditional_tree_permutations(tree)
  return tree if permutations.empty?

  alternatives = permutations.map.with_index do |variant, i|
    Node.new((i.zero? ? '(?:' : '|(?:'), variant, ')')
  end
  tree.update(children: alternatives)
end

.call(tree) ⇒ Object 



15
16
17
18
19
 
# File 'lib/js_regex/second_pass.rb', line 15

def call(tree)
  substitute_subexp_calls(tree)
  alternate_conditional_permutations(tree)
  tree
end

.captured_group_count(tree) ⇒ Object 



92
93
94
95
96
 
# File 'lib/js_regex/second_pass.rb', line 92

def captured_group_count(tree)
  count = 0
  crawl(tree) { |node| count += 1 if node.type.equal?(:captured_group)}
  count
end

.condition_permutations(conditions) ⇒ Object 



98
99
100
101
102
103
104
 
# File 'lib/js_regex/second_pass.rb', line 98

def condition_permutations(conditions)
  return [] if conditions.empty?

  condition_permutations = (0..(conditions.length)).inject([]) do |arr, n|
    arr += conditions.combination(n).to_a
  end
end

.conditional_tree_permutations(tree) ⇒ Object 



53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
 
# File 'lib/js_regex/second_pass.rb', line 53

def conditional_tree_permutations(tree)
  all_conditions = conditions(tree)
  return [] if all_conditions.empty?

  captured_groups_per_branch = captured_group_count(tree)

  condition_permutations(all_conditions).map.with_index do |truthy_conds, i|
    tree_permutation = tree.clone
    # find referenced groups and conditionals and make one-sided
    crawl(tree_permutation) do |node|
      truthy = truthy_conds.include?(node.reference)

      if node.type.equal?(:captured_group) &&
         all_conditions.include?(node.reference)
        truthy ? min_quantify(node) : null_quantify(node)
      elsif node.type.equal?(:conditional)
        branches = node.children[1...-1]
        if branches.count == 1
          truthy || null_quantify(branches.first)
        else
          null_quantify(truthy ? branches.last : branches.first)
        end
        node.update(type: :plain)
      elsif node.type.equal?(:backref_num)
        new_num = node.children[0].to_i + captured_groups_per_branch * i
        node.update(children: [new_num.to_s])
      end
    end
  end
end

.conditions(tree) ⇒ Object 



84
85
86
87
88
89
90
 
# File 'lib/js_regex/second_pass.rb', line 84

def conditions(tree)
  conditions = []
  crawl(tree) do |node|
    conditions << node.reference if node.type.equal?(:conditional)
  end
  conditions
end

.crawl(node) {|node| ... } ⇒ Object

Yields:

  • (node) 


30
31
32
33
34
 
# File 'lib/js_regex/second_pass.rb', line 30

def crawl(node, &block)
  return if node.instance_of?(String)
  yield(node)
  node.children.each { |child| crawl(child, &block) }
end

.find_group_by_reference(ref, in_node: nil) ⇒ Object 



46
47
48
49
50
51
 
# File 'lib/js_regex/second_pass.rb', line 46

def find_group_by_reference(ref, in_node: nil)
  crawl(in_node) do |node|
    return node if node.type == :captured_group && node.reference == ref
  end
  Node.new('()')
end

.min_quantify(node) ⇒ Object 



106
107
108
109
110
111
112
113
114
 
# File 'lib/js_regex/second_pass.rb', line 106

def min_quantify(node)
  return if (qtf = node.quantifier).nil? || qtf.min > 0

  if qtf.max.equal?(1) # any zero_or_one quantifier (?, ??, ?+)
    node.update(quantifier: nil)
  else
    node.update(quantifier: "{1,#{qtf.max}}#{'?' if qtf.reluctant?}")
  end
end

.null_quantify(node) ⇒ Object 



116
117
118
 
# File 'lib/js_regex/second_pass.rb', line 116

def null_quantify(node)
  node.update(quantifier: '{0}')
end

.substitute_subexp_calls(tree) ⇒ Object 



21
22
23
24
25
26
27
28
 
# File 'lib/js_regex/second_pass.rb', line 21

def substitute_subexp_calls(tree)
  crawl(tree) do |node|
    if node.type == :subexp_call
      called_group = find_group_by_reference(node.reference, in_node: tree)
      node.update(children: called_group.children, type: :captured_group)
    end
  end
end