Module: Oselvar::Var::Core::Matcher

Defined in:
lib/oselvar/var/core/matcher.rb

Overview

Cucumber-expression matching. Port of matcher.ts. cucumber-expressions' regexps are anchored (^...$) and its group offsets are code-point based, so we strip anchors for substring search and convert offsets to UTF-16.

Class Method Summary collapse

Class Method Details

.find_hits(sentence, registry) ⇒ Object

Every expression match found anywhere in sentence.



41
42
43
44
45
46
47
48
49
50
51
52
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
 
# File 'lib/oselvar/var/core/matcher.rb', line 41

def find_hits(sentence, registry)
  hits = []
  registry.steps.each do |step|
    pattern = unanchored_pattern(step)
    pos = 0
    while pos <= sentence.length
      m = pattern.match(sentence, pos)
      break if m.nil?

      matched_text = m[0]
      arguments = step.compiled.match(matched_text) || []
      args = arguments.map { |arg| arg.value(nil) }
      formats = arguments.map { |arg| registry.formats[arg.parameter_type.name] }

      # group.start/.end are code-point offsets within matched_text; add
      # m.begin(0) for the sentence-absolute code-point index, then to UTF-16.
      param_spans = arguments.filter_map do |arg|
        g = arg.group
        next unless g.start.is_a?(Integer) && g.end.is_a?(Integer)

        ParamSpan.new(
          start: Offsets.to_utf16_offset(sentence, m.begin(0) + g.start),
          end: Offsets.to_utf16_offset(sentence, m.begin(0) + g.end)
        )
      end

      hits << Hit.new(
        expression: step.expression,
        step_def: step,
        match_start: Offsets.to_utf16_offset(sentence, m.begin(0)),
        match_end: Offsets.to_utf16_offset(sentence, m.end(0)),
        args: args,
        param_spans: param_spans,
        formats: formats
      )

      pos = matched_text.empty? ? m.begin(0) + 1 : m.end(0)
    end
  end
  hits
end

.resolve_hits(hits) ⇒ Object

Select the best non-overlapping hits, or report ambiguities.



84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
 
# File 'lib/oselvar/var/core/matcher.rb', line 84

def resolve_hits(hits)
  return ResolvedSteps.new(kind: 'ok') if hits.empty?

  # Stable sort by (match_start asc, length desc); the original index
  # breaks ties so equal-key order follows registration order (Ruby's
  # sort_by is not stable, Python's sorted is).
  sorted = hits.each_with_index.sort_by do |h, i|
    [h.match_start, -(h.match_end - h.match_start), i]
  end.map(&:first)

  collisions = []
  i = 0
  while i < sorted.length
    here = sorted[i]
    here_len = here.match_end - here.match_start
    tied = [here]
    j = i + 1
    while j < sorted.length
      candidate = sorted[j]
      if candidate.match_start == here.match_start &&
         candidate.match_end - candidate.match_start == here_len
        tied << candidate
        j += 1
      else
        break
      end
    end
    if tied.length > 1
      collisions << AmbiguityCollision.new(
        match_start: here.match_start, match_end: here.match_end, candidates: tied
      )
    end
    i = j
  end

  return ResolvedSteps.new(kind: 'ambiguous', collisions: collisions) unless collisions.empty?

  steps = []
  cursor = -1
  sorted.each do |hit|
    next if hit.match_start < cursor

    steps << hit
    cursor = hit.match_end
  end
  ResolvedSteps.new(kind: 'ok', steps: steps)
end

.unanchored_pattern(step) ⇒ Object

A compiled, un-anchored pattern from the step's CucumberExpression.



32
33
34
35
36
37
38
 
# File 'lib/oselvar/var/core/matcher.rb', line 32

def unanchored_pattern(step)
  regexp = step.compiled.instance_variable_get(:@tree_regexp).regexp
  source = regexp.source
  source = source[1..] if source.start_with?('^')
  source = source[0...-1] if source.end_with?('$')
  Regexp.new(source, regexp.options)
end