Class: AiRedactor::TextRedactor

Inherits:
Object
  • Object
show all
Defined in:
lib/ai_redactor/text_redactor.rb

Constant Summary collapse

DEFAULT_MASK_CHAR = 
"*"
DEFAULT_MASK_LENGTH = 
8

Instance Method Summary collapse

Constructor Details

#initialize(options = {}) ⇒ TextRedactor

Returns a new instance of TextRedactor.



12
13
14
15
16
17
18
 
# File 'lib/ai_redactor/text_redactor.rb', line 12

def initialize(options = {})
  @mask_char = options[:mask_char] || DEFAULT_MASK_CHAR
  @mask_length = options[:mask_length] || DEFAULT_MASK_LENGTH
  @preserve_format = options[:preserve_format] || false
  @patterns = options[:patterns] || Patterns::ALL_PATTERNS.keys
  @case_sensitive = options[:case_sensitive] || false
end

Instance Method Details

#analyze(text) ⇒ Object 



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
82
83
84
85
86
87
88
89
90
 
# File 'lib/ai_redactor/text_redactor.rb', line 52

def analyze(text)
  return Report.new(text, [], text) if text.nil? || text.empty?

  detections = []
  masked_text = text.dup

  @patterns.each do |pattern_name|
    pattern = Patterns.get_pattern(pattern_name)
    next unless pattern

    flags = @case_sensitive ? 0 : Regexp::IGNORECASE
    regex = Regexp.new(pattern.source, flags)

    text.scan(regex) do
      match = Regexp.last_match
      start_pos = match.begin(0)
      end_pos = match.end(0)
      matched_text = match[0]

      detections << {
        type: pattern_name.to_s,
        original: matched_text,
        start_position: start_pos,
        end_position: end_pos,
        confidence: calculate_confidence(matched_text, pattern_name),
        masked_value: generate_mask(matched_text, pattern_name)
      }
    end
  end

  # Apply masking to create masked text
  detections.sort_by { |d| -d[:start_position] }.each do |detection|
    start_pos = detection[:start_position]
    end_pos = detection[:end_position]
    masked_text[start_pos...end_pos] = detection[:masked_value]
  end

  Report.new(text, detections, masked_text)
end

#mask(text) ⇒ Object 



20
21
22
23
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
49
50
 
# File 'lib/ai_redactor/text_redactor.rb', line 20

def mask(text)
  return "" if text.nil? || text.empty?

  masked_text = text.dup
  detections = []

  @patterns.each do |pattern_name|
    pattern = Patterns.get_pattern(pattern_name)
    next unless pattern

    flags = @case_sensitive ? 0 : Regexp::IGNORECASE
    regex = Regexp.new(pattern.source, flags)

    masked_text.gsub!(regex) do |match|
      start_pos = Regexp.last_match.begin(0)
      end_pos = Regexp.last_match.end(0)

      detections << {
        type: pattern_name.to_s,
        original: match,
        start_position: start_pos,
        end_position: end_pos,
        confidence: 1.0
      }

      generate_mask(match, pattern_name)
    end
  end

  masked_text
end