Class: Parse::Algorithm::Ver0_1_0

Inherits:
Object
  • Object
show all
Defined in:
lib/parse/algorithm/ver0_1_0.rb

Constant Summary collapse

NULL =

from bigml’s list

[ '', '-', '?', 'N/A', 'n/a', 'NULL', 'null', '#REF!', '#NAME?', 'NIL', 'nil', 'NA', 'na', '#VALUE!', '#NULL!', '00/00/00', '0000-00-00']
REGION_DATE_FORMAT = 
{
  euro: ['%d-%m-%Y', '%d-%m-%y'],
  us:   ['%m-%d-%Y', '%m-%d-%y'],
  iso:  ['%Y-%m-%d', '%y-%m-%d'], # second one is silly
}
DATE_DETECT = 
{
  %r{\A0*[12]\d\d\d[\-/](?:(?:0[1-9])|(?:1[0-2]))[\-/][1-9]\d\z} => :iso, # $1 will be delimiter
}
EMPTY_OPTIONS = 
{}

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(raw, options = nil) ⇒ Ver0_1_0

Returns a new instance of Ver0_1_0.



17
18
19
20
 
# File 'lib/parse/algorithm/ver0_1_0.rb', line 17

def initialize(raw, options = nil)
  @raw = raw
  @options = options || EMPTY_OPTIONS
end

Instance Attribute Details

#options ⇒ Object (readonly)

Returns the value of attribute options.



16
17
18
 
# File 'lib/parse/algorithm/ver0_1_0.rb', line 16

def options
  @options
end

#raw ⇒ Object (readonly)

Returns the value of attribute raw.



15
16
17
 
# File 'lib/parse/algorithm/ver0_1_0.rb', line 15

def raw
  @raw
end

Instance Method Details

#result ⇒ Object 



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
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
82
83
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
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
 
# File 'lib/parse/algorithm/ver0_1_0.rb', line 22

def result
  return raw unless raw.is_a? String
  
  memo = raw.strip

  return nil if NULL.include? memo
  return 1.0/0 if INFINITY.include? memo
  return -1.0/0 if NEG_INFINITY.include? memo
  return 0.0/0 if NAN.include? memo

  date_region = if options[:date]
    options[:date]
  else
    catch :hit do
      DATE_DETECT.each do |pattern, date_region|
        # binding.pry if memo.include?('2011-')
        if memo =~ pattern
          throw :hit, date_region
        end
      end
      nil
    end
  end

  if date_region.nil? and options[:type] == Date
    date_region = :iso
  end

  if date_region
    yyyy, yy = REGION_DATE_FORMAT.fetch date_region
    is_yyyy = memo =~ /[1-9]\d\d\d/
    memo.sub! /\A0+/, ''
    memo.gsub! '/', '-'
    if is_yyyy
      if memo.length < 10 and date_region == :iso
        return Date.parse(memo)
      else
        return Date.strptime(memo, yyyy)
      end
    else
      return Date.strptime(memo, yy)
    end
  end

  possible_numeric = nil
  not_numeric = nil
  certain_numeric = nil
  if [Numeric, Integer, Float].include?(options[:type])
    certain_numeric = true
    possible_numeric = true
    not_numeric = false
  else
    # not_numeric ||= memo =~ /[1-9][^)\d_,%.eE]/ # has a dash in the middle
    not_numeric ||= memo.include?('_')
    not_numeric ||= memo =~ %r{[1-9][/-]\d}
    not_numeric ||= memo =~ /,\d{1,2}(?:[.\D]|\z)/
    not_numeric ||= memo.scan(/[^\d_,%.eE]/).length > memo.scan(/[\d_,%.eE]/).length
    not_numeric ||= memo =~ /\A[^(+\-\$0-9%]/ # starts with letter or smth
    possible_numeric = !not_numeric
  end
  accounting_negative = nil
  percentage = nil
  if possible_numeric
    accounting_negative = memo =~ /\A[0$]*\([0$]*/
    percentage = memo.end_with?('%')
    memo.sub! /%\z/, '' if percentage
    memo.delete!('()') if accounting_negative # accounting negative
    # in yaml 1.1, anything starting with zero is treated as octal... in 1.2, it's 0o
    memo.sub!(/0+/, '') if memo =~ /\A[+\-]?0+[+\-\$]?[1-9]+/ # leading zeros
    memo.delete!('$') if memo =~ /\A[+\-]?0*\$/
    memo.sub!('D', 'e') if memo =~ /\A[+\-]?[\d.]+D[+\-]?[\d.]+\z/ # fortran double precision
    if memo.include?(',')
      a, b = memo.split('.', 2)
      a.delete! ','
      memo = b ? [a, b].join('.') : a
    end
  end

  if certain_numeric
    memo.gsub! /[a-z]/i, ''
  end

  not_safe_for_yaml = nil
  not_safe_for_yaml ||= memo =~ /\A(on|off)\z/i
  not_safe_for_yaml ||= memo.include?('#')
  not_safe_for_yaml ||= memo =~ /\A[@&,]/
  not_safe_for_yaml ||= not_numeric && memo.start_with?('0')
  not_safe_for_yaml ||= not_numeric && memo =~ /\A[^{\[]*\d[,_]/ #1,2,3, maybe a csv

  safe_for_yaml = !not_safe_for_yaml

  if safe_for_yaml
    begin
      memo = SafeYAML.load memo
    rescue Exception # Psych::SyntaxError will blow up plain rescue in 1.9.3
      $stderr.puts "#{memo.inspect} => #{$!}"
    end
  end

  if possible_numeric
    case memo
    when /\A[+\-]?[\d._]+[eE][+\-]?[\d._]+\z/
      # scientific notation
      memo = memo.to_f
    when /\A[+\-]?0o/
      # octal per yaml 1.2
      memo = memo.to_i 8
    end
  end
  
  if memo.is_a?(String)
    # compress whitespace
    memo.gsub! /\s+/, ' '
  end

  memo = memo / 100.0 if percentage
  memo = -memo if accounting_negative
  memo
rescue
  if options and options[:ignore_error]
    # nothing to see here
  else
    raise "#{memo.inspect} => #{$!}"
  end
end