Class: FakedCSV::Generator

Inherits:
Object
  • Object
show all
Defined in:
lib/faked_csv/generator.rb

Instance Attribute Summary collapse

Class Method Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(config) ⇒ Generator

Returns a new instance of Generator.



5
6
7
# File 'lib/faked_csv/generator.rb', line 5

def initialize(config)
    @config = config
end

Instance Attribute Details

#configObject (readonly)

Returns the value of attribute config.



3
4
5
# File 'lib/faked_csv/generator.rb', line 3

def config
  @config
end

#rowsObject (readonly)

Returns the value of attribute rows.



3
4
5
# File 'lib/faked_csv/generator.rb', line 3

def rows
  @rows
end

Class Method Details

.fake(type) ⇒ Object



235
236
237
# File 'lib/faked_csv/generator.rb', line 235

def self.fake(type)
    Fakerer.new(type).fake
end

.rand_char(length) ⇒ Object

individual random generators



185
186
187
188
# File 'lib/faked_csv/generator.rb', line 185

def self.rand_char(length)
    o = [('a'..'z'), ('A'..'Z'), (0..9)].map { |i| i.to_a }.flatten
    string = (0...length).map { o[rand(o.length)] }.join
end

.rand_float(min, max, precision) ⇒ Object



230
231
232
233
# File 'lib/faked_csv/generator.rb', line 230

def self.rand_float(min, max, precision)
    raise "min > max" if min > max
    (rand * (max - min) + min).round(precision)
end

.rand_formatted_char(format) ⇒ Object



190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
# File 'lib/faked_csv/generator.rb', line 190

def self.rand_formatted_char(format)
    res = []
    i = 0
    while i < format.size
        case a = format[i]
        when '/'
            i += 1
            res << single_rand_char(format[i])
        else
            res << a
        end
        i += 1
    end
    return res.join("")
end

.rand_int(min, max) ⇒ Object



225
226
227
228
# File 'lib/faked_csv/generator.rb', line 225

def self.rand_int(min, max)
    raise "min > max" if min > max
    min + rand(max - min + 1)
end

.single_rand_char(format) ⇒ Object



206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
# File 'lib/faked_csv/generator.rb', line 206

def self.single_rand_char(format)
    aa = nil
    case format
    when 'W' # A-Z
        aa = ('A'..'Z').to_a
    when 'w' # a-z
        aa = ('a'..'z').to_a
    when 'd' # 0-9
        aa = (0..9).to_a
    when 'D' # A-Za-z
        aa = [('a'..'z'), ('A'..'Z')].map { |i| i.to_a }.flatten
    when '@'
        aa = [('a'..'z'), ('A'..'Z'), (0..9)].map { |i| i.to_a }.flatten
    else
        raise "invalid format: #{format} in single_rand_char"
    end
    return aa[rand(aa.size)]
end

Instance Method Details

#_loopObject



173
174
175
176
177
178
179
180
181
# File 'lib/faked_csv/generator.rb', line 173

def _loop
    max_attempts = 1000_000_000_000
    i = 0
    (0...max_attempts).each do |j|
        yield
        i += 1
    end
    raise "max attempts reached" if i == max_attempts
end

#_random_distribution(total, parts) ⇒ Object



115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
# File 'lib/faked_csv/generator.rb', line 115

def _random_distribution(total, parts)
    raise "parts has to be greater than 0" unless parts > 0
    raise "parts should not be greater than total" if total < parts
    cuts = {}
    _loop do
        break if cuts.size == parts - 1
        cuts[rand(total - 1)] = true
    end
    arr = []
    part_index = 0
    (0...total).each do |i|
        arr << part_index
        part_index += 1 if cuts.has_key? i
    end
    arr.shuffle.each_with_index do |v, i|
        yield(i, v)
    end
end

#_random_inject(values, injects) ⇒ Object

inject <injects> into <values>



135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
# File 'lib/faked_csv/generator.rb', line 135

def _random_inject(values, injects)
    used_indexes = {}
    count = injects.size > values.size ? values.size : injects.size
    (0...count).each do |i|
        inj = injects[i]
        times_inject = rand(values.size / injects.size / 10)
        times_inject = 1 if times_inject < 1
        times_inject.times do
            rand_index = rand(values.size)
            _loop do
                break unless used_indexes.has_key? rand_index
                rand_index = rand(values.size)
            end
            used_indexes[rand_index] = true
            values[rand_index] = inj
        end
    end
end

#_random_value(field) ⇒ Object



154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
# File 'lib/faked_csv/generator.rb', line 154

def _random_value(field)
    case field[:type]
    when :rand_int
        return Generator.rand_int field[:min], field[:max]
    when :rand_float
        return Generator.rand_float field[:min], field[:max], field[:precision]
    when :rand_char
        if field[:format].nil?
            return Generator.rand_char field[:length]
        else
            return Generator.rand_formatted_char field[:format]
        end
    when :fixed
        return field[:values].sample
    else # faker
        return Generator.fake field[:type]
    end
end

#generateObject



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
# File 'lib/faked_csv/generator.rb', line 37

def generate
    puts "preparing values ..."
    prepare_values

    @config.fields.each do |field|
        puts "generating random data for #{field} ..."
        field[:data] = []

        # let's get some data!
        if field[:type] == :inc_int
            puts "generating increment values ..."
            i = field[:start]
            @config.row_count.times do
                field[:data] << i
                i += field[:step]
            end
        elsif field[:rotate].nil? || field[:type] == :fixed
            # not rotating? or fixed values? generate random value each time
            puts "calling generator ..."
            index = 0
            @config.row_count.times do
                field[:data] << _random_value(field)
                print '.' if index % 10000 == 0
                index += 1
            end
            puts

            # inject user values if given and not fixed type
            puts "injecting values ..."
            unless field[:type] == :fixed || field[:inject].nil?
                _random_inject(field[:data], field[:inject])
            end
        else
            # rotating? pick from prepared values
            puts "selecting data from rotation ..."
            _random_distribution(@config.row_count, field[:values].size) do |i, j|
                field[:data][i] = field[:values][j]
            end
        end
    end
end

#headersObject



9
10
11
# File 'lib/faked_csv/generator.rb', line 9

def headers
    @config.headers
end

#prepare_valuesObject



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
# File 'lib/faked_csv/generator.rb', line 79

def prepare_values
    @config.fields.each do |field|
        puts "prepare value for #{field} ..."
        # if it's fixed values or no rotate
        # we don't want to prepare values for this field
        if [:inc_int, :fixed].include?(field[:type]) || field[:rotate].nil?
            next
        end

        # we don't have enough integers for the rotate
        if field[:type] == :rand_int && field[:rotate] > field[:max] - field[:min] + 1
            raise "rotate should not be greater than the size of the range"
        end

        values = {}
        # let's first inject all user values if given
        puts "injecting user values ..."
        unless field[:inject].nil?
            field[:inject].each do |inj|
                values[inj] = true
                # truncate more inject values if we go over the rows count
                break if values.size == @config.row_count
            end
        end
        # then generate as many data as we need
        puts "looping to get enough values ..."
        _loop do
            # we want to get <rotate> unique values. stop when we got enough
            break if values.size >= field[:rotate]
            v = _random_value(field)
            values[v] = true
        end
        field[:values] = values.keys
    end
end


27
28
29
30
31
32
33
34
35
# File 'lib/faked_csv/generator.rb', line 27

def print_to(writer)
    (0...@config.row_count).each do |r|
        @config.fields.each_with_index do |field, index|
            suffix = (index == (@config.fields.size - 1)) ? '' : ','
            writer.write("#{field[:data][r]}#{suffix}")
        end
        writer.write("\n")
    end
end