Class: DatesFromString

Inherits:
Object
  • Object
show all
Includes:
Patterns
Defined in:
lib/dates_from_string.rb,
lib/dates_from_string/version.rb

Constant Summary collapse

VERSION = 
"1.3.0"

Constants included from Patterns

Patterns::DATE_COUNTRY_FORMAT, Patterns::PATTERNS

Instance Method Summary collapse

Constructor Details

#initialize(key_words = [], date_format: :default, ordinals: []) ⇒ DatesFromString

Returns a new instance of DatesFromString.



8
9
10
11
12
 
# File 'lib/dates_from_string.rb', line 8

def initialize(key_words = [], date_format: :default, ordinals: [])
  @key_words = key_words
  @date_format = date_format_by_country(date_format)
  @ordinals = Array(ordinals)
end

Instance Method Details

#add_to_structure(type, value, index, next_index, data_arr, key_word = nil) ⇒ Object 



198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
 
# File 'lib/dates_from_string.rb', line 198

def add_to_structure (type ,value, index, next_index, data_arr, key_word = nil)
  set_structura
  if value
    @first_index << index
    @structura[:type] = type
    @structura[:value] = value
  end

  if value && @main_arr.last
    @main_arr.last[:distance] = calc_index(index)
  end

  if @key_words && @key_words.include?(data_arr[next_index])
    @structura[:key_words] << data_arr[next_index]
  end

  if key_word
    @structura[:key_words] << key_word
  end

  if value
    @main_arr <<  @structura
    value = nil
  end
end

#calc_index(index) ⇒ Object 



224
225
226
227
228
229
230
231
232
233
234
 
# File 'lib/dates_from_string.rb', line 224

def calc_index(index)
  result = nil
  @indexs << index
  if @indexs.count > 1
    result = (index - @indexs[-2])
  elsif @first_index[0] < index
    result = (index - @first_index[0])
  else
    result = index
  end
end

#date_format_by_country(date_format) ⇒ Object 



27
28
29
 
# File 'lib/dates_from_string.rb', line 27

def date_format_by_country(date_format)
  DATE_COUNTRY_FORMAT[date_format.to_sym].call
end

#day_regex ⇒ Object 



31
32
33
34
35
36
37
38
39
40
 
# File 'lib/dates_from_string.rb', line 31

def day_regex
  @day_regex ||= begin
    if @ordinals.any?
      ordinal_regex = "(#{@ordinals.join('|')})?"
      /^\d{1,2}#{ordinal_regex},?$/
    else
      /^\d{1,2},?$/
    end
  end
end

#email_date(email) ⇒ Object 



23
24
25
 
# File 'lib/dates_from_string.rb', line 23

def email_date(email)
  email.match(/(?:(?:19|20)[0-9]{2})/).to_s
end

#find_date(string) ⇒ Object 



14
15
16
17
 
# File 'lib/dates_from_string.rb', line 14

def find_date(string)
  parsing_structure = ParsingStructure.new(get_structure(string))
  parsing_structure.start
end

#get_clear_text ⇒ Object 



19
20
21
 
# File 'lib/dates_from_string.rb', line 19

def get_clear_text
  @clear_text.strip
end

#get_dash_data(string) ⇒ Object 



165
166
167
168
169
170
171
172
173
 
# File 'lib/dates_from_string.rb', line 165

def get_dash_data(string)
  if (result = string.match(/\d{4}-\d{4}/))
    result.to_s.split("-")
  elsif (result = string.match(/\d{4}–\d{4}/))
    result.to_s.split("–")
  else
    nil
  end
end

#get_day(string) ⇒ Object 



156
157
158
159
160
161
162
 
# File 'lib/dates_from_string.rb', line 156

def get_day(string)
  if string =~ (day_regex)
    string.gsub(Regexp.union(@ordinals + [',']), '')
  else
    nil
  end
end

#get_full_date(string) ⇒ Object 



133
134
135
136
137
138
139
140
141
142
143
144
 
# File 'lib/dates_from_string.rb', line 133

def get_full_date(string)
  PATTERNS.keys.each do |patterns|
    patterns.each do |pattern|
      if (result = string.match(pattern))
        @clear_text.slice!(result.to_s)
        return PATTERNS[patterns].call result
      end
    end
  end

  nil
end

#get_month_by_list(string) ⇒ Object 



175
176
177
178
179
180
181
182
183
184
185
 
# File 'lib/dates_from_string.rb', line 175

def get_month_by_list(string)
  month = ['January','February','March','April','May','June','July','August','September','October','November','December']
  index = month.index(string.chomp(','))

  if index
    sprintf('%02d',(index+1))
  else
    nil
  end

end

#get_month_year_date(string) ⇒ Object 



146
147
148
149
150
151
152
153
154
 
# File 'lib/dates_from_string.rb', line 146

def get_month_year_date(string)
  if (result = string.match(/^\d{2}\.\d{4}$/))
    result.to_s.split(".").reverse
  elsif (result = string.match(/^\d{4}\.\d{2}$/))
    result.to_s.split(".")
  else
    nil
  end
end

#get_short_month(string) ⇒ Object 



187
188
189
190
191
192
193
194
195
196
 
# File 'lib/dates_from_string.rb', line 187

def get_short_month(string)
  short_month = ['Jan','Feb','Mar','Apr','May','Jun','Jul','Aug','Sep','Oct','Nov','Dec']
  short_index = short_month.index(string.chomp(','))

  if short_index
    sprintf('%02d',(short_index+1))
  else
    nil
  end
end

#get_structure(string) ⇒ Object 



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
 
# File 'lib/dates_from_string.rb', line 42

def get_structure(string)
  unless string.nil? || string.empty?
    @main_arr = []
    data_arr = string.split(" ")
    @indexs = []
    @first_index = []
    @clear_text = string.clone

    data_arr.each_with_index do |data, index|
      value_year = get_year(data)
      value_full_date = get_full_date(data)
      value_month_year_date = get_month_year_date(data)
      value_dash = get_dash_data(data)
      value_month = get_month_by_list(data)
      value_short_month = get_short_month(data)
      value_time = get_time(data)

      value_day = get_day(data)
      next_index = index + 1

      if value_year
        add_to_structure(:year ,value_year, index, next_index, data_arr)
      end

      if value_full_date
        if @main_arr.size == 0
          index = 0
        end
        add_to_structure(@date_format[0], value_full_date[0], index, next_index, data_arr)
        add_to_structure(@date_format[1], value_full_date[1], index, next_index, data_arr)
        add_to_structure(@date_format[2], value_full_date[2], index, next_index, data_arr)
      end

      if value_month_year_date
        add_to_structure(:year , value_month_year_date[0], index, next_index, data_arr)
        add_to_structure(:month ,value_month_year_date[1], index, next_index, data_arr)
      end

      if value_dash
        add_to_structure(:year ,value_dash[0], index, next_index, data_arr, '-')
        add_to_structure(:year ,value_dash[1], index, next_index, data_arr)
      end

      if value_month
        add_to_structure(:month ,value_month, index, next_index, data_arr)
      end

      if value_short_month
        add_to_structure(:month ,value_short_month, index, next_index, data_arr)
      end

      if value_day
        add_to_structure(:day ,value_day, index, next_index, data_arr)
      end

      if value_time
         add_to_structure(:time ,value_time, index, next_index, data_arr)
      end

    end

    return @main_arr
  else
    nil
  end
end

#get_time(string) ⇒ Object 



109
110
111
112
113
114
115
116
117
118
119
 
# File 'lib/dates_from_string.rb', line 109

def get_time(string)
  if (result = string.match(/\d{2}:\d{2}:\d{2}/))
    @clear_text.slice!(result.to_s)
    result.to_s
  elsif (result = string.match(/\d{2}:\d{2}/))
    @clear_text.slice!(result.to_s)
    result.to_s
  else
    nil
  end
end

#get_year(string) ⇒ Object 



121
122
123
124
125
126
127
128
129
130
131
 
# File 'lib/dates_from_string.rb', line 121

def get_year(string)
  if string =~ /^\d{4}$/
    string
  elsif string =~ /^\d{4}\.$/
    string.delete!('.')
  elsif string =~ /^\d{4}\,$/
    string.delete!(',')
  else
    nil
  end
end

#set_structura ⇒ Object 



236
237
238
 
# File 'lib/dates_from_string.rb', line 236

def set_structura
  @structura = {type: nil, value: nil, distance: 0, key_words: []}
end