Module: Worldwide::Zip

Extended by:
Zip
Included in:
Zip
Defined in:
lib/worldwide/zip.rb

Instance Method Summary collapse

Instance Method Details

#find_country(country_code: nil, zip:, min_confidence: 0) ⇒ Region

Suggest a country for a given postal code.

Some countries have an ambiguous dual state. For example, some consider Jersey to be a top-level country in its own right, while others consider it to be part of the United Kingdom. This leads to frustrated buyers receiving a validation error when they enter their address and the postal code isn’t valid for the selected “country”.

Also, in some cases, users are simply confused, and use a postal code that is obviously inappropriate for the selected country. Some examples:

- entering a postal code for Georgia, United States, but selecting the country Georgia
- using a USA FPO address, but selecting the non-US physical location of the US base as the country

This method attempts to heuristically suggest a country in some common cases where we see a lot of confusion. There’s no way to solve this problem for all cases, and we will often not have a suggestion (in which case this method will return ‘nil`).

Parameters:

  • country_code (String) (defaults to: nil) —

    The country code that the user thought this postal code was for (optional)

  • zip (String) —

    The postal code that the user entered (required)

  • min_confidence (Integer) (defaults to: 0) —

    The minimum confidence level (between 0-100) that is accepted from a suggestion (optional)

Returns:

  • (Region) —

    which is a “country” if we have a suggestion, or ‘nil` if we do not.



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/worldwide/zip.rb', line 52

def find_country(country_code: nil, zip:, min_confidence: 0)
  return nil unless Util.present?(zip)

  country = Worldwide.region(code: country_code) unless country_code.nil?
  return country if country&.valid_zip?(zip)

  adjusted_zip = zip.strip.upcase

  # Try to match based on the alleged country
  suggestion, confidence = find_country_using_alleged_country(country_code, adjusted_zip)
  return suggestion unless suggestion.nil? || confidence.nil? || confidence < min_confidence

  # If our postal code is wholly numeric, we can't make an intelligent suggestion without an alleged country.
  return nil unless adjusted_zip.match?(/[A-Z]/)

  # Try a broader-ranging match without considering the alleged country
  # We'll see if we have only a single suggestion and, if so, return it.
  # In cases where there's more than one possible match, we'll return nil.
  suggestions = find_country_using_zip_alone(adjusted_zip)
  suggestion = suggestions.first[0] unless Util.blank?(suggestions)
  confidence = suggestions.first[1] unless Util.blank?(suggestions)

  return suggestion if suggestions.length == 1 && confidence && confidence >= min_confidence

  nil
end

#gb_style?(country_code:) ⇒ Boolean

The United Kingdom has an unusual style of postcode. It consists of two halves (outcode and incode), separated by a space. The incode is always 3 characters, with 1 digit followed by 2 letters. The outcode varies from 2 to 4 characters, following certain patterns. A handful of other countries also allocate their codes within the UK “namespace”.

Returns:

  • (Boolean) 


28
29
30
 
# File 'lib/worldwide/zip.rb', line 28

def gb_style?(country_code:)
  GB_STYLE_ZIP_COUNTRIES.include?(country_code.to_s.upcase)
end

#gb_style_outcode(country_code:, input:) ⇒ Object 



215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
 
# File 'lib/worldwide/zip.rb', line 215

def gb_style_outcode(country_code:, input:)
  normalized = input

  m = /^([A-Z]{1,2})/.match(normalized)
  return normalized if m.nil?

  area = m[1]
  return normalized if area.nil?

  # Let's get the first, second and third characters following the area
  first = normalized[area.length]
  second = normalized[area.length + 1]
  third = normalized[area.length + 2]

  candidate = case area
  when "E", "EC", "N", "NW", "SE", "SW", "W", "WC"
    # These London districts may include a single-letter suffix as part of the district
    # In all such cases, the district is one digit + one letter

    if /[A-Z]/.match?(second)
      # This must be a district suffix, so disctrict consists of first + second
      "#{area}#{first}#{second}".strip
    elsif /[A-Z]/.match?(third)
      # Because third is a letter, second must be the sector, and the district must be a single digit
      "#{area}#{first}".strip
    else
      # Otherwise, we'll assume that we've got a two-digit district
      "#{area}#{first}#{second}".strip
    end
  else
    # The outcode is the area plus at most two digits following it

    if /[A-Z]/.match?(third)
      # Because third is a letter, second must be the sector, and the district must be a single digit
      "#{area}#{first}"
    else
      "#{area}#{first}#{second}".strip
    end
  end

  if /[0-9]/.match?(second) && !Worldwide::ExtantOutcodes.for_country(code: country_code).include?(candidate)
    # We've guessed a two-digit district that does not actually exist, so we'll assume a single-digit district.
    # For example, "GL71" can't be an outcode, so we'll assume the user meant outcode "GL7".
    "#{area}#{first}"
  else
    candidate
  end
end

#normalize(country_code:, zip:, allow_autofill: true, strip_extraneous_characters: false) ⇒ String

Normalizes the postal code into the format expected by the national postal authority.

Parameters:

  • country_code (String) —

    The country in which this postal code is located.

  • zip (String) —

    The postal code as the user has entered it.

  • allow_autofill (String) (defaults to: true) —

    Normalize the postal code for the candidate country prior to seeing whether the regular expression matches that (normalized) code.

  • strip_extraneous_characters (String) (defaults to: false) —

    Allow additional stripping of characters to either fully numeric or alphanumeric depending on the country.

Returns:

  • (String) —

    The postal code with spaces inserted/removed and other formatting fixes applied.



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
147
148
149
150
151
 
# File 'lib/worldwide/zip.rb', line 85

def normalize(country_code:, zip:, allow_autofill: true, strip_extraneous_characters: false)
  input = zip # preserve the original zip, in case we need to fall back to it

  return input if Util.blank?(country_code)

  country = Worldwide.region(code: country_code)
  return zip if country.nil? || NORMALIZATION_DISABLED_COUNTRIES.include?(country.iso_code)

  if allow_autofill
    autofill = country.autofill_zip
    return autofill if Util.present?(autofill)
  end

  return nil if zip.nil?

  # Convert to uppercase
  # Convert numeric and romaji full-width to half-width
  # Strip hyphens, dashes and the Japanese postcode marker
  zip = zip.upcase.tr("０-９ａ-ｚＡ-Ｚ", "0-9a-zA-Z")
    .delete("〒\u058a\u05be\u1806\u1b60\u200b\u2010\u2011\u2012\u2013\u2014\u2015\u2053\u2e17\u2e3a\u2e3b\u2212\u30fb\u30fc\ufe58\ufe63\uff0d\uff65(),./_~-")
  zip = add_prefix_if_required(country_code: country_code, zip: zip)
  if strip_extraneous_characters
    zip = strip_extraneous_characters(zip: zip, country_code: country_code)
  end

  return normalize_for_gb(zip: zip) if gb_style?(country_code: country.iso_code)

  # Remove both normal-width and double-width spaces
  zip.delete!(" 　")
  zip = replace_letters_and_numbers(country_code: country.iso_code, zip: zip)

  result = if "BD" == country.iso_code
    normalize_for_bd(zip: zip)
  elsif "FO" == country.iso_code
    normalize_for_fo(zip: zip)
  elsif "GH" == country.iso_code
    normalize_for_gh(zip: zip)
  elsif "HT" == country.iso_code
    normalize_for_ht(zip: zip)
  elsif "LK" == country.iso_code
    normalize_for_lk(zip: zip)
  elsif "MD" == country.iso_code
    normalize_for_md(zip: zip)
  elsif "MG" == country.iso_code
    normalize_for_mg(zip: zip)
  elsif "NG" == country.iso_code
    normalize_for_ng(zip: zip)
  elsif "SG" == country.iso_code
    normalize_for_sg(zip: zip)
  elsif "MA" == country.iso_code
    normalize_for_ma(zip: zip)
  elsif "XK" == country.iso_code
    normalize_for_xk(zip: zip)
  elsif "BR" == country.iso_code || "JP" == country.iso_code
    insert_spaces_and_hyphens_for_partial_code(country_code: country.iso_code, zip: zip)
  else
    insert_spaces_and_hyphens(country_code: country.iso_code, zip: zip)
  end

  if country.send(:valid_normalized_zip?, result)
    result
  elsif country.send(:valid_normalized_zip?, result, partial_match: true)
    result
  else
    input # fall back to the original input, because we don't seem to have generated anything sensible
  end
end

#numeric_only_zip?(country_code:) ⇒ Boolean

We want to show numeric keypad on mobile view for countries with only numeric postal codes Spaces or dashes are allowed.

Returns:

  • (Boolean) —

    true or false depending on whether postal code can contain only digits or not.



10
11
12
 
# File 'lib/worldwide/zip.rb', line 10

def numeric_only_zip?(country_code:)
  NUMERIC_ONLY_ZIP_COUNTRIES.include?(country_code.to_s.upcase)
end

#outcode(country_code:, zip:) ⇒ Object

Returns the “outcode” (first portion) of a postcode for a country that uses the UK style. Returns the “forward sortation area” (first portion) of a postal code for Canada. Returns the “routing key” (first portion) of a postal code for Ireland. Otherwise, returns the full zip.



157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
 
# File 'lib/worldwide/zip.rb', line 157

def outcode(country_code:, zip:)
  @split_code_countries ||= Set.new(GB_STYLE_ZIP_COUNTRIES).add("CA").add("IE")

  return zip unless @split_code_countries.include?(country_code.to_s.upcase)

  normalized_input = normalize(country_code: country_code, zip: zip).upcase
  normalized_country_code = country_code.to_s.upcase

  if Worldwide.region(code: country_code).valid_zip?(normalized_input)
    # We successfully normalized the postcode, so we can split on the space to determine the outcode
    normalized_input&.split(" ")&.first
  else
    # We have either an invalid or an incomplete postcode.  Let's try some fancy heuristics
    case country_code.to_s.upcase
    when "CA", "IE"
      # the FSA / routing code is the first 3 characters
      normalized_input[0..2]
    when "GB", "GG", "IM", "JE"
      gb_style_outcode(country_code: normalized_country_code, input: normalized_input)
    when "GI"
      "GX11"
    else
      normalized_input
    end
  end
end

#pure_numeric_only_zip?(country_code:) ⇒ Boolean

We want to show numeric keypad on mobile view for countries with only numeric postal codes Spaces or dashes aren’t allowed.

Returns:

  • (Boolean) —

    true or false depending on whether postal code can contain only digits or not.



17
18
19
20
21
 
# File 'lib/worldwide/zip.rb', line 17

def pure_numeric_only_zip?(country_code:)
  return false if SPACES_AND_HYPHENS.key?(country_code.to_s.upcase.to_sym)

  numeric_only_zip?(country_code: country_code)
end

#strip_optional_country_prefix(country_code:, zip:) ⇒ Object 



184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
 
# File 'lib/worldwide/zip.rb', line 184

def strip_optional_country_prefix(country_code:, zip:)
  return zip if Util.blank?(zip)

  unless OPTIONAL_PREFIX_COUNTRIES.include?(country_code&.to_sym)
    return zip
  end

  vehicle_code = OPTIONAL_PREFIX_COUNTRIES[country_code&.to_sym]

  stripped = zip.strip
  upcased = stripped.upcase

  if upcased.start_with?(country_code&.to_s)
    stripped = stripped[country_code&.to_s&.length..-1]
  elsif upcased.start_with?(vehicle_code)
    stripped = stripped[vehicle_code.length..-1]
  end

  if stripped.start_with?("-")
    stripped = stripped[1..-1]
  end

  m = stripped.match(/^\d/)

  if m.nil?
    zip
  else
    stripped
  end
end