normalize_country 0.2.1 → 0.3.2
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +5 -5
- data/README.rdoc +6 -0
- data/lib/normalize_country.rb +5 -3
- data/lib/normalize_country/countries/en.yml +341 -11
- data/lib/normalize_country/scanner.rb +120 -0
- data/spec/normalize_country_spec.rb +3 -9
- metadata +8 -7
@@ -0,0 +1,120 @@
|
|
1
|
+
module NormalizeCountry
|
2
|
+
class Tokenizer # :nodoc: all
|
3
|
+
# Need "-" to match emoji names
|
4
|
+
# Need regex to match emojis
|
5
|
+
PATTERN = /[[:word:]]+/
|
6
|
+
|
7
|
+
def initialize(s)
|
8
|
+
@scanner = StringScanner.new(s)
|
9
|
+
end
|
10
|
+
|
11
|
+
def scan
|
12
|
+
@scanner.scan_until(PATTERN)
|
13
|
+
@scanner.matched
|
14
|
+
end
|
15
|
+
|
16
|
+
def peek
|
17
|
+
match = scan
|
18
|
+
@scanner.unscan if match
|
19
|
+
match
|
20
|
+
end
|
21
|
+
|
22
|
+
def end?
|
23
|
+
peek.nil?
|
24
|
+
end
|
25
|
+
end
|
26
|
+
|
27
|
+
class Scanner
|
28
|
+
def initialize(options = nil)
|
29
|
+
options ||= {}
|
30
|
+
|
31
|
+
@to = options[:to] || NormalizeCountry.to
|
32
|
+
@table = lookup_table(options[:from] || NormalizeCountry.formats) # need aliases!!!
|
33
|
+
end
|
34
|
+
|
35
|
+
def convert(text)
|
36
|
+
s = Tokenizer.new(text)
|
37
|
+
|
38
|
+
matches = []
|
39
|
+
stack = []
|
40
|
+
match_position = @table
|
41
|
+
|
42
|
+
while !s.end?
|
43
|
+
word = s.peek.downcase
|
44
|
+
if !match_position[word]
|
45
|
+
s.scan
|
46
|
+
next
|
47
|
+
end
|
48
|
+
|
49
|
+
stack << s.scan
|
50
|
+
alternatives = match_position[stack[-1].downcase]
|
51
|
+
|
52
|
+
peek = s.peek
|
53
|
+
if alternatives && peek && alternatives[peek.downcase]
|
54
|
+
match_position = alternatives
|
55
|
+
next
|
56
|
+
end
|
57
|
+
|
58
|
+
if match_position[stack[-1].downcase][:match]
|
59
|
+
text = stack.join(" ")
|
60
|
+
matches << { :matched => text, :converted => NormalizeCountry(text, :to => @to) }
|
61
|
+
|
62
|
+
end
|
63
|
+
|
64
|
+
stack.clear
|
65
|
+
match_position = @table
|
66
|
+
end
|
67
|
+
|
68
|
+
matches
|
69
|
+
end
|
70
|
+
|
71
|
+
private
|
72
|
+
|
73
|
+
def lookup_table(from_formats)
|
74
|
+
table = {}
|
75
|
+
|
76
|
+
NormalizeCountry::Countries.values.uniq.each_with_object(table) do |country, o|
|
77
|
+
# country.name # eq all
|
78
|
+
# no way to get aliases
|
79
|
+
from_formats.each do |format|
|
80
|
+
name = country[format]
|
81
|
+
next unless name
|
82
|
+
|
83
|
+
head = o
|
84
|
+
parts = name.split(/[[:space:]]+/)
|
85
|
+
parts.each_with_index do |word, i|
|
86
|
+
# options[:case_sensitive_codes] = [x,y] # or true
|
87
|
+
# options[:case_sensitive_formats] = [x,y]
|
88
|
+
# options[:case_sensitive_formats] = true # alpha2, alpha3, fifa, ioc
|
89
|
+
word.downcase! #unless parts.size == 1 && ABBRV_FORMATS.include?(format)
|
90
|
+
|
91
|
+
# if head[word]
|
92
|
+
# if i == parts.size - 1
|
93
|
+
# head[word][:match] = true
|
94
|
+
# else
|
95
|
+
# head = head[word]
|
96
|
+
# end
|
97
|
+
# else
|
98
|
+
# head[word] = {}
|
99
|
+
# if i == parts.size - 1
|
100
|
+
# head[word][:match] = true
|
101
|
+
# else
|
102
|
+
# head = head[word]
|
103
|
+
# end
|
104
|
+
# end
|
105
|
+
|
106
|
+
head[word] = {} unless head[word]
|
107
|
+
|
108
|
+
if i == parts.size - 1
|
109
|
+
head[word][:match] = true
|
110
|
+
else
|
111
|
+
head = head[word]
|
112
|
+
end
|
113
|
+
end
|
114
|
+
end
|
115
|
+
|
116
|
+
o
|
117
|
+
end
|
118
|
+
end
|
119
|
+
end
|
120
|
+
end
|
@@ -4,7 +4,7 @@ require "minitest/autorun"
|
|
4
4
|
require "normalize_country"
|
5
5
|
|
6
6
|
describe NormalizeCountry do
|
7
|
-
COUNTRY_COUNT =
|
7
|
+
COUNTRY_COUNT = 249
|
8
8
|
|
9
9
|
it "normalizes to a country's ISO name by default" do
|
10
10
|
NormalizeCountry.convert("USA").must_equal("United States")
|
@@ -110,15 +110,9 @@ describe NormalizeCountry do
|
|
110
110
|
|
111
111
|
describe ".formats" do
|
112
112
|
it "returns a list of supported formats" do
|
113
|
-
expected = [:alpha2, :alpha3, :fifa, :ioc, :iso_name, :numeric, :official, :short, :simple, :emoji]
|
113
|
+
expected = [:alpha2, :alpha3, :fifa, :ioc, :iso_name, :numeric, :official, :short, :shortcode, :simple, :emoji]
|
114
114
|
formats = NormalizeCountry.formats
|
115
|
-
|
116
|
-
# Ugh, support this in 1.8.7 for a least one version
|
117
|
-
if Symbol < Comparable
|
118
|
-
formats.sort.must_equal(expected.sort)
|
119
|
-
else
|
120
|
-
formats.sort_by { |f| f.to_s }.must_equal(expected.sort_by { |f| f.to_s })
|
121
|
-
end
|
115
|
+
formats.sort.must_equal(expected.sort)
|
122
116
|
end
|
123
117
|
end
|
124
118
|
|
metadata
CHANGED
@@ -1,29 +1,29 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: normalize_country
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version: 0.2
|
4
|
+
version: 0.3.2
|
5
5
|
platform: ruby
|
6
6
|
authors:
|
7
7
|
- Skye Shaw
|
8
8
|
autorequire:
|
9
9
|
bindir: bin
|
10
10
|
cert_chain: []
|
11
|
-
date:
|
11
|
+
date: 2021-05-04 00:00:00.000000000 Z
|
12
12
|
dependencies:
|
13
13
|
- !ruby/object:Gem::Dependency
|
14
14
|
name: rake
|
15
15
|
requirement: !ruby/object:Gem::Requirement
|
16
16
|
requirements:
|
17
|
-
- - "
|
17
|
+
- - ">="
|
18
18
|
- !ruby/object:Gem::Version
|
19
|
-
version:
|
19
|
+
version: 12.3.3
|
20
20
|
type: :development
|
21
21
|
prerelease: false
|
22
22
|
version_requirements: !ruby/object:Gem::Requirement
|
23
23
|
requirements:
|
24
|
-
- - "
|
24
|
+
- - ">="
|
25
25
|
- !ruby/object:Gem::Version
|
26
|
-
version:
|
26
|
+
version: 12.3.3
|
27
27
|
- !ruby/object:Gem::Dependency
|
28
28
|
name: minitest
|
29
29
|
requirement: !ruby/object:Gem::Requirement
|
@@ -54,6 +54,7 @@ files:
|
|
54
54
|
- bin/normalize_country
|
55
55
|
- lib/normalize_country.rb
|
56
56
|
- lib/normalize_country/countries/en.yml
|
57
|
+
- lib/normalize_country/scanner.rb
|
57
58
|
- spec/normalize_country_spec.rb
|
58
59
|
homepage: http://github.com/sshaw/normalize_country
|
59
60
|
licenses:
|
@@ -75,7 +76,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
|
|
75
76
|
version: '0'
|
76
77
|
requirements: []
|
77
78
|
rubyforge_project:
|
78
|
-
rubygems_version: 2.
|
79
|
+
rubygems_version: 2.7.6
|
79
80
|
signing_key:
|
80
81
|
specification_version: 4
|
81
82
|
summary: Convert country names and codes to a standard
|