normalize_country 0.3.2 → 0.3.3
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +4 -4
- data/README.rdoc +6 -2
- data/lib/normalize_country/countries/en.yml +2 -0
- data/lib/normalize_country.rb +2 -2
- metadata +6 -8
- data/lib/normalize_country/scanner.rb +0 -120
checksums.yaml
CHANGED
|
@@ -1,7 +1,7 @@
|
|
|
1
1
|
---
|
|
2
2
|
SHA256:
|
|
3
|
-
metadata.gz:
|
|
4
|
-
data.tar.gz:
|
|
3
|
+
metadata.gz: 793c9aaa61fc028c685a5a27681a670e12675e3f2d5db2bb4343167da1d24d94
|
|
4
|
+
data.tar.gz: 3a8c27532a071ec3baf5114a0ff1c35c643b85944fc45f0cb69586261eb1616c
|
|
5
5
|
SHA512:
|
|
6
|
-
metadata.gz:
|
|
7
|
-
data.tar.gz:
|
|
6
|
+
metadata.gz: 800a8cd65c7ed894a1af4950e03e795a86a2debbec7641aba2d676d6ebd99061a28e0672590d6ce087078813ac365ef1eff3f4a19adbb925cc8d7b8022a7fd98
|
|
7
|
+
data.tar.gz: ec34c172d11d352a1107254f9e9382a8a6e50538cf6cdf4b9a354cb751d6f8cce1b422900182dc530ff051fd6f9e08b2093ba5f5571c43a8c0b1b765222840f6
|
data/README.rdoc
CHANGED
|
@@ -2,8 +2,7 @@
|
|
|
2
2
|
|
|
3
3
|
Convert country names and codes to a standard.
|
|
4
4
|
|
|
5
|
-
{<img src="https://
|
|
6
|
-
{<img src="https://codeclimate.com/github/sshaw/normalize_country.svg" />}[https://codeclimate.com/github/sshaw/normalize_country]
|
|
5
|
+
{<img src="https://github.com/sshaw/normalize_country/actions/workflows/ci.yml/badge.svg" alt="Build Status" />}[https://github.com/sshaw/normalize_country]
|
|
7
6
|
|
|
8
7
|
=== Overview
|
|
9
8
|
|
|
@@ -59,6 +58,11 @@ will convert to/from the following:
|
|
|
59
58
|
|
|
60
59
|
A list of valid formats can be obtained by calling +NormalizeCountry.formats+.
|
|
61
60
|
|
|
61
|
+
=== Custom Database
|
|
62
|
+
|
|
63
|
+
Set the +NORMALIZE_COUNTRY_DB+ environment variable.
|
|
64
|
+
See <code>lib/normalize_country/countries/en.yml</code> for the expected format.
|
|
65
|
+
|
|
62
66
|
=== Obtaining an Array or Hash
|
|
63
67
|
|
|
64
68
|
NormalizeCountry.to_a # Defaults to NormalizeCountry.to
|
data/lib/normalize_country.rb
CHANGED
|
@@ -3,7 +3,7 @@
|
|
|
3
3
|
require "yaml"
|
|
4
4
|
|
|
5
5
|
module NormalizeCountry
|
|
6
|
-
VERSION = "0.3.
|
|
6
|
+
VERSION = "0.3.3"
|
|
7
7
|
Countries = {}
|
|
8
8
|
|
|
9
9
|
class << self
|
|
@@ -78,7 +78,7 @@ module NormalizeCountry
|
|
|
78
78
|
end
|
|
79
79
|
end
|
|
80
80
|
|
|
81
|
-
path = File.join(File.dirname(__FILE__), "normalize_country", "countries", "en.yml")
|
|
81
|
+
path = ENV["NORMALIZE_COUNTRY_DB"] || File.join(File.dirname(__FILE__), "normalize_country", "countries", "en.yml")
|
|
82
82
|
data = YAML.load_file(path)
|
|
83
83
|
data.values.each do |mapping|
|
|
84
84
|
country = Country.new(mapping)
|
metadata
CHANGED
|
@@ -1,14 +1,14 @@
|
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
|
2
2
|
name: normalize_country
|
|
3
3
|
version: !ruby/object:Gem::Version
|
|
4
|
-
version: 0.3.
|
|
4
|
+
version: 0.3.3
|
|
5
5
|
platform: ruby
|
|
6
6
|
authors:
|
|
7
7
|
- Skye Shaw
|
|
8
|
-
autorequire:
|
|
8
|
+
autorequire:
|
|
9
9
|
bindir: bin
|
|
10
10
|
cert_chain: []
|
|
11
|
-
date:
|
|
11
|
+
date: 2025-02-22 00:00:00.000000000 Z
|
|
12
12
|
dependencies:
|
|
13
13
|
- !ruby/object:Gem::Dependency
|
|
14
14
|
name: rake
|
|
@@ -54,13 +54,12 @@ 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
|
|
58
57
|
- spec/normalize_country_spec.rb
|
|
59
58
|
homepage: http://github.com/sshaw/normalize_country
|
|
60
59
|
licenses:
|
|
61
60
|
- MIT
|
|
62
61
|
metadata: {}
|
|
63
|
-
post_install_message:
|
|
62
|
+
post_install_message:
|
|
64
63
|
rdoc_options: []
|
|
65
64
|
require_paths:
|
|
66
65
|
- lib
|
|
@@ -75,9 +74,8 @@ required_rubygems_version: !ruby/object:Gem::Requirement
|
|
|
75
74
|
- !ruby/object:Gem::Version
|
|
76
75
|
version: '0'
|
|
77
76
|
requirements: []
|
|
78
|
-
|
|
79
|
-
|
|
80
|
-
signing_key:
|
|
77
|
+
rubygems_version: 3.5.16
|
|
78
|
+
signing_key:
|
|
81
79
|
specification_version: 4
|
|
82
80
|
summary: Convert country names and codes to a standard
|
|
83
81
|
test_files:
|
|
@@ -1,120 +0,0 @@
|
|
|
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
|