phonetics 1.5.4 → 1.8.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- checksums.yaml +4 -4
- data/Rakefile +22 -11
- data/VERSION +1 -1
- data/bin/console +13 -0
- data/ext/c_levenshtein/levenshtein.c +104 -76
- data/ext/c_levenshtein/next_phoneme_length.c +1364 -0
- data/ext/c_levenshtein/next_phoneme_length.h +1 -0
- data/ext/c_levenshtein/phonemes.c +33 -0
- data/ext/c_levenshtein/phonemes.h +2 -0
- data/ext/c_levenshtein/phonetic_cost.c +134245 -42305
- data/ext/c_levenshtein/phonetic_cost.h +1 -1
- data/lib/phonetics.rb +2 -90
- data/lib/phonetics/code_generator.rb +285 -0
- data/lib/phonetics/levenshtein.rb +12 -21
- data/lib/phonetics/ruby_levenshtein.rb +5 -14
- metadata +8 -2
|
@@ -17,14 +17,13 @@ module Phonetics
|
|
|
17
17
|
attr_reader :str1, :str2, :len1, :len2, :matrix
|
|
18
18
|
|
|
19
19
|
def initialize(ipa_str1, ipa_str2, verbose = false)
|
|
20
|
-
@str1 = ipa_str1
|
|
21
|
-
@str2 = ipa_str2
|
|
22
|
-
@len1 =
|
|
23
|
-
@len2 =
|
|
20
|
+
@str1 = ipa_str1.each_char.select { |c| Phonetics.phonemes.include?(c) }.join
|
|
21
|
+
@str2 = ipa_str2.each_char.select { |c| Phonetics.phonemes.include?(c) }.join
|
|
22
|
+
@len1 = @str1.size
|
|
23
|
+
@len2 = @str2.size
|
|
24
24
|
@verbose = verbose
|
|
25
|
-
ensure_is_phonetic!
|
|
26
25
|
prepare_matrix
|
|
27
|
-
set_edit_distances(
|
|
26
|
+
set_edit_distances(@str1, @str2)
|
|
28
27
|
end
|
|
29
28
|
|
|
30
29
|
def distance
|
|
@@ -40,14 +39,6 @@ module Phonetics
|
|
|
40
39
|
|
|
41
40
|
private
|
|
42
41
|
|
|
43
|
-
def ensure_is_phonetic!
|
|
44
|
-
[str1, str2].each do |string|
|
|
45
|
-
string.chars.each do |char|
|
|
46
|
-
raise ArgumentError, "#{char.inspect} is not a character in the International Phonetic Alphabet. #{self.class.name} only works with IPA-transcribed strings" unless Phonetics.phonemes.include?(char)
|
|
47
|
-
end
|
|
48
|
-
end
|
|
49
|
-
end
|
|
50
|
-
|
|
51
42
|
def walk
|
|
52
43
|
res = []
|
|
53
44
|
i = len2
|
metadata
CHANGED
|
@@ -1,14 +1,14 @@
|
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
|
2
2
|
name: phonetics
|
|
3
3
|
version: !ruby/object:Gem::Version
|
|
4
|
-
version: 1.
|
|
4
|
+
version: 1.8.0
|
|
5
5
|
platform: ruby
|
|
6
6
|
authors:
|
|
7
7
|
- Jack Danger
|
|
8
8
|
autorequire:
|
|
9
9
|
bindir: bin
|
|
10
10
|
cert_chain: []
|
|
11
|
-
date: 2019-
|
|
11
|
+
date: 2019-09-05 00:00:00.000000000 Z
|
|
12
12
|
dependencies:
|
|
13
13
|
- !ruby/object:Gem::Dependency
|
|
14
14
|
name: bundler
|
|
@@ -115,11 +115,17 @@ files:
|
|
|
115
115
|
- VERSION
|
|
116
116
|
- _site/orthographic_example.png
|
|
117
117
|
- _site/phonetic_example.png
|
|
118
|
+
- bin/console
|
|
118
119
|
- ext/c_levenshtein/extconf.rb
|
|
119
120
|
- ext/c_levenshtein/levenshtein.c
|
|
121
|
+
- ext/c_levenshtein/next_phoneme_length.c
|
|
122
|
+
- ext/c_levenshtein/next_phoneme_length.h
|
|
123
|
+
- ext/c_levenshtein/phonemes.c
|
|
124
|
+
- ext/c_levenshtein/phonemes.h
|
|
120
125
|
- ext/c_levenshtein/phonetic_cost.c
|
|
121
126
|
- ext/c_levenshtein/phonetic_cost.h
|
|
122
127
|
- lib/phonetics.rb
|
|
128
|
+
- lib/phonetics/code_generator.rb
|
|
123
129
|
- lib/phonetics/levenshtein.rb
|
|
124
130
|
- lib/phonetics/ruby_levenshtein.rb
|
|
125
131
|
- lib/phonetics/version.rb
|