ruby-spacy 0.1.5.1 → 0.1.5.3
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +4 -4
- data/Gemfile.lock +34 -40
- data/lib/ruby-spacy/version.rb +1 -1
- data/lib/ruby-spacy.rb +10 -4
- metadata +3 -3
checksums.yaml
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
---
|
2
2
|
SHA256:
|
3
|
-
metadata.gz:
|
4
|
-
data.tar.gz:
|
3
|
+
metadata.gz: f24b8651e27c45774a7cc3ce945e085b76a4ef8cf62d2c622f049b4e73ca4e6d
|
4
|
+
data.tar.gz: 41e019d39764ab082d97af5a097e05f783774c542130bedd39ca1f798ad42477
|
5
5
|
SHA512:
|
6
|
-
metadata.gz:
|
7
|
-
data.tar.gz:
|
6
|
+
metadata.gz: 52e32fa278445492c3f5d1464b5bd8c61252bd5bff37c6af96d2ebafbc426ebf10e32640fbd2be274a4a90d165418b02ca31a29f84bfb98bcf662e705e4fabc9
|
7
|
+
data.tar.gz: 8d79c94e1715a327c96a4cf1fd55be535fb7b2e7cf1780545bc2ad433f08180e8b177b6f3b0e9d99a4279108b3ffac859d727c1dcfc55fbeaad5f83a9dee6ea1
|
data/Gemfile.lock
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
PATH
|
2
2
|
remote: .
|
3
3
|
specs:
|
4
|
-
ruby-spacy (0.1.5.
|
4
|
+
ruby-spacy (0.1.5.3)
|
5
5
|
numpy (~> 0.4.0)
|
6
6
|
pycall (~> 1.4.0)
|
7
7
|
terminal-table (~> 3.0.1)
|
@@ -15,34 +15,31 @@ GEM
|
|
15
15
|
diff-lcs (1.5.0)
|
16
16
|
e2mmap (0.1.0)
|
17
17
|
github-markup (4.0.1)
|
18
|
-
jaro_winkler (1.5.
|
18
|
+
jaro_winkler (1.5.6)
|
19
19
|
json (2.6.3)
|
20
20
|
kramdown (2.4.0)
|
21
21
|
rexml
|
22
22
|
kramdown-parser-gfm (1.1.0)
|
23
23
|
kramdown (~> 2.0)
|
24
|
-
|
25
|
-
|
26
|
-
|
27
|
-
|
28
|
-
|
29
|
-
nokogiri (1.14.0-arm64-darwin)
|
30
|
-
racc (~> 1.4)
|
31
|
-
nokogiri (1.14.0-x86_64-darwin)
|
32
|
-
racc (~> 1.4)
|
33
|
-
nokogiri (1.14.0-x86_64-linux)
|
24
|
+
language_server-protocol (3.17.0.3)
|
25
|
+
mini_portile2 (2.8.4)
|
26
|
+
minitest (5.19.0)
|
27
|
+
nokogiri (1.15.3)
|
28
|
+
mini_portile2 (~> 2.8.2)
|
34
29
|
racc (~> 1.4)
|
35
30
|
numpy (0.4.0)
|
36
31
|
pycall (>= 1.2.0.beta1)
|
37
|
-
parallel (1.
|
38
|
-
parser (3.2.
|
32
|
+
parallel (1.23.0)
|
33
|
+
parser (3.2.2.3)
|
39
34
|
ast (~> 2.4.1)
|
35
|
+
racc
|
40
36
|
pycall (1.4.2)
|
41
|
-
racc (1.
|
37
|
+
racc (1.7.1)
|
42
38
|
rainbow (3.1.1)
|
43
39
|
rake (13.0.6)
|
44
|
-
|
45
|
-
|
40
|
+
rbs (2.8.4)
|
41
|
+
redcarpet (3.6.0)
|
42
|
+
regexp_parser (2.8.1)
|
46
43
|
reverse_markdown (2.1.1)
|
47
44
|
nokogiri
|
48
45
|
rexml (3.2.5)
|
@@ -50,57 +47,54 @@ GEM
|
|
50
47
|
rspec-core (~> 3.12.0)
|
51
48
|
rspec-expectations (~> 3.12.0)
|
52
49
|
rspec-mocks (~> 3.12.0)
|
53
|
-
rspec-core (3.12.
|
50
|
+
rspec-core (3.12.2)
|
54
51
|
rspec-support (~> 3.12.0)
|
55
|
-
rspec-expectations (3.12.
|
52
|
+
rspec-expectations (3.12.3)
|
56
53
|
diff-lcs (>= 1.2.0, < 2.0)
|
57
54
|
rspec-support (~> 3.12.0)
|
58
|
-
rspec-mocks (3.12.
|
55
|
+
rspec-mocks (3.12.6)
|
59
56
|
diff-lcs (>= 1.2.0, < 2.0)
|
60
57
|
rspec-support (~> 3.12.0)
|
61
|
-
rspec-support (3.12.
|
62
|
-
rubocop (1.
|
58
|
+
rspec-support (3.12.1)
|
59
|
+
rubocop (1.55.0)
|
63
60
|
json (~> 2.3)
|
61
|
+
language_server-protocol (>= 3.17.0)
|
64
62
|
parallel (~> 1.10)
|
65
|
-
parser (>= 3.2.
|
63
|
+
parser (>= 3.2.2.3)
|
66
64
|
rainbow (>= 2.2.2, < 4.0)
|
67
65
|
regexp_parser (>= 1.8, < 3.0)
|
68
66
|
rexml (>= 3.2.5, < 4.0)
|
69
|
-
rubocop-ast (>= 1.
|
67
|
+
rubocop-ast (>= 1.28.1, < 2.0)
|
70
68
|
ruby-progressbar (~> 1.7)
|
71
69
|
unicode-display_width (>= 2.4.0, < 3.0)
|
72
|
-
rubocop-ast (1.
|
73
|
-
parser (>= 3.
|
74
|
-
ruby-progressbar (1.
|
75
|
-
solargraph (0.
|
70
|
+
rubocop-ast (1.29.0)
|
71
|
+
parser (>= 3.2.1.0)
|
72
|
+
ruby-progressbar (1.13.0)
|
73
|
+
solargraph (0.49.0)
|
76
74
|
backport (~> 1.2)
|
77
75
|
benchmark
|
78
|
-
bundler (
|
76
|
+
bundler (~> 2.0)
|
79
77
|
diff-lcs (~> 1.4)
|
80
78
|
e2mmap
|
81
79
|
jaro_winkler (~> 1.5)
|
82
80
|
kramdown (~> 2.3)
|
83
81
|
kramdown-parser-gfm (~> 1.1)
|
84
82
|
parser (~> 3.0)
|
85
|
-
|
86
|
-
|
83
|
+
rbs (~> 2.0)
|
84
|
+
reverse_markdown (~> 2.0)
|
85
|
+
rubocop (~> 1.38)
|
87
86
|
thor (~> 1.0)
|
88
87
|
tilt (~> 2.0)
|
89
88
|
yard (~> 0.9, >= 0.9.24)
|
90
89
|
terminal-table (3.0.2)
|
91
90
|
unicode-display_width (>= 1.1.1, < 3)
|
92
|
-
thor (1.2.
|
93
|
-
tilt (2.0
|
91
|
+
thor (1.2.2)
|
92
|
+
tilt (2.2.0)
|
94
93
|
unicode-display_width (2.4.2)
|
95
|
-
|
96
|
-
yard (0.9.28)
|
97
|
-
webrick (~> 1.7.0)
|
94
|
+
yard (0.9.34)
|
98
95
|
|
99
96
|
PLATFORMS
|
100
|
-
arm64-darwin-20
|
101
97
|
ruby
|
102
|
-
x86_64-darwin-20
|
103
|
-
x86_64-linux
|
104
98
|
|
105
99
|
DEPENDENCIES
|
106
100
|
bundler
|
@@ -117,4 +111,4 @@ DEPENDENCIES
|
|
117
111
|
yard
|
118
112
|
|
119
113
|
BUNDLED WITH
|
120
|
-
2.4.
|
114
|
+
2.4.13
|
data/lib/ruby-spacy/version.rb
CHANGED
data/lib/ruby-spacy.rb
CHANGED
@@ -7,7 +7,7 @@ require "pycall"
|
|
7
7
|
|
8
8
|
# This module covers the areas of spaCy functionality for _using_ many varieties of its language models, not for _building_ ones.
|
9
9
|
module Spacy
|
10
|
-
MAX_RETRIAL =
|
10
|
+
MAX_RETRIAL = 5
|
11
11
|
|
12
12
|
spacy = PyCall.import_module("spacy")
|
13
13
|
|
@@ -59,10 +59,16 @@ module Spacy
|
|
59
59
|
# @param nlp [Language] an instance of {Language} class
|
60
60
|
# @param py_doc [Object] an instance of Python `Doc` class
|
61
61
|
# @param text [String] the text string to be analyzed
|
62
|
-
def initialize(nlp, py_doc: nil, text: nil)
|
62
|
+
def initialize(nlp, py_doc: nil, text: nil, max_retrial: MAX_RETRIAL, retrial: 0)
|
63
63
|
@py_nlp = nlp
|
64
64
|
@py_doc = py_doc || @py_doc = nlp.call(text)
|
65
65
|
@text = @py_doc.text
|
66
|
+
rescue StandardError
|
67
|
+
retrial += 1
|
68
|
+
raise "Error: Failed to construct a Doc object" unless retrial <= max_retrial
|
69
|
+
|
70
|
+
sleep 0.5
|
71
|
+
initialize(nlp, py_doc: py_doc, text: text, max_retrial: max_retrial, retrial: retrial)
|
66
72
|
end
|
67
73
|
|
68
74
|
# Retokenizes the text merging a span into a single token.
|
@@ -212,7 +218,7 @@ module Spacy
|
|
212
218
|
|
213
219
|
# Creates a language model instance, which is conventionally referred to by a variable named `nlp`.
|
214
220
|
# @param model [String] A language model installed in the system
|
215
|
-
def initialize(model = "en_core_web_sm", max_retrial
|
221
|
+
def initialize(model = "en_core_web_sm", max_retrial: MAX_RETRIAL, retrial: 0)
|
216
222
|
@spacy_nlp_id = "nlp_#{model.object_id}"
|
217
223
|
PyCall.exec("import spacy; #{@spacy_nlp_id} = spacy.load('#{model}')")
|
218
224
|
@py_nlp = PyCall.eval(@spacy_nlp_id)
|
@@ -221,7 +227,7 @@ module Spacy
|
|
221
227
|
raise "Error: Pycall failed to load Spacy" unless retrial <= max_retrial
|
222
228
|
|
223
229
|
sleep 0.5
|
224
|
-
initialize(model, max_retrial, retrial)
|
230
|
+
initialize(model, max_retrial: max_retrial, retrial: retrial)
|
225
231
|
end
|
226
232
|
|
227
233
|
# Reads and analyze the given text.
|
metadata
CHANGED
@@ -1,14 +1,14 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: ruby-spacy
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version: 0.1.5.
|
4
|
+
version: 0.1.5.3
|
5
5
|
platform: ruby
|
6
6
|
authors:
|
7
7
|
- Yoichiro Hasebe
|
8
8
|
autorequire:
|
9
9
|
bindir: bin
|
10
10
|
cert_chain: []
|
11
|
-
date: 2023-
|
11
|
+
date: 2023-07-26 00:00:00.000000000 Z
|
12
12
|
dependencies:
|
13
13
|
- !ruby/object:Gem::Dependency
|
14
14
|
name: bundler
|
@@ -208,7 +208,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
|
|
208
208
|
- !ruby/object:Gem::Version
|
209
209
|
version: '0'
|
210
210
|
requirements: []
|
211
|
-
rubygems_version: 3.4.
|
211
|
+
rubygems_version: 3.4.12
|
212
212
|
signing_key:
|
213
213
|
specification_version: 4
|
214
214
|
summary: A wrapper module for using spaCy natural language processing library from
|