relaton-w3c 0.1.0 → 1.2.0
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +4 -4
- data/grammars/basicdoc.rng +986 -0
- data/grammars/biblio.rng +1237 -0
- data/lib/relaton_w3c/hash_converter.rb +7 -0
- data/lib/relaton_w3c/hit_collection.rb +3 -0
- data/lib/relaton_w3c/scrapper.rb +2 -2
- data/lib/relaton_w3c/version.rb +1 -1
- data/lib/relaton_w3c/w3c_bibliographic_item.rb +1 -1
- data/lib/relaton_w3c/w3c_bibliography.rb +3 -1
- data/lib/relaton_w3c/xml_parser.rb +6 -13
- data/relaton_w3c.gemspec +3 -3
- metadata +8 -6
data/lib/relaton_w3c/scrapper.rb
CHANGED
@@ -44,7 +44,7 @@ module RelatonW3c
|
|
44
44
|
titles << { content: title, type: "main" }
|
45
45
|
subtitle = doc.at("//h2[@id='subtitle']").text
|
46
46
|
titles << { content: subtitle, tipe: "subtitle" }
|
47
|
-
|
47
|
+
elsif hit["title"]
|
48
48
|
titles << { content: hit["title"], type: "main" }
|
49
49
|
end
|
50
50
|
titles.map do |t|
|
@@ -166,7 +166,7 @@ module RelatonW3c
|
|
166
166
|
|
167
167
|
hit = { "link" => link }
|
168
168
|
item = parse_page hit
|
169
|
-
[RelatonBib::DocumentRelation.new(type: "
|
169
|
+
[RelatonBib::DocumentRelation.new(type: "obsoletedBy", bibitem: item)]
|
170
170
|
end
|
171
171
|
|
172
172
|
# @param doc [Nokogiri::HTML::Document]
|
data/lib/relaton_w3c/version.rb
CHANGED
@@ -12,8 +12,8 @@ module RelatonW3c
|
|
12
12
|
if args[:doctype] && !TYPES.include?(args[:doctype])
|
13
13
|
warn "[relaton-w3c] invalid document type: #{args[:doctype]}"
|
14
14
|
end
|
15
|
-
@doctype = args.delete :doctype
|
16
15
|
super **args
|
16
|
+
# @doctype = args[:doctype]
|
17
17
|
end
|
18
18
|
|
19
19
|
# @param builder [Nokogiri::XML::Builder, NilClass]
|
@@ -1,5 +1,7 @@
|
|
1
1
|
# frozen_string_literal: true
|
2
2
|
|
3
|
+
require "net/http"
|
4
|
+
|
3
5
|
module RelatonW3c
|
4
6
|
# Class methods for search W3C standards.
|
5
7
|
class W3cBibliography
|
@@ -10,7 +12,7 @@ module RelatonW3c
|
|
10
12
|
HitCollection.new text
|
11
13
|
rescue SocketError, Timeout::Error, Errno::EINVAL, Errno::ECONNRESET,
|
12
14
|
EOFError, Net::HTTPBadResponse, Net::HTTPHeaderSyntaxError,
|
13
|
-
Net::ProtocolError,
|
15
|
+
Net::ProtocolError, Errno::ETIMEDOUT
|
14
16
|
raise RelatonBib::RequestError,
|
15
17
|
"Could not access #{HitCollection::DOMAIN}"
|
16
18
|
end
|
@@ -1,19 +1,6 @@
|
|
1
1
|
module RelatonW3c
|
2
2
|
class XMLParser < RelatonBib::XMLParser
|
3
3
|
class << self
|
4
|
-
# @param xml [String]
|
5
|
-
# @return [RelatonW3c::W3cBibliographicItem, NilClass]
|
6
|
-
def from_xml(xml)
|
7
|
-
doc = Nokogiri::XML xml
|
8
|
-
doc.remove_namespaces!
|
9
|
-
item = doc.at("/bibitem|/bibdata")
|
10
|
-
if item
|
11
|
-
W3cBibliographicItem.new(item_data(item))
|
12
|
-
else
|
13
|
-
warn "[relaton-w3c] can't find bibitem or bibdata element in the XML"
|
14
|
-
end
|
15
|
-
end
|
16
|
-
|
17
4
|
private
|
18
5
|
|
19
6
|
# Override RelatonBib::XMLParser.item_data method.
|
@@ -27,6 +14,12 @@ module RelatonW3c
|
|
27
14
|
data[:doctype] = ext.at("./doctype")&.text
|
28
15
|
data
|
29
16
|
end
|
17
|
+
|
18
|
+
# @param item_hash [Hash]
|
19
|
+
# @return [RelatonBib::BibliographicItem]
|
20
|
+
def bib_item(item_hash)
|
21
|
+
W3cBibliographicItem.new item_hash
|
22
|
+
end
|
30
23
|
end
|
31
24
|
end
|
32
25
|
end
|
data/relaton_w3c.gemspec
CHANGED
@@ -9,9 +9,9 @@ Gem::Specification.new do |spec|
|
|
9
9
|
spec.email = ["open.source@ribose.com"]
|
10
10
|
|
11
11
|
spec.summary = "RelatonIso: retrieve W3C Standards for bibliographic "\
|
12
|
-
"
|
12
|
+
"using the IsoBibliographicItem model"
|
13
13
|
spec.description = "RelatonIso: retrieve W3C Standards for bibliographic "\
|
14
|
-
"
|
14
|
+
"using the IsoBibliographicItem model"
|
15
15
|
spec.homepage = "https://github.com/relaton/relaton-wc3"
|
16
16
|
spec.license = "BSD-2-Clause"
|
17
17
|
spec.required_ruby_version = Gem::Requirement.new(">= 2.4.0")
|
@@ -39,5 +39,5 @@ Gem::Specification.new do |spec|
|
|
39
39
|
spec.add_development_dependency "vcr"
|
40
40
|
spec.add_development_dependency "webmock"
|
41
41
|
|
42
|
-
spec.add_dependency "relaton-bib", ">=
|
42
|
+
spec.add_dependency "relaton-bib", ">= 1.2.0"
|
43
43
|
end
|
metadata
CHANGED
@@ -1,14 +1,14 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: relaton-w3c
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version:
|
4
|
+
version: 1.2.0
|
5
5
|
platform: ruby
|
6
6
|
authors:
|
7
7
|
- Ribose Inc.
|
8
8
|
autorequire:
|
9
9
|
bindir: exe
|
10
10
|
cert_chain: []
|
11
|
-
date: 2020-
|
11
|
+
date: 2020-06-26 00:00:00.000000000 Z
|
12
12
|
dependencies:
|
13
13
|
- !ruby/object:Gem::Dependency
|
14
14
|
name: debase
|
@@ -114,15 +114,15 @@ dependencies:
|
|
114
114
|
requirements:
|
115
115
|
- - ">="
|
116
116
|
- !ruby/object:Gem::Version
|
117
|
-
version:
|
117
|
+
version: 1.2.0
|
118
118
|
type: :runtime
|
119
119
|
prerelease: false
|
120
120
|
version_requirements: !ruby/object:Gem::Requirement
|
121
121
|
requirements:
|
122
122
|
- - ">="
|
123
123
|
- !ruby/object:Gem::Version
|
124
|
-
version:
|
125
|
-
description: 'RelatonIso: retrieve W3C Standards for bibliographic
|
124
|
+
version: 1.2.0
|
125
|
+
description: 'RelatonIso: retrieve W3C Standards for bibliographic using the IsoBibliographicItem
|
126
126
|
model'
|
127
127
|
email:
|
128
128
|
- open.source@ribose.com
|
@@ -144,6 +144,8 @@ files:
|
|
144
144
|
- bin/console
|
145
145
|
- bin/rspec
|
146
146
|
- bin/setup
|
147
|
+
- grammars/basicdoc.rng
|
148
|
+
- grammars/biblio.rng
|
147
149
|
- lib/relaton_w3c.rb
|
148
150
|
- lib/relaton_w3c/hash_converter.rb
|
149
151
|
- lib/relaton_w3c/hit.rb
|
@@ -178,6 +180,6 @@ requirements: []
|
|
178
180
|
rubygems_version: 3.0.6
|
179
181
|
signing_key:
|
180
182
|
specification_version: 4
|
181
|
-
summary: 'RelatonIso: retrieve W3C Standards for bibliographic
|
183
|
+
summary: 'RelatonIso: retrieve W3C Standards for bibliographic using the IsoBibliographicItem
|
182
184
|
model'
|
183
185
|
test_files: []
|