relaton-w3c 1.3.1 → 1.5.0

Sign up to get free protection for your applications and to get access to all the features.
Files changed (7) hide show
  1. checksums.yaml +4 -4
  2. data/README.adoc +32 -22
  3. data/lib/relaton_w3c/hit_collection.rb +7 -2
  4. data/lib/relaton_w3c/scrapper.rb +12 -3
  5. data/lib/relaton_w3c/version.rb +1 -1
  6. data/relaton_w3c.gemspec +1 -1
  7. metadata +4 -4
checksums.yaml CHANGED
@@ -1,7 +1,7 @@
1
1
  ---
2
2
  SHA256:
3
- metadata.gz: 8591c4405cfb771a9cf41e2c74a7d36f21f66e02776389efdf670e639818524a
4
- data.tar.gz: be0e0ea3effefd0ab0a48d72cb42eb5da6c62537deaa6dca876d3ee000cc12c6
3
+ metadata.gz: da93038fdc3d0a5448818ddb7b32173f99e6552a54f2d19b7a88edbe20ac14d0
4
+ data.tar.gz: f9a0ccbbe5172ed6618a7d570f032feb8b8d16b915dc19b1533215134a18e66f
5
5
  SHA512:
6
- metadata.gz: de33be8c4b6a7d9b3ee165454f6988044f5e614239d35e5c9bfa4b8dfc724430f6366dfb90d9bd1c648c78ef1ef7a3701e4ac4461eff9d82be2e2502e588f8bf
7
- data.tar.gz: fa7c426893f99844fa702dcbd349bcc141800994e167da25657ea1d5143de442aa5be4be03242221635edf9836d4c1c678e4a522f52c964ddd50f851fdda8d4a
6
+ metadata.gz: 806960b3a8ada23fa2015fb88c9d4e9095c790ada8a0ed758d58cf7d99f974c2c25b2e0453ecf3f9e10a518f1b5ed4bdc2a829e33821ef7f86ddb3ad2ef894cf
7
+ data.tar.gz: d546bf3674eef23bfc90484471dd9130cc849b74557145ae674cf5eba5b1a127938db7f65321846eb669e45e68dc1b921a2c0716a4f8a1f8ee7666a5d945fc62
data/README.adoc CHANGED
@@ -10,7 +10,7 @@ Add this line to your application's Gemfile:
10
10
 
11
11
  [source,ruby]
12
12
  ----
13
- gem 'relaton_w3c'
13
+ gem 'relaton-w3c'
14
14
  ----
15
15
 
16
16
  And then execute:
@@ -19,14 +19,16 @@ And then execute:
19
19
 
20
20
  Or install it yourself as:
21
21
 
22
- $ gem install relaton_w3c
22
+ $ gem install relaton-w3c
23
23
 
24
24
  == Usage
25
25
 
26
26
  === Search for a standard using keywords
27
27
 
28
+ [source,ruby]
28
29
  ----
29
30
  require 'relaton_w3c'
31
+ => true
30
32
 
31
33
  hits = RelatonW3c::W3cBibliography.search("W3C JSON-LD 1.1")
32
34
  => <RelatonW3c::HitCollection:0x007f93b5e4ff48 @ref=W3C JSON-LD 1.1 @fetched=false>
@@ -38,40 +40,43 @@ item = hits[0].fetch
38
40
 
39
41
  === XML serialization
40
42
 
43
+ [source,ruby]
41
44
  ----
42
45
  item.to_xml
43
- => "<bibitem type="standard">
44
- <fetched>2020-04-07</fetched>
45
- <title type="main" format="text/plain" language="en" script="Latn">JSON-LD 1.1</title>
46
- <title format="text/plain" language="en" script="Latn">A JSON-based Serialization for Linked Data</title>
47
- <uri type="src">https://www.w3.org/TR/2020/CR-json-ld11-20200316/</uri>
48
- <date type="published">
49
- <on>2020</on>
50
- </date>
51
- ...
46
+ => "<bibitem id="CR-json-ld11-20200316" type="standard">
47
+ <fetched>2020-04-07</fetched>
48
+ <title type="main" format="text/plain" language="en" script="Latn">JSON-LD 1.1</title>
49
+ <title format="text/plain" language="en" script="Latn">A JSON-based Serialization for Linked Data</title>
50
+ <uri type="src">https://www.w3.org/TR/2020/CR-json-ld11-20200316/</uri>
51
+ <date type="published">
52
+ <on>2020</on>
53
+ </date>
54
+ ...
52
55
  </bibitem>"
53
56
  ----
54
57
 
55
58
  With argument `bibdata: true` it ouputs XML wrapped by `bibdata` element and adds flavour `ext` element.
56
59
 
60
+ [source,ruby]
57
61
  ----
58
62
  item.to_xml bibdata: true
59
63
  => "<bibdata type="standard">
60
- <fetched>2020-04-07</fetched>
61
- <title type="main" format="text/plain" language="en" script="Latn">JSON-LD 1.1</title>
62
- <title format="text/plain" language="en" script="Latn">A JSON-based Serialization for Linked Data</title>
63
- <uri type="src">https://www.w3.org/TR/2020/CR-json-ld11-20200316/</uri>
64
- <date type="published">
65
- <on>2020</on>
66
- </date>
67
- ...
68
- <ext>
69
- <doctype>candidateRecommendation</doctype>
70
- </ext>
64
+ <fetched>2020-04-07</fetched>
65
+ <title type="main" format="text/plain" language="en" script="Latn">JSON-LD 1.1</title>
66
+ <title format="text/plain" language="en" script="Latn">A JSON-based Serialization for Linked Data</title>
67
+ <uri type="src">https://www.w3.org/TR/2020/CR-json-ld11-20200316/</uri>
68
+ <date type="published">
69
+ <on>2020</on>
70
+ </date>
71
+ ...
72
+ <ext>
73
+ <doctype>candidateRecommendation</doctype>
74
+ </ext>
71
75
  </bibdata>"
72
76
  ----
73
77
 
74
78
  === Get document by title
79
+ [source,ruby]
75
80
  ----
76
81
  RelatonW3c::W3cBibliography.get "W3C JSON-LD 1.1"
77
82
  [relaton-w3c] ("W3C JSON-LD 1.1") fetching...
@@ -81,6 +86,7 @@ RelatonW3c::W3cBibliography.get "W3C JSON-LD 1.1"
81
86
  ----
82
87
 
83
88
  === Get document by title and type
89
+ [source,ruby]
84
90
  ----
85
91
  RelatonW3c::W3cBibliography.get "W3C Candidate Recommendation JSON-LD 1.1"
86
92
  [relaton-w3c] ("W3C Candidate Recommendation JSON-LD 1.1") fetching...
@@ -90,6 +96,7 @@ RelatonW3c::W3cBibliography.get "W3C Candidate Recommendation JSON-LD 1.1"
90
96
  ----
91
97
 
92
98
  === Get document by title and short type
99
+ [source,ruby]
93
100
  ----
94
101
  RelatonW3c::W3cBibliography.get "W3C CR JSON-LD 1.1"
95
102
  [relaton-w3c] ("W3C CR JSON-LD 1.1") fetching...
@@ -99,6 +106,7 @@ RelatonW3c::W3cBibliography.get "W3C CR JSON-LD 1.1"
99
106
  ----
100
107
 
101
108
  === Get document by title, type, and date
109
+ [source,ruby]
102
110
  ----
103
111
  RelatonW3c::W3cBibliography.get "W3C WD JSON-LD 1.1 2019-10-18"
104
112
  [relaton-w3c] ("W3C WD JSON-LD 1.1 2019-10-18") fetching...
@@ -108,6 +116,7 @@ RelatonW3c::W3cBibliography.get "W3C WD JSON-LD 1.1 2019-10-18"
108
116
  ----
109
117
 
110
118
  === Create bibliographic item from XML
119
+ [source,ruby]
111
120
  ----
112
121
  RelatonW3c::XMLParser.from_xml File.read('spec/fixtures/cr_json_ld11.xml')
113
122
  => #<RelatonW3c::W3cBibliographicItem:0x007f9381efce98
@@ -115,6 +124,7 @@ RelatonW3c::XMLParser.from_xml File.read('spec/fixtures/cr_json_ld11.xml')
115
124
  ----
116
125
 
117
126
  === Create bibliographic item from YAML
127
+ [source,ruby]
118
128
  ----
119
129
  hash = YAML.load_file 'spec/fixtures/cr_json_ld11.yml'
120
130
  => {"title"=>
data/lib/relaton_w3c/hit_collection.rb CHANGED
@@ -45,7 +45,12 @@ module RelatonW3c
45
45
  /(?<title>.+)\s(?<date>\d{4}-\d{2}-\d{2})$/ =~ title_date
46
46
  title ||= title_date
47
47
  result = data.select do |hit|
48
- hit["title"] == title && type_date_filter(hit, type, date)
48
+ (hit["title"].casecmp?(title) ||
49
+ hit["link"].split("/").last.match?(/-#{title}-/)) &&
50
+ type_date_filter(hit, type, date)
51
+ end
52
+ if result.empty?
53
+ result = data.select { |h| h["link"].split("/").last.match? /#{title}/ }
49
54
  end
50
55
  result.map { |h| Hit.new(h, self) }
51
56
  end
@@ -54,7 +59,7 @@ module RelatonW3c
54
59
  # @param type [String]
55
60
  # @param date [String]
56
61
  # @return [TrueClass, FalseClass]
57
- def type_date_filter(hit, type, date)
62
+ def type_date_filter(hit, type, date) # rubocop:disable Metrics/AbcSize
58
63
  if type && hit["type"] != short_type(type) || date && hit["date"] != date
59
64
  history = get_history hit, type, date
60
65
  return false unless history.any?
data/lib/relaton_w3c/scrapper.rb CHANGED
@@ -18,6 +18,7 @@ module RelatonW3c
18
18
  doc = resp.code == "200" ? Nokogiri::HTML(resp.body) : nil
19
19
  W3cBibliographicItem.new(
20
20
  type: "standard",
21
+ docid: fetch_docid(hit),
21
22
  fetched: Date.today.to_s,
22
23
  language: ["en"],
23
24
  script: ["Latn"],
@@ -34,19 +35,27 @@ module RelatonW3c
34
35
 
35
36
  private
36
37
 
38
+ # @param hit [Hash]
39
+ # @return [Array<RelatonBib::DocumentIdentifier>]
40
+ def fetch_docid(hit)
41
+ id = hit["link"].split("/").last
42
+ [RelatonBib::DocumentIdentifier.new(id: id, type: "W3C")]
43
+ end
44
+
37
45
  # @param hit [Hash]
38
46
  # @param doc [Nokogiri::HTML::Document]
39
47
  # @return [Array<RelatonBib::TypedTitleString>]
40
48
  def fetch_title(hit, doc) # rubocop:disable Metrics/AbcSize, Metrics/CyclomaticComplexity, Metrics/MethodLength, Metrics/PerceivedComplexity
41
49
  titles = []
42
50
  if doc
43
- title = doc.at("//h1[contains(@id, 'title')]")&.text
44
- titles << { content: title, type: "main" } if title
51
+ title = doc.at("//*[contains(@id, 'title')]")&.text
52
+ titles << { content: title, type: "main" } if title && !title.empty?
45
53
  subtitle = doc.at(
46
54
  "//h2[@id='subtitle']|//p[contains(@class, 'subline')]"
47
55
  )&.text
48
56
  titles << { content: subtitle, tipe: "subtitle" } if subtitle
49
- elsif hit["title"]
57
+ end
58
+ if titles.empty? && hit["title"]
50
59
  titles << { content: hit["title"], type: "main" }
51
60
  end
52
61
  titles.map do |t|
data/lib/relaton_w3c/version.rb CHANGED
@@ -1,3 +1,3 @@
1
1
  module RelatonW3c
2
- VERSION = "1.3.1".freeze
2
+ VERSION = "1.5.0".freeze
3
3
  end
data/relaton_w3c.gemspec CHANGED
@@ -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", ">= 1.3.0"
42
+ spec.add_dependency "relaton-bib", ">= 1.5.0"
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: 1.3.1
4
+ version: 1.5.0
5
5
  platform: ruby
6
6
  authors:
7
7
  - Ribose Inc.
8
8
  autorequire:
9
9
  bindir: exe
10
10
  cert_chain: []
11
- date: 2020-08-31 00:00:00.000000000 Z
11
+ date: 2020-10-09 00:00:00.000000000 Z
12
12
  dependencies:
13
13
  - !ruby/object:Gem::Dependency
14
14
  name: debase
@@ -114,14 +114,14 @@ dependencies:
114
114
  requirements:
115
115
  - - ">="
116
116
  - !ruby/object:Gem::Version
117
- version: 1.3.0
117
+ version: 1.5.0
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: 1.3.0
124
+ version: 1.5.0
125
125
  description: 'RelatonIso: retrieve W3C Standards for bibliographic using the IsoBibliographicItem
126
126
  model'
127
127
  email: