relaton-bib 0.4.1 → 0.5.0
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +4 -4
- data/.gitignore +2 -1
- data/Gemfile.lock +24 -19
- data/README.adoc +185 -136
- data/bin/rspec +29 -0
- data/docs/hash.adoc +130 -20
- data/lib/relaton_bib/bibliographic_item.rb +176 -9
- data/lib/relaton_bib/bibtex_parser.rb +236 -0
- data/lib/relaton_bib/contribution_info.rb +1 -1
- data/lib/relaton_bib/hash_converter.rb +12 -6
- data/lib/relaton_bib/place.rb +37 -0
- data/lib/relaton_bib/series.rb +2 -2
- data/lib/relaton_bib/typed_title_string.rb +1 -1
- data/lib/relaton_bib/version.rb +1 -1
- data/lib/relaton_bib/xml_parser.rb +10 -5
- data/relaton-bib.gemspec +2 -1
- metadata +33 -16
@@ -0,0 +1,236 @@
|
|
1
|
+
require "bibtex"
|
2
|
+
require "iso639"
|
3
|
+
|
4
|
+
module RelatonBib
|
5
|
+
class BibtexParser
|
6
|
+
class << self
|
7
|
+
# @param bibtex [String]
|
8
|
+
# @return [Hash{String=>RelatonBib::BibliographicItem}]
|
9
|
+
def from_bibtex(bibtex)
|
10
|
+
BibTeX.parse(bibtex).reduce({}) do |h, bt|
|
11
|
+
h[bt.key] = BibliographicItem.new(
|
12
|
+
id: bt.key,
|
13
|
+
docid: fetch_docid(bt),
|
14
|
+
fetched: fetch_fetched(bt),
|
15
|
+
type: fetch_type(bt),
|
16
|
+
title: fetch_title(bt),
|
17
|
+
contributor: fetch_contributor(bt),
|
18
|
+
date: fetch_date(bt),
|
19
|
+
place: fetch_place(bt),
|
20
|
+
biblionote: fetch_note(bt),
|
21
|
+
relation: fetch_relation(bt),
|
22
|
+
extent: fetch_extent(bt),
|
23
|
+
edition: bt["edition"]&.to_s,
|
24
|
+
series: fetch_series(bt),
|
25
|
+
link: fetch_link(bt),
|
26
|
+
language: fetch_language(bt),
|
27
|
+
classification: fetch_classification(bt)
|
28
|
+
)
|
29
|
+
h
|
30
|
+
end
|
31
|
+
end
|
32
|
+
|
33
|
+
private
|
34
|
+
|
35
|
+
# @param bibtex [BibTeX::Entry]
|
36
|
+
# @return [Array<RelatonBib::DocumentIdentifier>]
|
37
|
+
def fetch_docid(bibtex)
|
38
|
+
docid = []
|
39
|
+
docid << DocumentIdentifier.new(id: bibtex.isbn.to_s, type: "isbn") if bibtex["isbn"]
|
40
|
+
docid << DocumentIdentifier.new(id: bibtex.lccn.to_s, type: "lccn") if bibtex["lccn"]
|
41
|
+
docid << DocumentIdentifier.new(id: bibtex.issn.to_s, type: "issn") if bibtex["issn"]
|
42
|
+
docid
|
43
|
+
end
|
44
|
+
|
45
|
+
# @param bibtex [BibTeX::Entry]
|
46
|
+
# @return [String, NilClass]
|
47
|
+
def fetch_fetched(bibtex)
|
48
|
+
Date.parse(bibtex.timestamp.to_s) if bibtex["timestamp"]
|
49
|
+
end
|
50
|
+
|
51
|
+
# @param bibtex [BibTeX::Entry]
|
52
|
+
# @return [String]
|
53
|
+
def fetch_type(bibtex)
|
54
|
+
case bibtex.type
|
55
|
+
when :mastersthesis, :phdthesis then "thesis"
|
56
|
+
when :conference then "inproceedings"
|
57
|
+
when :misc then "standard"
|
58
|
+
else bibtex.type.to_s
|
59
|
+
end
|
60
|
+
end
|
61
|
+
|
62
|
+
# @param bibtex [BibTeX::Entry]
|
63
|
+
# @return [Array<Hash>]
|
64
|
+
def fetch_place(bibtex)
|
65
|
+
bibtex["address"] ? [bibtex.address.to_s] : []
|
66
|
+
end
|
67
|
+
|
68
|
+
# @param bibtex [BibTeX::Entry]
|
69
|
+
# @return [Array<Hash>]
|
70
|
+
def fetch_title(bibtex)
|
71
|
+
title = []
|
72
|
+
title << { type: "main", content: bibtex.title.to_s } if bibtex["title"]
|
73
|
+
title << { type: "main", content: bibtex.subtitle.to_s } if bibtex["subtitle"]
|
74
|
+
title
|
75
|
+
end
|
76
|
+
|
77
|
+
# @param bibtex [BibTeX::Entry]
|
78
|
+
# @return [Array<Hash>]
|
79
|
+
def fetch_contributor(bibtex)
|
80
|
+
contributor = []
|
81
|
+
fetch_person(bibtex["author"]).each do |entity|
|
82
|
+
contributor << { entity: entity, role: [{ type: "author" }] }
|
83
|
+
end
|
84
|
+
|
85
|
+
fetch_person(bibtex["editor"]).each do |entity|
|
86
|
+
contributor << { entity: entity, role: [{ type: "editor" }] }
|
87
|
+
end
|
88
|
+
|
89
|
+
if bibtex["publisher"]
|
90
|
+
contributor << { entity: { name: bibtex.publisher.to_s }, role: [{ type: "publisher" }] }
|
91
|
+
end
|
92
|
+
|
93
|
+
if bibtex["institution"]
|
94
|
+
contributor << {
|
95
|
+
entity: { name: bibtex.institution.to_s },
|
96
|
+
role: [{ type: "distributor", description: ["sponsor"] }]
|
97
|
+
}
|
98
|
+
end
|
99
|
+
|
100
|
+
if bibtex["organization"]
|
101
|
+
contributor << {
|
102
|
+
entity: { name: bibtex.organization.to_s },
|
103
|
+
role: [{ type: "distributor", description: ["sponsor"] }]
|
104
|
+
}
|
105
|
+
end
|
106
|
+
|
107
|
+
if bibtex["school"]
|
108
|
+
contributor << {
|
109
|
+
entity: { name: bibtex.school.to_s },
|
110
|
+
role: [{ type: "distributor", description: ["sponsor"] }]
|
111
|
+
}
|
112
|
+
end
|
113
|
+
contributor
|
114
|
+
end
|
115
|
+
|
116
|
+
# @param bibtex [BibTeX::Entry]
|
117
|
+
# @return [Array<RelatonBib::Person>]
|
118
|
+
def fetch_person(person)
|
119
|
+
return [] unless person
|
120
|
+
|
121
|
+
person.map do |name|
|
122
|
+
parts = name.split ", "
|
123
|
+
surname = LocalizedString.new parts.first
|
124
|
+
fname = parts.size > 1 ? parts[1].split : []
|
125
|
+
forename = fname.map { |fn| LocalizedString.new fn }
|
126
|
+
Person.new name: FullName.new(surname: surname, forename: forename)
|
127
|
+
end
|
128
|
+
end
|
129
|
+
|
130
|
+
# @param bibtex [BibTeX::Entry]
|
131
|
+
# @return [Array<Hash>]
|
132
|
+
def fetch_date(bibtex)
|
133
|
+
date = []
|
134
|
+
if bibtex["year"]
|
135
|
+
on = Date.new(bibtex.year.to_i, bibtex["month_numeric"]&.to_i || 1).to_s
|
136
|
+
date << { type: "published", on: on }
|
137
|
+
end
|
138
|
+
|
139
|
+
if bibtex["urldate"]
|
140
|
+
date << { type: "accessed", on: Date.parse(bibtex.urldate.to_s).to_s }
|
141
|
+
end
|
142
|
+
|
143
|
+
date
|
144
|
+
end
|
145
|
+
|
146
|
+
# @param bibtex [BibTeX::Entry]
|
147
|
+
# @return [Array<RelatonBib::BiblioNote>]
|
148
|
+
def fetch_note(bibtex)
|
149
|
+
note = []
|
150
|
+
note << BiblioNote.new(type: "annote", content: bibtex.annote.to_s) if bibtex["annote"]
|
151
|
+
|
152
|
+
if bibtex["howpublished"]
|
153
|
+
note << BiblioNote.new(type: "howpublished", content: bibtex.howpublished.to_s)
|
154
|
+
end
|
155
|
+
|
156
|
+
note << BiblioNote.new(type: "comment", content: bibtex.comment.to_s) if bibtex["comment"]
|
157
|
+
note << BiblioNote.new(content: bibtex.note.to_s) if bibtex["note"]
|
158
|
+
note << BiblioNote.new(type: "tableOfContents", content: bibtex["content"]) if bibtex["content"]
|
159
|
+
note
|
160
|
+
end
|
161
|
+
|
162
|
+
# @param bibtex [BibTeX::Entry]
|
163
|
+
# @return [Array<Hash>]
|
164
|
+
def fetch_relation(bibtex)
|
165
|
+
return [] unless bibtex["booktitle"]
|
166
|
+
|
167
|
+
title = TypedTitleString.new(type: "main", content: bibtex.booktitle.to_s)
|
168
|
+
[{ type: "partOf", bibitem: BibliographicItem.new(title: [title]) }]
|
169
|
+
end
|
170
|
+
|
171
|
+
# @param bibtex [BibTeX::Entry]
|
172
|
+
# @return [Array<RelatonBib::BibItemLocality>]
|
173
|
+
def fetch_extent(bibtex)
|
174
|
+
extent = []
|
175
|
+
extent << BibItemLocality.new("chapter", bibtex.chapter.to_s) if bibtex["chapter"]
|
176
|
+
|
177
|
+
if bibtex["pages"]
|
178
|
+
from, to = bibtex.pages.to_s.split "-"
|
179
|
+
extent << BibItemLocality.new("page", from, to)
|
180
|
+
end
|
181
|
+
|
182
|
+
extent << BibItemLocality.new("volume", bibtex.volume.to_s) if bibtex["volume"]
|
183
|
+
extent
|
184
|
+
end
|
185
|
+
|
186
|
+
# @param bibtex [BibTeX::Entry]
|
187
|
+
# @return [Array<RelatonBib::Series>]
|
188
|
+
def fetch_series(bibtex)
|
189
|
+
series = []
|
190
|
+
if bibtex["journal"]
|
191
|
+
series << Series.new(
|
192
|
+
type: "journal",
|
193
|
+
title: TypedTitleString.new(content: bibtex.journal.to_s),
|
194
|
+
number: bibtex["number"]&.to_s
|
195
|
+
)
|
196
|
+
end
|
197
|
+
|
198
|
+
if bibtex["series"]
|
199
|
+
title = TypedTitleString.new content: bibtex.series.to_s
|
200
|
+
series << Series.new(title: title)
|
201
|
+
end
|
202
|
+
series
|
203
|
+
end
|
204
|
+
|
205
|
+
# @param bibtex [BibTeX::Entry]
|
206
|
+
# @return [Array<RelatonBib::TypedUri>]
|
207
|
+
def fetch_link(bibtex)
|
208
|
+
link = []
|
209
|
+
link << TypedUri.new(type: "src", content: bibtex.url.to_s) if bibtex["url"]
|
210
|
+
link << TypedUri.new(type: "doi", content: bibtex.doi.to_s) if bibtex["doi"]
|
211
|
+
link << TypedUri.new(type: "file", content: bibtex.file2.to_s) if bibtex["file2"]
|
212
|
+
link
|
213
|
+
end
|
214
|
+
|
215
|
+
# @param bibtex [BibTeX::Entry]
|
216
|
+
# @return [Array<String>]
|
217
|
+
def fetch_language(bibtex)
|
218
|
+
return [] unless bibtex["language"]
|
219
|
+
|
220
|
+
[Iso639[bibtex.language.to_s].alpha2]
|
221
|
+
end
|
222
|
+
|
223
|
+
# @param bibtex [BibTeX::Entry]
|
224
|
+
# @return [RelatonBib::Classification, NilClass]
|
225
|
+
def fetch_classification(bibtex)
|
226
|
+
cls = []
|
227
|
+
cls << Classification.new(type: "type", value: bibtex["type"].to_s) if bibtex["type"]
|
228
|
+
cls << Classification.new(type: "keyword", value: bibtex.keywords.to_s) if bibtex["keywords"]
|
229
|
+
if bibtex["mendeley-tags"]
|
230
|
+
cls << Classification.new(type: "mendeley", value: bibtex["mendeley-tags"].to_s)
|
231
|
+
end
|
232
|
+
cls
|
233
|
+
end
|
234
|
+
end
|
235
|
+
end
|
236
|
+
end
|
@@ -22,7 +22,7 @@ module RelatonBib
|
|
22
22
|
# @param description [Array<String>]
|
23
23
|
def initialize(**args)
|
24
24
|
if args[:type] && !TYPES.include?(args[:type])
|
25
|
-
|
25
|
+
warn %{[relaton-bib] Contributor's type "#{args[:type]}" is invalid.}
|
26
26
|
end
|
27
27
|
|
28
28
|
@type = args[:type]
|
@@ -84,7 +84,9 @@ module RelatonBib
|
|
84
84
|
def place_hash_to_bib(ret)
|
85
85
|
return unless ret[:place]
|
86
86
|
|
87
|
-
ret[:place] = array(ret[:place])
|
87
|
+
ret[:place] = array(ret[:place]).map do |pl|
|
88
|
+
pl.is_a?(String) ? Place.new(name: pl) : Place.new(pl)
|
89
|
+
end
|
88
90
|
end
|
89
91
|
|
90
92
|
def accesslocation_hash_to_bib(ret)
|
@@ -130,10 +132,14 @@ module RelatonBib
|
|
130
132
|
|
131
133
|
ret[:biblionote] = array(ret[:biblionote])
|
132
134
|
(ret[:biblionote])&.each_with_index do |n, i|
|
133
|
-
ret[:biblionote][i] =
|
134
|
-
|
135
|
-
|
136
|
-
|
135
|
+
ret[:biblionote][i] = if n.is_a?(String)
|
136
|
+
BiblioNote.new content: n
|
137
|
+
else
|
138
|
+
BiblioNote.new(
|
139
|
+
content: n[:content], type: n[:type], language: n[:language],
|
140
|
+
script: n[:script], format: n[:format]
|
141
|
+
)
|
142
|
+
end
|
137
143
|
end
|
138
144
|
end
|
139
145
|
|
@@ -308,7 +314,7 @@ module RelatonBib
|
|
308
314
|
# ret[:classification][i] = RelatonBib::Classification.new(c)
|
309
315
|
# end
|
310
316
|
if ret[:classification]
|
311
|
-
ret[:classification] =
|
317
|
+
ret[:classification] = array(ret[:classification]).map { |cls| Classification.new cls }
|
312
318
|
end
|
313
319
|
end
|
314
320
|
|
@@ -0,0 +1,37 @@
|
|
1
|
+
module RelatonBib
|
2
|
+
class Place
|
3
|
+
# @return [String]
|
4
|
+
attr_reader :name
|
5
|
+
|
6
|
+
# @return [String, NilClass]
|
7
|
+
attr_reader :uri, :region
|
8
|
+
|
9
|
+
# @param name [String]
|
10
|
+
# @param uri [String, NilClass]
|
11
|
+
# @param region [String, NilClass]
|
12
|
+
def initialize(name:, uri: nil, region: nil)
|
13
|
+
@name = name
|
14
|
+
@uri = uri
|
15
|
+
@region = region
|
16
|
+
end
|
17
|
+
|
18
|
+
# @param builder [Nologiri::XML::Builder]
|
19
|
+
def to_xml(builder)
|
20
|
+
xml = builder.place name
|
21
|
+
xml[:uri] = uri if uri
|
22
|
+
xml[:region] = region if region
|
23
|
+
end
|
24
|
+
|
25
|
+
# @return [Hash]
|
26
|
+
def to_hash
|
27
|
+
if uri || region
|
28
|
+
hash = { name: name }
|
29
|
+
hash["uri"] = uri if uri
|
30
|
+
hash["region"] = region if region
|
31
|
+
hash
|
32
|
+
else
|
33
|
+
name
|
34
|
+
end
|
35
|
+
end
|
36
|
+
end
|
37
|
+
end
|
data/lib/relaton_bib/series.rb
CHANGED
@@ -55,7 +55,7 @@ module RelatonBib
|
|
55
55
|
end
|
56
56
|
|
57
57
|
if args[:type] && !TYPES.include?(args[:type])
|
58
|
-
|
58
|
+
warn "[relaton-bib] Series type is invalid: #{args[:type]}"
|
59
59
|
end
|
60
60
|
|
61
61
|
@type = args[:type] # if %w[main alt].include? args[:type]
|
@@ -67,7 +67,7 @@ module RelatonBib
|
|
67
67
|
@from = args[:from]
|
68
68
|
@to = args[:to]
|
69
69
|
@number = args[:number]
|
70
|
-
@partnumber
|
70
|
+
@partnumber = args[:partnumber]
|
71
71
|
end
|
72
72
|
# rubocop:enable Metrics/MethodLength
|
73
73
|
|
@@ -17,7 +17,7 @@ module RelatonBib
|
|
17
17
|
# @param script [String]
|
18
18
|
def initialize(**args)
|
19
19
|
if args[:type] && !TITLE_TYPES.include?(args[:type])
|
20
|
-
|
20
|
+
warn %{[relaton-bib] title type "#{args[:type]}" is invalid.}
|
21
21
|
end
|
22
22
|
|
23
23
|
unless args[:title] || args[:content]
|
data/lib/relaton_bib/version.rb
CHANGED
@@ -38,7 +38,7 @@ module RelatonBib
|
|
38
38
|
relation: fetch_relations(bibitem),
|
39
39
|
series: fetch_series(bibitem),
|
40
40
|
medium: fetch_medium(bibitem),
|
41
|
-
place: bibitem
|
41
|
+
place: fetch_place(bibitem),
|
42
42
|
extent: fetch_extent(bibitem),
|
43
43
|
accesslocation: bibitem.xpath("./accesslocation").map(&:text),
|
44
44
|
classification: fetch_classification(bibitem),
|
@@ -56,6 +56,12 @@ module RelatonBib
|
|
56
56
|
RelatonBib::BibliographicItem::Version.new revision_date, draft
|
57
57
|
end
|
58
58
|
|
59
|
+
def fetch_place(item)
|
60
|
+
item.xpath("./place").map do |pl|
|
61
|
+
Place.new(name: pl.text, uri: pl[:uri], region: pl[:region])
|
62
|
+
end
|
63
|
+
end
|
64
|
+
|
59
65
|
def fetch_note(item)
|
60
66
|
item.xpath("./note").map do |n|
|
61
67
|
BiblioNote.new(
|
@@ -113,10 +119,9 @@ module RelatonBib
|
|
113
119
|
end
|
114
120
|
|
115
121
|
def fetch_classification(item)
|
116
|
-
|
117
|
-
|
118
|
-
|
119
|
-
Classification.new type: cls[:type], value: cls.text
|
122
|
+
item.xpath("classification").map do |cls|
|
123
|
+
Classification.new type: cls[:type], value: cls.text
|
124
|
+
end
|
120
125
|
end
|
121
126
|
|
122
127
|
def fetch_validity(item)
|
data/relaton-bib.gemspec
CHANGED
@@ -23,7 +23,6 @@ Gem::Specification.new do |spec|
|
|
23
23
|
spec.require_paths = ["lib"]
|
24
24
|
spec.required_ruby_version = Gem::Requirement.new(">= 2.4.0")
|
25
25
|
|
26
|
-
spec.add_development_dependency "bundler", "~> 2.0"
|
27
26
|
spec.add_development_dependency "debase"
|
28
27
|
spec.add_development_dependency "equivalent-xml", "~> 0.6"
|
29
28
|
spec.add_development_dependency "rake", "~> 10.0"
|
@@ -34,5 +33,7 @@ Gem::Specification.new do |spec|
|
|
34
33
|
spec.add_development_dependency "byebug"
|
35
34
|
|
36
35
|
spec.add_dependency "addressable"
|
36
|
+
spec.add_dependency "bibtex-ruby"
|
37
|
+
spec.add_dependency "iso639"
|
37
38
|
spec.add_dependency "nokogiri"
|
38
39
|
end
|
metadata
CHANGED
@@ -1,29 +1,15 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: relaton-bib
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version: 0.
|
4
|
+
version: 0.5.0
|
5
5
|
platform: ruby
|
6
6
|
authors:
|
7
7
|
- Ribose Inc.
|
8
8
|
autorequire:
|
9
9
|
bindir: exe
|
10
10
|
cert_chain: []
|
11
|
-
date:
|
11
|
+
date: 2020-01-11 00:00:00.000000000 Z
|
12
12
|
dependencies:
|
13
|
-
- !ruby/object:Gem::Dependency
|
14
|
-
name: bundler
|
15
|
-
requirement: !ruby/object:Gem::Requirement
|
16
|
-
requirements:
|
17
|
-
- - "~>"
|
18
|
-
- !ruby/object:Gem::Version
|
19
|
-
version: '2.0'
|
20
|
-
type: :development
|
21
|
-
prerelease: false
|
22
|
-
version_requirements: !ruby/object:Gem::Requirement
|
23
|
-
requirements:
|
24
|
-
- - "~>"
|
25
|
-
- !ruby/object:Gem::Version
|
26
|
-
version: '2.0'
|
27
13
|
- !ruby/object:Gem::Dependency
|
28
14
|
name: debase
|
29
15
|
requirement: !ruby/object:Gem::Requirement
|
@@ -150,6 +136,34 @@ dependencies:
|
|
150
136
|
- - ">="
|
151
137
|
- !ruby/object:Gem::Version
|
152
138
|
version: '0'
|
139
|
+
- !ruby/object:Gem::Dependency
|
140
|
+
name: bibtex-ruby
|
141
|
+
requirement: !ruby/object:Gem::Requirement
|
142
|
+
requirements:
|
143
|
+
- - ">="
|
144
|
+
- !ruby/object:Gem::Version
|
145
|
+
version: '0'
|
146
|
+
type: :runtime
|
147
|
+
prerelease: false
|
148
|
+
version_requirements: !ruby/object:Gem::Requirement
|
149
|
+
requirements:
|
150
|
+
- - ">="
|
151
|
+
- !ruby/object:Gem::Version
|
152
|
+
version: '0'
|
153
|
+
- !ruby/object:Gem::Dependency
|
154
|
+
name: iso639
|
155
|
+
requirement: !ruby/object:Gem::Requirement
|
156
|
+
requirements:
|
157
|
+
- - ">="
|
158
|
+
- !ruby/object:Gem::Version
|
159
|
+
version: '0'
|
160
|
+
type: :runtime
|
161
|
+
prerelease: false
|
162
|
+
version_requirements: !ruby/object:Gem::Requirement
|
163
|
+
requirements:
|
164
|
+
- - ">="
|
165
|
+
- !ruby/object:Gem::Version
|
166
|
+
version: '0'
|
153
167
|
- !ruby/object:Gem::Dependency
|
154
168
|
name: nokogiri
|
155
169
|
requirement: !ruby/object:Gem::Requirement
|
@@ -184,6 +198,7 @@ files:
|
|
184
198
|
- README.adoc
|
185
199
|
- Rakefile
|
186
200
|
- bin/console
|
201
|
+
- bin/rspec
|
187
202
|
- bin/setup
|
188
203
|
- docs/hash.adoc
|
189
204
|
- lib/relaton_bib.rb
|
@@ -192,6 +207,7 @@ files:
|
|
192
207
|
- lib/relaton_bib/biblio_version.rb
|
193
208
|
- lib/relaton_bib/bibliographic_date.rb
|
194
209
|
- lib/relaton_bib/bibliographic_item.rb
|
210
|
+
- lib/relaton_bib/bibtex_parser.rb
|
195
211
|
- lib/relaton_bib/classification.rb
|
196
212
|
- lib/relaton_bib/contribution_info.rb
|
197
213
|
- lib/relaton_bib/contributor.rb
|
@@ -210,6 +226,7 @@ files:
|
|
210
226
|
- lib/relaton_bib/medium.rb
|
211
227
|
- lib/relaton_bib/organization.rb
|
212
228
|
- lib/relaton_bib/person.rb
|
229
|
+
- lib/relaton_bib/place.rb
|
213
230
|
- lib/relaton_bib/series.rb
|
214
231
|
- lib/relaton_bib/typed_title_string.rb
|
215
232
|
- lib/relaton_bib/typed_uri.rb
|