relaton-iso 1.0.3 → 1.2.1
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +4 -4
- data/.rubocop.yml +2 -2
- data/lib/relaton_iso/iso_bibliography.rb +3 -2
- data/lib/relaton_iso/scrapper.rb +30 -19
- data/lib/relaton_iso/version.rb +1 -1
- data/relaton_iso.gemspec +2 -2
- metadata +10 -10
checksums.yaml
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
---
|
2
2
|
SHA256:
|
3
|
-
metadata.gz:
|
4
|
-
data.tar.gz:
|
3
|
+
metadata.gz: 9e3b6cecb41e4d2c51cecad3b906e25fc1202ff2f9d69261d23a199c9084f0d8
|
4
|
+
data.tar.gz: 8a3f4c0a7857ef1c8668b1c1624f22573817b333a6bc674a77d729f2034098d9
|
5
5
|
SHA512:
|
6
|
-
metadata.gz:
|
7
|
-
data.tar.gz:
|
6
|
+
metadata.gz: 379e71ee50015f6ee0d012f3f157c1a355d5a69a670f3ed2a6dc7be6dafd86b4376808c8263f15c8deea3c85617ed7368dee72770c5d13e3f6e8948a1bfd9cea
|
7
|
+
data.tar.gz: d66f8ede9750f437adbc36682c4ae591ea800f887c8df54d36ea48f57ee1a09cd98e6e4b1b22dfc8653afdef2a29d7cf678782d41600110fca1e59e5074c77cc
|
data/.rubocop.yml
CHANGED
@@ -38,9 +38,10 @@ module RelatonIso
|
|
38
38
|
|
39
39
|
if year.nil?
|
40
40
|
/^(?<code1>[^\s]+(\s\w+)?\s[\d-]+)(:(?<year1>\d{4}))?(?<code2>\s\w+)?/ =~ code
|
41
|
+
/:(?<year2>\d{4})/ =~ corr
|
41
42
|
unless code1.nil?
|
42
43
|
code = code1 + code2.to_s
|
43
|
-
year = year1
|
44
|
+
year = year2 || year1
|
44
45
|
end
|
45
46
|
end
|
46
47
|
opts[:all_parts] ||= code !~ %r{^[^\s]+\s\d+-\d+} && opts[:all_parts].nil? && code2.nil?
|
@@ -145,7 +146,7 @@ module RelatonIso
|
|
145
146
|
def isobib_results_filter(result, year, opts)
|
146
147
|
missed_years = []
|
147
148
|
hits = result.reduce!([]) do |hts, h|
|
148
|
-
if !year || %r{:(?<iyear>\d{4})} =~ h.hit["docRef"] && iyear == year
|
149
|
+
if !year || %r{:(?<iyear>\d{4})(?!.*:\d{4})} =~ h.hit["docRef"] && iyear == year
|
149
150
|
hts << h
|
150
151
|
else
|
151
152
|
missed_years << iyear
|
data/lib/relaton_iso/scrapper.rb
CHANGED
@@ -33,7 +33,21 @@ module RelatonIso
|
|
33
33
|
"40" => "DIS",
|
34
34
|
"50" => "FDIS",
|
35
35
|
"60" => { "00" => "PRF", "60" => "FINAL" },
|
36
|
-
}
|
36
|
+
}.freeze
|
37
|
+
|
38
|
+
PUBLISHERS = {
|
39
|
+
"IEC" => { name: "International Electrotechnical Commission",
|
40
|
+
url: "www.iec.ch" },
|
41
|
+
"ISO" => { name: "International Organization for Standardization",
|
42
|
+
url: "www.iso.org" },
|
43
|
+
"IEEE" => { name: "Institute of Electrical and Electronics Engineers",
|
44
|
+
url: "www.ieee.org" },
|
45
|
+
"SAE" => { name: "SAE International", url: "www.sae.org" },
|
46
|
+
"CIE" => { name: " International Commission on Illumination",
|
47
|
+
url: "cie.co.at" },
|
48
|
+
"ASME" => { name: "American Society of Mechanical Engineers",
|
49
|
+
url: "www.asme.org" },
|
50
|
+
}.freeze
|
37
51
|
|
38
52
|
class << self
|
39
53
|
# Parse page.
|
@@ -90,7 +104,7 @@ module RelatonIso
|
|
90
104
|
d = l[:path] ? get_page(l[:path])[0] : doc
|
91
105
|
unless d.at("//h5[@class='help-block'][.='недоступно на русском языке']")
|
92
106
|
s << l
|
93
|
-
titles
|
107
|
+
titles += fetch_title(d, l[:lang])
|
94
108
|
|
95
109
|
# Fetch abstracts.
|
96
110
|
abstract_content = d.css("div[itemprop='description'] p").text
|
@@ -273,12 +287,14 @@ module RelatonIso
|
|
273
287
|
# Fetch titles.
|
274
288
|
# @param doc [Nokogiri::HTML::Document]
|
275
289
|
# @param lang [String]
|
276
|
-
# @return [
|
290
|
+
# @return [Array<RelatonBib::TypedTitleString>]
|
277
291
|
def fetch_title(doc, lang)
|
278
292
|
content = doc.at(
|
279
|
-
"//nav[contains(@class,'
|
280
|
-
)&.text
|
281
|
-
|
293
|
+
"//nav[contains(@class,'heading-condensed')]/h2 | //nav[contains(@class,'heading-condensed')]/h3",
|
294
|
+
)&.text&.gsub(/\u2014/, "-")
|
295
|
+
return [] unless content
|
296
|
+
|
297
|
+
RelatonBib::TypedTitleString.from_string content, lang, script(lang)
|
282
298
|
end
|
283
299
|
|
284
300
|
# Return ISO script code.
|
@@ -320,17 +336,12 @@ module RelatonIso
|
|
320
336
|
end
|
321
337
|
|
322
338
|
def fetch_contributors(ref)
|
323
|
-
ref.sub(/\s.*/, "").split("/").
|
324
|
-
|
325
|
-
|
326
|
-
|
327
|
-
|
328
|
-
|
329
|
-
name = "International Organization for Standardization"
|
330
|
-
url = "www.iso.org"
|
331
|
-
end
|
332
|
-
{ entity: { name: name, url: url, abbreviation: abbrev },
|
333
|
-
role: [type: "publisher"] }
|
339
|
+
ref.sub(/\s.*/, "").split("/").reduce([]) do |mem, abbrev|
|
340
|
+
publisher = PUBLISHERS[abbrev]
|
341
|
+
next mem unless publisher
|
342
|
+
|
343
|
+
publisher[:abbreviation] = abbrev
|
344
|
+
mem << { entity: publisher, role: [type: "publisher"] }
|
334
345
|
end
|
335
346
|
end
|
336
347
|
# rubocop:enable Metrics/MethodLength
|
@@ -364,14 +375,14 @@ module RelatonIso
|
|
364
375
|
# Fetch copyright.
|
365
376
|
# @param ref [String]
|
366
377
|
# @param doc [Nokogiri::HTML::Document]
|
367
|
-
# @return [Hash]
|
378
|
+
# @return [Array<Hash>]
|
368
379
|
def fetch_copyright(ref, doc)
|
369
380
|
owner_name = ref.match(/.*?(?=\s)/).to_s
|
370
381
|
from = ref.match(/(?<=:)\d{4}/).to_s
|
371
382
|
if from.empty?
|
372
383
|
from = doc.xpath("//span[@itemprop='releaseDate']").text.match(/\d{4}/).to_s
|
373
384
|
end
|
374
|
-
{ owner: { name: owner_name }, from: from }
|
385
|
+
[{ owner: [{ name: owner_name }], from: from }]
|
375
386
|
end
|
376
387
|
end
|
377
388
|
end
|
data/lib/relaton_iso/version.rb
CHANGED
data/relaton_iso.gemspec
CHANGED
@@ -37,6 +37,6 @@ Gem::Specification.new do |spec|
|
|
37
37
|
spec.add_development_dependency "vcr"
|
38
38
|
spec.add_development_dependency "webmock"
|
39
39
|
|
40
|
-
spec.add_dependency "relaton-iec", "
|
41
|
-
spec.add_dependency "relaton-iso-bib", "
|
40
|
+
spec.add_dependency "relaton-iec", "~> 1.2.0"
|
41
|
+
spec.add_dependency "relaton-iso-bib", "~> 1.2.0"
|
42
42
|
end
|
metadata
CHANGED
@@ -1,14 +1,14 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: relaton-iso
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version: 1.
|
4
|
+
version: 1.2.1
|
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: 2020-08-20 00:00:00.000000000 Z
|
12
12
|
dependencies:
|
13
13
|
- !ruby/object:Gem::Dependency
|
14
14
|
name: byebug
|
@@ -154,30 +154,30 @@ dependencies:
|
|
154
154
|
name: relaton-iec
|
155
155
|
requirement: !ruby/object:Gem::Requirement
|
156
156
|
requirements:
|
157
|
-
- - "
|
157
|
+
- - "~>"
|
158
158
|
- !ruby/object:Gem::Version
|
159
|
-
version: 1.0
|
159
|
+
version: 1.2.0
|
160
160
|
type: :runtime
|
161
161
|
prerelease: false
|
162
162
|
version_requirements: !ruby/object:Gem::Requirement
|
163
163
|
requirements:
|
164
|
-
- - "
|
164
|
+
- - "~>"
|
165
165
|
- !ruby/object:Gem::Version
|
166
|
-
version: 1.0
|
166
|
+
version: 1.2.0
|
167
167
|
- !ruby/object:Gem::Dependency
|
168
168
|
name: relaton-iso-bib
|
169
169
|
requirement: !ruby/object:Gem::Requirement
|
170
170
|
requirements:
|
171
|
-
- - "
|
171
|
+
- - "~>"
|
172
172
|
- !ruby/object:Gem::Version
|
173
|
-
version: 1.0
|
173
|
+
version: 1.2.0
|
174
174
|
type: :runtime
|
175
175
|
prerelease: false
|
176
176
|
version_requirements: !ruby/object:Gem::Requirement
|
177
177
|
requirements:
|
178
|
-
- - "
|
178
|
+
- - "~>"
|
179
179
|
- !ruby/object:Gem::Version
|
180
|
-
version: 1.0
|
180
|
+
version: 1.2.0
|
181
181
|
description: 'RelatonIso: retrieve ISO Standards for bibliographic use using the IsoBibliographicItem
|
182
182
|
model'
|
183
183
|
email:
|