relaton-iso 0.6.3 → 0.6.4
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- checksums.yaml +4 -4
- data/Gemfile.lock +4 -4
- data/lib/relaton_iso/hit_collection.rb +2 -17
- data/lib/relaton_iso/iso_bibliography.rb +1 -1
- data/lib/relaton_iso/scrapper.rb +6 -15
- data/lib/relaton_iso/version.rb +1 -1
- metadata +2 -2
checksums.yaml
CHANGED
|
@@ -1,7 +1,7 @@
|
|
|
1
1
|
---
|
|
2
2
|
SHA1:
|
|
3
|
-
metadata.gz:
|
|
4
|
-
data.tar.gz:
|
|
3
|
+
metadata.gz: 66e86b23d321882dcc79e8f5fb47fbde454b238e
|
|
4
|
+
data.tar.gz: c9c1e37df8b4054e274ed5a77737a7e909380873
|
|
5
5
|
SHA512:
|
|
6
|
-
metadata.gz:
|
|
7
|
-
data.tar.gz:
|
|
6
|
+
metadata.gz: 6f70fc424d1fcbbd786b1389412a170f7e48d589cf030771ea6847824f8cf62b379f34ba22df82b94f7b5663981a166654eed16b08eeab69fc2cc14ee971fa15
|
|
7
|
+
data.tar.gz: e2fc938d41c9c1d8d59d81388d6b653757c66f4e2eb6158284f540a8394427bb68adc3d0fe342742a98ea11e9485a20ea70000f42566d72003d9773b365d8f47
|
data/Gemfile.lock
CHANGED
|
@@ -1,7 +1,7 @@
|
|
|
1
1
|
PATH
|
|
2
2
|
remote: .
|
|
3
3
|
specs:
|
|
4
|
-
relaton-iso (0.6.
|
|
4
|
+
relaton-iso (0.6.4)
|
|
5
5
|
relaton-iec (~> 0.4.0)
|
|
6
6
|
relaton-iso-bib (~> 0.3.0)
|
|
7
7
|
|
|
@@ -36,13 +36,13 @@ GEM
|
|
|
36
36
|
pry (~> 0.10)
|
|
37
37
|
public_suffix (3.1.1)
|
|
38
38
|
rake (10.5.0)
|
|
39
|
-
relaton-bib (0.3.
|
|
39
|
+
relaton-bib (0.3.5)
|
|
40
40
|
addressable
|
|
41
41
|
nokogiri (~> 1.10)
|
|
42
|
-
relaton-iec (0.4.
|
|
42
|
+
relaton-iec (0.4.3)
|
|
43
43
|
addressable
|
|
44
44
|
relaton-iso-bib (~> 0.3.0)
|
|
45
|
-
relaton-iso-bib (0.3.
|
|
45
|
+
relaton-iso-bib (0.3.4)
|
|
46
46
|
isoics (~> 0.1.6)
|
|
47
47
|
relaton-bib (~> 0.3.0)
|
|
48
48
|
ruby_deep_clone (~> 0.8.0)
|
|
@@ -27,10 +27,8 @@ module RelatonIso
|
|
|
27
27
|
search << "docNumber=#{num}"
|
|
28
28
|
search << "docPartNo=#{part}" if part
|
|
29
29
|
q = search.join "&"
|
|
30
|
-
resp = http.get(
|
|
31
|
-
"/
|
|
32
|
-
{ 'Accept' => 'application/json, text/plain, */*' }
|
|
33
|
-
)
|
|
30
|
+
resp = http.get("/cms/render/live/en/sites/isoorg.advancedSearch.do?#{q}",
|
|
31
|
+
"Accept" => "application/json, text/plain, */*")
|
|
34
32
|
return if resp.body.empty?
|
|
35
33
|
|
|
36
34
|
json = JSON.parse resp.body
|
|
@@ -44,19 +42,6 @@ module RelatonIso
|
|
|
44
42
|
end
|
|
45
43
|
end
|
|
46
44
|
|
|
47
|
-
# @return [RelatonIso::HitCollection]
|
|
48
|
-
# def fetch
|
|
49
|
-
# workers = RelatonBib::WorkersPool.new 4
|
|
50
|
-
# workers.worker(&:fetch)
|
|
51
|
-
# each do |hit|
|
|
52
|
-
# workers << hit
|
|
53
|
-
# end
|
|
54
|
-
# workers.end
|
|
55
|
-
# workers.result
|
|
56
|
-
# @fetched = true
|
|
57
|
-
# self
|
|
58
|
-
# end
|
|
59
|
-
|
|
60
45
|
def to_s
|
|
61
46
|
inspect
|
|
62
47
|
end
|
|
@@ -116,7 +116,7 @@ module RelatonIso
|
|
|
116
116
|
end
|
|
117
117
|
|
|
118
118
|
def try_stages(result, corr)
|
|
119
|
-
%w[NP WD CD DIS FDIS PRF IS].each do |st| # try stages
|
|
119
|
+
%w[NP WD CD DIS FDIS PRF IS AWI].each do |st| # try stages
|
|
120
120
|
warn "Attempting #{st} stage retrieval"
|
|
121
121
|
c = yield st
|
|
122
122
|
res = search_code result, c, corr
|
data/lib/relaton_iso/scrapper.rb
CHANGED
|
@@ -26,21 +26,6 @@ module RelatonIso
|
|
|
26
26
|
}.freeze
|
|
27
27
|
|
|
28
28
|
class << self
|
|
29
|
-
# @param text [String]
|
|
30
|
-
# @return [Array<Hash>]
|
|
31
|
-
# def get(text)
|
|
32
|
-
# iso_workers = RelatonBib::WorkersPool.new 4
|
|
33
|
-
# iso_workers.worker { |hit| iso_worker(hit, iso_workers) }
|
|
34
|
-
# algolia_workers = start_algolia_search(text, iso_workers)
|
|
35
|
-
# iso_docs = iso_workers.result
|
|
36
|
-
# algolia_workers.end
|
|
37
|
-
# algolia_workers.result
|
|
38
|
-
# iso_docs
|
|
39
|
-
# rescue
|
|
40
|
-
# warn "Could not connect to http://www.iso.org"
|
|
41
|
-
# []
|
|
42
|
-
# end
|
|
43
|
-
|
|
44
29
|
# Parse page.
|
|
45
30
|
# @param hit [Hash]
|
|
46
31
|
# @return [Hash]
|
|
@@ -58,6 +43,7 @@ module RelatonIso
|
|
|
58
43
|
RelatonIsoBib::IsoBibliographicItem.new(
|
|
59
44
|
fetched: Date.today.to_s,
|
|
60
45
|
docid: fetch_docid(doc),
|
|
46
|
+
docnumber: fetch_docnumber(doc),
|
|
61
47
|
edition: edition,
|
|
62
48
|
language: langs(doc).map { |l| l[:lang] },
|
|
63
49
|
script: langs(doc).map { |l| script(l[:lang]) }.uniq,
|
|
@@ -203,6 +189,11 @@ module RelatonIso
|
|
|
203
189
|
[RelatonBib::DocumentIdentifier.new(id: item_ref.text, type: "ISO")]
|
|
204
190
|
end
|
|
205
191
|
|
|
192
|
+
def fetch_docnumber(doc)
|
|
193
|
+
id = doc.at("//nav[contains(@class, 'heading-condensed')]/h1")&.text
|
|
194
|
+
id&.match(/\d+/)&.to_s
|
|
195
|
+
end
|
|
196
|
+
|
|
206
197
|
# @param doc [Nokogiri::HTML::Document]
|
|
207
198
|
def fetch_structuredidentifier(doc)
|
|
208
199
|
item_ref = doc.at("//nav[contains(@class, 'heading-condensed')]/h1")
|
data/lib/relaton_iso/version.rb
CHANGED
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: 0.6.
|
|
4
|
+
version: 0.6.4
|
|
5
5
|
platform: ruby
|
|
6
6
|
authors:
|
|
7
7
|
- Ribose Inc.
|
|
8
8
|
autorequire:
|
|
9
9
|
bindir: exe
|
|
10
10
|
cert_chain: []
|
|
11
|
-
date: 2019-08-
|
|
11
|
+
date: 2019-08-20 00:00:00.000000000 Z
|
|
12
12
|
dependencies:
|
|
13
13
|
- !ruby/object:Gem::Dependency
|
|
14
14
|
name: bundler
|