relaton-bib 1.11.4 → 1.11.7

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 66d703af6f525fba6b4cd25e054ab2f5f64e0d5040b75430c04296990a6a12ca
-  data.tar.gz: c0c85e99f1e3b511fbc329b9b631349ed63afee1a1779a2147a1756e6dbf0559
+  metadata.gz: 9dffbd3f78b2468d48d601dd2cb1fa883c94d16d0b6c16a213219f85bbeef4c3
+  data.tar.gz: b04d3b2db08c74e0ba9599c36a35ebd00dc62501ce0847bbce046cbcc9892fa8
 SHA512:
-  metadata.gz: d8f4ef1047ad4d3ab345c3d67079ca66e986b707871f243c11eb8c52ba5792202bf046bbcb1b3722d957d3e882e3ce645b30d9f5a422a8f8ccd2ac38eb7a7d12
-  data.tar.gz: 2974513d8a823bc2efbbfe51e4bc9a88c7e7f77b3aa117e2835b5e13d6b33f366b0c15e7987afcb2dc1b5f5aa9f63d5a45e0df384ffb321132df30d2ed6ebe09
+  metadata.gz: adf68e1aa4a1c7f8fc719cf51bfc1b74d0e8ae49374647d70140021d5250908c52b0728219cb5eda9aed442eaf3f12674648ac1e61af744ffce345d0e81c7b91
+  data.tar.gz: 29690620969a335e11ab867163f33b530072cea8ceaab4d2980c60eadf14c9eaddb2976c02826cdf13ad677dd2ef781539419a014929481bce0e90274cc6dad7

data/.github/workflows/rake.yml

@@ -10,23 +10,4 @@ on:
 
 jobs:
   rake:
-    name: Test on Ruby ${{ matrix.ruby }} ${{ matrix.os }}
-    runs-on: ${{ matrix.os }}
-    continue-on-error: ${{ matrix.experimental }}
-    strategy:
-      fail-fast: false
-      matrix:
-        ruby: [ '3.0', '2.7', '2.6' ]
-        os: [ ubuntu-latest, windows-latest, macos-latest ]
-        experimental: [ false ]
-    steps:
-      - uses: actions/checkout@v2
-        with:
-          submodules: true
-
-      - uses: ruby/setup-ruby@v1
-        with:
-          ruby-version: ${{ matrix.ruby }}
-          bundler-cache: true
-
-      - run: bundle exec rake
+    uses: relaton/support/.github/workflows/rake.yml@master

data/.rubocop.yml

@@ -7,6 +7,6 @@ require: rubocop-rails
 inherit_from:
   - https://raw.githubusercontent.com/riboseinc/oss-guides/master/ci/rubocop.yml
 AllCops:
-  TargetRubyVersion: 2.5
+  TargetRubyVersion: 2.6
 Rails:
   Enabled: false

data/lib/relaton_bib/bibliographic_item.rb

@@ -37,7 +37,9 @@ module RelatonBib
     TYPES = %W[article book booklet conference manual proceedings presentation
                thesis techreport standard unpublished map electronic\sresource
                audiovisual film video broadcast graphic_work music patent
-               inbook incollection inproceedings journal].freeze
+               inbook incollection inproceedings journal software website
+               webresource dataset archival social_media alert message
+               conversation misc].freeze
 
     # @return [TrueClass, FalseClass, NilClass]
     attr_accessor :all_parts

data/lib/relaton_bib/bibxml_parser.rb

@@ -7,7 +7,7 @@ module RelatonBib
     FLAVOR = nil
 
     ORGNAMES = {
-      "IEEE" => "Istitute of Electrical and Electronics Engineers",
+      "IEEE" => "Institute of Electrical and Electronics Engineers",
       "W3C" => "World Wide Web Consortium",
       "3GPP" => "3rd Generation Partnership Project",
     }.freeze
@@ -194,7 +194,7 @@ module RelatonBib
     # @return [Array<RelatonBib::FormattedString>]
     def abstracts(ref)
       ref.xpath("./front/abstract").map do |a|
-        c = a.children.to_s.gsub(/\s*(<\/?)t(>)\s*/, '\1p\2')
+        c = a.inner_html.gsub(/\s*(<\/?)t(>)\s*/, '\1p\2')
           .gsub(/[\t\n]/, " ").squeeze " "
         FormattedString.new(content: c, language: language(ref), script: "Latn",
                             format: "text/html")

data/lib/relaton_bib/localized_string.rb

@@ -56,12 +56,65 @@ module RelatonBib
       else
         builder.parent["language"] = language.join(",") if language&.any?
         builder.parent["script"]   = script.join(",") if script&.any?
-        builder.parent << escaped_content # .encode(xml: :text)
+        builder.parent.inner_html = encode content
       end
     end
 
-    def escaped_content
-      content.encode("UTF-8") # .gsub(/&/, "&amp;").gsub(/"/, "&quot;").gsub(/'/, "&apos;")
+    #
+    # Encode content.
+    #
+    # @param [String] cnt content
+    #
+    # @return [String] encoded content
+    #
+    def encode(cnt) # rubocop:disable Metrics/MethodLength
+      regex = /(?<prf>.*?)(?<xml><(?<tag>\w+)>.*<\/\k<tag>>)(?<sfx>.*)/m
+      if cnt.match(regex)
+        prf = Regexp.last_match(:prf).lstrip
+        xml = Regexp.last_match[:xml]
+        sfx = Regexp.last_match(:sfx).rstrip
+        parts = xml.scan(/\s*<(?<tago>\w+)>(?<cnt1>.*?)(?=<\/?\w+>)|(?<cnt2>.*?)<\/(?<tagc>\w+)>/)
+        out = scan_xml parts
+        "#{escp(prf)}#{out}#{escp(sfx)}"
+      else
+        escp cnt
+      end
+    end
+
+    #
+    # Scan XML and escape HTML entities.
+    #
+    # @param [Array<Array<String,nik>>] parts XML parts
+    #
+    # @return [String] output string
+    #
+    def scan_xml(parts) # rubocop:disable Metrics/AbcSize,Metrics/CyclomaticComplexity,Metrics/PerceivedComplexity,Metrics/MethodLength
+      return "" unless parts.any? && parts.first[0]
+
+      tago, cnt1, = parts.shift
+      if tago && tago == parts.first[3]
+        _, _, cnt2, tagc = parts.shift
+        "<#{tago}>#{escp(cnt1)}#{escp(cnt2)}</#{tagc}>"
+      elsif tago
+        inr = scan_xml parts
+        _, _, cnt2, tagc = parts.shift
+        if tago == tagc
+          "<#{tago}>#{escp(cnt1)}#{inr}#{escp(cnt2)}</#{tagc}>"
+        else
+          "#{escp("<#{tago}>#{cnt1}")}#{inr}#{escp("#{cnt2}</#{tagc}>")}"
+        end
+      end + scan_xml(parts)
+    end
+
+    #
+    # Escope HTML entities.
+    #
+    # @param [String] str input string
+    #
+    # @return [String] output string
+    #
+    def escp(str)
+      HTMLEntities.new.encode str
     end
 
     # @return [Hash]

data/lib/relaton_bib/version.rb

@@ -1,3 +1,3 @@
 module RelatonBib
-  VERSION = "1.11.4".freeze
+  VERSION = "1.11.7".freeze
 end

data/lib/relaton_bib/xml_parser.rb

@@ -354,7 +354,7 @@ module RelatonBib
       # @return [Array<RelatonBib::FormattedString>]
       def fetch_abstract(item)
         item.xpath("./abstract").map do |a|
-          c = a.children.to_s
+          c = a.inner_html(encoding: "utf-8").strip
           FormattedString.new(content: c, language: a[:language],
                               script: a[:script], format: a[:format])
         end

data/lib/relaton_bib.rb

@@ -1,5 +1,6 @@
 require "forwardable"
 require "yaml"
+require "htmlentities"
 require "relaton_bib/version"
 require "relaton_bib/deep_dup"
 require "relaton_bib/localized_string"

data/relaton-bib.gemspec

@@ -35,6 +35,7 @@ Gem::Specification.new do |spec|
 
   spec.add_dependency "addressable"
   spec.add_dependency "bibtex-ruby"
+  spec.add_dependency "htmlentities"
   spec.add_dependency "iso639"
   spec.add_dependency "nokogiri", "~> 1.13.0"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: relaton-bib
 version: !ruby/object:Gem::Version
-  version: 1.11.4
+  version: 1.11.7
 platform: ruby
 authors:
 - Ribose Inc.
 autorequire:
 bindir: exe
 cert_chain: []
-date: 2022-05-10 00:00:00.000000000 Z
+date: 2022-06-01 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: byebug
@@ -164,6 +164,20 @@ dependencies:
     - - ">="
       - !ruby/object:Gem::Version
         version: '0'
+- !ruby/object:Gem::Dependency
+  name: htmlentities
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
 - !ruby/object:Gem::Dependency
   name: iso639
   requirement: !ruby/object:Gem::Requirement