linkheaders-processor 0.1.14 → 0.1.16

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 4df2bc9ea84550e0de7be8110f49121ea84962ac797b01860858a445213787d5
-  data.tar.gz: ea148084a6ee71bb8e5509ae94c1dde4714dc66e609527b8c186362cc71c7c76
+  metadata.gz: ef79d8492d08ea54cc8c23067d4f3c443dee0941f5d901007928d62df93d8073
+  data.tar.gz: ec81315a8baedf8ea2b0ffdc1c882d44647ca45b0a2a5a1f71734c47ce3d9f64
 SHA512:
-  metadata.gz: 9b66ae4ca8d0b48f71d4d33f2bccc1575abc5efe65ebe7bc756cab9fb47836d06fe9dc011ae0e55b0a3be6db331e3018e62d006bb8140b181538b65dc66cbab1
-  data.tar.gz: 8c2ae044bd1a111f915ea419932b431227b344386ad1824d4e1fc6f86269c2c513d72918d693addf0fa6e0ae0a3fcc2cabb9fa9c27542b3f5bed4d1d5edbf3d9
+  metadata.gz: 0b510601d2b51739b893b07c546d1ed723dd2bef296c3af59b7fdc3dbfdcfdc08f8e20741c6f2f9ca0957675c02c9ddfb6d5a21201976f256b1b2a45ee7dde2b
+  data.tar.gz: d20ed0c9b8b04607b104aad8ce5a07c26cc12ae16b925c9ccd3e9feca78f691592976a352ac301a907d517eb55b8972960b7cec4fc3913316c9453f1d2549098

data/Gemfile.lock

@@ -1,7 +1,7 @@
 PATH
   remote: .
   specs:
-    linkheaders-processor (0.1.14)
+    linkheaders-processor (0.1.16)
       json (~> 2.0)
       json-ld (~> 3.2)
       json-ld-preloaded (~> 3.2)
@@ -118,7 +118,7 @@ GEM
       diff-lcs (>= 1.2.0, < 2.0)
       rspec-support (~> 3.11.0)
     rspec-support (3.11.0)
-    rubocop (1.32.0)
+    rubocop (1.33.0)
       json (~> 2.3)
       parallel (~> 1.10)
       parser (>= 3.1.0.0)

data/lib/linkheaders/processor/version.rb

@@ -3,6 +3,6 @@
 
 module LinkHeaders
   class Processor
-    VERSION = "0.1.14"
+    VERSION = "0.1.16"
   end
 end

data/lib/linkheaders/processor.rb

@@ -1,7 +1,6 @@
 # frozen_string_literal: true
 
 require_relative 'processor/version'
-require_relative 'constants'
 require_relative 'link'
 require_relative 'web_utils'
 require 'link_header'
@@ -63,7 +62,7 @@ module LinkHeaders
       newlinks = parse_http_link_headers(head) # pass guid to check against anchors in linksets
       warn "HTTPlinks #{newlinks.inspect}"
 
-      HTML_FORMATS['html'].each do |format|
+      ['text/html','text/xhtml+xml', 'application/xhtml+xml'].each do |format|
         if head[:content_type] and head[:content_type].match(format)
           warn "found #{format} content - parsing"
           htmllinks = parse_html_link_headers(body) # pass html body to find HTML link headers

data/lib/linkheaders/web_utils.rb

@@ -1,4 +1,4 @@
-def lhfetch(url, headers = ACCEPT_STAR_HEADER)
+def lhfetch(url, headers = {accept: "*/*"})
   # warn "In fetch routine now.  "
 
   # warn "executing call over the Web to #{url.to_s}"

data/spec/linkheader/parser_spec.rb

@@ -1,7 +1,37 @@
 # frozen_string_literal: true
+require_relative '../../lib/linkheaders/processor'
+require 'rest-client'
+
+
+url1 = "https://w3id.org/a2a-fair-metrics/22-http-html-citeas-describedby-mixed/"
+p = LinkHeaders::Processor.new(default_anchor: url1)
+r = RestClient.get(url1)
+p.extract_and_parse(response: r)
+factory = p.factory  # LinkHeaders::LinkFactory
+
 
 RSpec.describe LinkHeaders::Processor do
+
   it 'has a version number' do
     expect(LinkHeaders::Processor::VERSION).not_to be nil
   end
+
+  it "should find PURL citeas which has described-by and cite-as in mixed HTTP and HTML headers" do
+    expect(factory.all_links.length).to eq 5
+  end
+  it "should find find href on all links" do
+    expect(factory.all_links.select{|l| l.href}.length).to eq 5
+  end
+  it "should find find href on all links" do
+    expect(factory.all_links.select{|l| l.anchor}.length).to eq 5
+  end
+  it "should find 5 links in mixed HTTP and HTML headers" do
+    expect(factory.all_links.select{|l| l.relation}.length).to eq 5
+  end
+  it "should find one citeas in mixed HTTP and HTML headers" do
+    expect(factory.all_links.select{|l| l.relation == 'cite-as'}.length).to eq 1
+  end
+  it "should find described-by in mixed HTTP and HTML headers" do
+    expect(factory.all_links.select{|l| l.relation == 'describedby'}.length).to eq 1
+  end
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: linkheaders-processor
 version: !ruby/object:Gem::Version
-  version: 0.1.14
+  version: 0.1.16
 platform: ruby
 authors:
 - Mark Wilkinson
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2022-08-04 00:00:00.000000000 Z
+date: 2022-08-05 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rspec
@@ -137,7 +137,6 @@ files:
 - README.md
 - Rakefile
 - launch.json
-- lib/linkheaders/constants.rb
 - lib/linkheaders/link.rb
 - lib/linkheaders/processor.rb
 - lib/linkheaders/processor/version.rb

data/lib/linkheaders/constants.rb

@@ -1,29 +0,0 @@
-ACCEPT_ALL_HEADER = {'Accept' => 'text/turtle, application/ld+json, application/rdf+xml, text/xhtml+xml, application/n3, application/rdf+n3, application/turtle, application/x-turtle, text/n3, text/turtle, text/rdf+n3, text/rdf+turtle, application/n-triples' }
-ACCEPT_STAR_HEADER = {'Accept' => '*/*'}
-
-TEXT_FORMATS = {
-    'text' => ['text/plain',],
-}
-
-RDF_FORMATS = {
-  'jsonld'  => ['application/ld+json', 'application/vnd.schemaorg.ld+json'],  # NEW FOR DATACITE
-  'turtle'  => ['text/turtle','application/n3','application/rdf+n3',
-               'application/turtle', 'application/x-turtle','text/n3','text/turtle',
-               'text/rdf+n3', 'text/rdf+turtle'],
-  #'rdfa'    => ['text/xhtml+xml', 'application/xhtml+xml'],
-  'rdfxml'  => ['application/rdf+xml'],
-  'triples' => ['application/n-triples','application/n-quads', 'application/trig']
-}
-
-XML_FORMATS = {
-  'xml' => ['text/xhtml','text/xml',]
-}
-
-HTML_FORMATS = {
-  'html' => ['text/html','text/xhtml+xml', 'application/xhtml+xml']
-}
-
-JSON_FORMATS = {
-            'json' => ['application/json',]
-}
-