tao_rdfizer 0.11.3 → 0.11.4

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 122108e353debbc7d153a9aaa4f4a7b1d0d964ab1f24697443d2e65977e61efa
-  data.tar.gz: 2a62939c0ac1bf15e1225a1512d3d797c379ddb563bfeb422996c752b1865f61
+  metadata.gz: 59d7999a921ef33104728a1f86edbeb7f20161641b37a53435daf316772ffed8
+  data.tar.gz: d6cd66e38eececa7b7cbb251686d148a4924da16ad751eea3af2cef21b00121a
 SHA512:
-  metadata.gz: 6b25ad6199e58d03b7e8acdf4c60a8a8c9c87b634940edddebb73b696a5e4bb37b96b7d74cb818acb93b951ea5977ae72ec8273c3ec264d3c73d6e381efe1092
-  data.tar.gz: 5085e04e204ef0eb8f1282e0f659a2cc60eae7ef70283c658072b310683f016cf2c1806ff71cad66311950b246eb99a402e502d799f998acd34434ea9d3bc4ff
+  metadata.gz: 4e98ac931a0f2913f44e8e234222f55c93ff9bbd76173787393fcf1e866a14b7707168c8446f33cfd00f208599439cacdd67b4b6296c9ab8a1288a3963fd4f0c
+  data.tar.gz: 788aaf5c427d44783edadfc30f84a8d2cc7aac99641575dd3999f780520c374cf5c443fc9a5dd656424a4291940861fa268a37b6ca4f3dab9a2ae196f6df2bc8

data/lib/tao_rdfizer/tao_rdfizer.rb

@@ -14,14 +14,14 @@ class TAO::RDFizer
 		else
 			ERB_ANNOTATIONS_TTL
 		end
-		@tao_ttl_erb = ERB.new(template, nil, '-')
-		@prefix_ttl_erb = ERB.new(ERB_PREFIXES_TTL, nil, '-')
+		@tao_ttl_erb = ERB.new(template, trim_mode: '-')
+		@prefix_ttl_erb = ERB.new(ERB_PREFIXES_TTL, trim_mode: '-')
 	end
 
 	def rdfize(annotations_col, options = nil)
 		options ||= {}
-		only_prefixes = options.has_key?(:only_prefixes) ? options[:only_prefixes] == true : false
-		with_prefixes = options.has_key?(:with_prefixes) ? options[:with_prefixes] == true : true
+		only_prefixes = options[:only_prefixes] == true
+		with_prefixes = options[:with_prefixes] != true
 
 		# check the format
 		annotations_col.each do |annotations|
@@ -29,35 +29,33 @@ class TAO::RDFizer
 		end
 
 		# namespaces
-		namespaces = {}
-
-		anns = annotations_col.first
-		anns[:namespaces].each {|n| namespaces[n[:prefix]] = n[:uri]} unless anns[:namespaces].nil?
+		namespaces = get_namespaces(annotations_col.first)
 
 		prefixes_ttl = @prefix_ttl_erb.result_with_hash(namespaces:namespaces) if only_prefixes || with_prefixes
 
-		if only_prefixes
-			prefixes_ttl
-		else
-			annotations_ttl = get_annotations_ttl(annotations_col, namespaces)
-			if with_prefixes
-				prefixes_ttl + annotations_ttl
-			else
-				annotations_ttl
-			end
-		end
+		return prefixes_ttl if only_prefixes
+
+		annotations_ttl = get_annotations_ttl(annotations_col, namespaces)
+		with_prefixes ? prefixes_ttl + annotations_ttl : annotations_ttl
 	end
 
 	private
 
-	def get_annotations_ttl(annotations_col, namespaces)
-		anns = annotations_col.first
+	def get_namespaces(annotations)
+		return {} if annotations[:namespaces].nil?
+
+		annotations[:namespaces].each_with_object({}) do |n, namespaces|
+			namespaces[n[:prefix]] = n[:uri]
+		end
+	end
 
+	def get_annotations_ttl(annotations_col, namespaces)
 		unless @mode ==:spans
+			anns = annotations_col.first
 			raise ArgumentError, "A project name has to be specified." unless anns.has_key?(:project)
 			prefix_for_this = anns[:project].downcase.gsub(/ /, '_')
 			raise ArgumentError, "'#{prefix_for_this}' is a reserved prefix for this project." if namespaces.has_key?(prefix_for_this)
-			project_uri = 'http://pubannotation.org/projects/' + anns[:project]
+			project_uri = 'https://pubannotation.org/projects/' + anns[:project]
 			namespaces[prefix_for_this] = project_uri + '/'
 		end
 
@@ -70,8 +68,8 @@ class TAO::RDFizer
 			text = annotations[:text]
 			text_uri = annotations[:target]
 			text_id = begin
-				sourcedb, sourceid, divid = get_target_info(text_uri)
-				divid.nil? ? "#{sourcedb}-#{sourceid}" : "#{sourcedb}-#{sourceid}-#{divid}"
+				sourcedb, sourceid = get_target_info(text_uri)
+				"#{sourcedb}-#{sourceid}"
 			end
 
 			# denotations and relations
@@ -153,37 +151,29 @@ class TAO::RDFizer
 					end
 				end
 
-				_spans.sort!{|a, b| (a[:begin] <=> b[:begin]).nonzero? || b[:end] <=> a[:end]}
+				_spans.sort_by! { |s| [s[:begin], -s[:end]] }
 
 				## begin indexing
 				len = text.length
 				num = _spans.length
 
-				# initilaize the index
-				(0 ... num).each do |i|
-					_spans[i][:followings] = []
-					_spans[i][:children] = []
-				end
+				# create the indexes for followings and children
+				_spans.each_with_index do |span, i|
+					span[:followings] = []
+					span[:children] = []
 
-				(0 ... num).each do |i|
-					# find the first following span
 					j = i + 1
-
-					while j < num && _spans[j][:begin] < _spans[i][:end]
-						unless include_parent?(_spans[i][:children], _spans[j])
-							_spans[i][:children] << _spans[j]
-						end
+					while j < num && _spans[j][:begin] < span[:end]
+						span[:children] << _spans[j] unless include_parent?(span[:children], _spans[j])
 						j += 1
 					end
 
-					# find adjacent positions
-					fp = _spans[i][:end]
+					fp = span[:end]
 					fp += 1 while fp < len && text[fp].match(/\s/)
 					next if fp == len
 
-					# index adjacent spans
 					while j < num && _spans[j][:begin] == fp
-						_spans[i][:followings] << _spans[j]
+						span[:followings] << _spans[j]
 						j += 1
 					end
 				end
@@ -206,28 +196,26 @@ class TAO::RDFizer
 	def get_target_info (text_uri)
 		sourcedb = (text_uri =~ %r|/sourcedb/([^/]+)|)? $1 : nil
 		sourceid = (text_uri =~ %r|/sourceid/([^/]+)|)? $1 : nil
-		divid    = (text_uri =~ %r|/divs/([^/]+)|)? $1 : nil
 
-		return sourcedb, sourceid, divid
+		return sourcedb, sourceid
 	end
 
-	def find_uri (label, namespaces, prefix_for_this)
+	def find_uri(label, namespaces, prefix_for_this)
+		@uri_cache ||= {}
+
+		return @uri_cache[label] if @uri_cache.key?(label)
+
 		if label.match(/\s/)
-			# raise ArgumentError, "A label including a whitespace character found: #{label}."
-			label.gsub(/\s/, '_')
+			label.gsub!(/\s/, '_')
 		end
 		delimiter_position = label.index(':')
-		if !delimiter_position.nil? && namespaces.keys.include?(label[0...delimiter_position])
-			label.gsub('(', '\(').gsub(')', '\)') # brackets have to be escaped
+		if delimiter_position && namespaces.key?(label[0...delimiter_position])
+			@uri_cache[label] = label.gsub('(', '\(').gsub(')', '\)')
 		elsif label =~ %r[^https?://]
-			"<#{label}>"
+			@uri_cache[label] = "<#{label}>"
 		else
-			clabel = if label.match(/^\W+$/)
-				'SYM'
-			else
-				label.sub(/^\W+/, '').sub(/[^a-zA-Z0-9_)]+$/, '').gsub(/ +/, '_')
-			end
-			namespaces.has_key?('_base') ? "<#{clabel}>" : "#{prefix_for_this}:#{clabel.gsub('(', '\(').gsub(')', '\)')}"
+			clabel = label.match(/^\W+$/) ? 'SYM' : label.sub(/^\W+/, '').sub(/[^a-zA-Z0-9_)]+$/, '').gsub(/ +/, '_')
+			@uri_cache[label] = namespaces.key?('_base') ? "<#{clabel}>" : "#{prefix_for_this}:#{clabel.gsub('(', '\(').gsub(')', '\)')}"
 		end
 	end
 
@@ -274,7 +262,7 @@ class TAO::RDFizer
 	# variable: namespaces
 	ERB_PREFIXES_TTL = <<~HEREDOC
 		@prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#> .
-		@prefix tao: <http://pubannotation.org/ontology/tao.owl#> .
+		@prefix tao: <https://pubannotation.org/ontology/tao.owl#> .
 		<%# namespaces -%>
 		<% namespaces.each_key do |p| -%>
 		<% if p == '_base' -%>

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: tao_rdfizer
 version: !ruby/object:Gem::Version
-  version: 0.11.3
+  version: 0.11.4
 platform: ruby
 authors:
 - Jin-Dong Kim
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2021-05-23 00:00:00.000000000 Z
+date: 2024-09-18 00:00:00.000000000 Z
 dependencies: []
 description: It uses TAO (text annotation ontology) for representation of annotations
   to text.
@@ -33,14 +33,14 @@ required_ruby_version: !ruby/object:Gem::Requirement
   requirements:
   - - ">="
     - !ruby/object:Gem::Version
-      version: '0'
+      version: 3.0.0
 required_rubygems_version: !ruby/object:Gem::Requirement
   requirements:
   - - ">="
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.0.9
+rubygems_version: 3.5.11
 signing_key: 
 specification_version: 4
 summary: A RDF statement generator for annotations in the PubAnnotation JSON format.