tao_rdfizer 0.9.3 → 0.9.4

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: aa880e0cce2aae4ec41d6e3c2de0a5cafa8b53a9
-  data.tar.gz: 91b43e74c2cdd085ed5ba9c50c269987a0ca6516
+  metadata.gz: 4614f8d18eaae8dd3861464004f1885935a2e375
+  data.tar.gz: bd2ae206163a9b74bb8c3338b75d6c6c296faa9e
 SHA512:
-  metadata.gz: c9c83a025384fcb45bc15543183a2bab00b3b2c185280c5c7d3a2e88ef0f2fef2783c21894fea5d0b7464e7dc1d2dff957324ace72d5767f82cb6db8677dbff5
-  data.tar.gz: 95a19fc8ca13d66cc154b7968f75e9256d69970d0ead660b8271a7156e548903df5217a6d97d6ab8081151d9a83e059753ba4bad430e4458e332a89141ba0e8e
+  metadata.gz: 7ba0a1fb7a42873b77e04f6d79b81f3d75f926a0a91f0d737c8694a94cf09098844bf12f30e45298a9c8283942ebffc443660e6eb1d1423547eb58c1c85ec765
+  data.tar.gz: a54949e67f7d2fad1ce73cd87f7eacdcddfef732d66273d0933b18dd1cbaa931bf81f1c086dcaa77bc50ef34f9fc8ff10d465a50192774ecea4f06b85e9f2e25

data/bin/tao_rdfizer

@@ -2,11 +2,37 @@
 require 'tao_rdfizer'
 require 'json'
 
+mode = nil
+
+## command line option processing
+require 'optparse'
+optparse = OptionParser.new do |opts|
+  opts.banner = "Usage: tao_rdfizer [options] an_annotation_file.json"
+
+  opts.on('-a', '--annotations', 'rdfize only the annotations.') do
+    mode = :annotations
+  end
+
+  opts.on('-s', '--spans', 'rdfize only the spans.') do
+    mode = :spans
+  end
+
+  opts.on('-h', '--help', 'displays this screen.') do
+    puts opts
+    exit
+  end
+end
+
+optparse.parse!
+
+unless ARGV.length > 0
+	puts optparse.help 
+	exit
+end
+
 begin
 	annotations = JSON.parse File.read(ARGV[0]), :symbolize_names => true
 	annotations = [annotations] unless annotations.class == Array
-	mode = :annotations
-	# mode = :spans
 	rdfizer = TAO::RDFizer.new(mode)
 	puts rdfizer.rdfize(annotations)
 rescue ArgumentError, IOError => e

data/lib/tao_rdfizer/tao_rdfizer.rb

@@ -73,14 +73,6 @@ class TAO::RDFizer
 			unless @mode == :annotations
 				# collect spans
 				_spans = _denotations.map{|d| d[:span]}
-
-				# # collect virtual spans
-				# position = 0
-				# annotations[:text].scan(/[^\W]*\W/).each do |tok|
-				# 	_spans << {:begin => position, :end => position + tok.index(/\W/)}
-				# 	position += tok.length
-				# end
-
 				_spans.uniq!
 
 				# add_infomation
@@ -111,30 +103,21 @@ class TAO::RDFizer
 
 				# initilaize the index
 				(0 ... num).each do |i|
-					_spans[i][:followings] = []
 					_spans[i][:precedings] = []
-					_spans[i][:children] = []
 				end
 
 				(0 ... num).each do |i|
-					# index the embedded spans
+					# 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]
-							_spans[j][:parent] = _spans[i]
-						end
-						j += 1
-					end
+					j += 1 while j < num && _spans[j][:begin] < _spans[i][:end]
 
-					# find the following position
+					# find adjacent positions
 					fp = _spans[i][:end]
 					fp += 1 while fp < len && text[fp].match(/\s/)
 					next if fp == len
 
-					# index the following spans
+					# index adjacent spans
 					while j < num && _spans[j][:begin] == fp
-						_spans[i][:followings] << _spans[j]
 						_spans[j][:precedings] << _spans[i]
 						j += 1
 					end 
@@ -195,16 +178,13 @@ class TAO::RDFizer
 	ERB_SPANS_TTL = <<~HEREDOC
 		<% spans.each do |s| -%>
 		<%= s[:span_uri] %> rdf:type tao:Text_span ;
+			tao:has_text "<%= s[:text] %>" .
 			tao:belongs_to <<%= s[:source_uri] %>> ;
 			tao:begins_at <%= s[:begin] %> ;
 			tao:ends_at <%= s[:end] %> ;
 		<% s[:precedings].each do |s| -%>
 			tao:follows <%= s[:span_uri] %> ;
 		<% end -%>
-		<% s[:children].each do |s| -%>
-			tao:contains <%= s[:span_uri] %> ;
-		<% end -%>
-			tao:has_text "<%= s[:text] %>" .
 		<% end -%>
 	HEREDOC
 

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: tao_rdfizer
 version: !ruby/object:Gem::Version
-  version: 0.9.3
+  version: 0.9.4
 platform: ruby
 authors:
 - Jin-Dong Kim
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2017-09-27 00:00:00.000000000 Z
+date: 2017-11-07 00:00:00.000000000 Z
 dependencies: []
 description: It uses TAO (text annotation ontology) for representation of annotations
   to text.