ttl2html 0.0.6 → 0.1.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 02d0284dc6455fa2b571b8c6ff8ef730889327ca819dd5f05ee39514867fcd7c
-  data.tar.gz: '038ab96245d4ff3a5688716d41416bb9ca1f7c65b544283990faa0d8009799fe'
+  metadata.gz: 3285a319273819d00281486dadbdea6c82c162856157075db553e18b659b16cc
+  data.tar.gz: 2f66f22a4ad30bd4dc6e0e9a44f24be37cf2e5c11396f4a2c85414becb1dbe6a
 SHA512:
-  metadata.gz: d8591544285b417add4522e69deb85618c8e883058fad0d4080b4306cd7948d55d57b3c626dc47531952c448a874db3773184f6c7fd1ab0122d69f7813b76bd2
-  data.tar.gz: c1f4a606959b5214584b3fdaef8b7c429a7b2a6429f3a5954d111203bf5e59cdcdbeebc1a11739ea95cd40d2896d5561d901f4545e4ccc0932c475bae6f9888f
+  metadata.gz: 4301d623df7599b6344806d66c97c71efb0633f45f663be45a460d5ceb23c1fe452d4ed2b01b2df4f02109d4168b8a31acb2dccb21a7f4e4234926cc59e4f4f5
+  data.tar.gz: 930b188d2f0d89601ba6af9199fd0ce87d40b3a00bd8a6711c14a82cb7dc52dae90d871772cd4e39dd4680ef8894e4c9245651e2a158c33f690f887a267f597a

data/bin/xlsx2shape

@@ -0,0 +1,11 @@
+#!/usr/bin/env ruby
+
+require "xlsx2shape"
+
+if ARGV.size == 0
+  puts "USAGE: xlsx2shape file.xlsx"
+  exit
+end
+
+turtle = XLSX2Shape.xlsx2shape(ARGV[0])
+puts turtle

data/lib/ttl2html/version.rb

@@ -1 +1 @@
-TTL2HTML::VERSION = "0.0.6
+TTL2HTML::VERSION = "0.1.0"

data/lib/xlsx2shape.rb

@@ -0,0 +1,122 @@
+#!/usr/bin/env ruby
+
+require "roo"
+
+module XLSX2Shape
+  def xlsx2shape(filename)
+    shapes = {}
+    prefix = {}
+    xlsx = Roo::Excelx.new(filename)
+    xlsx.each_with_pagename do |name, sheet|
+      if name =~ /\Aprefix\z/i
+        sheet.each do |row|
+          prefix[row[0]] = row[1] if not row[1].empty?
+        end
+      else
+        headers = sheet.row(1)
+        uri = headers.first
+        shapes[uri] = ["<#{uri}> a sh:NodeShape"]
+        order = 1
+        sheet.each_with_index do |row, idx|
+          row_h = map_xlsx_row_headers(row, headers)
+          case row.first
+          when "sh:targetClass"
+            shapes[uri] << "#{format_property("sh:targetClass", row[1])}" if row[1]
+          when "sh:property"
+            prop_values = []
+            headers[1..-1].each do |prop|
+              next if row_h[prop].empty?
+              case prop
+              when /\@(\w+)\z/
+                lang = $1
+                property_name = prop.sub(/\@(\w+)\z/, "")
+                prop_values << format_property(property_name, row_h[prop], lang)
+              when "sh:minCount", "sh:maxCount"
+                prop_values << format_property(prop, row_h[prop].to_i)
+              when "sh:languageIn"
+                prop_values << "  sh:languageIn (#{row_h[prop].split.map{|e| format_pvalue(e) }.join(" ")})"
+              when "sh:uniqueLang"
+                case row_h[prop]
+                when "true"
+                  prop_values << "  sh:uniqueLang true"
+                when "false"
+                  prop_values << "  sh:uniqueLang false"
+                else
+                  logger.warn "sh:uniqueLang value unknown: #{row_h[prop]} at #{uri}"
+                end
+              else
+                prop_values << format_property(prop, row_h[prop])
+              end
+            end
+            prop_values << format_property("sh:order", order)
+            order += 1
+            str = prop_values.join(";\n  ")
+            shapes[uri] << "  sh:property [\n  #{str}\n  ]"
+          when "sh:or"
+            shapes[uri] << "  sh:or (#{row[1..-1].select{|e| not e.empty? }.map{|e| format_pvalue(e) }.join(" ")})"
+          end
+        end
+      end
+    end
+    result = ""
+    prefix.sort.each do |prefix, val|
+      result << "@prefix #{prefix} <#{val}>.\n"
+    end
+    shapes.sort_by{|uri, val| uri }.each do |uri, val|
+      result << "\n"
+      result << shapes[uri].join(";\n")
+      result << ".\n"
+    end
+    result
+  end
+
+  def map_xlsx_row_headers(data_row, headers)
+    hash = {}
+    headers.each_with_index do |h, idx|
+      hash[h] = data_row[idx].to_s
+    end
+    hash
+  end
+  def format_pvalue(value, lang = nil)
+    str = ""
+    if value.is_a? Hash
+      result = ["["]
+      array = []
+      value.keys.sort.each do |k|
+        array << format_property(k, value[k])
+      end
+      result << array.join(";\n")
+      result << "  ]"
+      str = result.join("\n")
+    elsif value.is_a? Integer
+      str = value
+    elsif value =~ /\Ahttps?:\/\//
+      str = %Q|<#{value}>|
+    elsif value =~ /\A\w+:[\w\-\.]+\Z/
+      str = value
+    elsif value =~ /\A(.+?)\^\^(\w+:\w+)\z/
+      str = %Q|"#{escape_turtle($1)}"^^#{$2}|
+    elsif lang
+      str = %Q|"#{escape_turtle(value)}"@#{lang}|
+    else
+      str = %Q|"#{escape_turtle(value)}"|
+    end
+    str
+  end
+  def format_property(property, value, lang = nil)
+    if value.is_a? Array
+      value = value.sort_by{|e|
+        format_pvalue(e)
+      }.map do |e|
+        format_pvalue(e)
+      end
+      %Q|  #{property} #{ value.join(", ") }|
+    else
+      value = format_pvalue(value, lang)
+      %Q|  #{property} #{value}|
+    end
+  end
+  def escape_turtle(str)
+    str.gsub(/\\/){ '\\\\' }.gsub(/"/){ '\"' }
+  end
+end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: ttl2html
 version: !ruby/object:Gem::Version
-  version: 0.0.6
+  version: 0.1.0
 platform: ruby
 authors:
 - Masao Takaku
@@ -38,6 +38,20 @@ dependencies:
     - - ">="
       - !ruby/object:Gem::Version
         version: '0'
+- !ruby/object:Gem::Dependency
+  name: roo
+  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: ruby-progressbar
   requirement: !ruby/object:Gem::Requirement
@@ -98,13 +112,16 @@ description: Static site generator for RDF/Turtle
 email: tmasao@acm.org
 executables:
 - ttl2html
+- xlsx2shape
 extensions: []
 extra_rdoc_files: []
 files:
 - bin/ttl2html
+- bin/xlsx2shape
 - lib/ttl2html.rb
 - lib/ttl2html/template.rb
 - lib/ttl2html/version.rb
+- lib/xlsx2shape.rb
 - templates/default.html.erb
 - templates/index.html.erb
 - templates/layout.html.erb