oddb2xml 1.0.5 → 1.0.6

data/Manifest.txt

@@ -1,6 +1,8 @@
 .gitignore
+.rspec
 Gemfile
 History.txt
+LICENSE
 Manifest.txt
 README.md
 Rakefile
@@ -13,3 +15,14 @@ lib/oddb2xml/downloader.rb
 lib/oddb2xml/extractor.rb
 lib/oddb2xml/version.rb
 oddb2xml.gemspec
+spec/builder_spec.rb
+spec/cli_spec.rb
+spec/compressor_spec.rb
+spec/data/XMLPublications.zip
+spec/data/oddb_article.xml
+spec/data/oddb_product.xml
+spec/data/swissindex.xml
+spec/data/wsdl.xml
+spec/downloader_spec.rb
+spec/extractor_spec.rb
+spec/spec_helper.rb

data/bin/oddb2xml

@@ -5,26 +5,44 @@ require 'pathname'
 root = Pathname.new(__FILE__).realpath.parent.parent
 $:.unshift root.join('lib') if $0 == __FILE__
 
+require 'optparse'
 require 'oddb2xml'
 
-# TODO: Parse option
-args = ARGV.dup
+def help
+  <<EOS
+#$0 ver.#{Oddb2xml::VERSION}
+Usage:
+  oddb2xml [option]
+    -c F, --compress=F  Compress format F. Currently only 'tar.gz' is available.
+    -a T, --append=T    Additional target. T is only 'nonpharma' available.
+    -h,   --help        Show this help message.
+EOS
+end
+
+parser = OptionParser.new
 opts = {
-  :compress => nil
+  :compress  => nil,
+  :nonpharma => false
 }
-if args == ['-c', 'tar.gz']
-  opts[:compress] = 'tar.gz'
-  args = {}
+
+parser.on('-c v', '--compress v', /tar\.gz/) {|v| opts[:compress] = v }
+parser.on('-a v', '--append v', /nonpharma/) {|v| opts[:nonpharma] = v }
+parser.on_tail('-h', '--help') { puts help; exit }
+
+args = ARGV.dup
+begin
+  parser.parse!(args)
+rescue OptionParser::MissingArgument,
+       OptionParser::InvalidArgument,
+       OptionParser::InvalidOption
+  puts help
+  exit
 end
 
 ui = Oddb2xml::Cli.new(opts)
-unless args.empty?
-  ui.help
-else
-  begin
-    ui.run
-  rescue Interrupt
-    puts
-    exit
-  end
+begin
+  ui.run
+rescue Interrupt
+  puts
+  exit
 end

data/lib/oddb2xml/cli.rb

@@ -13,20 +13,12 @@ module Oddb2xml
     def initialize(args)
       @options = args
       #@mutex = Mutex.new
-      @items = {} # Preparations.xml in BAG
-      @index = {}
+      @items = {} # Items from Preparations.xml in BAG
+      @index = {} # Base index from swissINDEX
       LANGUAGES.each do |lang|
-        @index[lang] = {} # swissINDEX
+        @index[lang] = {}
       end
     end
-    def help
-      puts <<EOS
-#$0 ver.#{Oddb2xml::VERSION}
-Usage:
-  oddb2xml [option]
-    -c    compress option. currently only 'tar.gz' is available.
-EOS
-    end
     def run
       # Sometimes nokogiri crashes with ruby in Threads.
       #threads = []
@@ -34,20 +26,20 @@ EOS
       #threads << Thread.new do
         downloader = BagXmlDownloader.new
         xml = downloader.download
-        extractor = BagXmlExtractor.new(xml)
-        @items = extractor.to_hash
+        @items = BagXmlExtractor.new(xml).to_hash
       #end
-      LANGUAGES.map do |lang|
+      LANGUAGES.each do |lang|
         # swissindex
-        #threads << Thread.new do
-          downloader = SwissIndexDownloader.new
-          xml = downloader.download_by(lang)
-          extractor = SwissIndexExtractor.new(xml)
-          index = extractor.to_hash
-          #@mutex.synchronize do
-            @index[lang] = index
+        types.each do |type|
+          #threads << Thread.new do
+            downloader = SwissIndexDownloader.new(type)
+            xml = downloader.download_by(lang)
+            hsh = SwissIndexExtractor.new(xml, type).to_hash
+            #@mutex.synchronize do
+              @index[lang][type] = hsh
+            #end
           #end
-        #end
+        end
       end
       #threads.map(&:join)
       build
@@ -60,9 +52,16 @@ EOS
       begin
         files.each_pair do |sbj, file|
           builder = Builder.new do |builder|
+            index = {}
+            LANGUAGES.each do |lang|
+              index[lang] = {} unless index[lang]
+              types.each do |type|
+                index[lang].merge!(@index[lang][type])
+              end
+            end
             builder.subject = sbj
-            builder.index   = @index
-            builder.items   = @items
+            builder.index = index
+            builder.items = @items
           end
           xml = builder.to_xml
           File.open(file, 'w:utf-8'){ |fh| fh << xml }
@@ -84,7 +83,22 @@ EOS
       end
     end
     def report
-      # pass
+      lines = []
+      LANGUAGES.each do |lang|
+        lines << lang
+        types.each do |type|
+          key = (type == :nonpharma ? 'NonPharma' : 'Pharma')
+          lines << sprintf("\t#{key} products: %i", @index[lang][type].values.length)
+        end
+      end
+      puts lines.join("\n")
+    end
+    def types # swissindex
+      if @options[:nonpharma]
+        [:pharma, :nonpharma]
+      else
+        [:pharma]
+      end
     end
   end
 end

data/lib/oddb2xml/compressor.rb

@@ -24,8 +24,9 @@ module Oddb2xml
             File.unlink file
           end
         end
-      rescue
-        if File.exits? @compressed_file
+      rescue => error
+        puts error
+        if File.exists? @compressed_file
           File.unlink @compressed_file
         end
         return false

data/lib/oddb2xml/downloader.rb

@@ -52,8 +52,12 @@ module Oddb2xml
     end
   end
   class SwissIndexDownloader < Downloader
+    def initialize(type=:pharma)
+      @type = (type == :pharma ? 'Pharma' : 'NonPharma')
+      url = "https://index.ws.e-mediat.net/Swissindex/#{@type}/ws_#{@type}_V101.asmx?WSDL"
+      super(url)
+    end
     def init
-      @url ||= 'https://index.ws.e-mediat.net/Swissindex/Pharma/ws_Pharma_V101.asmx?WSDL'
       Savon.configure do |config|
         config.log_level    = :info
         config.log          = false # $stdout
@@ -66,12 +70,13 @@ module Oddb2xml
         wsdl.document             = @url
       end
       begin
+        type = @type
         response = client.request :download_all do
           soap.xml = <<XML
 <?xml version="1.0" encoding="utf-8"?>
 <soap:Envelope xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xsd="http://www.w3.org/2001/XMLSchema" xmlns:soap="http://schemas.xmlsoap.org/soap/envelope/">
 <soap:Body>
-  <lang xmlns="http://swissindex.e-mediat.net/SwissindexPharma_out_V101">#{lang}</lang>
+  <lang xmlns="http://swissindex.e-mediat.net/Swissindex#{type}_out_V101">#{lang}</lang>
 </soap:Body>
 </soap:Envelope>
 XML

data/lib/oddb2xml/extractor.rb

@@ -11,9 +11,7 @@ module Oddb2xml
   end
   class BagXmlExtractor < Extractor
     def to_hash
-      #File.open('../bagxml.xml', 'r:ASCII-8BIT') do |f|
-      #  @xml = f.read
-      #end
+      #File.open('../bagxml.xml', 'r:ASCII-8BIT') {|f| @xml = f.read }
       # pharmacode => sequence
       data = {}
       doc = Nokogiri::XML(@xml)
@@ -90,15 +88,17 @@ module Oddb2xml
     end
   end
   class SwissIndexExtractor < Extractor
+    def initialize(xml, type)
+      @type = (type == :pharma ? 'PHARMA' : 'NONPHARMA')
+      super(xml)
+    end
     def to_hash
-      #File.open('../swissindex.xml', 'r:ASCII-8BIT') do |f|
-      #  @xml = f.read
-      #end
+      #File.open("../swissindex_#{@type.downcase}.xml", 'r:ASCII-8BIT'){|f| @xml = f.read }
       # pharmacode => package
       data = {}
       doc = Nokogiri::XML(@xml)
       doc.remove_namespaces!
-      doc.xpath('//Envelope/Body/PHARMA/ITEM').each do |pac|
+      doc.xpath("//Envelope/Body/#{@type}/ITEM").each do |pac|
         item = {}
         item[:ean]             = (gtin = pac.at_xpath('.//GTIN'))   ? gtin.text : ''
         item[:pharmacode]      = (phar = pac.at_xpath('.//PHAR'))   ? phar.text : ''

data/lib/oddb2xml/version.rb

@@ -1,3 +1,3 @@
 module Oddb2xml
-  VERSION = "1.0.5"
+  VERSION = "1.0.6"
 end

data/oddb2xml.gemspec

@@ -22,4 +22,8 @@ Gem::Specification.new do |gem|
   gem.add_dependency 'mechanize'
   gem.add_dependency 'nokogiri'
   gem.add_dependency 'savon'
+
+  gem.add_development_dependency 'rspec'
+  gem.add_development_dependency 'webmock'
+  gem.add_development_dependency 'ZenTest'
 end

data/spec/builder_spec.rb

@@ -0,0 +1,8 @@
+# encoding: utf-8
+
+require 'spec_helper'
+
+describe Oddb2xml::Builder do
+  it 'pending'
+end
+

data/spec/cli_spec.rb

@@ -0,0 +1,63 @@
+# encoding: utf-8
+
+require 'spec_helper'
+
+shared_examples_for 'any interface' do
+  it { cli.should respond_to(:run)  }
+  it 'should run successfully' do
+    $stdout.should_receive(:puts).with(/product/)
+    cli.run
+  end
+end
+
+describe Oddb2xml::Cli do
+  include ServerMockHelper
+  before(:each) do
+    setup_server_mocks
+  end
+  context "when -c tar.gz option is given" do
+    let(:cli) { Oddb2xml::Cli.new({:compress => 'tar.gz', :nonpharma => false}) }
+    it { cli.instance_variable_get(:@options).should == {:compress => 'tar.gz', :nonpharma => false} }
+    it_behaves_like 'any interface'
+    it "should create tar.gz file" do
+      $stdout.should_receive(:puts).with(/Pharma/)
+      cli.run
+      file = Dir.glob('oddb_*.tar.gz').first
+      File.exists?(file).should be_true
+    end
+    it "should not create any xml file" do
+      $stdout.should_receive(:puts).with(/Nonpharma/)
+      cli.run
+      Dir.glob('oddb_*.xml').each do |file|
+        File.exists?(file).should be_nil
+      end
+    end
+    after(:each) do
+      Dir.glob('oddb_*.tar.gz').each do |file|
+        File.unlink(file) if File.exists?(file)
+      end
+    end
+  end
+  context "when -a nonpharma option is given" do
+    let(:cli) { Oddb2xml::Cli.new({:compress => nil, :nonpharma => true}) }
+    it { cli.instance_variable_get(:@options).should == {:compress => nil, :nonpharma => true} }
+    it_behaves_like 'any interface'
+    it "should not create any compressed file" do
+      $stdout.should_receive(:puts).with(/Pharma/)
+      cli.run
+      Dir.glob('oddb_*.tar.gz').first.should be_nil
+    end
+    it "should create 2 xml files" do
+      $stdout.should_receive(:puts).with(/Nonpharma/)
+      cli.run
+      Dir.glob('oddb_*.xml').each do |file|
+        File.exists?(file).should be_true
+      end
+    end
+    after(:each) do
+      Dir.glob('oddb_*.xml').each do |file|
+        File.unlink(file) if File.exists?(file)
+      end
+    end
+  end
+end

data/spec/compressor_spec.rb

@@ -0,0 +1,37 @@
+# encoding: utf-8
+
+require 'spec_helper'
+
+describe Oddb2xml::Compressor do
+  context "when tar.gz ext is given at initialize" do
+    before(:each) do
+      @compressor = Oddb2xml::Compressor.new('tar.gz')
+    end
+    it 'should have formated filename with datetime' do
+      @compressor.instance_variable_get(:@compressed_file).
+        should =~ /oddb_xml_\d{2}.\d{2}.\d{4}_\d{2}.\d{2}.tar\.gz/
+    end
+    it 'should have empty contents as array' do
+      @compressor.contents.should be_a Array
+      @compressor.contents.should be_empty
+    end
+  end
+  context 'when finalize! is called' do
+    before(:each) do
+      @compressor = Oddb2xml::Compressor.new()
+    end
+    context 'unexpectedly' do
+      it 'should fail with no contents' do
+        @compressor.finalize!.should == false
+      end
+      it 'should fail with invalid file' do
+        @compressor.contents << '../invalid_file'
+        @compressor.finalize!.should == false
+      end
+    end
+    context 'successfully' do
+      it 'pending'
+    end
+  end
+end
+

data/spec/data/swissindex.xml

@@ -0,0 +1,56 @@
+<?xml version="1.0" encoding="utf-8"?>
+<soap:Envelope
+  xmlns:soap="http://schemas.xmlsoap.org/soap/envelope/"
+  xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
+  xmlns:xsd="http://www.w3.org/2001/XMLSchema">
+  <soap:Body>
+    <PHARMA
+      CREATION_DATETIME="2012-10-18T06:31:19.4687619+02:00"
+      xmlns="http://example.com/test">
+      <ITEM DT="2012-10-18T00:00:00">
+        <GTIN>0000000000000</GTIN>
+        <PHAR>00000</PHAR>
+        <STATUS>A</STATUS>
+        <STDATE>2002-05-23T00:00:00</STDATE>
+        <LANG>DE</LANG>
+        <DSCR>Foo 1000 mg</DSCR>
+        <ADDSCR>Amp</ADDSCR>
+        <ATC>A00AA00</ATC>
+        <COMP>
+          <NAME>EXAMPLE.COM AG</NAME>
+          <GLN>0000000000000</GLN>
+        </COMP>
+      </ITEM>
+      <ITEM DT="2012-10-18T00:00:00">
+        <GTIN>0000000000001</GTIN>
+        <PHAR>00001</PHAR>
+        <STATUS>A</STATUS>
+        <STDATE>2002-05-23T00:00:00</STDATE>
+        <LANG>DE</LANG>
+        <DSCR>Bar 10 Stk</DSCR>
+        <ADDSCR>10 Stk</ADDSCR>
+        <ATC>A02AA02</ATC>
+        <COMP>
+          <NAME>EXAMPLE.ORG AG</NAME>
+        </COMP>
+      </ITEM>
+      <ITEM DT="2012-10-18T00:00:00">
+        <GTIN>0000000000002</GTIN>
+        <PHAR>00002</PHAR>
+        <STATUS>A</STATUS>
+        <STDATE>2002-05-23T00:00:00</STDATE>
+        <LANG>DE</LANG>
+        <DSCR>Baz 10 Tablet</DSCR>
+        <ADDSCR>5 Stk</ADDSCR>
+        <ATC>A03AA03</ATC>
+        <COMP>
+          <NAME>EXAMPLE.AC AG</NAME>
+        </COMP>
+      </ITEM>
+      <RESULT>
+        <OK_ERROR>OK</OK_ERROR>
+        <NBR_RECORD>3</NBR_RECORD>
+      </RESULT>
+    </PHARMA>
+  </soap:Body>
+</soap:Envelope>

data/spec/data/wsdl.xml

@@ -0,0 +1,55 @@
+<?xml version="1.0" encoding="utf-8"?>
+<wsdl:definitions
+  xmlns:s="http://www.w3.org/2001/XMLSchema"
+  xmlns:soap12="http://schemas.xmlsoap.org/wsdl/soap12/"
+  xmlns:mime="http://schemas.xmlsoap.org/wsdl/mime/"
+  xmlns:tns="http://example.com/test"
+  xmlns:soap="http://schemas.xmlsoap.org/wsdl/soap/"
+  xmlns:tm="http://microsoft.com/wsdl/mime/textMatching/"
+  xmlns:http="http://schemas.xmlsoap.org/wsdl/http/"
+  xmlns:soapenc="http://schemas.xmlsoap.org/soap/encoding/"
+  targetNamespace="http://example.com/test"
+  xmlns:wsdl="http://schemas.xmlsoap.org/wsdl/">
+  <wsdl:types>
+    <s:schema elementFormDefault="qualified" targetNamespace="http://example.com/test">
+    </s:schema>
+  </wsdl:types>
+  <wsdl:message name="DownloadAllSoapIn">
+    <wsdl:part name="lang" element="tns:lang" />
+  </wsdl:message>
+  <wsdl:message name="DownloadAllSoapOut">
+    <wsdl:part name="DownloadAllResult" element="tns:PHARMA" />
+  </wsdl:message>
+  <wsdl:binding name="ws_Pharma_V101Soap" type="tns:ws_Pharma_V101Soap">
+    <soap:binding transport="http://schemas.xmlsoap.org/soap/http" />
+    <wsdl:operation name="DownloadAll">
+      <soap:operation soapAction="http://example.com/DownloadAll" style="document" />
+      <wsdl:input>
+        <soap:body use="literal" />
+      </wsdl:input>
+      <wsdl:output>
+        <soap:body use="literal" />
+      </wsdl:output>
+    </wsdl:operation>
+  </wsdl:binding>
+  <wsdl:binding name="ws_Pharma_V101Soap12" type="tns:ws_Pharma_V101Soap">
+    <soap12:binding transport="http://schemas.xmlsoap.org/soap/http" />
+    <wsdl:operation name="DownloadAll">
+      <soap12:operation soapAction="http://example.com/DownloadAll" style="document" />
+      <wsdl:input>
+        <soap12:body use="literal" />
+      </wsdl:input>
+      <wsdl:output>
+        <soap12:body use="literal" />
+      </wsdl:output>
+    </wsdl:operation>
+  </wsdl:binding>
+  <wsdl:service name="ws_Pharma_V101">
+    <wsdl:port name="ws_Pharma_V101Soap" binding="tns:ws_Pharma_V101Soap">
+      <soap:address location="https://example.com/test" />
+    </wsdl:port>
+    <wsdl:port name="ws_Pharma_V101Soap12" binding="tns:ws_Pharma_V101Soap12">
+      <soap12:address location="https://example.com/test" />
+    </wsdl:port>
+  </wsdl:service>
+</wsdl:definitions>

data/spec/downloader_spec.rb

@@ -0,0 +1,64 @@
+# encoding: utf-8
+
+require 'spec_helper'
+
+shared_examples_for 'any downloader' do
+  # this takes 5 sec. by call for sleep
+  it 'should count retry times as retrievable or not' do
+    expect {
+      Array.new(3).map do
+        Thread.new do
+          @downloader.send(:retrievable?).should be(true)
+        end
+      end.map(&:join)
+    }.to change {
+      @downloader.instance_variable_get(:@retry_times)
+    }.from(3).to(0)
+  end
+end
+
+describe Oddb2xml::BagXmlDownloader do
+  include ServerMockHelper
+  before(:each) do
+    setup_bag_xml_server_mock
+    @downloader = Oddb2xml::BagXmlDownloader.new()
+  end
+  it_behaves_like 'any downloader'
+  context 'when download is called' do
+    let(:xml) { @downloader.download }
+    it 'should parse zip to string' do
+      xml.should be_a String
+      xml.length.should_not == 0
+    end
+    it 'should return valid xml' do
+      xml.should =~ /xml\sversion="1.0"/
+      xml.should =~ /Preparations/
+      xml.should =~ /DescriptionDe/
+    end
+    it 'should cleanup current directory' do
+      xml.should_not raise_error(Timeout::Error)
+      File.exist?('XMLPublications.zip').should be(false)
+    end
+  end
+end
+
+describe Oddb2xml::SwissIndexDownloader do
+  include ServerMockHelper
+  before(:each) do
+    setup_swiss_index_server_mock
+    @downloader = Oddb2xml::SwissIndexDownloader.new()
+  end
+  it_behaves_like 'any downloader'
+  context 'when download_by is called with DE' do
+    let(:xml) { @downloader.download_by('DE') }
+    it 'should parse hash to xml' do
+      xml.should be_a String
+      xml.length.should_not == 0
+    end
+    it 'should return valid xml' do
+      xml.should =~ /xml\sversion="1.0"/
+      xml.should =~ /ITEM/
+      xml.should =~ /PHAR/
+    end
+  end
+end

data/spec/extractor_spec.rb

@@ -0,0 +1,7 @@
+# encoding: utf-8
+
+require 'spec_helper'
+
+describe Oddb2xml::Extractor do
+  it "pending"
+end

data/spec/spec_helper.rb

@@ -0,0 +1,77 @@
+# This file was generated by the `rspec --init` command. Conventionally, all
+# specs live under a `spec` directory, which RSpec adds to the `$LOAD_PATH`.
+# Require this file using `require "spec_helper"` to ensure that it is only
+# loaded once.
+#
+# See http://rubydoc.info/gems/rspec-core/RSpec/Core/Configuration
+#
+$:.unshift File.join(File.dirname(__FILE__), '..', 'lib')
+$:.unshift File.dirname(__FILE__)
+
+require 'bundler/setup'
+Bundler.require
+
+require 'rspec'
+require 'webmock/rspec'
+require 'oddb2xml'
+
+module ServerMockHelper
+  def setup_server_mocks
+    setup_bag_xml_server_mock
+    setup_swiss_index_server_mock
+  end
+  def setup_bag_xml_server_mock
+    # zip
+    stub_wsdl_url = 'http://bag.e-mediat.net/SL2007.Web.External/File.axd?file=XMLPublications.zip'
+    stub_response = File.read(File.expand_path('../data/XMLPublications.zip', __FILE__))
+    stub_request(:get, stub_wsdl_url).
+      with(:headers => {
+        'Accept'          => '*/*',
+        'Accept-Encoding' => 'gzip,deflate,identity',
+        'Host'            => 'bag.e-mediat.net',
+      }).
+      to_return(
+        :status  => 200,
+        :headers => {'Content-Type' => 'application/zip; charset=utf-8'},
+        :body    => stub_response)
+  end
+  def setup_swiss_index_server_mock
+    # wsdl
+    stub_wsdl_url = 'https://index.ws.e-mediat.net/Swissindex/Pharma/ws_Pharma_V101.asmx?WSDL'
+    stub_response = File.read(File.expand_path('../data/wsdl.xml', __FILE__))
+    stub_request(:get, stub_wsdl_url).
+      with(:headers => {
+        'Accept'     => '*/*',
+        'User-Agent' => 'Ruby'}).
+      to_return(
+        :status  => 200,
+        :headers => {'Content-Type' => 'text/xml; charset=utf-8'},
+        :body    => stub_response)
+    # soap (dummy)
+    stub_soap_url = 'https://example.com/test'
+    stub_response = File.read(File.expand_path('../data/swissindex.xml', __FILE__))
+    stub_request(:post, stub_soap_url).
+      with(:headers => {
+        'Accept'     => '*/*',
+        'User-Agent' => 'Ruby'}).
+      to_return(
+        :status  => 200,
+        :headers => {'Content-Type' => 'text/xml; chaprset=utf-8'},
+        :body    => stub_response)
+  end
+end
+
+RSpec.configure do |config|
+  config.treat_symbols_as_metadata_keys_with_true_values = true
+  config.run_all_when_everything_filtered = true
+  config.filter_run :focus
+
+  # Run specs in random order to surface order dependencies. If you find an
+  # order dependency and want to debug it, you can fix the order by providing
+  # the seed, which is printed after each run.
+  #     --seed 1234
+  config.order = 'random'
+
+  # Helper
+  config.include(ServerMockHelper)
+end