oddb2xml 1.1.0 → 1.1.1

data/History.txt

@@ -1,3 +1,8 @@
+=== 1.1.1 / 07.01.2013
+
+* Create oddb_fi.xml file with using swissmedicinfo
+    -needs -o or --optional fi
+
 === 1.1.0 / 27.12.2012
 
 * Add new additional file limitation.xml

data/bin/oddb2xml

@@ -15,6 +15,7 @@ Usage:
   oddb2xml [option]
     -c F, --compress=F   Compress format F. {tar.gz|zip}
     -a T, --append=T     Additional target. T, only 'nonpharma' is available.
+    -o O, --optional=O   Optional output. O, only 'fi' is available.
     -t S, --tag-suffix=S XML tag suffix S. Default is none. [A-z0-9_]
                          If S is given, it is also used as prefix of filename.
     -h,   --help         Show this help message.
@@ -28,8 +29,9 @@ opts = {
   :tag_suffix   => nil,
 }
 
-parser.on('-c v', '--compress v', /tar\.gz|zip/) {|v| opts[:compress_ext] = v }
-parser.on('-a v', '--append v', /nonpharma/) {|v| opts[:nonpharma] = v }
+parser.on('-c v', '--compress v',   /tar\.gz|zip/)   {|v| opts[:compress_ext] = v }
+parser.on('-a v', '--append v',     /nonpharma/)     {|v| opts[:nonpharma] = true }
+parser.on('-o v', '--optional v',   /fi/)            {|v| opts[:fi] = true }
 parser.on('-t v', '--tag-suffix v', /^[A-z0-9_]*$/i) {|v| opts[:tag_suffix] = v.upcase }
 parser.on_tail('-h', '--help') { puts help; exit }
 

data/lib/oddb2xml/builder.rb

@@ -17,12 +17,14 @@ end
 
 module Oddb2xml
   class Builder
-    attr_accessor :subject, :index, :items, :orphans, :fridges,
+    attr_accessor :subject, :index, :items, :infos,
+                  :orphans, :fridges,
                   :tag_suffix
     def initialize
       @subject    = nil
       @index      = {}
       @items      = {}
+      @infos      = {}
       @orphans    = []
       @fridges    = []
       @tag_suffix = nil
@@ -446,5 +448,47 @@ module Oddb2xml
       end
       _builder.to_xml
     end
+    def build_fi
+      _builder = Nokogiri::XML::Builder.new(:encoding => 'utf-8') do |xml|
+        xml.doc.tag_suffix = @tag_suffix
+        datetime = Time.new.strftime('%FT%T.%7N%z')
+        xml.KOMPENDIUM(
+          'xmlns:xsd'         => 'http://www.w3.org/2001/XMLSchema',
+          'xmlns:xsi'         => 'http://www.w3.org/2001/XMLSchema-instance',
+          'xmlns'             => 'http://wiki.oddb.org/wiki.php?pagename=Swissmedic.Datendeklaration',
+          'CREATION_DATETIME' => datetime,
+          'PROD_DATE'         => datetime,
+          'VALID_DATE'        => datetime,
+        ) {
+          length = 0
+          %w[de fr].each do |lang|
+            length += @infos[lang].length
+            @infos[lang].each_with_index do |info, i|
+              xml.ART(
+                'MONTYPE' => 'fi', # only
+                'LANG'    => lang.upcase,
+                'DT'      => '',
+              ) {
+                unless info[:name].empty?
+                  xml.name { xml.p info[:name] }
+                end
+                unless info[:owner].empty?
+                  xml.owner { xml.p info[:owner] }
+                end
+                xml.monid info[:monid] unless info[:monid].empty?
+                xml.paragraph { xml.cdata info[:paragraph] unless info[:paragraph].empty? }
+              }
+            end
+          end
+          xml.RESULT {
+            xml.OK_ERROR   'OK'
+            xml.NBR_RECORD length
+            xml.ERROR_CODE ''
+            xml.MESSAGE    ''
+          }
+        }
+      end
+      _builder.to_xml
+    end
   end
 end

data/lib/oddb2xml/cli.rb

@@ -10,12 +10,14 @@ module Oddb2xml
   class Cli
     SUBJECTS  = %w[product article]
     ADDITIONS = %w[substance limitation]
+    OPTIONALS = %w[fi]
     LANGUAGES = %w[DE FR] # EN does not exist
     def initialize(args)
       @options = args
       @mutex = Mutex.new
       @items = {} # Items from Preparations.xml in BAG
       @index = {} # Base index from swissINDEX
+      @infos = {} # FI from SwissmedicInfo
       @orphans = [] # Orphaned drugs from Swissmedic xls
       @fridges = [] # ReFridge drugs from Swissmedic xls
       LANGUAGES.each do |lang|
@@ -25,6 +27,14 @@ module Oddb2xml
     def run
       threads = []
       # swissmedic
+      threads << Thread.new do
+        downloader = SwissmedicInfoDownloader.new
+        xml = downloader.download
+        @mutex.synchronize do
+          hsh = SwissmedicInfoExtractor.new(xml).to_hash
+          @infos = hsh
+        end
+      end
       [:orphans, :fridges].each do |type|
         threads << Thread.new do
           downloader = SwissmedicDownloader.new
@@ -68,14 +78,17 @@ module Oddb2xml
             LANGUAGES.each do |lang|
               index[lang] = {} unless index[lang]
               types.each do |type|
-                index[lang].merge!(@index[lang][type])
+                index[lang].merge!(@index[lang][type]) if @index[lang][type]
               end
             end
-            builder.subject    = sbj
-            builder.index      = index
-            builder.items      = @items
-            builder.orphans    = @orphans
-            builder.fridges    = @fridges
+            builder.subject = sbj
+            builder.index   = index
+            builder.items   = @items
+            # additions
+            builder.orphans = @orphans
+            builder.fridges = @fridges
+            # optionals
+            builder.infos = @infos
             builder.tag_suffix = @options[:tag_suffix]
           end
           xml = builder.to_xml
@@ -102,7 +115,9 @@ module Oddb2xml
     def files
       unless @_files
         @_files = {}
-        (ADDITIONS + SUBJECTS).each do|sbj|
+        _files = (ADDITIONS + SUBJECTS)
+        _files += OPTIONALS if @options[:fi]
+        _files.each do|sbj|
           @_files[sbj] = "#{prefix}_#{sbj.to_s}.xml"
         end
       end
@@ -117,7 +132,10 @@ module Oddb2xml
         lines << lang
         types.each do |type|
           key = (type == :nonpharma ? 'NonPharma' : 'Pharma')
-          lines << sprintf("\t#{key} products: %i", @index[lang][type].values.length)
+          if @index[lang][type]
+            lines << sprintf(
+              "\t#{key} products: %i", @index[lang][type].values.length)
+          end
         end
       end
       puts lines.join("\n")

data/lib/oddb2xml/downloader.rb

@@ -58,21 +58,19 @@ module Oddb2xml
       super(url)
     end
     def init
-      Savon.configure do |config|
-        config.log_level    = :info
-        config.log          = false # $stdout
-        config.raise_errors = true
-      end
+      @config = {
+        :log_level       => :info,
+        :log             => false, # $stdout
+        :raise_errors    => true,
+        :ssl_verify_mode => :none,
+        :wsdl            => @url
+      }
     end
     def download_by(lang = 'DE')
-      client = Savon::Client.new do |wsdl, http|
-        http.auth.ssl.verify_mode = :none
-        wsdl.document             = @url
-      end
+      client = Savon::Client.new(@config)
       begin
         type = @type
-        response = client.request :download_all do
-          soap.xml = <<XML
+        soap = <<XML
 <?xml version="1.0" encoding="utf-8"?>
 <soap:Envelope xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xsd="http://www.w3.org/2001/XMLSchema" xmlns:soap="http://schemas.xmlsoap.org/soap/envelope/">
 <soap:Body>
@@ -80,7 +78,7 @@ module Oddb2xml
 </soap:Body>
 </soap:Envelope>
 XML
-        end
+        response = client.call(:download_all, :xml => soap)
         if response.success?
           if xml = response.to_xml
             return xml
@@ -128,4 +126,44 @@ XML
       end
     end
   end
+  class SwissmedicInfoDownloader < Downloader
+    def init
+      @url ||= "http://download.swissmedicinfo.ch/Accept.aspx?ReturnUrl=%2f"
+    end
+    def download
+      file = "swissmedic_info.zip"
+      begin
+        response = nil
+        agent = Mechanize.new
+        agent.ignore_bad_chunking = true
+        if home = agent.get(@url)
+          form = home.form_with(:id => 'Form1')
+          bttn = form.button_with(:name => 'ctl00$MainContent$btnOK')
+          if page = form.submit(bttn)
+            form = page.form_with(:id => 'Form1')
+            bttn = form.button_with(:name => 'ctl00$MainContent$BtnYes')
+            response = form.submit(bttn)
+          end
+        end
+        if response
+          response.save_as file
+        end
+        xml = ''
+        Zip::ZipFile.foreach(file) do |entry|
+          if entry.name =~ /^AipsDownload_/iu
+            entry.get_input_stream{ |io| xml = io.read }
+          end
+        end
+        return xml
+      rescue Timeout::Error
+        retrievable? ? retry : raise
+      rescue NoMethodError => e
+        # pass
+      ensure
+        if File.exists? file
+          File.unlink file
+        end
+      end
+    end
+  end
 end

data/lib/oddb2xml/extractor.rb

@@ -172,4 +172,32 @@ module Oddb2xml
       end
     end
   end
+  class SwissmedicInfoExtractor < Extractor
+    def to_hash
+      data = Hash.new{|h,k| h[k] = [] }
+      doc = Nokogiri::XML(@xml)
+      doc.xpath("//medicalInformations/medicalInformation[@type='fi']").each do |fi|
+        lang = fi.attr('lang').to_s
+        next unless lang =~ /de|fr/
+        item = {}
+        item[:name]  = (name = fi.at_xpath('.//title')) ? name.text : ''
+        item[:owner] = (ownr = fi.at_xpath('.//authHolder')) ? ownr.text : ''
+        if content = fi.at_xpath('.//content').children.detect{|child| child.cdata? }
+          html = Nokogiri::HTML(content.to_s)
+          # all HTML contents without MonTitle and ownerCompany
+          item[:paragraph] =  "<title><p>#{item[:name]}</p></title>" +
+             ((paragraph = html.xpath("///div[@class='paragraph']")) ? paragraph.to_s : '')
+          if text = html.xpath("///div[@id='Section7750']/p").text
+            if text =~ /(\d{5})[,\s]*(\d{5})?/ 
+              [$1, $2].compact.each do |n|
+                item[:monid] = n
+                data[lang] << item
+              end
+            end
+          end
+        end
+      end
+      data
+    end
+  end
 end

data/lib/oddb2xml/version.rb

@@ -1,3 +1,3 @@
 module Oddb2xml
-  VERSION = "1.1.0"
+  VERSION = "1.1.1"
 end

data/oddb2xml.gemspec

@@ -21,7 +21,7 @@ Gem::Specification.new do |gem|
   gem.add_dependency 'archive-tar-minitar'
   gem.add_dependency 'mechanize'
   gem.add_dependency 'nokogiri'
-  gem.add_dependency 'savon'
+  gem.add_dependency 'savon', '>= 2.0'
   gem.add_dependency 'spreadsheet'
 
   gem.add_development_dependency 'rspec'

data/spec/spec_helper.rb

@@ -20,6 +20,7 @@ module ServerMockHelper
     setup_bag_xml_server_mock
     setup_swiss_index_server_mock
     setup_swissmedic_server_mock
+    setup_swissmedic_info_server_mock
   end
   def setup_bag_xml_server_mock
     # zip
@@ -95,6 +96,34 @@ module ServerMockHelper
           :body    => stub_response)
     end
   end
+  def setup_swissmedic_info_server_mock
+    # html (dummy)
+    stub_html_url = "http://download.swissmedicinfo.ch/Accept.aspx?ReturnUrl=%2f"
+    stub_response = File.read(File.expand_path("../data/swissmedic_info.html", __FILE__))
+    stub_request(:get, stub_html_url).
+      with(:headers => {
+        'Accept' => '*/*',
+        'Host'   => 'download.swissmedicinfo.ch',
+      }).
+      to_return(
+        :status  => 200,
+        :headers => {'Content-Type' => 'text/html; charset=utf-8'},
+        :body    => stub_response)
+    # zip
+    stub_zip_url = "http://download.swissmedicinfo.ch/"
+    stub_response = File.read(File.expand_path('../data/swissmedic_info.zip', __FILE__))
+    stub_request(:get, stub_zip_url).
+      with(:headers => {
+        'Accept'          => '*/*',
+        'Accept-Encoding' => 'gzip,deflate,identity',
+        'Host'            => 'download.swisssmedicinfo.ch',
+      }).
+      to_return(
+        :status  => 200,
+        :headers => {'Content-Type' => 'application/zip; charset=utf-8'},
+        :body    => stub_response)
+
+  end
 end
 
 RSpec.configure do |config|

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: oddb2xml
 version: !ruby/object:Gem::Version
-  version: 1.1.0
+  version: 1.1.1
   prerelease: 
 platform: ruby
 authors:
@@ -9,11 +9,11 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2012-12-27 00:00:00.000000000 Z
+date: 2013-01-07 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rdoc
-  requirement: &11850420 !ruby/object:Gem::Requirement
+  requirement: &18836140 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ~>
@@ -21,10 +21,10 @@ dependencies:
         version: '3.10'
   type: :development
   prerelease: false
-  version_requirements: *11850420
+  version_requirements: *18836140
 - !ruby/object:Gem::Dependency
   name: hoe
-  requirement: &11849560 !ruby/object:Gem::Requirement
+  requirement: &18834840 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ~>
@@ -32,7 +32,7 @@ dependencies:
         version: '2.13'
   type: :development
   prerelease: false
-  version_requirements: *11849560
+  version_requirements: *18834840
 description: ''
 email:
 - yasaka@ywesee.com,  zdavatz@ywesee.com