ddi-parser 0.0.5 → 0.1.0

data/lib/ddi-parser.rb

@@ -8,19 +8,30 @@ require 'models/study'
 require 'models/study_date'
 require 'models/ddi_variable'
 require 'models/summary_stat'
-
+require 'logger'
 module DDI
   class Parser
     
+    attr_accessor :logger
+        
+    def initialize
+      @logger = Logger.new('ddi-parser.log')
+    end
+    
     #Given a DDI metadata file, parse it and return study information
     #
     #Returns a Nesstar::Study object
     def parse ddi_file
+      @logger.info 'Parsing DDI file ' + ddi_file
       catalog = DDI::Catalog.new
       study = DDI::Study.new
       study_info_hash = Hash.new
 #TODO This will not work on windows since it depends on the unix tool file need to use a different way. Possibly use rchardet instead
-      encode_type = `file --mime -br #{ddi_file}`.gsub(/\n/,"").split(';')[1].split('=')[1]
+      begin
+        encode_type = `file --mime -br #{ddi_file}`.gsub(/\n/,"").split(';')[1].split('=')[1]
+      rescue Exception => e
+        
+      end
       #have to convert to UTF-8 for libxml
       contents = File.open(ddi_file).read
       output = Iconv.conv("UTF-8", encode_type, contents)
@@ -33,13 +44,11 @@ module DDI
       studynodes = doc.find('//stdyDscr')
       abstracts = studynodes[0].find('//abstract')
       abstract = ""
-      abstracts.each do |ab|
-        abstract << ab.content.strip
-      end
+      abstracts.each {|ab| abstract << ab.content.strip}
       abstract.strip!
       study.abstract = abstract
-      study.title = studynodes[0].find('//stdyDscr/citation/titlStmt/titl')[0].first.content.strip
-      study.id = studynodes[0].find('//IDNo')[0].first.content.strip
+      study.title = studynodes[0].find('//stdyDscr/citation/titlStmt/titl')[0].first.content.strip unless studynodes[0].find('//stdyDscr/citation/titlStmt/titl')[0] == nil
+      study.id = studynodes[0].find('//IDNo')[0].first.content.strip unless studynodes[0].find('//IDNo')[0] == nil
       
       #start and finish dates for study
       dates = []
@@ -54,7 +63,10 @@ module DDI
         end
       end
       study.dates = dates
-      study.sampling_procedure = studynodes[0].find('//sampProc')[0].first.content.strip unless studynodes[0].find('//sampProc')[0].children.size == 0
+      studynodes[0].find('//sampProc')[0] ? samp_node = studynodes[0].find('//sampProc')[0] : samp_node = nil
+      unless samp_node == nil
+        study.sampling_procedure = samp_node.first.content.strip unless samp_node.first == nil
+      end
       # study.weight = studynodes[0].find('//sampProc')[0].first.content
       study.ddi_variables = get_variable_information doc
       return study
@@ -74,7 +86,7 @@ module DDI
         groups = a.get_attribute('var')
         if groups != nil
           groups = a.get_attribute('var')
-          variable_info_hash[vargroup.find('./labl')[0].first.content] = groups.value.split(' ')
+          variable_info_hash[vargroup.find('./labl')[0].first.content] = groups.value.split(' ') unless vargroup.find('./labl')[0] == nil
         # else
         #             variable_info_hash[vargroup.find('./labl')[0].first.content] = groups.value.split(' ')
         end
@@ -85,17 +97,25 @@ module DDI
         var_attr = var.attributes
         variable.id = var_attr.get_attribute('ID').value.strip unless var_attr.get_attribute('ID') == nil
         variable.name = var_attr.get_attribute('name').value.strip unless var_attr.get_attribute('name') == nil
+        #if there is no name then it has no meaning or context
+        next if variable.name == nil
         variable.file = var_attr.get_attribute('files').value.strip unless var_attr.get_attribute('files') == nil
         variable.interval = var_attr.get_attribute('intrvl').value.strip unless var_attr.get_attribute('intrvl') == nil
         variable.label = var.find('./labl')[0].content.strip unless var.find('./labl')[0] == nil 
+        #these things never seem consistent with the schema, might be an inner element, might be an attribute!
+        if var.find('./labl')[0] == nil
+          variable.label = var_attr.get_attribute('labl').value.strip unless var_attr.get_attribute('labl') == nil
+        end
         rng = var.find('./valrng')
         if rng != nil
           if rng[0] != nil
-            range_attr = rng[0].first.attributes
-            max_val = range_attr.get_attribute('max')
-            variable.max = max_val.value.strip unless max_val == nil
-            min_val = range_attr.get_attribute('min')
-            variable.min = min_val.value.strip unless min_val == nil
+            unless rng[0].first == nil
+              range_attr = rng[0].first.attributes
+              max_val = range_attr.get_attribute('max')
+              variable.max = max_val.value.strip unless max_val == nil
+              min_val = range_attr.get_attribute('min')
+              variable.min = min_val.value.strip unless min_val == nil
+            end
           end
         end
         q = var.find('./qstn')

data/lib/models/version.rb

@@ -1,5 +1,5 @@
 module DDI
   module Parser
-    VERSION = "0.0.5"
+    VERSION = "0.1.0"
   end
 end

metadata

@@ -1,34 +1,49 @@
---- !ruby/object:Gem::Specification
+--- !ruby/object:Gem::Specification 
 name: ddi-parser
-version: !ruby/object:Gem::Version
-  version: 0.0.5
+version: !ruby/object:Gem::Version 
+  hash: 27
   prerelease: 
+  segments: 
+  - 0
+  - 1
+  - 0
+  version: 0.1.0
 platform: ruby
-authors:
+authors: 
 - Ian Dunlop
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2012-04-16 00:00:00.000000000 Z
-dependencies:
-- !ruby/object:Gem::Dependency
+
+date: 2012-06-14 00:00:00 +01:00
+default_executable: 
+dependencies: 
+- !ruby/object:Gem::Dependency 
   name: libxml-ruby
-  requirement: &12850280 !ruby/object:Gem::Requirement
+  prerelease: false
+  requirement: &id001 !ruby/object:Gem::Requirement 
     none: false
-    requirements:
-    - - ! '>='
-      - !ruby/object:Gem::Version
+    requirements: 
+    - - ">="
+      - !ruby/object:Gem::Version 
+        hash: 7
+        segments: 
+        - 2
+        - 2
+        - 0
         version: 2.2.0
   type: :runtime
-  prerelease: false
-  version_requirements: *12850280
+  version_requirements: *id001
 description: This gem parses ddi metadata files
-email:
+email: 
 - ian.dunlop@manchester.ac.uk
 executables: []
+
 extensions: []
+
 extra_rdoc_files: []
-files:
+
+files: 
 - lib/ddi-parser.rb
 - lib/models/version.rb
 - lib/models/summary_stat.rb
@@ -38,28 +53,39 @@ files:
 - lib/models/category_statistic.rb
 - lib/models/ddi_variable.rb
 - lib/models/study_date.rb
+has_rdoc: true
 homepage: http://github.com/mygrid/ddi-parser
 licenses: []
+
 post_install_message: 
 rdoc_options: []
-require_paths:
+
+require_paths: 
 - lib
-required_ruby_version: !ruby/object:Gem::Requirement
+required_ruby_version: !ruby/object:Gem::Requirement 
   none: false
-  requirements:
-  - - ! '>='
-    - !ruby/object:Gem::Version
-      version: '0'
-required_rubygems_version: !ruby/object:Gem::Requirement
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      hash: 3
+      segments: 
+      - 0
+      version: "0"
+required_rubygems_version: !ruby/object:Gem::Requirement 
   none: false
-  requirements:
-  - - ! '>='
-    - !ruby/object:Gem::Version
-      version: '0'
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      hash: 3
+      segments: 
+      - 0
+      version: "0"
 requirements: []
+
 rubyforge_project: ddi-parser
-rubygems_version: 1.8.11
+rubygems_version: 1.6.2
 signing_key: 
 specification_version: 3
 summary: API for parsing ddi metadata files and returning results
 test_files: []
+