biomart 0.1.1 → 0.1.2

data/History.txt

@@ -5,3 +5,16 @@
     * Basic gathering of information about a Biomart server.
     * Counting functionality.
     * Basic searching of a single dataset.
+
+=== 0.1.1 2009-10-21
+
+* 1 minor enhancement:
+  * Addition of an .alive? function to the Server and Dataset classes 
+    to allow a user to ping a biomart server to make sure it is online 
+    and functioning as expected.
+
+=== 0.1.2 2009-10-29
+
+* 1 major bugfix:
+  * Added in code to handle poorly formatted tab-separated data 
+    coming back from a biomart query.

data/biomart.gemspec

@@ -2,11 +2,11 @@
 
 Gem::Specification.new do |s|
   s.name = %q{biomart}
-  s.version = "0.1.1"
+  s.version = "0.1.2"
 
   s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
   s.authors = ["Darren Oakley"]
-  s.date = %q{2009-10-21}
+  s.date = %q{2009-10-29}
   s.description = %q{A ruby API for interacting with Biomart services.}
   s.email = ["daz.oakley@gmail.com"]
   s.extra_rdoc_files = ["History.txt", "Manifest.txt", "README.rdoc"]

data/lib/biomart.rb

@@ -7,7 +7,7 @@ require "rubygems"
 require "builder"
 
 module Biomart
-  VERSION = "0.1.1"
+  VERSION = "0.1.2"
   
   # This is the base Biomart error/exception class. Rescue it if 
   # you want to catch any exceptions that this code might raise.

data/lib/biomart/dataset.rb

@@ -175,7 +175,8 @@ module Biomart
       # Utility function to transform the tab-separated data retrieved 
       # from the Biomart search query into a ruby object.
       def process_tsv( args, tsv )
-        headers = []
+        headers     = []
+        parsed_data = []
 
         if args[:attributes]
           args[:attributes].each do |attribute|
@@ -189,12 +190,56 @@ module Biomart
           end
         end
 
+        begin
+          parsed_data = CSV.parse( tsv, "\t" )
+        rescue CSV::IllegalFormatError => e
+          parsed_data = parse_tsv_line_by_line( headers.size, tsv )
+        end
+        
         return {
           :headers => headers,
-          :data    => CSV.parse( tsv, "\t" )
+          :data    => parsed_data
         }
       end
-
+      
+      # Utility function to process TSV formatted data that raises errors. (Biomart 
+      # has a habit of serving out this...) First attempts to use the CSV modules 
+      # 'parse_line' function to read in the data, if that fails, tries to use split 
+      # to recover the data.
+      def parse_tsv_line_by_line( expected_row_size, tsv )
+        parsed_data = []
+        
+        data_by_line = tsv.split("\n")
+        data_by_line.each do |line|
+          elements = CSV::parse_line( line, "\t" )
+          
+          if elements.size == 0
+            # This is a bad line (causing the above Exception), try and use split to recover.
+            # Alse add an empty value as split will miss the final value...
+            elements = line.split("\t")
+            elements.push(nil)
+            
+            # Substitute blank strings for nils
+            elements.map! do |elem|
+              if elem === ""
+                nil
+              else
+                elem
+              end
+            end
+            
+            # Add a safety clause...
+            if elements.size === expected_row_size
+              parsed_data.push(elements)
+            end
+          else
+            parsed_data.push(elements)
+          end
+        end
+        
+        return parsed_data
+      end
+      
       # Utility function to quickly convert a search result into an array of hashes
       # (keyed by the attribute name) for easier processing - this is not done by 
       # default on all searches as this can cause a large overhead on big data returns.

data/test/test_biomart.rb

@@ -92,6 +92,25 @@ class BiomartTest < Test::Unit::TestCase
       assert( search2.first.is_a?( Hash ), "Biomart::Dataset.search (filters and attributes defined with processing) is not returning an array of hashes." )
       assert( search2.first["marker_symbol"] == "Cbx1", "Biomart::Dataset.search (filters and attributes defined with processing) is not returning the correct info." )
     end
+    
+    should "handle search queries that will generate poorly formatted TSV data" do
+      search = @htgt_targ.search(
+        :filters => { "mgi_accession_id" => [ "MGI:1921569", "MGI:1913402", "MGI:1913300" ] },
+        :attributes => [
+          "is_eucomm", "is_komp_csd", "is_komp_regeneron", "is_norcomm",
+          "is_mgp", "mgi_accession_id", "marker_symbol", "ensembl_gene_id",
+          "status", "status_code", "status_type", "status_description",
+          "status_sequence", "pipeline_stage", "htgt_project_id", "bac",
+          "design_id", "design_plate", "design_well", "intvec_plate",
+          "intvec_well", "intvec_distribute", "targvec_plate", "targvec_well",
+          "targvec_distribute", "backbone", "cassette", "allele_name",
+          "escell_clone_name", "escell_distribute", "es_cell_line", "colonies_picked",
+          "is_latest_for_gene", "is_targeted_non_cond"
+        ]
+      )
+      assert( search.is_a?( Hash ), "Biomart::Dataset.search (no options) is not returning a hash." )
+      assert( search[:data].is_a?( Array ), "Biomart::Dataset.search[:data] (no options) is not returning an array." )
+    end
   end
   
   context "The Biomart module" do

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification 
 name: biomart
 version: !ruby/object:Gem::Version 
-  version: 0.1.1
+  version: 0.1.2
 platform: ruby
 authors: 
 - Darren Oakley
@@ -9,7 +9,7 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2009-10-21 00:00:00 +01:00
+date: 2009-10-29 00:00:00 +00:00
 default_executable: 
 dependencies: 
 - !ruby/object:Gem::Dependency