dataset_for.rb in bio-publisci-0.0.7

- old
+ new

@@ -1,8 +1,9 @@
-module R2RDF
+require 'open-uri'
+module PubliSci
   class Dataset
-    extend R2RDF::Interactive
+    extend PubliSci::Interactive
 
     def self.for(object, options={}, ask_on_ambiguous=true)
 
       if options == false || options == true
         ask_on_ambiguous = options
@@ -21,22 +22,31 @@
 
           case extension
           when ".RData"
             r_object(object, options, ask_on_ambiguous)
           when /.csv/i
-            R2RDF::Reader::CSV.new.automatic(object,nil,options,ask_on_ambiguous)
+            PubliSci::Reader::CSV.new.automatic(object,nil,options,ask_on_ambiguous)
           end
+        elsif object =~ %r{http[s]://.+}
+          self.for(download(object).path, options, ask_on_ambiguous)
         else
-          raise "Unable to find reader for File or String"
+          raise "Unable to find reader for File or String #{object}"
         end
       elsif object.is_a? Rserve::REXP
         r_object(object, options, ask_on_ambiguous)
       else
         raise "not recognize Ruby objects of this type yet (#{object})"
       end
     end
 
+    def self.download(uri)
+      out = Tempfile.new(uri.split('/').last)
+      out.write open(uri).read
+      out.close
+      out
+    end
+
     def self.r_object(object, options={}, ask_on_ambiguous=true)
       if object.is_a? String
         con = Rserve::Connection.new
         vars = con.eval("load('#{File.absolute_path object}')")
         if vars.to_ruby.size > 1 && ask_on_ambiguous
@@ -47,11 +57,11 @@
         end
 
         r_classes = con.eval("class(#{var})").to_ruby
 
         if r_classes.include? "data.frame"
-          df = R2RDF::Reader::Dataframe.new
+          df = PubliSci::Reader::Dataframe.new
           unless options[:dimensions] || !ask_on_ambiguous
             dims = con.eval("names(#{var})").to_ruby
             puts "Which dimensions? #{dims}"
             selection = gets.chomp
             if selection.size > 0
@@ -68,11 +78,11 @@
           end
 
           df.generate_n3(con.eval(var),var,options)
 
         elsif r_classes.include? "cross"
-          bc = R2RDF::Reader::RCross.new
+          bc = PubliSci::Reader::RCross.new
 
           unless options[:measures] || !ask_on_ambiguous
             pheno_names = con.eval("names(#{var}$pheno)").to_ruby
             puts "Which phenotype traits? #{pheno_names}"
             selection = gets.chomp
@@ -89,11 +99,11 @@
           end
 
           bc.generate_n3(con, var, base, options)
 
         elsif r_classes.include? "matrix"
-          mat = R2RDF::Reader::RMatrix.new
+          mat = PubliSci::Reader::RMatrix.new
 
           unless options[:measures] || !ask_on_ambiguous
             puts "Row label"
             rows = gets.chomp
             rows = "row" unless rows.size > 0
@@ -116,16 +126,16 @@
             base = var unless base.size > 0
           end
 
           mat.generate_n3(con, var, base, options)
         else
-          raise "no R2RDF::Reader found for #{r_classes}"
+          raise "no PubliSci::Reader found for #{r_classes}"
         end
 
       elsif object.is_a? Rserve::REXP
         if object.attr.payload["class"].payload.first
 
-          df = R2RDF::Reader::Dataframe.new
+          df = PubliSci::Reader::Dataframe.new
 
           var = nil
 
           if ask_on_ambiguous
             var = interact("Dataset name?",nil)
\ No newline at end of file