New shell plugin ncbirest.rb to provide efetch etc.

 * New shell plugin lib/bio/shell/plugin/ncbirest.rb, providing
   "efetch", "einfo", "esearch", and "esearch_count" methods.
   They act the same as those defined in Bio::NCBI::REST, except
   that efetch automatically selects databases if no options
   are given and only 1 id is given.
 * A shell command "getobj" is changed to use "efetch" method
   when "gb" or some variant is specified as the database.

lib/bio/shell.rb

@@ -5,7 +5,7 @@
 #               Toshiaki Katayama <k@bioruby.org>
 # License::     The Ruby License
 #
-# $Id: shell.rb,v 1.20 2007/07/09 11:17:09 k Exp $
+# $Id:$
 #
 
 require 'bio'
@@ -36,6 +36,7 @@ module Bio::Shell
   require 'bio/shell/plugin/emboss'
   require 'bio/shell/plugin/blast'
   require 'bio/shell/plugin/psort'
+  require 'bio/shell/plugin/ncbirest'
 
   extend Ghost
 

lib/bio/shell/plugin/entry.rb

@@ -5,7 +5,7 @@
 #               Toshiaki Katayama <k@bioruby.org>
 # License::     The Ruby License
 #
-# $Id: entry.rb,v 1.10 2007/04/05 23:35:41 trevor Exp $
+# $Id:$
 #
 
 module Bio::Shell
@@ -88,8 +88,14 @@ def getent(arg)
 
       # KEGG API at http://www.genome.jp/kegg/soap/
       else
-        puts "Retrieving entry from KEGG API (#{arg})"
-        entry = bget(arg)
+        case db.to_s.downcase
+        when 'genbank', 'gb', 'embl', 'emb', 'ddbj', 'dbj'
+          puts "Retrieving entry from NCBI eUtils"
+          entry = efetch(entry_id)
+        else
+          puts "Retrieving entry from KEGG API (#{arg})"
+          entry = bget(arg)
+        end
       end
     end
 

lib/bio/shell/plugin/ncbirest.rb

@@ -0,0 +1,66 @@
+#
+# = bio/shell/plugin/ncbirest.rb - plugin for NCBI eUtils
+#
+# Copyright::   Copyright (C) 2009
+#               Naohisa Goto <ng@bioruby.org>
+# License::     The Ruby License
+#
+# $Id:$
+#
+
+module Bio::Shell
+
+  private
+
+  # NCBI eUtils EFetch service.
+  # When two or more arguments are given, or multiple accession numbers
+  # are given it acts the same as Bio::NCBI::REST.efetch.
+  # Otherwise, assumes nucleotide or protein accessin is given, and
+  # automatically tries several databases.
+  def efetch(ids, *arg)
+    if !arg.empty? or ids.kind_of?(Array) or /\,/ =~ ids then
+      return Bio::NCBI::REST.efetch(ids, *arg)
+    end
+
+    rettype = 'gb'
+    prot_dbs = [ 'protein' ]
+    nucl_dbs = [ 'nuccore', 'nucleotide', 'nucgss', 'nucest' ]
+
+    case ids
+    when /\A[A-Z][A-Z][A-Z][0-9]+(\.[0-9]+)?\z/i,
+      /\A[OPQ][A-Z0-9]+(\.[0-9]+)?\z/i
+      # protein accession
+      dbs = prot_dbs
+    when /\A[0-9]+\z/, /\A[A-Z0-9]+\_[A-Z0-9]+\z/i
+      # NCBI GI or UniProt accession (with fail-safe)
+      dbs = prot_dbs + nucl_dbs
+    else
+      # nucleotide accession
+      dbs = nucl_dbs
+    end
+    result = nil
+    dbs.each do |db|
+      hash = { 'db' => db, 'rettype' => 'gb' }
+      result = Bio::NCBI::REST.efetch(ids, hash)
+      break if result and !result.empty?
+    end
+    result
+  end
+
+  # NCBI eUtils EInfo
+  def einfo
+    Bio::NCBI::REST.einfo
+  end
+
+  # NCBI eUtils ESearch
+  def esearch(str, *arg)
+    Bio::NCBI::REST.esearch(str, *arg)
+  end
+
+  # Same as Bio::NCBI::REST.esearch_count
+  def esearch_count(str, *arg)
+    Bio::NCBI::REST.esearch_count(str, *arg)
+  end
+
+end
+