From 4f56b82e17e1f20c55de3263ea3bfcdb24e1d5c6 Mon Sep 17 00:00:00 2001
From: Greg Von Kuster <greg@bx.psu.edu>
Date: Fri, 22 Aug 2008 11:22:44 -0400
Subject: [PATCH] Fix for dataset_metadata_in_file validator: all locally
 cached data files that are columnar must have columns separated by tabs
 characters, not white space.  I've eliminated the 'split' param from the
 validator as it was causing problems.  Extract genomic DNA tool should now
 finally work correctly.

---
 lib/galaxy/tools/parameters/validation.py | 7 +++----
 tools/extract/extract_genomic_dna.xml     | 2 +-
 2 files changed, 4 insertions(+), 5 deletions(-)

diff --git a/lib/galaxy/tools/parameters/validation.py b/lib/galaxy/tools/parameters/validation.py
index 3a43a8c47f1..81c7cca5776 100644
--- a/lib/galaxy/tools/parameters/validation.py
+++ b/lib/galaxy/tools/parameters/validation.py
@@ -235,18 +235,17 @@ class MetadataInFileColumnValidator( Validator ):
             metadata_name = metadata_name.strip()
         metadata_column = int( elem.get( "metadata_column", 0 ) )
         message = elem.get( "message", "Value for metadata %s was not found in %s." % ( metadata_name, filename ) )
-        split = elem.get( "split", None )
         line_startswith = elem.get( "line_startswith", None  )
         if line_startswith:
             line_startswith = line_startswith.strip()
-        return cls( filename, metadata_name, metadata_column, message, split, line_startswith )
-    def __init__( self, filename, metadata_name, metadata_column, message="Value for metadata not found.", split=None, line_startswith=None ):
+        return cls( filename, metadata_name, metadata_column, message, line_startswith )
+    def __init__( self, filename, metadata_name, metadata_column, message="Value for metadata not found.", line_startswith=None ):
         self.metadata_name = metadata_name
         self.message = message
         self.valid_values = []
         for line in open( filename ):
             if line_startswith is None or line.startswith( line_startswith ):
-                fields = line.split( split )
+                fields = line.split( '\t' )
                 if metadata_column < len( fields ):
                     self.valid_values.append( fields[metadata_column] )
     def validate( self, value, history = None ):
diff --git a/tools/extract/extract_genomic_dna.xml b/tools/extract/extract_genomic_dna.xml
index 1f8f1a47d2f..f30e18de90b 100644
--- a/tools/extract/extract_genomic_dna.xml
+++ b/tools/extract/extract_genomic_dna.xml
@@ -4,7 +4,7 @@
   <inputs>
     <param format="interval" name="input" type="data" label="Fetch sequences corresponding to Query">
       <validator type="unspecified_build" />
-      <validator type="dataset_metadata_in_file" filename="alignseq.loc" metadata_name="dbkey" metadata_column="1" message="Sequences are not currently available for the specified build." split="\t" line_startswith="seq" />
+      <validator type="dataset_metadata_in_file" filename="alignseq.loc" metadata_name="dbkey" metadata_column="1" message="Sequences are not currently available for the specified build." line_startswith="seq" />
     </param>
 	<param name="out_format" type="select" label="Output data type">
 	<option value="fasta">FASTA</option>