Allow setting dataset attributes of additional primary datasets through galaxy.json.

lib/galaxy/tools/__init__.py

@@ -2623,10 +2623,19 @@ class Tool:
         Find any additional datasets generated by a tool and attach (for 
         cases where number of outputs is not known in advance).
         """
-        primary_datasets = {}
+        new_primary_datasets = {}
+        try:
+            json_file = open( os.path.join( job_working_directory, jobs.TOOL_PROVIDED_JOB_METADATA_FILE ), 'r' )
+            for line in json_file:
+                line = simplejson.loads( line )
+                if line.get( 'type' ) == 'new_primary_dataset':
+                    new_primary_datasets[ os.path.split( line.get( 'filename' ) )[-1] ] = line
+        except Exception, e:
+            log.debug( "Error opening galaxy.json file: %s" % e )
         # Loop through output file names, looking for generated primary 
         # datasets in form of:
         #     'primary_associatedWithDatasetID_designation_visibility_extension(_DBKEY)'
+        primary_datasets = {}
         for name, outdata in output.items():
             filenames = []
             if 'new_file_path' in self.app.config.collect_outputs_from:
@@ -2664,8 +2673,6 @@ class Tool:
                 primary_data.info = outdata.info
                 primary_data.init_meta( copy_from=outdata )
                 primary_data.dbkey = dbkey
-                primary_data.set_meta()
-                primary_data.set_peek()
                 # Associate new dataset with job
                 job = None
                 for assoc in outdata.creating_job_associations:
@@ -2677,6 +2684,15 @@ class Tool:
                     self.sa_session.add( assoc )
                     self.sa_session.flush()
                 primary_data.state = outdata.state
+                #add tool/metadata provided information
+                new_primary_datasets_attributes = new_primary_datasets.get( os.path.split( filename )[-1] )
+                if new_primary_datasets_attributes:
+                    dataset_att_by_name = dict( ext='extension' )
+                    for att_set in [ 'name', 'info', 'ext', 'dbkey' ]:
+                        dataset_att_name = dataset_att_by_name.get( att_set, att_set )
+                        setattr( primary_data, dataset_att_name, new_primary_datasets_attributes.get( att_set, getattr( primary_data, dataset_att_name ) ) )           
+                primary_data.set_meta()
+                primary_data.set_peek()
                 self.sa_session.add( primary_data )
                 self.sa_session.flush()
                 outdata.history.add_dataset( primary_data )