Commented out call to Tabular().set_meta() since it is causing concurrent update problems in the database.

This commit is contained in:
Greg Von Kuster
2007-09-06 16:56:19 +00:00
parent 97fda63c73
commit 21423969e5
+7
View File
@@ -223,10 +223,17 @@ class JobWrapper( object ):
is a temporary work-around to ensure columns metadata attribute is set.
This code will be eliminated soon...
"""
"""
gvk: commented out on 9/6/2007 - This code seems to be causing concurrent update problems in the database.
At the current time, tabular datasets are missing the 'column_types' metadata attribute, which is currently
used only for the ColumnListParameter tool parameter. To be safe, I'm calling Tabular().set_meta() there
anyway, so we don't need this. Also, when we run Dan's script again, it will eliminate the need for this.
for dataset_assoc in job.input_datasets:
dataset = dataset_assoc.dataset
if issubclass(type(dataset.datatype), type(self.app.datatypes_registry.get_datatype_by_extension('tabular'))) and dataset.missing_meta():
Tabular().set_meta(dataset)
"""
# Resore input / output data lists
inp_data = dict( [ ( da.name, da.dataset ) for da in job.input_datasets ] )
out_data = dict( [ ( da.name, da.dataset ) for da in job.output_datasets ] )