[PIG-2680] TOBAG output schema reporting

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 0.10.0
• Fix Version/s: 0.11, 0.10.1
• Component/s: internal-udfs
• Patch Info: Patch Available
• Assignee: Unassigned
• Reporter: Andy Schlaikjer
• Created: 03/May/12 05:09
• Updated: 06/Jan/13 23:56
• Resolved: 08/May/12 22:52

Description

TOBAG only reports an output schema beyond {{

{(NULL)}

}} when all input field schemas match deeply, including field schema aliases. This seems wrong to me. Shouldn't it just require recursive type equality?

For relevant code, see:

http://svn.apache.org/viewvc/pig/tags/release-0.9.2/src/org/apache/pig/builtin/TOBAG.java?view=markup#l142

    private boolean nullEquals(Schema currentSchema, Schema newSchema) {
        if(currentSchema == null){
            if(newSchema != null){
                return false;
            }
            return true;
        }
        return currentSchema.equals(newSchema);
    }

The included patch modifies the return line to use Schema.equals(currentSchema, newSchema, false, true) to avoid alias matching requirement.