Details
-
Type:
Improvement
-
Status: Won't Fix
-
Resolution: Done
-
Fix Version/s: None
-
Component/s: pipe_base
-
Labels:
-
Story Points:2
-
Epic Link:
-
Sprint:Alert Production F17 - 11
-
Team:Alert Production
Description
At present, the object returned by parseAndRun has a resultList field for any Tasks created by the method, whose metadata field contains the metadata for each instance of the top-level Task. However, there appears to be no way to access the metadata for subtasks from this object; i.e., there is nothing analogous to Task.getFullMetadata.
Would it be possible to modify the output of parseAndRun to return the full metadata?
Attachments
Issue Links
Activity
On inspection, the API for TaskRunner appears badly broken – of the many subclasses of TaskRunner that override TaskRunner.__call__, only four conform to its return spec, while the rest return a simplified Struct, a bare result from run, or None.
Given that the return value cannot be relied on to contain any particular data, modifying the return value will not solve the problem of extracting full metadata from parseAndRun for a generic command-line task. I'm therefore closing the ticket and working around it in DM-12432.
Krzysztof Findeisen
added a comment - - edited On inspection, the API for TaskRunner appears badly broken – of the many subclasses of TaskRunner that override TaskRunner.__call__ , only four conform to its return spec, while the rest return a simplified Struct , a bare result from run , or None .
Given that the return value cannot be relied on to contain any particular data, modifying the return value will not solve the problem of extracting full metadata from parseAndRun for a generic command-line task. I'm therefore closing the ticket and working around it in DM-12432 . 
Claiming this ticket, as it blocks completion of ap_verify.