AGC cannot handle talking to a WES server that does not encode AWS Batch job IDs into WES task names with a pipe | separator

[adamnovak]

As noted here:

amazon-genomics-cli/packages/wes_adapter/amazon_genomics/wes/adapters/BatchAdapter.py

Lines 249 to 250 in 2b93a2c

	# The CLI expects a pipe-delimited string of name + string in the response
	task_name = f"{job_details['jobName']}|{job_details['jobId']}"

AGC needs to see a | and then an AWS Batch job ID at the end of a task name reported from a WES server it is talking to. This is an AGC-specific WES extension; if AGC is put in touch with a merely standard WES server which reports task names and those names do not follow this format, AGC will crash.

AGC probably ought to learn to handle task names that do not follow this format. A WES server might want to report tasks that have not yet launched, or which were run under the same AWS Batch job as another task, or which were handled entirely by the engine without running a Batch job.