when running in a condor style, container universe should only be associated with a job if the Pegasus job has a container associated with it

[vahi]

Describe the current behavior

As part of #2063 , users can run workflows using htcondor container universe. In order to enable this, we require the user to specify a condor profile universe that is set to container in their site catalog for the execution site.

https://pegasus.isi.edu/documentation/user-guide/containers.html#configuring-workflows-to-use-htcondor-container-universe

- name: condorpool
  arch: x86_64
  os.type: linux
  directories: []
  profiles:
    condor: {universe: container}
    pegasus: {style: condor, clusters.num: 1}

However, if we run a workflow with jobs that don't have a container associated with them, then condor submission fails with the following error message

ULOG_SUBMIT_FAILURE for HTCondor Node hello_ID0000001 (-1.-1.-1) {07/21/26 23:09:58} failed to create job classad: Submit:-1:container jobs require a container_image or docker_image
07/21/26 23:09:58 Job submit try 4/6 failed, will try again in >= 8 seconds.
07/21/26 23:10:09 Submitting HTCondor Node hello_ID0000001 job(s)...[4:25 PM]

Describe the improvement

Pegasus should set universe = vanilla automatically if a job does not have a container associated with it. Even though from the profiles it gets universe set to container,

Additional context
Add any other context, mockups, or screenshots about the improvement here.