handling PrimaryDataset in TaskChain and StepChain

[vlimant]

not sure if this is a feature or a bug, but it does lead to inconsistent production behaviour, and while there's little chance to have a fix, I still want to leave a trace as "known issue".
Somehow PrimaryDataset is not handled the same way in TaskChain and StepChain. Taking the example of:

https://cmsweb.cern.ch/reqmgr2/fetch?rid=pdmvserv_task_NPS-RunIII2024Summer24wmLHEGS-00157__v1_T_250925_225054_3750

"OutputDatasets": [
    "/NMSSMCascade-HBB_Par-mH-125-mSUSY-1000_TuneCP5_13p6TeV_madgraphMLM-pythia8/RunIII2024Summer24MiniAODv6-150X_mcRun3_2024_realistic_v2-v1/MINIAODSIM",
    "/NMSSMCascade-HBB_Par-mH-125-mSUSY-1000_TuneCP5_13p6TeV_madgraphMLM-pythia8/RunIII2024Summer24NanoAODv15-150X_mcRun3_2024_realistic_v2-v1/NANOAODSIM"
  ],

while it's StepChain clone, with the exact Step<->Task PrimaryDataset assignment

https://cmsweb.cern.ch/reqmgr2/fetch?rid=cmsunified_task_NPS-RunIII2024Summer24wmLHEGS-00157__v1_T_250926_213301_1302

"OutputDatasets": [
    "/NMSSMCascade-HBB_Par-mH-110-mSUSY-1000_TuneCP5_13p6TeV_madgraphMLM-pythia8/RunIII2024Summer24MiniAODv6-150X_mcRun3_2024_realistic_v2-v2/MINIAODSIM",
    "/NMSSMCascade-HBB_Par-mH-110-mSUSY-1000_TuneCP5_13p6TeV_madgraphMLM-pythia8/RunIII2024Summer24NanoAODv15-150X_mcRun3_2024_realistic_v2-v2/NANOAODSIM"
  ],

Note the mH-125 <-> mH-110 difference.

This means that TaskChain does take into account the PrimaryDataset specified in each Task, while StepChain uses only the one specified in the Step1.

In the long run, StepChain should use the PrimaryDataset specified in each Step