I overlooked the fact that this doesn't work for running over datasets, since it instrinsically opens input files to grab the metadata. This is especially problematic when submitting prun jobs, since it tries to grab the metadata before it submits the prun job, which sort of defeats the functionality of automating this metadata reading during runtime. Would you have any ideas on how to make this work?
Hello,
I implemented a patch to read metadata via athena based configs while running over files locally - https://github.com/UCATLAS/xAODAnaHelpers/commit/99174eb02c58d52bbb1fb3c1e62ef407b9172a58
I overlooked the fact that this doesn't work for running over datasets, since it instrinsically opens input files to grab the metadata. This is especially problematic when submitting prun jobs, since it tries to grab the metadata before it submits the prun job, which sort of defeats the functionality of automating this metadata reading during runtime. Would you have any ideas on how to make this work?
Thanks for the help! Sagar