CMS: scripts to prepare the derived data records

[katilp]

CMS has now some quite extensive and systematically produced derived datasets under

/eos/opendata/cms/derived-data/POET/23-Jul-22/
/eos/opendata/cms/derived-data/NanoAODRun1/01-Jul-22/

produced, respectively with

• https://github.com/cms-opendata-analyses/PhysObjectExtractorTool/tree/odws2022-ttbaljets-prod
• https://github.com/cms-opendata-analyses/NanoAODRun1ProducerTool

These were used at the CMS OD workshop lessons

• https://cms-opendata-workshop.github.io/workshop2022-lesson-ttbarljetsanalysis/
• https://cms-opendata-workshop.github.io/workshop2022-lesson-run2-adlcl/
• https://cms-opendata-workshop.github.io/workshop2022-lesson-run1example/

New CODP records can now be built for these derived datasets.

To do

• review if the current metadata fields are adequate, see e.g. http://opendata.cern.ch/record/12353 or https://github.com/cernopendata/opendata.cern.ch/blob/master/cernopendata/modules/fixtures/data/records/cms-derived-nanoaod-outreach-higgstautau.json
  • dataset semantics? ideally as in the standard automated doc pages such as this for a NanoAOD version or the ones prepared for NanoAODRun1 but that's not yet done for POET afaik
• write scripts to prepare the CODP records for them in a new cms-YYYY-derived-data directory under https://github.com/cernopendata/data-curation
• run the script to produce new cms-derived-data-poet.json and cms-derived-data-nanoaodrun1.json and make a PR to the qa branch of https://github.com/cernopendata/opendata.cern.ch