Set allow_rawbank_same_source_id_only_for_bad_2023_data in write_pass and write_hlt2

Looks like this was just an oversight and would have been needed for lhcb-dpa/prod-requests#56 (comment 8247863).

cc @nskidmor

added RTA label

thank you for this @cburr

/ci-test

added ci-test-triggered label

• [2025-02-02 15:20] Validation started with lhcb-2024-patches-mr#1909

resolved all threads

Just to be sure : is this one really to be merged to 2024-patches ? If yes, then I let @msaur dealing with it. Otherwise, please rebase on master

As the original author of this modification to CombineRawBankViewsToRawEvent/SelectiveCombineRawBankViewsToRawEvent, this was implemented due to a bug in 2023 HLT2 processing and should be relevant only for sprucing (implemented at https://gitlab.cern.ch/lhcb/Moore/-/blob/master/Hlt/Moore/python/Moore/stream_writers.py#L400).

In practice this flag can be applied to HLT1 and HLT2 step as well, but there is no real need for that, as far I am aware.

@cburr Was there any special reason to add it to HLT steps as this is relevant only for data?

I probably just did it for consistency though I have no recollection about this so all I know is the contents of the descriptions.

Unless you are aware of some specific problems, I would propose to close this.

Only place where it could matter would be if we would decide to re-run HLT2 on already processed HLT2 files, but I don't think there are even any thoughts about it or need for that. If that would even happen, we could modify settings and I think that description at allow_rawbank_same_source_id_only_for_bad_2023_data is good enough.

Isn’t write pass used by the sprucing, see: lhcb-dpa/prod-requests#56 (comment 8247863)

mentioned in issue #910 (closed)

mentioned in issue #915 (closed)

mentioned in issue #927 (closed)

mentioned in issue #936