Update runDataProcessing.cxx to allow for - and _ in (Process?)Configurable<bool> names

[cholmcc]

No description provided.

[github-actions]

REQUEST FOR PRODUCTION RELEASES:
To request your PR to be included in production software, please add the corresponding labels called "async-" to your PR. Add the labels directly (if you have the permissions) or add a comment of the form (note that labels are separated by a ",")

+async-label <label1>, <label2>, !<label3> ...

This will add <label1> and <label2> and removes <label3>.

The following labels are available
async-2023-pbpb-apass
async-2023-pp-apass1
async-2022-pp-apass6
async-2022-pp-apass4
async-mc
async-data

[cholmcc]

build/O2/o2-dataflow-cs8 failure seems unrelated.

[alibuild]

Error while checking build/O2/fullCI for 7f4bcf3 at 2024-02-29 07:25:

## sw/BUILD/O2-latest/log
c++: error: unrecognized command-line option '--rtlib=compiler-rt'
c++: error: unrecognized command-line option '--rtlib=compiler-rt'


## sw/BUILD/O2-full-system-test-latest/log
task timeout reached .. killing all processes


## sw/BUILD/o2checkcode-latest/log
--
========== List of errors found ==========
++ GRERR=0
++ grep -v clang-diagnostic-error error-log.txt
++ grep ' error:'
++ GRERR=1
++ [[ 1 == 0 ]]
++ mkdir -p /sw/INSTALLROOT/69888d193c896a6906a834795f31d2bdeb4133e9/slc8_x86-64/o2checkcode/1.0-local138/etc/modulefiles
++ cat
--

Full log here.

[ktf]

@aalkin I think you were saying privately that this happens because we do have some convention on the process options which is not respected? Can you comment here?

[aalkin]

@ktf indeed, this code is only applied to process switches, and by default those inherit their names after the process function which cannot contain dashes due to them being C++ identifiers. However, this is not a strict requirement as long as the names are OK to be put in JSON.

@cholmcc could you please change the title to say ProcessConfigurable ?

[jgrosseo]

I am not against this PR but I don't understand why you, Christian, use instead of normal configurables the ones of process switches?

[cholmcc]

@ktf indeed, this code is only applied to process switches, and by default those inherit their names after the process function which cannot contain dashes due to them being C++ identifiers. However, this is not a strict requirement as long as the names are OK to be put in JSON.

I'm afraid that the above is not entirely correct. The code in question is also called for regular switches as far as I can tell.

Take the O2Physics:PWGMM/Rivet/Tasks/rivet.cxx workflow, which among other things defines regular (not process) boolean switches like hepmc-recenter. The code in question is called on these options too. That may not be the intention, but it is what happens, I believe.

@cholmcc could you please change the title to say ProcessConfigurable ?

I'm not sure it only applies to ProcessConfigurable only - but will change to something similar.

I am not against this PR but I don't understand why you, Christian, use instead of normal configurables the ones of process switches?

I don't. i use one ProcessConfigurable in O2Physics:PWGMM/Rivet/Tasks/rivet.cxx called hepmc-no-aux and that is it. This is defined not via the relevant pre-processor macro exactly to allow a name scoping to the "family" of "hepmc" (i.e., "hepmc-no-aux" instead of "hepmcNoAux" in line with f.ex. "hepmc-recenter").

Yours,

Christian

[cholmcc]

BTW, if the fundamental requirement is that option name can be put into a JSON key, then perhaps the regular expression for the option should capture that requirement - perhaps

([^"']+)

That is, anything not a quote or double quote, since JSON keys can be any UTF-8 character string - including white-space and so on. One could even allow for '" if they are properly escaped.

Just a though.

Yours,
Christian

[aalkin]

The code in question is also called for regular switches as far as I can tell.

@cholmcc absolutely not, this is specifically made to extract configuration parameter name from the string metadata attached to inputs, that is placed there based on process switches only. You can look at O2Physics codebase and find plenty of configurables with dashes and underscores.

[github-actions]

This PR did not have any update in the last 30 days. Is it still needed? Unless further action in will be closed in 5 days.