need to enter the output event file before creating and writing

[tomeichlersmith]

ROOT requires us to be in the correct directory when we create
the output TTree for us to have it written into the correct
location.

In order to allow downstream processors to getHistoDirectory()
and enter that directory for the histograms we need to enter
the output event file again here so that our run tree ends up
in the correct location.

[tomeichlersmith]

Alright, I added a fake analyzer to the framework-testbench that just calls getHistoDirectory() during processing and a config that can test using this analyzer both in a production run and a recon run.

I can confirm the issues observed in LDMX-Software/ldmx-sw#1255 and LDMX-Software/ldmx-sw#1256 by running this analyzer with Framework v1.4.1 (current trunk).

# inside the denv for framework-testbench
tom@framework-testbench:~$ fire config/histogramming.py 
---- LDMXSW: Loading configuration --------
---- LDMXSW: Configuration load complete  --------
---- LDMXSW: Starting event processing --------
 [ Process ] 1 : Processing 1 Run 1 Event 1  (2024-01-30 15:33:34.436424000+0000)
 [ Process ] 1 : Processing 2 Run 1 Event 2  (2024-01-30 15:33:34.461226000+0000)
 [ Process ] 1 : Processing 3 Run 1 Event 3  (2024-01-30 15:33:34.461311001+0000)
 [ Process ] 1 : RunHeader { run: 1, numTries: 3, detectorName: , description: 
  intParameters: 
    RandomNumberMasterSeed[hist] = 1
  floatParameters: 
  stringParameters: 
}
---- LDMXSW: Event processing complete  --------
tom@framework-testbench:~$ rootls test_processormove_*.root:*
test_processormove_events.root :
  LDMX_Events
test_processormove_hist.root :
  LDMX_Run  LDMX_Run

Re-compiling with this branch and re-running this config resolves this issue:

tom@framework-testbench:~$ fire config/histogramming.py 
---- LDMXSW: Loading configuration --------
---- LDMXSW: Configuration load complete  --------
---- LDMXSW: Starting event processing --------
 [ Process ] 1 : Processing 1 Run 1 Event 1  (2024-01-30 15:34:52.268631000+0000)
 [ Process ] 1 : Processing 2 Run 1 Event 2  (2024-01-30 15:34:52.286868000+0000)
 [ Process ] 1 : Processing 3 Run 1 Event 3  (2024-01-30 15:34:52.286953001+0000)
 [ Process ] 1 : RunHeader { run: 1, numTries: 3, detectorName: , description: 
  intParameters: 
    RandomNumberMasterSeed[hist] = 1
  floatParameters: 
  stringParameters: 
}
---- LDMXSW: Event processing complete  --------
tom@framework-testbench:~$ rootls test_processormove_*.root:*
test_processormove_events.root :
  LDMX_Events  LDMX_Run
test_processormove_hist.root :
tom@framework-testbench:~$ fire config/produce.py 
---- LDMXSW: Loading configuration --------
---- LDMXSW: Configuration load complete  --------
---- LDMXSW: Starting event processing --------
 [ Process ] 1 : Processing 1 Run 1 Event 1  (2024-01-30 15:36:26.325298000+0000)
 [ Process ] 1 : Processing 2 Run 1 Event 2  (2024-01-30 15:36:26.341166000+0000)
 [ Process ] 1 : Processing 3 Run 1 Event 3  (2024-01-30 15:36:26.341330001+0000)
 [ Process ] 1 : RunHeader { run: 1, numTries: 3, detectorName: , description: 
  intParameters: 
    RandomNumberMasterSeed[prod] = 1
  floatParameters: 
  stringParameters: 
}
---- LDMXSW: Event processing complete  --------
tom@framework-testbench:~$ fire config/histogramming.py test_produce.root 
---- LDMXSW: Loading configuration --------
---- LDMXSW: Configuration load complete  --------
---- LDMXSW: Starting event processing --------
 [ Process ] 1 : Opening file test_produce.root
 [ Process ] 1 : Got new run header from 'test_processormove_events.root' ...
RunHeader { run: 1, numTries: 3, detectorName: , description: 
  intParameters: 
    RandomNumberMasterSeed[prod] = 1
  floatParameters: 
  stringParameters: 
}
 [ Process ] 1 : Processing 1 Run 1 Event 1  (2024-01-30 15:36:32.621931000+0000)
 [ Process ] 1 : Processing 2 Run 1 Event 2  (2024-01-30 15:36:32.624864000+0000)
 [ Process ] 1 : Processing 3 Run 1 Event 3  (2024-01-30 15:36:32.624957000+0000)
 [ Process ] 1 : Closing file test_produce.root
---- LDMXSW: Event processing complete  --------
tom@framework-testbench:~$ rootls test_processormove_*.root:*
test_processormove_events.root :
  LDMX_Events  LDMX_Run
test_processormove_hist.root :
tom@framework-testbench:~$ rootls test_produce.root:*
LDMX_Events  LDMX_Run

[tomeichlersmith]

Alright, I've checked that the different configs in framework-testbench are functional and this doesn't introduce a regression of any kind.

I also checked to see if we could move the writeRunTree call to within the destructor and the answer is No. Because we are just holding an address to a RunHeader that is owned by Process during a production run, there can be a seg vio at the end of running if the RunHeader is de-allocated before the EventFile is. It is just easier to call writeRunTree when we are cleanly ending a run (production or recon).