GetRefChild debug pollutes output

[veprbl]

Environment: (where does this bug occur, have you tried other environments)

• Which branch (often main for latest released): main?
• Which version (or HEAD for the most recent on git): HEAD?
• Any specific OS or system where the issue occurs? linux
• Any special versions of ROOT or Geant4? no

Steps to reproduce: (give a step by step account of how to trigger the bug)

1. Run reconstruction, look at the output

or look at logs in CI

1. https://github.com/eic/EICrecon/actions/runs/4170888923/jobs/7220477374

Expected Result: (what do you expect when you execute the steps above)

Remove/hide this output or make it more clear to a regular user.

Actual Result: (what do you get when you execute the steps above)

GetRefChild: Failed to find child with name: HcalEndcapPInsertHits Map contains 25 elements.
   0  B0ECalHits
   0  B0TrackerHits
   0  DIRCBarHits
   0  DRICHHits
   0  EcalBarrelSciGlassHits
   0  EcalEndcapNHits
   0  EcalEndcapPHits
   0  ForwardOffMTrackerHits
   0  ForwardRomanPotHits
   0  HcalBarrelHits
   0  HcalEndcapNHits
   0  HcalEndcapPHits
   0  LumiSpecCALHits
   0  LumiSpecTrackerHits
   0  MPGDBarrelHits
   0  MPGDDIRCHits
   0  MRICHHits
   0  SiBarrelHits
   0  TOFBarrelHits
   0  TOFEndcapHits
   0  TaggerTrackerHits
   0  TrackerEndcapHits
   0  VertexBarrelHits
   0  ZDCEcalHits
   0  ZDCHcalHits

[veprbl]

Somehow grep GetRefChild doesn't return anything for EICrecon

[veprbl]

Apparently it comes from DD4hep
https://github.com/AIDASoft/DD4hep/blob/84f2e45ec8a26efbfb6da07258be895310e652bf/DDCore/src/DetectorImp.cpp#L649

[veprbl]

Filed AIDASoft/DD4hep#1102

But we should still figure out how to not ignore those exceptions in a loop.

[wdconinc]

One related issue could be that we don't throw these runtime errors up to jana2 anymore, https://github.com/eic/EICrecon/blob/2a3f2680d102d8babd7952ca5ed18c7245783b49/src/algorithms/calorimetry/CalorimeterHitDigi.cc. That means jana2 doesn't know to start treating the algorithm as bad and we get GetRefChild errors for every event.

[veprbl]

AIDASoft/DD4hep#1106
cc @wdconinc

[veprbl]

Silencing the excessive output is an immediate issue for the upcoming production. The upstream PR above addresses that. The issue with exception handling in EICrecon would need to be filed and addressed separately.

[veprbl]

Okay, it would seem there will be rebase conflicts due to recent changes. Here is a patch tested with 01-25-01:

diff --git a/DDCore/src/DetectorImp.cpp b/DDCore/src/DetectorImp.cpp
index 12262bb7..c938ad6a 100644
--- a/DDCore/src/DetectorImp.cpp
+++ b/DDCore/src/DetectorImp.cpp
@@ -631,17 +631,22 @@ vector<DetElement> DetectorImp::detectors(const string& type1,
 }
 
 Handle<NamedObject> DetectorImp::getRefChild(const HandleMap& e, const string& name, bool do_throw) const {
-  HandleMap::const_iterator i = e.find(name);
-  if (i != e.end()) {
-    return (*i).second;
+  HandleMap::const_iterator it = e.find(name);
+  if (it != e.end()) {
+    return it->second;
   }
   if (do_throw) {
-    int cnt = 0;
-    cout << "GetRefChild: Failed to find child with name: " << name
-         << " Map contains " << e.size() << " elements." << endl;
-    for(i=e.begin(); i!=e.end(); ++i)
-      cout << "   " << cnt << "  " << (*i).first << endl;
-    throw runtime_error("Cannot find a child with the reference name:" + name);
+    std::stringstream err;
+    err << "getRefChild: Failed to find child with name: " << name
+        << " Map contains " << e.size() << " elements: {";
+    for (it = e.begin(); it != e.end(); ++it) {
+      if (it != e.begin()) {
+        err << ", " << endl;
+      }
+      err << it->first << endl;
+    }
+    err << "}";
+    throw runtime_error(err.str());
   }
   return 0;
 }

[wdconinc]

huh

/usr/bin/patch: **** malformed patch at line 34:  

not sure what's wrong

[veprbl]

Could it be that you are missing a trailing newline?

[wdconinc]

That's what it seems like. I'm going through permutations and rediffing... On my phone...

[wdconinc]

I think this can be closed, right?

[veprbl]

Fixed in DD4hep