More messenger commands for physics lists

[tdixon97]

I need to test it

[tdixon97]

Any idea why my new macro command isn't picked up by rmg-doc-dump? I did run it

[ManuelHu]

That would be a very curious case, as the CI also just runs make remage-doc-dump... (just running remage-doc-dump without make is not enough)

[ManuelHu]

there is also a way to have parameters with units, that would be nicer. Or, at least mention the unit this is in. Currently this just renders as "default value - 1" without any unit information in the docs.

[tdixon97]

I am working on it, but I had to change this into two macro commands

[tdixon97]

Now this is done

[ManuelHu]

You don't have to split it in two commands... I have some example code how to use G4UIcmdWithADoubleAndUnit with extra parameters:

diff --git a/include/RMGHardwareMessenger.hh b/include/RMGHardwareMessenger.hh
index 8ee2873..bcceed3 100644
--- a/include/RMGHardwareMessenger.hh
+++ b/include/RMGHardwareMessenger.hh
@@ -17,6 +17,7 @@
 #define _RMG_HARDWARE_MESSENGER_HH_
 
 #include "G4UImessenger.hh"
+#include "G4UIcmdWithADoubleAndUnit.hh"
 
 class RMGHardware;
 class RMGHardwareMessenger : public G4UImessenger {
@@ -31,7 +32,7 @@ class RMGHardwareMessenger : public G4UImessenger {
   private:
 
     RMGHardware* fHardware;
-    G4UIcommand* fRegisterCmd;
+    G4UIcmdWithADoubleAndUnit* fRegisterCmd;
 
     void RegisterDetectorCmd(const std::string& parameters);
 };
diff --git a/src/RMGHardwareMessenger.cc b/src/RMGHardwareMessenger.cc
index aa200af..1a4bdd8 100644
--- a/src/RMGHardwareMessenger.cc
+++ b/src/RMGHardwareMessenger.cc
@@ -17,14 +17,19 @@
 
 #include "G4Tokenizer.hh"
 #include "G4UIcommand.hh"
+#include "G4UIcmdWithADoubleAndUnit.hh"
 
 #include "RMGHardware.hh"
 #include "RMGTools.hh"
 
 RMGHardwareMessenger::RMGHardwareMessenger(RMGHardware* hw) : fHardware(hw) {
-  fRegisterCmd = new G4UIcommand("/RMG/Geometry/RegisterDetector", this);
+  fRegisterCmd = new G4UIcmdWithADoubleAndUnit("/RMG/Geometry/RegisterDetector", this);
   fRegisterCmd->SetGuidance("register a sensitive detector");
 
+  fRegisterCmd->SetParameterName("test", false);
+  fRegisterCmd->SetDefaultValue(5);
+  fRegisterCmd->SetUnitCategory("Length");
+
   auto p_type = new G4UIparameter("type", 's', false);
   p_type->SetParameterCandidates(RMGTools::GetCandidates<RMGDetectorType>().c_str());
   p_type->SetGuidance("Detector type");
@@ -59,9 +64,15 @@ void RMGHardwareMessenger::SetNewValue(G4UIcommand* command, G4String newValues)
 }
 
 void RMGHardwareMessenger::RegisterDetectorCmd(const std::string& parameters) {
+  RMGLog::Out(RMGLog::error, "params ", parameters);
   G4Tokenizer next(parameters);
 
+  auto x = next();
+  auto y = next();
+  auto num = G4UIcmdWithADoubleAndUnit::GetNewDoubleValue((x + " " + y).c_str());
+
   auto type_str = next();
+  RMGLog::Out(RMGLog::fatal, " num ", num);
   auto type = RMGTools::ToEnum<RMGDetectorType>(std::string(type_str), "detector type");
   auto pv_name = next();
   const int uid = std::stoi(next());

[tdixon97]

The step limiting seems to be working, we could have a nicer interface (eg. specifying with regex) but I am not sure since I doubt its something we will use.

Here is an example with 10 um step limits. Of course now the simulation is extremely slow

[gipert]

Seems good to me.

[tdixon97]

I would wait until its more tested to merge.