Clean up EventSaver module order dependancies

[ckierans]

After selecting the strip pairing module in Nuclearizer, the allowable modules are shown below:

This means that someone can select saving an event in L1b/L2 format without running the DepthCalibration module. This needs to be changed in the MModuleSaverMeasurementsFITS.cxx file:

153 void MModuleSaverMeasurementsFITS::ConfigurePreceedingModules()
154 {
155   m_PreceedingModules.clear();
156   m_PreceedingModulesHardRequirement.clear();
157 
158   AddPreceedingModuleType(MAssembly::c_EventLoader);
159   if (m_OutputDataLevel != 0) {
160     AddPreceedingModuleType(MAssembly::c_StripPairing);
161   }
162 }

Also, there is an GUI option "Save events to FITS files (L1b/L2)" and then individual options to "Save events to L1a FITS" and "Save events to L1b FITS" etc, so this is redundant and the XmlTags need to be cleaned up.

As a reminder, this is the data level definition:

• L1a contains the raw telemetered data (a) event time, event ID, event type, number of interactions, detector number, detector side, strip number, pulse height ADC, and arrival time TAC.
• L1b contains data after calibration and event reconstruction
• L2 contains a cleaned version of the event data from Level 1b

The dependancies should be as follows (I think!):

• Save events to L0 binary (DD packets): after either cosima simulation loader + DEE or any measurement loaders, but before any calibration modules.
• Save events to L1a FITS: after either cosima simulation loader + DEE or any measurement loaders, but before any calibration modules.
• Save events to L1b FITS: only after depth calibration. But I think we want this option both before or after Event reconstruction (revan)? The only thing that is modified after reconstruction is the sequence of hit information, so that can be empty in the L1b file if the reconstruction hasn't been performed, I assume?
• Save events to L2 FITS: only after Event reconstruction (revan)