#############################################################################
#                                                                           # 
# SiStripDigi+SiStripClus - digitization and clusterizer packages           #
#                                                                           #
#  version 1.0.0                                                            #
#                                                                           #
# Z. Drasal, Charles University Prague                                      #
#                                                                           #
#############################################################################


 Instructions for building SiStripDigi with CMake:
 -------------------------------------------------

  - Make sure that cmake is in your PATH environment!
  
  - Make sure that you have already installed GEAR, LCIO, MARLIN, ROOT 
    and CMakeModules! If not, you can download them from ilcsoft home page:
    http://ilcsoft.desy.de/portal/software_packages. (Each package has 
    its own howto.)

  - Copy original template file ./CMakeLists.txt.orig to
    ./CMakeLists.txt and edit it if needed (for experts only)  

  - Copy original template file .config/BuildSetup.cmake.orig to file 
    ./config/BuildSetup.cmake, edit it and change paths according to your 
    system setup, i.e. change paths to GEAR, LCIO, MARLIN or use shell
    environment variables (see setup.sh) ... 

    Afterwards go to SiStripDigi directory and type:

    $ mkdir build
    $ cd build
    $ cmake -C ../config/BuildSetup.cmake ..
    $ make install
    
  - Don't forget to set plugins to be loaded by Marlin:
    
    export MARLIN_DLL=$MARLIN_DLL:$SISTRIPDIGI/lib/libSiStripDigi.so
    
  - If you like, use doxygen to generate documentation: 
     
    $ doxygen Doxyfile  
   
 How SiStripDigi works:
 ----------------------
 
  SiStripDigi (a Marlin processor) represents a detailed strip detector digitizer 
  that's mainly intended for detailed simulation studies; but in principle, can 
  be used for full physics studies. The geometry information required for data 
  processing is being accessed on the fly via a geometry interface: SiStripGeom 
  that simply utilizes the information obtained from given GEAR *.xml file. As 
  regards the solid-state physics, the processor naturally implements the following 
  processes for e-h pairs created by a high-energy particle: drift in electric field 
  (simplified just as electric field of a p-n junction - oriented in negative direction 
  of X axis of local reference system, i.e. electrons drift in positive direction X, holes 
  vice versa), diffusion due to multiple collisions and Lorentz shift in magnetic
  field; furthermore, it also takes into account electronics processes: mutual micro-strip 
  crosstalks (dependent on AC/DC coupling), electronics noise, the type of current collected 
  by strips and set if there are floating string in between real read-out strips. As DSSDs 
  are assumed to be digitized, both electrons and holes are collected. Electrons drift to 
  strips oriented in R-Phi direction, holes to strips oriented in Z direction. A user can 
  naturally influence these effects when changing processor parameters: sensor depletion 
  voltage (set by default to 60 V), sensor bias voltage (set by default to 150 V), temperature 
  of a silicon wafer (set by default to 300 K), interstrip capacitance (set by default to 6 pF), 
  strip-to-backplane capacitance (set by default to 0 pF), coupling capacitance (set by default 
  to 120 pF) and sigma of CMS-like noise (set by default to 1200 electrons). Finally, to 
  achieve higher effectivity one can influence the simulation precision using these parameters:
  space precision (set by default to 20 um), relative Lorentz angle precision (set by
  default to 1%) and relative drift time precision (set by default to 1%). As the space
  precision might be different from the one set in Geant 4, the Landau fluctuations
  of deposited energy are simulated here too, using internal fluctuator SiEnergyFluct,
  which exactly follows a Geant4 class called G4UniversalFluctuation. If the particle is
  regarded as low energy (below MIP beta*gamma factor), the energy is distributed
  uniformly using the info from Geant 4, otherwise internal fluctuation is used. As an
  output one obtains a standard LCIO collection of TrackerPulses, so to obtain hits,
  one still must perform clustering, for instance SiStripClus processor can do it ...
  
  For more details: see generated documentation ...
  
 How SiStripClus works:
 ----------------------
 
  SiStripClus (a Marlin processor) provides a cluster finding algorithm based on 
  COG (centre of gravity) method (cluster size is lower than 3) or on head-tail 
  analog method (cluster-size is equal or higher than 3) and thus transforms electric 
  pulses (TrackerPulses - obtained from SiStripDigi processor) into real 2D hits 
  (RPhi + Z) (resp. 3D hits, where the 3rd component is calculated as a center of 
  a sensor). First, a seed strip (which is a strip with signal higher than roughly 
  5 * noise level - set by user) in Z direction is found, than the processor looks 
  for neighbouring strips (with signals higher than roughly 2 * noise level - set 
  by user) and thus finds 1D clusters in Z. The same is performed in R-Phi direction 
  (pitch in R-Phi might differ in forward region based on current Z position); final 
  2D hits are calculated (including "blind" hits, so-called ghosts). All the calculations 
  are performed in local reference system (using information from geometry interface: 
  SiStripGeom) and final 3D hits are transformed back into global ref. system. In 
  order to obtain quasi-realistic covariance matrices for each hit, the
  resolution can be simulated using this processor (use #define ROOT_OUTPUT)
  and then put as input Marlin resolution parameters ...
  
  For more details: see generated documentation ...
   
 What's new?
 -----------
    
  v1.0.0
  ------
  - Added possibility to digitize forward slanted sensors with non-uniform pitch in RPhi (along Z-axiz), Z. Drasal Jul 2009
  
  v1.0.1
  ------
  - Implemented new parameter for cut on time to emulate the integration time, K. Prothmann Dec. 2009
  
  v1.0.2
  ------
  - Added possibility to digitize floating strips (in RPhi or Z) too, Z. Drasal Feb 2010
  
  v1.0.3
  ------
  - Added detailed landau fluctuations in Si based on Geant 4 G4UniversalFluctuation, Z. Drasal Apr 2010
  - Improved clustering procedure, two different algorithms implemented: COG & analog head-tail, Z. Drasal Apr 2010