Mass Doxygen comment fixes

See merge request !539

LArCalibTools: Fix gcc6 inconsistent indentation warnings.

See merge request !550

JetUncertainties: Fix gcc6 warning and factor superfluous FP operations

See merge request !551

HadronicRecoilData: Comply with ATLAS naming conventions.

See merge request !552

TileRec: Fix references to deprecated gaudi methods.

See merge request !553

GeoSpecialShapes: gcc7 warnings

See merge request !554

StorageSvc: Fix ubsan warning.

See merge request !524

Fix linking of TrigT1Monitoring in dbg build.

See merge request !525

Avoid gcc7 switch fallthrough warnings.

ntupleService -> ntupleSvc.

Local variables should not start with m_.

Factor multiple divisions to a single one.
Remove FP multiplications by 1.

gcc6 warning: Misleading indentation.

gcc6 warnings: Inconsistent indentation.

Fix linking of TrigT1CaloTools in dbg build.

See merge request !528

  Fixed doxygen @class capitalization
  Fixed doxygen @warning capitalization
  Fixed doxygen parem typo
  Fixed doxygen details typo
  Fixed doxygen formula syntax
  misc doxygen changes

FTK_RecTools: Fix linking in dbg build

See merge request !529

PanTauAlgs: Various fixes

See merge request !534

Private data members should start with m_, and other identifiers should not.

clang warnings: unused members.

Generic WorkDir project creation (2017.03.22.)

See merge request !510

Fix linking of TrigT1CaloCalibTools in dbg build.

See merge request !527

TrigL2LongLivedParticles: Comply with ATLAS naming conventions.

See merge request !530

Printout fixes when using the Athena project (2017.03.22.)

See merge request !507

tauRecTools: Various fixes

See merge request !504

JetCalibTools: Various minor fixes

See merge request !502

Local variables should not start with m_.

Use IIBLParameterSvc instead of IBLParameterSvc to avoid a link dependency
on the (component-only) library PixelGeoModel.

TrigT1CaloTools directly references L1CaloCondSvc from TrigT1CaloCondSvc,
which is a class with out-of-line virtual functions.  Thus, it should
link against the library where these functions are defined in order
to see the typeinfo (which it needs) and vtable (which it doesn't
really need).  However, TrigT1CaloCondSvc was built only as a component
library.

In a previous MR, TrigT1CaloCondSvc was changed to dual use library.
This changes TrigT1CalibTools to link against it, resolving the error.

A better solution would be to rework L1CaloCondSvc in terms of an abstract
interface, so that linking against TrigT1CaloCondSvc would not be
necessary.  However, L1CaloCondSvc has non-trivial templated functions
accessing the class internals, so this is not trivial.
Leaving that for someone else to enjoy doing...

TrigT1CaloCalibTools directly references L1CaloCondSvc from TrigT1CaloCondSvc,
which is a class with out-of-line virtual functions.  Thus, it should
link against the library where these functions are defined in order
to see the typeinfo (which it needs) and vtable (which it doesn't
really need).  However, TrigT1CaloCondSvc was built only as a component
library.

In a previous MR, TrigT1CaloCondSvc was changed to dual use library.
This changes TrigT1CaloCalibTools to link against it, resolving the error.

A better solution would be to rework L1CaloCondSvc in terms of an abstract
interface, so that linking against TrigT1CaloCondSvc would not be
necessary.  However, L1CaloCondSvc has non-trivial templated functions
accessing the class internals, so this is not trivial.
Leaving that for someone else to enjoy doing...

TrigT1Monitoring directly references L1CaloCondSvc from TrigT1CaloCondSvc,
which is a class with out-of-line virtual functions.  Thus, it should
link against the library where these functions are defined in order
to see the typeinfo (which it needs) and vtable (which it doesn't
really need).  However, TrigT1CaloCondSvc was built only as a component
library.

In a previous MR, TrigT1CaloCondSvc was changed to a dual use library.
We now link to this library to resolve this problem.    

A better solution would be to rework L1CaloCondSvc in terms of an abstract
interface, so that linking against TrigT1CaloCondSvc would not be
necessary.  However, L1CaloCondSvc has non-trivial templated functions
accessing the class internals, so this is not trivial.
Leaving that for someone else to enjoy doing...

Don't cross-cast unless it is known that the dynamic type is correct.

So that the same code could possibly be cherry-picked back into 21.0 without
having to set up a separate MR for that.

Migrate BeamEffectsAlg to use HandleKeys rather than Handles as properties.

See merge request !496

egammaMVACalib: Various fixes

See merge request !505

CaloTPCnv: Adjust use of DataPool to work properly with locking.

See merge request !500

add GoogleTestTools to AtlasTest

See merge request !501

ElectronPhotonSelectorTools: Comply with ATLAS naming conventions.

See merge request !503

deal with performing a sparse build on top of any base project. So far
only tested with Athena and AnalysisBase though.