Truth-match objects beyond the standard HLT2 reconstruction in Tesla

See merge request !246

(cherry picked from commit 89ed5446)

86cfcf02 Truth match additional Track and ProtoParticle containers.
5d0b2b2c Hard-code the list of extra locations.
1b95995a Clarify NeverFail logic.

Add script to check TCK-TurboStreamProd compatibility

See merge request !238

(cherry picked from commit 00fd8dcd)

c2f82f73 Add script to check TCK-TurboStreamProd compatibility

Execute legacy behaviour in Tesla only for legacy data types

See merge request !198

(cherry picked from commit bea25013)

b01d0f7f Execute legacy behaviour only for legacy data types.

Fix some Tesla tests

See merge request !201

(cherry picked from commit e0eb7de1)

da1180cf Delete old tests.
1230deda Kill input Turbo data when re-running Tesla.
42b92bce Always set RootInTES for Turbo DaVinci jobs.
758fe257 Query DecReports rather than SelReports.

Fix python error when using the DaVinci configurable with the Upgrade DataType and no specific list of detectors.

Support 2018 Turbo data processing

See merge request !192

(cherry picked from commit aa02ef5b)

fdb472f9 Support 2018 Turbo data processing.

This reverts merge request !186

This reverts merge request !180

Copy Particle to RelatedInfo maps in Tesla.

See merge request !179

(cherry picked from commit 5e437277)

bfd6c1ad Copy Particle to RelatedInfo maps.

Fixed Tesla tests for latest version of DaVinci configurable

See merge request !162

(cherry picked from commit 1fff4e3a)

86ff5887 Fixed Tesla tests for latest version of DaVinci configurable.
4b88a734 Updates of references for common particles and tagging tests in DaVinciTests.

Fix StrippingCache for 2018-patches

See merge request !159

(cherry picked from commit 9a54f9bc)

55e0f88a Fix StrippingCache by switching it to build S29r2; update stripping test to run S29r2.

DaVinci:minor adjustment for Turbo MC processing

See merge request !155

(cherry picked from commit 0dd24b73)

6f35fe49 DaVinci:
f0f62ef0 A bit more doc in DaVinci's Configuration.py.
d30e22ba  remove commented line and  suppress fake warning message

Propagate Simulation to TurboConf

See merge request !153

(cherry picked from commit f80e7c42)

ab32f1b2 Propagate Simulation to TurboConf.

Trying to get TurboSP and MC with truthmatching to work

See merge request !101

(cherry picked from commit bf52f8bf)

87e70041 First commit of Configuration file, trying to get TurboSP MC to run with truthmatching
8e646c9a fix some typos
6fc9415d Commit latest try for neutral association in TurboSP
d6b201cf Add working neutral-matching logic, tidy up.
67d09999 Working version.
c4797d9e Remove temporary comments.

Make CaloDigit cloning behaviour consistent in Tesla

See merge request !139

(cherry picked from commit ac98d6c8)

d29cbdb4 Make CaloDigit cloning behaviour consistent, and also always-on for PersistReco.
fb4d7326 Add missing imports, fix typo.

- replaced `DECLARE_*_FACTORY` with `DECLARE_COMPONENT`
- made constructors public
- removed some unnecessary _disabling_ of copy constructors
- removed some trivial destructors

See gaudi/Gaudi!420

 - Set "CondDB ( Upgdade = True)"  for DataType='Upgrade'

base them on the needed data rather than on the decoder name itself and use decoderToLocation

This reverts commit 07f4e043.

The major differences are:

* Turbo processing generally requires `DaVinci().Turbo = True`;
* The packers now require the locations of packed containers that should
always be created to be explicitly specified, and emit a warning if
these are not specified.

The objects are: VeloCluster, STCluster (for the TT and IT), CaloDigit,
and CaloAdc.

By declaring the root library as a system one in the CMakeLists.txt

This places Rec/Summary in stream-dependent location, which is useful
for TurCal productions where the Brunel summary already exists at
/Event/Rec/Summary, so the Moore summary must go somewhere else.

Addresses !78 (comment 295430).

Occasionally, HLT2 did not write `Hlt2/SelReports` for very busy
events. Previously, the `TeslaReportAlgo` was skipping such events
happily. Here we restore this behaviour, such that if
`Hlt2/SelReports` cannot be decoded, no event is written out.

Ideally, this would already be done in the main cmake file in LHCb.