Feature/antares in memory

[JasonMarechal25]

Introduce "study mode". A way to run xpansion and problem generation with a study folder as parameter. In this case Antares will not be run as a stand-alone but as a library by problem generation.

• New ADRs
• Add a "new" data set for lpnamer E2E tests: SmallTestFiveCandidates, a copy of the example provided by Xpansion.
• Refactor test_lpnamerEndToEnd.py : when possible prefer the use multiple @parametrize instead of writing the combinations in a list.
• Refactor test_lpnamerEndToEnd.py : copy the test data in a temporary folder and works there. Prevent creating outputs in source code.
• Add a test case for Study mode in test_lpnamerEndToEnd.py
• Add a new E2E test: short memory
• Refactor ProblemGenerationExeOptionsTest to test several combinations of mutually exclusive parameters
• In SolverFactory.cpp call Init() on solver objects before returning them. Prevent temporal coupling for Xpress solver where a call to init() was mandatory.
• Add Antares as a build dependency

Missing pieces (every thing is open to discussion):

• Handle Antares error Properly handle Antares simulator errors #888
• Update general data ini
• Revert general data ini Revert general data ini when error occur with full run "antares in memory" #889
• Write ADR
• Update sequence diagram Update technical documentation for antares in memory #890
• Maybe update C4 Update technical documentation for antares in memory #890
• Use enum mode in LPFileExtractor [API] Better handling of execution modes #797
• find/define a way to log before lp dir exist (for antares error for exemple) Properly handle Antares simulator errors #888
• Handle variables.txt directly at Lps level (SignificationMetierDesVariables)
• Factorise variables name and "calculate" time step with week number on xpansion side (A faire dans Antares bien et vite)
• Cleanup python to only support memory mode (See ADR, we keep some support for now)
• antares named problèmes force mps writing. Peut etre pas besoin de l'option named sachant que les variables sont nommées dans la structure LP
• Mettre à jour les docs: notamment comment build, dépendances, etc Update documentation to use Antares in memory #893
• Investigate to use full run executable with --study option and use fullrun driver Use antares in memory with fullrun #892
• Handle PBGen step with memory
• Add entry in multisolver to rename a range of col/var Feature/antares in memory #753 (comment) Add entry in multisolver to rename a range of col/var #891
• Run Antares in parallele mode

Following other PR:

• Don't overly rename problem in API mode

[a-zakir]

there are a lot unnecessary changes in tests data, ex most (or all) *.ini files

[JasonMarechal25]

Intentional.
Data where missing some constraints

[tbittar]

Donc pour debugger en step by step, file mode, il faudra lancer directement la bonne commande C++ et pas antares-xpansion-launcher -i <study> --step antares (actuellement je me sers beaucoup de cette dernière commande) ?

[tbittar]

Is this file useful ?

[JasonMarechal25]

No I need to remove it

[tbittar]

This was just needed to for testing purposes. In production, we want to keep full data precision

[sonarcloud]

Quality Gate passed

Issues
10 New issues
0 Accepted issues

Measures
0 Security Hotspots
13.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarCloud