Cnh/replace simulation conf reducer by slice

[kmer2016]

close #5273
close #5274

Key Updates:

• Introduction of Contexts: We have added two new contexts, useOsrdConfAction and useOsrdConfSelectors. These enable components that utilize osrdconf to accurately select the appropriate selectors/actions (either stdcmConf or simulationConf).
• wrapping stdcm and operational-studies routes with the the osrdConf context passing the right parameters.

Pre-Merge Tasks:

Currently, components can only utilize osrdConf actions/selectors if they are encapsulated within OsrdConfContextLayout. This limitation causes issues in components that require access to osrdConf from the store but are not appropriately wrapped. For instance, the Home component in the application editor, which utilizes getInfraID, is not encapsulated in osrdConf. To address these issues:

• We need to incorporate an infraID field for osrdConf within the map slice.
• The editor's matching route should be wrapped with the OsrdConfContextLayout, using the correct parameters.
  These adjustments must be applied consistently across all osrdConf fields usages outside the stdcm/* and operational-studies routes.
  Additionally, it's worth mentioning the presence of functions like buildCommonConfReducers and buildCommonConfSelectors, which broadly generate reducers and selectors for managing osrdConf parameters. it can help to generate osrdconf parameters needed inside others slices.

Testing

• test component which use selectors/actions of stdcm/simulation slice inside modal
• test destination/origin component as they have been refactored.
• test store persistence
• Check if need to save projectId, studyId and scenarioId into store anymore

[codecov]

Codecov Report

Attention: 1781 lines in your changes are missing coverage. Please review.

Comparison is base (84d57c2) 27.30% compared to head (a4085f5) 26.85%.

Files	Patch %	Lines
front/src/common/osrdContext.tsx	38.79%	71 Missing ⚠️
...ainschedule/components/ManageTrainSchedule/Map.tsx	0.00%	68 Missing ⚠️
...lications/editor/tools/pointEdition/components.tsx	0.00%	55 Missing and 1 partial ⚠️
...mulationResult/components/SimulationResultsMap.tsx	0.00%	55 Missing and 1 partial ⚠️
front/src/main/app.jsx	0.00%	45 Missing ⚠️
front/src/common/Map/Layers/GeoJSONs.tsx	0.00%	42 Missing and 1 partial ⚠️
...ools/routeEdition/components/EditRouteMetadata.tsx	0.00%	40 Missing and 1 partial ⚠️
...ainSchedule/helpers/adjustConfWithTrainToModify.ts	0.00%	40 Missing ⚠️
...ManageTrainSchedule/SubmitConfAddTrainSchedule.tsx	0.00%	36 Missing ⚠️
...or/tools/routeEdition/components/EditRoutePath.tsx	0.00%	33 Missing and 1 partial ⚠️
... and 100 more

Additional details and impacted files

@@             Coverage Diff              @@
##                dev    #5583      +/-   ##
============================================
- Coverage     27.30%   26.85%   -0.46%     
  Complexity     2136     2136              
============================================
  Files           952      952              
  Lines        124900   124095     -805     
  Branches       2690     2574     -116     
============================================
- Hits          34106    33327     -779     
- Misses        89189    89279      +90     
+ Partials       1605     1489     -116     

Flag	Coverage Δ
core	78.88% <ø> (ø)
editoast	75.36% <ø> (ø)
front	8.69% <13.41%> (-0.79%)	⬇️
gateway	2.54% <ø> (ø)
tests	81.77% <ø> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[SharglutDev]

Thank you for this amazing work ! Reviewed only the code for now, will test it asap.

Left comments. 90% of them are related with this recent front guideline for type imports.

[SharglutDev]

Tested the app in operational studies, stdcm with modals, store is still updated and persistent, good job !

One more request regarding one of my previous comment : you can remove the updateFeatureInfoClick action in the map reducer (the one updating an id), its only used at one place in the code and doesn't do anything as there isn't any selector for the related state. (approved by @nicolaswurtz)

[SharglutDev]

Lgtm, thank you for the hard work ! I approve for my part but I'll let @clarani check it too in case I missed something
(Good luck for the many many conflicts :) )

[clarani]

Here are some comments, I'm still reviewing the PR

[clarani]

I've just finished to review the reducers files ! :)

[clarani]

Thanks, here are some global comments:

• check the imports to 'Store' (it is invalid, the import should occur from 'store')
• rename the following types to pascal case and remove the -Type:
  • OperationalStudiesConfSliceType (=> OperationalStudiesConfSlice)
  • OperationalStudiesConfSliceActionsType
  • OperationalStudiesConfSelectorsType
  • mapSliceActionsType
  • StdcmConfSliceType
  • StdcmConfSelectorsType
  • editorSliceType
  • EditorSliceActionsType (=> EditorSliceAction)
  • EditorActionsType (=> EditorAction)
  • EditorSelectorsType
  • mapViewerSliceType
  • MapViewerStateType
  • CommonConfReducersType
  • InfraStateReducersType (=> InfraStateReducer)
  • InfraStateType
  • MapViewerSelectorsType
  • ... (maybe I forgot some)
• stdcmMode is not used anymore, so you can remove it from the store

[clarani]

Some other comments:

• do you think it would be relevant to throw an error if we use useInfraId and infraId is undefined ? This way, we can type it as number
• when using useOsrdActions(), no need to cast now (see 20 occurences approximately). If we use an action specific to a slice, we can directly import it (no need to go through useOsrdActions())
• same thing for useOsrdConfSelectors()

[clarani]

Some last comments:

• can you also remove the cast for all the useOsrdConfActions()

[clarani]

Almost good, I only found a warning when testing. Once this warning is solved, we can merge this PR 🎉

[kmer2016]

Apparently, the warning is linked to our store becoming increasingly large and, by default, Redux traverses the entire store to check the immutability of each field. Therefore, this check can take some time. If this time exceeds the threshold of 32ms, it triggers the warning.
To resolve this simply, we can adjust the ImmutableStateInvariantMiddleware middleware:

• either we disable it
• we increase the threshold at which the warning is displayed.

I opted for the first solution because, as mentioned here, in any case, we use Immer to create our store slices (using createSlice), so we are already ensured of having immutable fields. Therefore, it's not really necessary to check the entire store each time. And as indicated by one of the maintainers here, disabling this option can improve performance.

[clarani]

LGTM, tested and it's more fluid now ✅
I tried to check everything, we will have to stay vigilant about it in the coming days