DB update and Request processor

[juliendiot42]

Hello, @timflutre (cc. @chabrault ) I've made some big updates to the game:

1. Better usage of the database

Now the app fully use the power of the data-base rather than storing information in
files. The genetic values calculation have been optimized too. Rather than calculating
those before each phenotyping, it is calculated after the individuals creation and stored
in the DB. This doesn't affect much the plant material creation (the genotypes of the new
individuals is already in memory, and only 3 values are stored per individuals in the DB)
but it make the phenotyping request much faster as it can grab the genetic values directly
rather than loading the genotypes (rather slow) and calculating it.

2. Enhanced UI

Additional UI improvements have been made:

• Phenotype visualization: Interactive data tables with advanced filtering (for all) and
  graphic plots (currently for tester and admin only, but I planned to implement a better system to select which player can have access to this).
• Request tracking: Real-time progress indicators of the requests (cf. point 3)
• Plant material informations: All players can now see: parent names, cross types,
  phenotypic values, and selection status
• Genotyping interface: Redesigned download page with better request separation

3. Asynchronous Request Processing

A background worker process (./src/request_worker.R) now execute the pheno/geno/pltmat
requests asynchronously. So the app is not frozen for all the users when one someone make a request.

When a player send a request, the Shiny app only store the requests information in the database
(which is fast). The worker (which automatically start with the application) regularly fetch the database
for unprocessed requests execute them one by one. The players can see the progress of their request with progress bar in the app.

To monitor the worker, the admin page have a new tab with:

• the worker status (with a button to restart it if it unexpectedly crashed)
• the list of all the requests (ie. the requests queue, currently processing request, and
  processed requests)
• the worker's log displaying information about what the worker is doing

---

To test, I have deployed the docker image with this changes with the development tag eg.:

docker run -d --rm --name plantbreedgame -p 80:3838 \
    -v /host/path/to/plantBreedGameData:/var/lib/plantBreedGame \
    juliendiot/plantbreedgame:development

[timflutre]

Looks great Julien! I will notify Cédric Goby from Institut Agro Montpellier who has been deploying it recently on INRAE's SK8 service, so that he can test it.