f/shashank-website -- Website Documentation -- Creation of Website using Python

[Shashankjain12]

Pull Request for Python Website solving issues #42 #44 #45 which is used for building website documentation using Python and removing Java dependencies.

[StevenAWhite]

Ok, we need to start with a reorganization. Move this project to project/website. Consolidate all source to simply project/website/src and simplify your CMakeLists.txt to a minimal clean cmake file. There are lots of extra directives in your two files that are just carrying over legacy structure that is not needed for this new addition.

Once your done with that I'll fully review the python source.

[StevenAWhite]

I also want you to remove all references to JAVA in the build system. Delete the project/java and any cmake commands related to it.

[StevenAWhite]

You likely do not need to keep this block of variables. They are left overs from an older way I propagated -I flags in C++ from file to file. As this is a python based project I doubt they are used anywhere outside of this file.

[StevenAWhite]

Add https://cmake.org/cmake/help/v3.14/module/FindLATEX.html to this. Its required for the full generation of the website to generate the BibTex entries might as well help CMake point this out.

[StevenAWhite]

We can afford to simplify this file a little. I know it breaks convention, but given we only have SRCS we can just refer to that variable in python projects.

[StevenAWhite]

This has no command so its should be removed

[StevenAWhite]

Do you use this variable ?

[StevenAWhite]

If these variables are not referenced go ahead and remove them.

[StevenAWhite]

Do we have ant files in this folder? If not remove this call. I don't see anything generating files up to this point

[StevenAWhite]

This command does nothing. Remove it

[StevenAWhite]

I do not believe you use this anywhere. So this target is not needed.

[StevenAWhite]

It appears that at least two of your python files are completely unused. Remove any dead files from your python source. Description.py and CDMSerializer.py or explain there purpose. If they are functionality for a future pull request leave them out of this pull request and include them when they do have use.

[Shashankjain12]

@StevenAWhite I have created the pull request with all of the suggested changes and adding the virtual environment based target as well.

[StevenAWhite]

If you have the plotter working you need to remove the comments in the command. If the comments are no longer needed you should remove them.

[StevenAWhite]

Don't override this here. Just use the paramaters in the command

[StevenAWhite]

Ok the next challenge is to let the plotter fall back when looking to results to baselines. A baseline is a zip file which contains Results files from a previous run of BioGears. CMake determines if a baseline exist by

find_path (
  ${ROOT_PROJECT_NAME}_Baseline_DIR
  NAME Scenarios
  PATHS ${PROJECT_SOURCE_DIR}/../verification/core
        ${PROJECT_SOURCE_DIR}/../validation/core
        ${PROJECT_SOURCE_DIR}/../baselines/core
        ${PROJECT_SOURCE_DIR}/../verification
        ${PROJECT_SOURCE_DIR}/../validation
        ${PROJECT_SOURCE_DIR}/../baselines
  DOC "Optional directory path to hiarchial filetree of baseline archives. This will be used by Python Plotters to find Baselines"
)

Other then the Biogears_Baseline_DIR you can also look in a subdirectory baselines/ in the same folder that you expect to find a given *Results.csv file. I'm attaching a baseline for review its very simple. All a Baseline Dir is an exsact directory structure match to runtime/Scenarios where each folder has a baselines/ director and those baseline directories contain zip files of the Results.csv and Results.log file.

sawhite@SED-SAWHITE:/mnt/d/remotes/github/Shashankjain12/core/projects/website$ tree -L 4  -d /mnt/d/remotes/sed-stash/validation/
/mnt/d/remotes/sed-stash/validation/
├── core
│   ├── Scenarios
│   │   ├── AnesthesiaMachine
│   │   │   └── baselines
│   │   ├── Combined
│   │   │   └── baselines
│   │   ├── Compartments
│   │   │   └── baselines
│   │   ├── ConsumeMeal
│   │   │   └── baselines
│   │   ├── Drug
│   │   │   └── baselines
│   │   ├── EnergyEnvironment
│   │   │   └── baselines
│   │   ├── Miscellaneous
│   │   │   └── baselines
│   │   ├── Patient
│   │   │   └── baselines
│   │   ├── Showcase
│   │   │   └── baselines
│   │   ├── Validation
│   │   │   └── baselines
│   │   └── baselines
│   └── UnitTests
│       ├── BioGearsTests
│       │   └── baselines
│       └── CDMTests
│           └── baselines
└── lite
    └── Scenarios
        ├── AnesthesiaMachine
        │   └── baselines
        ├── Combined
        │   └── baselines
        ├── Compartments
        │   └── baselines
        ├── ConsumeMeal
        │   └── baselines
        ├── Drug
        │   └── baselines
        ├── EnergyEnvironment
        │   └── baselines
        ├── Miscellaneous
        │   └── baselines
        ├── Patient
        │   └── baselines
        ├── Showcase
        │   └── baselines
        └── Validation
            └── baselines```

[EndotrachealTubeLeakVariedResults.zip](https://github.com/BioGearsEngine/core/files/5031824/EndotrachealTubeLeakVariedResults.zip)

The behavior I want is 
Provide the baseline argument when BioGears_Baseline_DIR is valid

For each request plot
 Plotter will try to find the correct Results.csv file 
 if Results.csv is not found it will try baselines/Results.zip 
 if baselines/Results.zip is not found Biogears_Baseline_DIR/baselines/Results.zip will be searched  

The goal of this fallback option is to allow our docker scripts to checkout biogears and our verification repos then run the website generation with out ever needing to compile biogears and run verification as the verification should be the same as teh checked in baselines for a given release. 

[StevenAWhite]

I like what I'm seeing. I want to see a thread pool added to the plotter to speed up its execution time. We can decide a default number of threads for CMake to pass latter, but for now I just want proof the threadpool works and we can plot say 3 plots at a time.

Really minor syntax clean up on your write calls just to clean up how that looks to other maintainers. The only other thing which isn't a huge deal is to teach the PlotDriver to fallback to baselines if it doesn't find a fresh results file. You likely want to log when you do this, but 90% of the time these tools will be used in an automated system that doesn't have 48 hours to run verification so we want to run against historical data if its there.

[StevenAWhite]

Accepted