Improve logging routine in runtime engine

Debug log messages need to be improved by providing additional information, such as TID, Category, and provide a better verbosity overall. This will allow us to spot bugs more easily when reported by users, as well as to pinpoint potential bottlenecks in the engine.

The improved logging should provide more granular control of the logs, or just categorization of the different execution paths in the runtime. It can be modeled after the Tracing routine in .NET -> https://msdn.microsoft.com/en-us/library/sk36c28t(v=vs.110).aspx

Trace (Log) statements throughout the runtime code (Instrumentation of the code) are categorized with a relevant preset category. If a listener is attached for that category, messages will be output as per the Listener's implementation. - Default LogCat implementation should be available for all categories.

Can be enabled by setting the "android": { "diagnosticTracing": true } in the project's package.json

Micro benchmarking can also be enabled to record time spent in execution of most key methods (that are measurable, can be measured with little to no error added by resolution timing - time spent to record start and end time). The benchmarking routine should output logs in a parsable manner for easier visualization.

Can be enabled by setting the "android": { "diagnosticBenchmarking": true } in the project's package.json

Enabling tracing could and probably WILL impose a performance penalty

Commencez par localiser le code existant de logging et d’instrumentation du moteur d’exécution, puis examinez la gestion de package.json pour android.diagnosticTracing et android.diagnosticBenchmarking. Utilisez le modèle .NET Tracing lié comme contexte. Le travail est considéré comme terminé lorsque le tracing catégorisé et configurable ainsi que des logs de benchmarking analysables sont disponibles via ces paramètres, et que la pénalité de performance est documentée.