[FEA] Profiling tool should detect if there is "file.encoding=UTF-8"

[viadea]

I wish Profiling tool should detect if there is "file.encoding=UTF-8 ​​​​​" from event log.
If it is empty or set to some other non-UTF8 encoding, we should raise the warning.

This is because, our regexp related functions might fallback to CPU if it is non-UTF8.
Such as below driver log message might show up:

!Expression <RegExpReplace> regexp_replace(xxx) cannot run on GPU because regular expression support is disabled because the GPU only supports the UTF-8 charset when using regular expressions

[amahussein]

This property is part of the environment information in the eventlog.
It should be one of the "System Properties"

For example

{
   "Event":"SparkListenerEnvironmentUpdate",
   "JVM Information":{
      "Java Home":"/usr/lib/jvm/java-8-openjdk-amd64/jre",
      "Java Version":"1.8.0_232 (Private Build)",
      "Scala Version":"version 2.12.10"
   },
   "Spark Properties":{
      
   },
   "Hadoop Properties":{
      
   },
   "System Properties":{
      "file.encoding":"UTF-8"
   },
   "Classpath Entries":{
      
   }
}

[amahussein]

@viadea How do you want this information to be reported?
Do you want it to be a log warning message? or part of the application output info?

[amahussein]

Does this mean we need to update the qualification to consider the regex not supported if the non UTF-8 files are used?