add a way to pass env vars to custom spark images for worker and master

[jkremser]

apiVersion: v1
kind: ConfigMap
metadata:
  name: sparky-cluster                                # compulsory
  labels:
    radanalytics.io/kind: cluster                     # compulsory
data:
  config: |-
    workerNodes: "2"                                  # optional defaults to 1
    masterNodes: "1"                                  # optional defaults to 1
    customImage: ""                                   # optional defaults to jkremser/openshift-spark:2.3-latest
    env:                                              # optional
    - name: SPARK_WORKER_CORES
      value: 2
    - name: FOO
      value: bar
    sparkConfigurationMap: my-config                  # optional defaults to ${name}-config
                                                      # kubectl create configmap my-config --from-file=example/config.conf
    sparkConfiguration:                               # optional, it overrides the config defined above, the cm must exist
    - name: spark.executor.memory
      value: 2g
    - name: spark.sql.conf.autoBroadcastJoinThreshold
      value: 20971520
    downloadData:                                     # optional, it downloads these files to each node
    - url: https://raw.githubusercontent.com/Jiri-Kremser/spark-operator/master/README.md
      to: /tmp/

[jkremser]

done in pr #51