Setting spark.app.id to a more intuitive name

[edingroot]

1. Setting spark app id to same as the spark app name
2. Replacing -, ., , characters to _ for app id

Test: test starting a new spark session and call the Spark's API at http://<driver_ip>:<ui_port>/api/v1/applications:

[ {
  "id" : "jupyterhub_chi_test_spark_39091_1692232285",
  "name" : "jupyterhub_chi_test-spark_39091_1692232285",
  "attempts" : [ {
    "startTime" : "2023-08-17T00:31:28.283GMT",
    "endTime" : "1969-12-31T23:59:59.999GMT",
    "lastUpdated" : "2023-08-17T00:31:28.283GMT",
    "duration" : 1038136,
    "sparkUser" : "chi",
    "completed" : false,
    "appSparkVersion" : "3.2.4",
    "startTimeEpoch" : 1692232288283,
    "endTimeEpoch" : -1,
    "lastUpdatedEpoch" : 1692232288283
  } ]
} ]

[edingroot]

This is just for moving the following lines down a bit for better code readability:

# app_name from env is already appended port and time to make it unique
app_name = (spark_opts_from_env or {}).get('spark.app.name')
if not app_name:
# We want to make the app name more unique so that we can search it
# from history server.
app_name = f'{app_base_name}_{ui_port}_{int(time.time())}'

[88manpreet]

Sorry I missed it earlier.
There is a character limit of 63 for app id and I think 253 for app name. app id needs to be trimmed to make sure it is within the limit.
Secondly, would likely be useful to have app name as: service__instance_timestamp or service__job__id__action_name_timestamp or adhoc/tron jobs.