closes #12443

- Add GlueVersion, WorkerType, JobCommandName enum-like classes for future-proofing these values
- Add PythonVersion as a normal enum as it's less likely to change
- Add Job construct
  - CloudFormation base resource is a complex resource with almost all properties being optional and allowed in specific combinations based on command, glue version and worker types
  - It supports multiple job types e.g. ETL, Streaming and Python Shell and languages e.g. Scala and Python. This requires different combinations of props
  - One possibility is to try to create different constructs for the different types but I've opted to do just one base on resembling CloudFormation which does not prevent us from introducing more specialized types later
  - Documentation is not accurate, based on integ test deployments which led me to not having any assertions on valid props combinations
    - https://docs.aws.amazon.com/glue/latest/dg/add-job.html
    - https://docs.aws.amazon.com/glue/latest/dg/aws-glue-api-jobs-job.html
  - Not exposing some CloudFormation props
    - allocatedCapacity as it's deprecated in favour of maxCapacity
    - logUri which is not used and is reserved for future use

- glue job does not emit success or failure metrics to cloudwatch metrics. Instead, it emits events to cloudwatch events
- add JobEventState enum for known job states in https://docs.aws.amazon.com/AmazonCloudWatch/latest/events/EventTypes.html#glue-event-types
- add utility methods to create event rules and cloudwatch metrics based on those roles supporting the most common use cases; success, failure and timeout

https://docs.aws.amazon.com/glue/latest/dg/monitoring-awsglue-with-cloudwatch-metrics.html

https://docs.aws.amazon.com/glue/latest/dg/aws-glue-programming-etl-glue-arguments.html
https://docs.aws.amazon.com/glue/latest/dg/monitor-spark-ui-jobs.html

- introduce JobBase to contain common logic between Job and Import
  - JobBase now handles CloudWatch Event Rules and Rule-based metrics methods
- address the rename comments
- use the private constructor + static of() pattern where relevant
- Make JobProps.glueVersion required

- introduce JobExecutable
- drop JobCommand
- rename JobCommandName to JobType

https://aws.amazon.com/blogs/big-data/introducing-aws-glue-3-0-with-optimized-apache-spark-3-1-runtime-for-faster-data-integration/

Co-authored-by: Ben Chaimberg <[email protected]>

- move AssetCode directory check to constructor
- rename SparkUIConfig to SparkUILoggingLocation
- rename filter to quiet
- rename jobStateRule to onStateChange and make it public
- check no reserved arguments used
- rename JobProps.numberOfWorkers to JobProps.workerCount
- move JobExecutableConfig to the bottom of the file
- rename PythonJobExecutableProps to PythonSparkJobExecutableProps

Co-authored-by: Ben Chaimberg <[email protected]>

- update README
- use const instead of let and beforeEach where relevant
- assert on stack assets metadata
- add test case of 2 jobs reusing asset
- other minor things

- fix bug with code bucket permissions
  - Glue role did not have permissions to read from the code bucket which manifested as DeniedAccess exceptions when attempting to run jobs
  - Update Code.bind to take a Job instead of Construct so it can grant it the necessary read permissions
- Python shell failed saying Glue 2.0 is not supported
  - Console UI for adding a new Python shell jobs shows Glue 1.0 as the only supported version
  - Modified tests to reject using Glue 0.9 but didn't exclude Glue 2.0 or 3.0 for any future support by the new versions