add repos block-ingest command

[kebroad]

Implements the blockIngest() mutation and adds a subcommand to block ingest for one or all repos (in parallel).

Crowdstrike has recommended we block all ingest prior to disruptive changes to our Logscale instance, so this command will make it easy for us to do so.

⇒ bin/humioctl repos block-ingest --help
Block ingest for one or all repositories for a specified number of seconds.

Examples:
  # Block ingest for a specific repository
  humioctl repos block-ingest myrepo 3600

  # Block ingest for all repositories
  humioctl repos block-ingest --all 3600

Usage:
  humioctl repos block-ingest [flags] [repository] <seconds>

[bderrly]

There is an Example field in the cobra.Command struct where you could move these example invocations.

[bderrly]

It might be nicer to have seconds be a flag rather than a positional argument. One reason is that you can define the flag as a time.Duration so one could say --duration=4h. Another is that it will simplify some of the logic around the positional args.

[SaaldjorMike]

The most recent changes refactors how we interact with GraphQL for the humioctl tool. Going forward, you have to

1. update https://github.com/humio/cli/tree/master/internal/api/humiographql with any new interaction with the LogScale GraphQL API
2. generate Go-code based on the updated GraphQL queries/mutations: go generate ./...
3. update humioctl within https://github.com/humio/cli/tree/master/cmd/humioctl so it uses the newly generated code

Right now the code path is as follows: https://github.com/humio/cli/tree/master/cmd/humioctl calls https://github.com/humio/cli/tree/master/internal/api which in turn calls the autogenerated code in https://github.com/humio/cli/blob/master/internal/api/humiographql/humiographql.go.