asmcli install has multiple errors when installing to Anthos on AWS

[Jumwah]

The asmcli script assumes in multiple places that the cluster is a GCP cluster, and uses the gcloud container clusters subcommand instead of gcloud container aws clusters subcommand. This means the validate subcommand is not usable for Anthos on AWS, and the install subcommand has multiple errors. Presumably this is the case for all non-GCP based Anthos clusters.

For example, when trying to determine if labels are required:

asmcli: Running: '/usr/bin/gcloud container clusters describe  --zone= --project=XXX --format=value(resourceLabels)[delimiter=","]'
asmcli: -------------
ERROR: (gcloud.container.clusters.describe) One of [--zone, --region] must be supplied: Please specify location.

Most can be worked around by using the various --enable_XXX flags, but there is one which means I can't enable CNI:

asmcli: Running: '/usr/bin/gcloud container node-pools list --project=XXX --region  --cluster  --filter     config.machineType.split(sep="-").slice(-1:) >= 0  --format=json'
asmcli: -------------
ERROR: (gcloud.container.node-pools.list) One of [--zone, --region] must be supplied: Please specify location.
asmcli: Running: '/asm-1.14.1-asm.3/kpt cfg set asm anthos.servicemesh.managed-cni.cni-enable-install false'

Modifying the asmcli script to remove the line installs ASM with CNI in AWS successfully.