How to construct list of ARNs from multiple source resources?

Terraform Providers AWS
1

Hi,
How do i use this earlier solution, if i have multiple resources?

data "aws_iam_policy_document" "assume" {
   statement {
   sid    = "AssumeIntoChildren"
   effect = "Allow"

   actions = [
      "sts:AssumeRole"
   ]

  resources = [
      for id in data.aws_organizations_organization.all_accounts.accounts[*].id :
     "arn:aws:iam::${id}:role/assume-into-me",
     "arn:aws:iam::${id}:role/assume-into-me1",
     "arn:aws:iam::${id}:role/assume-into-me2"
     ] 
    }
 }
2

Hi @palaparthis833,

If you are asking how you can include multiple different lists of ARNs into resources at once, I think the flatten function will help:

  resources = flatten(
    [
      for id in data.aws_organizations_organization.all_accounts.accounts[*].id : [
        "arn:aws:iam::${id}:role/assume-into-me",
        "arn:aws:iam::${id}:role/assume-into-me1",
        "arn:aws:iam::${id}:role/assume-into-me2",
     ]
   ],
   [
     "arn:aws:iam::foo::role/other-hardcoded-one",
   ],
  )

The purpose flatten is serving here is to flatten all of the nested lists, including deeply-nested ones as for expressions can often create, so that the result is just a flat list of strings. In the above case, perhaps something like this:

[
  "arn:aws:iam::first-id:role/assume-into-me",
  "arn:aws:iam::first-id:role/assume-into-me1",
  "arn:aws:iam::first-id:role/assume-into-me2",
  "arn:aws:iam::second-id:role/assume-into-me",
  "arn:aws:iam::second-id:role/assume-into-me1",
  "arn:aws:iam::second-id:role/assume-into-me2",
  "arn:aws:iam::foo::role/other-hardcoded-one",
]

Each of the elements of data.aws_organizations_organization.all_accounts.accounts here would’ve produced its own separate nested list before flatten, but using that function brings all of the nested strings up to the top-level so that the data structure is the shape that the resources argument expects.