feat: Use the latest, custom Terraform parser

[francescomari]

The snyk CLI uses the version of the Terraform parser implemented by the terraform package in this library. The ParseHCL2 function, exported by this library and used by snyk-iac-rules, still uses plain hcl2json.

Because of this difference, users of snyk-iac-rules and users of snyk don't have a consistent developer experience. A user writing a custom rule, and testing it via snyk-iac-rules, might be unable to run that rule in snyk. An example of this incompatibility is the handling of the jsonencode function in Terraform code.

Given this input file:

resource "aws_iam_role_policy" "test_policy" {
  name = "test_policy"
  role = aws_iam_role.test_role.id
  policy = jsonencode({
    Version = "2012-10-17"
    Statement = [
      {
        Action = [
          "ec2:*",
        ]
        Effect   = "Allow"
        Resource = "*"
      },
    ]
  })
}

snyk-iac-rules, which is still based on hcl2json, produces this:

{
  "resource": {
    "aws_iam_role_policy": {
      "test_policy": {
        "name": "test_policy",
        "policy": "${jsonencode({\r\n    Version = \"2012-10-17\"\r\n    Statement = [\r\n      {\r\n        Action = [\r\n          \"ec2:*\",\r\n        ]\r\n        Effect   = \"Allow\"\r\n        Resource = \"*\"\r\n      },\r\n    ]\r\n  })}",
        "role": "${aws_iam_role.test_role.id}"
      }
    }
  }
}

The parser in the snyk CLI, based on the terraform package, produces this:

{
  "resource": {
    "aws_iam_role_policy": {
      "test_policy": {
        "name": "test_policy",
        "policy": "{\"Statement\":[{\"Action\":[\"ec2:*\"],\"Effect\":\"Allow\",\"Resource\":\"*\"}],\"Version\":\"2012-10-17\"}",
        "role": "${aws_iam_role.test_role.id}"
      }
    }
  }
}

This PR fixes this incompatibility by using the same parser everywhere, consistently. Moreover, this PR stacks on top of #31, which should be merged first.

[CLAassistant]

All committers have signed the CLA.