Data sharing

docs/resources/datashare.md

@@ -0,0 +1,128 @@
+---
+# generated by https://github.com/hashicorp/terraform-plugin-docs
+page_title: "redshift_datashare Resource - terraform-provider-redshift"
+subcategory: ""
+description: |-
+  Defines a Redshift datashare. Datashares allows a Redshift cluster (the "consumer") to
+  read data stored in another Redshift cluster (the "producer"). For more information, see
+  https://docs.aws.amazon.com/redshift/latest/dg/datashare-overview.html
+  The redshift_datashare resource should be defined on the producer cluster.
+  Note: Data sharing is only supported on certain Redshift instance families,
+  such as RA3.
+---
+
+# redshift_datashare (Resource)
+
+Defines a Redshift datashare. Datashares allows a Redshift cluster (the "consumer") to
+read data stored in another Redshift cluster (the "producer"). For more information, see
+https://docs.aws.amazon.com/redshift/latest/dg/datashare-overview.html
+
+The redshift_datashare resource should be defined on the producer cluster.
+
+Note: Data sharing is only supported on certain Redshift instance families,
+such as RA3.
+
+## Example Usage
+
+```terraform
+# Example: Datashare that can only be consumed by a non-public Redshift cluster.
+resource "redshift_datashare" "private_datashare" {
+  name = "my_private_datashare" # Required
+  owner = "my_user" # Optional
+
+  # Example of adding a schema to a data share in "auto" mode.
+  # All tables/views and functions in the schema are added to the datashare,
+  # and redshift will automatically add newly-created tables/views and functions
+  # to the datashare without needing to re-run terraform.
+  schema {
+    name = "public" # Required
+    mode = "auto" # Required
+  }
+
+  # Example of ading a schema to a data share in "manual" mode.
+  # Only the specified tables/views and functions will be added to the data share.
+  schema {
+    name = "other" # Required
+    mode = "manual" # Required
+    tables = [ # Optional. If unspecified then no tables/views will be added.
+      "my_table",
+      "my_view",
+      "my_late_binding_view",
+      "my_materialized_view",
+    ]
+    functions = [ # Optional. If unspecified then no functions will be added.
+      "my_sql_udf",
+    ]
+  }
+}
+
+# Example: Datashare that can be shared with publicly available consumer clusters.
+resource "redshift_datashare" "publicly_accessible_datashare" {
+  name = "my_public_datashare" # Required
+  publicly_accessible = true # Optional. Default is `false`
+
+  # Example of adding a schema to a data share in "auto" mode.
+  # All tables/views and functions in the schema are added to the datashare,
+  # and redshift will automatically add newly-created tables/views and functions
+  # to the datashare without needing to re-run terraform.
+  schema {
+    name = "public" # Required
+    mode = "auto" # Required
+  }
+
+  # Example of ading a schema to a data share in "manual" mode.
+  # Only the specified tables/views and functions will be added to the data share.
+  schema {
+    name = "other" # Required
+    mode = "manual" # Required
+    tables = [ # Optional. If unspecified then no tables/views will be added.
+      "my_table",
+      "my_view",
+      "my_late_binding_view",
+      "my_materialized_view",
+    ]
+    functions = [ # Optional. If unspecified then no functions will be added.
+      "my_sql_udf",
+    ]
+  }
+}
+```
+
+<!-- schema generated by tfplugindocs -->
+## Schema
+
+### Required
+
+- **name** (String) The name of the datashare.
+
+### Optional
+
+- **id** (String) The ID of this resource.
+- **owner** (String) The user who owns the datashare.
+- **publicly_accessible** (Boolean) Specifies whether the datashare can be shared to clusters that are publicly accessible. Default is `false`.
+- **schema** (Block Set) Defines which objects in the specified schema are exposed to the data share (see [below for nested schema](#nestedblock--schema))
+
+### Read-Only
+
+- **created** (String) The date when datashare was created
+- **producer_account** (String) The ID for the datashare producer account.
+- **producer_namespace** (String) The unique cluster identifier for the datashare producer cluster.
+
+<a id="nestedblock--schema"></a>
+### Nested Schema for `schema`
+
+Required:
+
+- **mode** (String) Configures how schema objects will be exposed to the datashare. Must be either `auto` or `manual`.
+
+  In `auto` mode, all tables, views, and UDFs will be exposed to the datashare, and Redshift will automatically expose new tables, views, and functions in the schema to the datashare (without requiring `terraform apply` to be run again).
+
+  In `manual` mode, only the `tables` and `functions` explicitly declared in the `schema` block will be exposed to the datashare.
+- **name** (String) The name of the schema
+
+Optional:
+
+- **functions** (Set of String) UDFs that are to exposed to the datashare. You should configure this attribute explicitly when using `manual` mode. When using `auto` mode, this is treated as a computed attribute and you should not explicitly declare it.
+- **tables** (Set of String) Tables and views that are exposed to the datashare. You should configure this attribute explicitly when using `manual` mode. When using `auto` mode, this is treated as a computed attribute and you should not explicitly declare it.
+
+

examples/resources/redshift_datashare/resource.tf

@@ -0,0 +1,61 @@
+# Example: Datashare that can only be consumed by a non-public Redshift cluster.
+resource "redshift_datashare" "private_datashare" {
+  name = "my_private_datashare" # Required
+  owner = "my_user" # Optional
+
+  # Example of adding a schema to a data share in "auto" mode.
+  # All tables/views and functions in the schema are added to the datashare,
+  # and redshift will automatically add newly-created tables/views and functions
+  # to the datashare without needing to re-run terraform.
+  schema {
+    name = "public" # Required
+    mode = "auto" # Required
+  }
+
+  # Example of ading a schema to a data share in "manual" mode.
+  # Only the specified tables/views and functions will be added to the data share.
+  schema {
+    name = "other" # Required
+    mode = "manual" # Required
+    tables = [ # Optional. If unspecified then no tables/views will be added.
+      "my_table",
+      "my_view",
+      "my_late_binding_view",
+      "my_materialized_view",
+    ]
+    functions = [ # Optional. If unspecified then no functions will be added.
+      "my_sql_udf",
+    ]
+  }
+}
+
+# Example: Datashare that can be shared with publicly available consumer clusters.
+resource "redshift_datashare" "publicly_accessible_datashare" {
+  name = "my_public_datashare" # Required
+  publicly_accessible = true # Optional. Default is `false`
+
+  # Example of adding a schema to a data share in "auto" mode.
+  # All tables/views and functions in the schema are added to the datashare,
+  # and redshift will automatically add newly-created tables/views and functions
+  # to the datashare without needing to re-run terraform.
+  schema {
+    name = "public" # Required
+    mode = "auto" # Required
+  }
+
+  # Example of ading a schema to a data share in "manual" mode.
+  # Only the specified tables/views and functions will be added to the data share.
+  schema {
+    name = "other" # Required
+    mode = "manual" # Required
+    tables = [ # Optional. If unspecified then no tables/views will be added.
+      "my_table",
+      "my_view",
+      "my_late_binding_view",
+      "my_materialized_view",
+    ]
+    functions = [ # Optional. If unspecified then no functions will be added.
+      "my_sql_udf",
+    ]
+  }
+}

redshift/helpers.go

@@ -128,11 +128,12 @@ func isRetryablePQError(code string) bool {
 	return ok
 }
 
-func splitCsvAndTrim(raw string) ([]string, error) {
+func splitCsvAndTrim(raw string, delimiter rune) ([]string, error) {
 	if raw == "" {
 		return []string{}, nil
 	}
 	reader := csv.NewReader(strings.NewReader(raw))
+	reader.Comma = delimiter
 	rawSlice, err := reader.Read()
 	if err != nil {
 		return nil, err
@@ -146,3 +147,21 @@ func splitCsvAndTrim(raw string) ([]string, error) {
 	}
 	return result, nil
 }
+
+func setToMap(set *schema.Set, key string) map[string]map[string]interface{} {
+	result := make(map[string]map[string]interface{})
+	for _, s := range set.List() {
+		m := s.(map[string]interface{})
+		id := m[key].(string)
+		result[id] = m
+	}
+	return result
+}
+
+func mapToSet(mapOfMaps map[string]map[string]interface{}, hashFunction schema.SchemaSetFunc) *schema.Set {
+	result := schema.NewSet(hashFunction, nil)
+	for _, m := range mapOfMaps {
+		result.Add(m)
+	}
+	return result
+}

redshift/provider.go

@@ -121,6 +121,7 @@ func Provider() *schema.Provider {
 			"redshift_schema":    redshiftSchema(),
 			"redshift_privilege": redshiftPrivilege(),
 			"redshift_database":  redshiftDatabase(),
+			"redshift_datashare": redshiftDatashare(),
 		},
 		DataSourcesMap: map[string]*schema.Resource{
 			"redshift_user":     dataSourceRedshiftUser(),