Merge remote-tracking branch 'upstream/master' into master

* upstream/master:
  Add a random string to ad directory so tests can be run in parallel (hashicorp#3802) (hashicorp#6905)
  Set context early on in LoadAndValidate to prevent NPE (hashicorp#3800) (hashicorp#6903)
  Add tests for DICOM store pubsub topic removal. (hashicorp#3442) (hashicorp#6893)
  Bump Bigtable version from 1.1.0 to 1.5.0 (hashicorp#6890)
  Fix colon in doc notes (hashicorp#3796) (hashicorp#6888)
  Fix bug: diff being detected for source_repo_repository even when there are no changes (hashicorp#3786) (hashicorp#6886)
  Use proper prefix inside router_interface test (hashicorp#3795) (hashicorp#6883)
  Update Bigtable Num of Clusters Block Docs (hashicorp#3792) (hashicorp#6881)

.changelog/3442.txt

@@ -0,0 +1,3 @@
+```release-note:none
+healthcare: added tests for pubsub topic removal in `google_healthcare_dicom_store`
+```

.changelog/3786.txt

@@ -0,0 +1,3 @@
+```release-note:bug
+sourcerepo: fixed perma-diff in `google_sourcerepo_repository`
+```

.changelog/3792.txt

@@ -0,0 +1,3 @@
+```release-note:none
+
+```

.changelog/3795.txt

@@ -0,0 +1,3 @@
+```release-note:none
+
+```

.changelog/3796.txt

@@ -0,0 +1,3 @@
+```release-note:none
+
+```

.changelog/3800.txt

@@ -0,0 +1,3 @@
+```release-note:bug
+Fixed crash due to nil context when loading credentials
+```

.changelog/3802.txt

@@ -0,0 +1,3 @@
+```release-note:none
+
+```

go.mod

@@ -1,7 +1,7 @@
 module github.com/terraform-providers/terraform-provider-google
 
 require (
-	cloud.google.com/go/bigtable v1.1.0
+	cloud.google.com/go/bigtable v1.5.0
 	github.com/apparentlymart/go-cidr v1.0.1
 	github.com/client9/misspell v0.3.4
 	github.com/davecgh/go-spew v1.1.1
@@ -20,9 +20,9 @@ require (
 	github.com/sirupsen/logrus v1.2.0 // indirect
 	github.com/stoewer/go-strcase v1.0.2
 	github.com/terraform-providers/terraform-provider-random v0.0.0-20190925200408-30dac3233094
-	golang.org/x/net v0.0.0-20200324143707-d3edc9973b7e
+	golang.org/x/net v0.0.0-20200625001655-4c5254603344
 	golang.org/x/oauth2 v0.0.0-20200107190931-bf48bf16ab8d
-	google.golang.org/api v0.26.0
+	google.golang.org/api v0.29.0
 )
 
 go 1.14

google/config.go

@@ -287,6 +287,8 @@ func (c *Config) LoadAndValidate(ctx context.Context) error {
 		c.Scopes = DefaultClientScopes
 	}
 
+	c.context = ctx
+
 	tokenSource, err := c.getTokenSource(c.Scopes)
 	if err != nil {
 		return err
@@ -318,7 +320,6 @@ func (c *Config) LoadAndValidate(ctx context.Context) error {
 	userAgent := fmt.Sprintf("%s %s", tfUserAgent, providerVersion)
 
 	c.client = client
-	c.context = ctx
 	c.userAgent = userAgent
 
 	// This base path and some others below need the version and possibly more of the path

google/resource_active_directory_domain_update_test.go

@@ -1,6 +1,7 @@
 package google
 
 import (
+	"fmt"
 	"testing"
 
 	"github.com/hashicorp/terraform-plugin-sdk/helper/resource"
@@ -9,8 +10,9 @@ import (
 func TestAccActiveDirectoryDomain_update(t *testing.T) {
 	t.Parallel()
 
+	domain := fmt.Sprintf("mydomain%s.org1.com", randString(t, 5))
 	context := map[string]interface{}{
-		"domain":        "mydomain.org1.com",
+		"domain":        domain,
 		"resource_name": "ad-domain",
 	}
 

google/resource_bigtable_instance.go

@@ -52,7 +52,7 @@ func resourceBigtableInstance() *schema.Resource {
 				Type:        schema.TypeList,
 				Optional:    true,
 				Computed:    true,
-				Description: `A block of cluster configuration options. This can be specified 1 or 2 times.`,
+				Description: `A block of cluster configuration options. This can be specified at least once, and up to 4 times.`,
 				Elem: &schema.Resource{
 					Schema: map[string]*schema.Schema{
 						"cluster_id": {

google/resource_cloud_scheduler_job.go

@@ -322,7 +322,7 @@ a message to the provided topic`,
 							Required: true,
 							ForceNew: true,
 							Description: `The full resource name for the Cloud Pub/Sub topic to which
-messages will be published when a job is delivered. ~>**NOTE**:
+messages will be published when a job is delivered. ~>**NOTE:**
 The topic name must be in the same format as required by PubSub's
 PublishRequest.name, e.g. 'projects/my-project/topics/my-topic'.`,
 						},

google/resource_compute_forwarding_rule.go

@@ -80,7 +80,7 @@ forwarding rule. By default, if this field is empty, an ephemeral
 internal IP address will be automatically allocated from the IP range
 of the subnet or network configured for this forwarding rule.
 
-An address must be specified by a literal IP address. ~> **NOTE**: While
+An address must be specified by a literal IP address. ~> **NOTE:** While
 the API allows you to specify various resource paths for an address resource
 instead, Terraform requires this to specifically be an IP address to
 avoid needing to fetching the IP address from resource paths on refresh

google/resource_compute_router_interface_test.go

@@ -256,7 +256,7 @@ resource "google_compute_network" "foobar" {
 }
 
 resource "google_compute_subnetwork" "foobar" {
-  name          = "router-interface-test-subnetwork-%s"
+  name          = "tf-test-router-interface-subnetwork-%s"
   network       = google_compute_network.foobar.self_link
   ip_cidr_range = "10.0.0.0/16"
   region        = "us-central1"
@@ -317,7 +317,7 @@ resource "google_compute_network" "foobar" {
 }
 
 resource "google_compute_subnetwork" "foobar" {
-  name          = "router-interface-test-subnetwork-%s"
+  name          = "tf-test-router-interface-subnetwork-%s"
   network       = google_compute_network.foobar.self_link
   ip_cidr_range = "10.0.0.0/16"
   region        = "us-central1"

google/resource_deployment_manager_deployment.go

@@ -161,7 +161,7 @@ actually change the deployment, just how it is updated. Default value: "DELETE"
 that are not actually instantiated. This allows you to preview a
 deployment. It can be updated to false to actually deploy
 with real resources.
- ~>**NOTE**: Deployment Manager does not allow update
+ ~>**NOTE:** Deployment Manager does not allow update
 of a deployment in preview (unless updating to preview=false). Thus,
 Terraform will force-recreate deployments if either preview is updated
 to true or if other fields are updated while preview is true.`,

google/resource_healthcare_dicom_store_test.go

@@ -101,15 +101,14 @@ func TestAccHealthcareDicomStore_basic(t *testing.T) {
 				ImportState:       true,
 				ImportStateVerify: true,
 			},
-			// TODO(b/148536607): Uncomment once b/148536607 is fixed.
-			// {
-			// 	Config: testGoogleHealthcareDicomStore_basic(dicomStoreName, datasetName),
-			// },
-			// {
-			// 	ResourceName:      resourceName,
-			// 	ImportState:       true,
-			// 	ImportStateVerify: true,
-			// },
+			{
+				Config: testGoogleHealthcareDicomStore_basic(dicomStoreName, datasetName),
+			},
+			{
+				ResourceName:      resourceName,
+				ImportState:       true,
+				ImportStateVerify: true,
+			},
 		},
 	})
 }

google/resource_healthcare_fhir_store.go

@@ -211,7 +211,7 @@ value 2. The maximum depth allowed is 5.`,
 						"resource_types": {
 							Type:     schema.TypeList,
 							Optional: true,
-							Description: `Supply a FHIR resource type (such as "Patient" or "Observation"). See 
+							Description: `Supply a FHIR resource type (such as "Patient" or "Observation"). See
 https://www.hl7.org/fhir/valueset-resource-types.html for a list of all FHIR resource types. The server treats
 an empty list as an intent to stream all the supported resource types in this FHIR store.`,
 							Elem: &schema.Schema{

google/resource_source_repo_repository.go

@@ -15,17 +15,36 @@
 package google
 
 import (
+	"bytes"
 	"fmt"
 	"log"
 	"reflect"
 	"strconv"
 	"strings"
 	"time"
 
+	"github.com/hashicorp/terraform-plugin-sdk/helper/hashcode"
 	"github.com/hashicorp/terraform-plugin-sdk/helper/schema"
 	"github.com/hashicorp/terraform-plugin-sdk/helper/validation"
 )
 
+func resourceSourceRepoRepositoryPubSubConfigsHash(v interface{}) int {
+	if v == nil {
+		return 0
+	}
+
+	var buf bytes.Buffer
+	m := v.(map[string]interface{})
+
+	buf.WriteString(fmt.Sprintf("%s-", GetResourceNameFromSelfLink(m["topic"].(string))))
+	buf.WriteString(fmt.Sprintf("%s-", m["message_format"].(string)))
+	if v, ok := m["service_account_email"]; ok {
+		buf.WriteString(fmt.Sprintf("%s-", v.(string)))
+	}
+
+	return hashcode.String(buf.String())
+}
+
 func resourceSourceRepoRepository() *schema.Resource {
 	return &schema.Resource{
 		Create: resourceSourceRepoRepositoryCreate,
@@ -81,6 +100,7 @@ If unspecified, it defaults to the compute engine default service account.`,
 						},
 					},
 				},
+				Set: resourceSourceRepoRepositoryPubSubConfigsHash,
 			},
 			"size": {
 				Type:        schema.TypeInt,

website/docs/d/kms_secret.html.markdown

@@ -15,7 +15,7 @@ within your resource definitions.
 For more information see
 [the official documentation](https://cloud.google.com/kms/docs/encrypt-decrypt).
 
-~> **NOTE**: Using this data provider will allow you to conceal secret data within your
+~> **NOTE:** Using this data provider will allow you to conceal secret data within your
 resource definitions, but it does not take care of protecting that data in the
 logging output, plan output, or state output.  Please take care to secure your secret
 data outside of resource definitions.

website/docs/d/kms_secret_ciphertext.html.markdown

@@ -17,7 +17,7 @@ ciphertext within your resource definitions.
 For more information see
 [the official documentation](https://cloud.google.com/kms/docs/encrypt-decrypt).
 
-~> **NOTE**: Using this data source will allow you to conceal secret data within your
+~> **NOTE:** Using this data source will allow you to conceal secret data within your
 resource definitions, but it does not take care of protecting that data in the
 logging output, plan output, or state output.  Please take care to secure your secret
 data outside of resource definitions.

website/docs/guides/provider_reference.html.markdown

@@ -115,7 +115,7 @@ Values are expected to include the version of the service, such as
 `https://www.googleapis.com/compute/v1/`.
 
 * `batching` - (Optional) This block controls batching GCP calls for groups of specific resource types. Structure is documented below.
-~>**NOTE**: Batching is not implemented for the majority or resources/request types and is bounded by two values. If you are running into issues with slow batches
+~>**NOTE:** Batching is not implemented for the majority or resources/request types and is bounded by two values. If you are running into issues with slow batches
 resources, you may need to adjust one or both of 1) the core [`-parallelism`](https://www.terraform.io/docs/commands/apply.html#parallelism-n) flag, which controls how many concurrent resources are being operated on and 2) `send_after`, the time interval after which a batch is sent.
 
 * `request_timeout` - (Optional) A duration string controlling the amount of time

website/docs/r/app_engine_application.html.markdown

@@ -39,7 +39,7 @@ resource "google_app_engine_application" "app" {
 The following arguments are supported:
 
 * `project` - (Required) The project ID to create the application under.
-   ~>**NOTE**: GCP only accepts project ID, not project number. If you are using number,
+   ~>**NOTE:** GCP only accepts project ID, not project number. If you are using number,
    you may get a "Permission denied" error.
 
 * `location_id` - (Required) The [location](https://cloud.google.com/appengine/docs/locations)

website/docs/r/bigquery_table.html.markdown

@@ -113,7 +113,7 @@ The following arguments are supported:
 * `labels` - (Optional) A mapping of labels to assign to the resource.
 
 * `schema` - (Optional) A JSON schema for the table.
-    ~>**NOTE**: Because this field expects a JSON string, any changes to the
+    ~>**NOTE:** Because this field expects a JSON string, any changes to the
     string will create a diff, even if the JSON itself hasn't changed.
     If the API returns a different value for the same schema, e.g. it
     switched the order of values or replaced `STRUCT` field type with `RECORD`
@@ -166,7 +166,7 @@ The `external_data_configuration` block supports:
 * `schema` - (Optional) A JSON schema for the external table. Schema is required
     for CSV and JSON formats if autodetect is not on. Schema is disallowed
     for Google Cloud Bigtable, Cloud Datastore backups, Avro, ORC and Parquet formats.
-    ~>**NOTE**: Because this field expects a JSON string, any changes to the
+    ~>**NOTE:** Because this field expects a JSON string, any changes to the
     string will create a diff, even if the JSON itself hasn't changed.
     Furthermore drift for this field cannot not be detected because BigQuery
     only uses this schema to compute the effective schema for the table, therefore

website/docs/r/bigtable_instance.html.markdown

@@ -63,7 +63,8 @@ The following arguments are supported:
 
 * `name` - (Required) The name (also called Instance Id in the Cloud Console) of the Cloud Bigtable instance.
 
-* `cluster` - (Required) A block of cluster configuration options. This can be specified 1 or 2 times. See structure below.
+* `cluster` - (Required) A block of cluster configuration options. This can be specified at least once, and up to 4 times.
+See structure below.
 
 -----
 

website/docs/r/bigtable_table.html.markdown

@@ -13,7 +13,7 @@ Creates a Google Cloud Bigtable table inside an instance. For more information s
 [the official documentation](https://cloud.google.com/bigtable/) and
 [API](https://cloud.google.com/bigtable/docs/go/reference).
 
--> **Note**: It is strongly recommended to set `lifecycle { prevent_destroy = true }`
+-> **Note:** It is strongly recommended to set `lifecycle { prevent_destroy = true }`
 on tables in order to prevent accidental data loss. See
 [Terraform docs](https://www.terraform.io/docs/configuration/resources.html#prevent_destroy)
 for more information on lifecycle parameters.

website/docs/r/cloud_scheduler_job.html.markdown

@@ -288,7 +288,7 @@ The `pubsub_target` block supports:
 * `topic_name` -
   (Required)
   The full resource name for the Cloud Pub/Sub topic to which
-  messages will be published when a job is delivered. ~>**NOTE**:
+  messages will be published when a job is delivered. ~>**NOTE:**
   The topic name must be in the same format as required by PubSub's
   PublishRequest.name, e.g. `projects/my-project/topics/my-topic`.
 

website/docs/r/compute_forwarding_rule.html.markdown

@@ -386,7 +386,7 @@ The following arguments are supported:
   forwarding rule. By default, if this field is empty, an ephemeral
   internal IP address will be automatically allocated from the IP range
   of the subnet or network configured for this forwarding rule.
-  An address must be specified by a literal IP address. ~> **NOTE**: While
+  An address must be specified by a literal IP address. ~> **NOTE:** While
   the API allows you to specify various resource paths for an address resource
   instead, Terraform requires this to specifically be an IP address to
   avoid needing to fetching the IP address from resource paths on refresh

website/docs/r/deployment_manager_deployment.html.markdown

@@ -186,7 +186,7 @@ The `imports` block supports:
   that are not actually instantiated. This allows you to preview a
   deployment. It can be updated to false to actually deploy
   with real resources.
-   ~>**NOTE**: Deployment Manager does not allow update
+   ~>**NOTE:** Deployment Manager does not allow update
   of a deployment in preview (unless updating to preview=false). Thus,
   Terraform will force-recreate deployments if either preview is updated
   to true or if other fields are updated while preview is true.

website/docs/r/healthcare_fhir_store.html.markdown

@@ -225,7 +225,7 @@ The `stream_configs` block supports:
 
 * `resource_types` -
   (Optional)
-  Supply a FHIR resource type (such as "Patient" or "Observation"). See 
+  Supply a FHIR resource type (such as "Patient" or "Observation"). See
   https://www.hl7.org/fhir/valueset-resource-types.html for a list of all FHIR resource types. The server treats
   an empty list as an intent to stream all the supported resource types in this FHIR store.
 

website/docs/r/kms_secret_ciphertext.html.markdown

@@ -25,7 +25,7 @@ description: |-
 Encrypts secret data with Google Cloud KMS and provides access to the ciphertext.
 
 
-~> **NOTE**: Using this resource will allow you to conceal secret data within your
+~> **NOTE:** Using this resource will allow you to conceal secret data within your
 resource definitions, but it does not take care of protecting that data in the
 logging output, plan output, or state output.  Please take care to secure your secret
 data outside of resource definitions.

website/docs/r/spanner_database_iam.html.markdown

@@ -126,6 +126,6 @@ IAM policy imports use the identifier of the resource in question, e.g.
 $ terraform import google_spanner_database_iam_policy.database project-name/instance-name/database-name
 ```
 
--> **Custom Roles**: If you're importing a IAM resource with a custom role, make sure to use the
+-> **Custom Roles:** If you're importing a IAM resource with a custom role, make sure to use the
  full name of the custom role, e.g. `[projects/my-project|organizations/my-org]/roles/my-custom-role`.
 

website/docs/r/sql_database_instance.html.markdown

@@ -122,7 +122,7 @@ resource "google_sql_database_instance" "postgres" {
 ```
 
 ### Private IP Instance
-~> **NOTE**: For private IP instance setup, note that the `google_sql_database_instance` does not actually interpolate values from `google_service_networking_connection`. You must explicitly add a `depends_on`reference as shown below.
+~> **NOTE:** For private IP instance setup, note that the `google_sql_database_instance` does not actually interpolate values from `google_service_networking_connection`. You must explicitly add a `depends_on`reference as shown below.
 
 ```hcl
 resource "google_compute_network" "private_network" {