本文為英文版的機器翻譯版本，如內容有任何歧義或不一致之處，概以英文版為準。

Amazon DataZone 快速入門範例指令碼

您可以透過 DataZone 管理入口網站或 Amazon DataZone 資料入口網站存取 Amazon，或使用 Amazon 以程式設計方式存取 Amazon DataZone HTTPSAPI，這可讓您直接向 服務發出HTTPS請求。本節包含叫用 Amazon DataZone APIs 的範例指令碼，可用來完成下列常見任務：

建立 Amazon DataZone 網域和資料入口網站

您可以使用下列範例指令碼來建立 Amazon DataZone 網域。如需 Amazon DataZone 網域的詳細資訊，請參閱 Amazon DataZone 術語和概念。

import sys
import boto3

// Initialize datazone client
region = 'us-east-1'
dzclient = boto3.client(service_name='datazone', region_name='us-east-1')

// Create DataZone domain
def create_domain(name):
    return dzclient.create_domain(
        name = name,
        description = "this is a description",
        domainExecutionRole = "arn:aws:iam::<account>:role/AmazonDataZoneDomainExecutionRole",
    )
      

建立發佈專案

您可以使用下列範例指令碼，在 Amazon 中建立發佈專案 DataZone。

// Create Project
def create_project(domainId):
    return dzclient.create_project(
        domainIdentifier = domainId,
        name = "sample-project"
    )

建立環境設定檔

您可以使用下列範例指令碼，在 Amazon 中建立環境設定檔 DataZone。

叫用 CreateEnvironmentProfile API 時，會使用此範例承載：

Sample Payload
{
    "Content":{
        "project_name": "Admin_project",
        "domain_name": "Drug-Research-and-Development",
        "blueprint_account_region": [
            {
                "blueprint_name": "DefaultDataLake",
                "account_id": ["066535990535",
                "413878397724",
                "676266385322", 
                "747721550195", 
                "755347404384"
                ],
                "region": ["us-west-2", "us-east-1"]
            },
            {
                "blueprint_name": "DefaultDataWarehouse",
                "account_id": ["066535990535",
                "413878397724",
                "676266385322", 
                "747721550195", 
                "755347404384"
                ],
                "region":["us-west-2", "us-east-1"]
            }
        ]
    }
}
      

此範例指令碼會叫用 CreateEnvironmentProfile API：

def create_environment_profile(domain_id, project_id, env_blueprints)    
        try:
            response = dz.list_environment_blueprints(
                domainIdentifier=domain_id,
                managed=True
            )
            env_blueprints = response.get("items")
            env_blueprints_map = {}
            for i in env_blueprints:
                env_blueprints_map[i["name"]] = i['id']
            
            print("Environment Blueprint map", env_blueprints_map)
            for i in blueprint_account_region:
                print(i)
                for j in i["account_id"]:
                    for k in i["region"]:
                        print("The env blueprint name is", i['blueprint_name'])
                        dz.create_environment_profile(
                            description='This is a test environment profile created via lambda function',
                            domainIdentifier=domain_id,
                            awsAccountId=j,
                            awsAccountRegion=k,
                            environmentBlueprintIdentifier=env_blueprints_map.get(i["blueprint_name"]),
                            name=i["blueprint_name"] + j + k + "_profile",
                            projectIdentifier=project_id
                        )
        except Exception as e:
            print("Failed to created Environment Profile")
            raise e 
      

這是叫用 CreateEnvironmentProfile API 時的範例輸出承載：

{
    "Content":{
        "project_name": "Admin_project",
        "domain_name": "Drug-Research-and-Development",
        "blueprint_account_region": [
            {
                "blueprint_name": "DefaultDataWarehouse",
                "account_id": ["111111111111"],
                "region":["us-west-2"],
                "user_parameters":[
                    {
                        "name": "dataAccessSecretsArn",
                        "value": ""
                    }
                ] 
            }
        ]
    }
}
      

建立環境

您可以使用下列範例指令碼，在 Amazon 中建立環境 DataZone。

def create_environment(domain_id, project_id,blueprint_account_region ):
         try:
            #refer to get_domain_id and get_project_id for fetching ids using names.
            sts_client = boto3.client("sts")
            # Get the current account ID
            account_id = sts_client.get_caller_identity()["Account"]
            print("Fetching environment profile ids")
            env_profile_map = get_env_profile_map(domain_id, project_id)

            for i in blueprint_account_region:
                for j in i["account_id"]:
                    for k in i["region"]:
                        print(" env blueprint name", i['blueprint_name'])
                        profile_name = i["blueprint_name"] + j + k + "_profile"
                        env_name = i["blueprint_name"] + j + k + "_env"
                        description = f'This is environment is created for {profile_name}, Account {account_id} and region {i["region"]}'
                        try:
                            dz.create_environment(
                                description=description,
                                domainIdentifier=domain_id,
                                environmentProfileIdentifier=env_profile_map.get(profile_name),
                                name=env_name,
                                projectIdentifier=project_id
                            )
                            print(f"Environment created - {env_name}")
                        except:
                            dz.create_environment(
                                description=description,
                                domainIdentifier=domain_id,
                                environmentProfileIdentifier=env_profile_map.get(profile_name),
                                name=env_name,
                                projectIdentifier=project_id,
                                userParameters= i["user_parameters"] 
                            )
                            print(f"Environment created - {env_name}")
        except Exception as e:
            print("Failed to created Environment")
            raise e  
      

從 AWS Glue 收集中繼資料

您可以使用此範例指令碼從 AWS Glue 收集中繼資料。此指令碼會以標準排程執行。您可以從範例指令碼擷取參數，並將其設為全域。使用標準函數擷取專案、環境和網域 ID。 AWS Glue 資料來源是在標準時間建立和執行，可在指令碼的 cron 區段中更新。

def crcreate_data_source(domain_id, project_id,data_source_name)
        print("Creating Data Source")
        data_source_creation = dz.create_data_source(
            # Define data source : Customize the data source to which you'd like to connect
            # define the name of the Data source to create, example: name ='TestGlueDataSource'
            name=data_source_name,
            # give a description for the datasource (optional), example: description='This is a dorra test for creation on DZ datasources'
            description=data_source_description,
            # insert the domain identifier corresponding to the domain to which the datasource will belong, example: domainIdentifier= 'dzd_6f3gst5jjmrrmv'
            domainIdentifier=domain_id,
            # give environment identifier , example: environmentIdentifier= '3weyt6hhn8qcvb'
            environmentIdentifier=environment_id,
            # give corresponding project identifier, example: projectIdentifier= '6tl4csoyrg16ef',
            projectIdentifier=project_id,
            enableSetting="ENABLED",
            # publishOnImport used to select whether assets are added to the inventory and/or discovery catalog .
            # publishOnImport = True : Assets will be added to project's inventory as well as published to the discovery catalog
            # publishOnImport = False : Assets will only be added to project's inventory.
            # You can later curate the metadata of the assets and choose subscription terms to publish them from the inventory to the discovery catalog.
            publishOnImport=False,
            # Automated business name generation : Use AI to automatically generate metadata for assets as they are published or updated by this data source run.
            # Automatically generated metadata can be be approved, rejected, or edited by data publishers.
            # Automatically generated metadata is badged with a small icon next to the corresponding metadata field.
            recommendation={"enableBusinessNameGeneration": True},
            type="GLUE",
            configuration={
                "glueRunConfiguration": {
                    "dataAccessRole": "arn:aws:iam::"
                    + account_id
                    + ":role/service-role/AmazonDataZoneGlueAccess-"
                    + current_region
                    + "-"
                    + domain_id
                    + "",
                    "relationalFilterConfigurations": [
                        {
                            #
                            "databaseName": glue_database_name,
                            "filterExpressions": [
                                {"expression": "*", "type": "INCLUDE"},
                            ],
                            #    "schemaName": "TestSchemaName",
                        },
                    ],
                },
            },
            # Add metadata forms to the data source (OPTIONAL).
            # Metadata forms will be automatically applied to any assets that are created by the data source.
            # assetFormsInput=[
            #     {
            #         "content": "string",
            #         "formName": "string",
            #         "typeIdentifier": "string",
            #         "typeRevision": "string",
            #     },
            # ],
            schedule={
                "schedule": "cron(5 20 * * ? *)",
                "timezone": "UTC",
            },
        )
        # This is a suggested syntax to return values
        #        return_values["data_source_creation"] = data_source_creation["items"]
        print("Data Source Created")


//This is the sample response payload after the CreateDataSource API is invoked:

{
    "Content":{
        "project_name": "Admin",
        "domain_name": "Drug-Research-and-Development",
        "env_name": "GlueEnvironment",
        "glue_database_name": "test",
        "data_source_name" : "test",
        "data_source_description" : "This is a test data source"
    }
}
      

整理和發佈資料資產

您可以使用下列範例指令碼，在 Amazon 中策劃和發佈資料資產 DataZone。

您可以使用下列指令碼來建立自訂表單類型：

 
def create_form_type(domainId, projectId):
    return dzclient.create_form_type(
        domainIdentifier = domainId,
        name = "customForm",
        model = {
            "smithy": "structure customForm { simple: String }"
        },
        owningProjectIdentifier = projectId,
        status = "ENABLED"
    )
        

您可以使用下列範例指令碼來建立自訂資產類型：

def create_custom_asset_type(domainId, projectId):
    return dzclient.create_asset_type(
        domainIdentifier = domainId,
        name = "userCustomAssetType",
        formsInput = {
            "Model": {
                "typeIdentifier": "customForm",
                "typeRevision": "1",
                "required": False
            }
        },
        owningProjectIdentifier = projectId,
    )
      

您可以使用下列範例指令碼來建立自訂資產：

def create_custom_asset(domainId, projectId):
    return dzclient.create_asset(
        domainIdentifier = domainId,
        name = 'custom asset',
        description = "custom asset",
        owningProjectIdentifier = projectId,
        typeIdentifier = "userCustomAssetType",
        formsInput = [
            {
                "formName": "UserCustomForm",
                "typeIdentifier": "customForm",
                "content": "{\"simple\":\"sample-catalogId\"}"
            }
        ]
    )
      

您可以使用下列範例指令碼來建立詞彙表：

def create_glossary(domainId, projectId):
    return dzclient.create_glossary(
        domainIdentifier = domainId,
        name = "test7",
        description = "this is a test glossary",
        owningProjectIdentifier = projectId
    )
      

您可以使用下列範例指令碼來建立詞彙表術語：

def create_glossary_term(domainId, glossaryId):
    return dzclient.create_glossary_term(
        domainIdentifier = domainId,
        name = "soccer",
        shortDescription = "this is a test glossary",
        glossaryIdentifier = glossaryId,
    )
      

您可以使用下列範例指令碼，使用系統定義的資產類型建立資產：

def create_asset(domainId, projectId):
    return dzclient.create_asset(
        domainIdentifier = domainId,
        name = 'sample asset name',
        description = "this is a glue table asset",
        owningProjectIdentifier = projectId,
        typeIdentifier = "amazon.datazone.GlueTableAssetType",
        formsInput = [
            {
                "formName": "GlueTableForm",
                "content": "{\"catalogId\":\"sample-catalogId\",\"columns\":[{\"columnDescription\":\"sample-columnDescription\",\"columnName\":\"sample-columnName\",\"dataType\":\"sample-dataType\",\"lakeFormationTags\":{\"sample-key1\":\"sample-value1\",\"sample-key2\":\"sample-value2\"}}],\"compressionType\":\"sample-compressionType\",\"lakeFormationDetails\":{\"lakeFormationManagedTable\":false,\"lakeFormationTags\":{\"sample-key1\":\"sample-value1\",\"sample-key2\":\"sample-value2\"}},\"primaryKeys\":[\"sample-Key1\",\"sample-Key2\"],\"region\":\"us-east-1\",\"sortKeys\":[\"sample-sortKey1\"],\"sourceClassification\":\"sample-sourceClassification\",\"sourceLocation\":\"sample-sourceLocation\",\"tableArn\":\"sample-tableArn\",\"tableDescription\":\"sample-tableDescription\",\"tableName\":\"sample-tableName\"}"
            }
        ]
    )
      

您可以使用下列範例指令碼來建立資產修訂並連接詞彙表術語：

def create_asset_revision(domainId, assetId):
    return dzclient.create_asset_revision(
        domainIdentifier = domainId,
        identifier = assetId,
        name = 'glue table asset 7',
        description = "glue table asset description update",
        formsInput = [
            {
                "formName": "GlueTableForm",
                "content": "{\"catalogId\":\"sample-catalogId\",\"columns\":[{\"columnDescription\":\"sample-columnDescription\",\"columnName\":\"sample-columnName\",\"dataType\":\"sample-dataType\",\"lakeFormationTags\":{\"sample-key1\":\"sample-value1\",\"sample-key2\":\"sample-value2\"}}],\"compressionType\":\"sample-compressionType\",\"lakeFormationDetails\":{\"lakeFormationManagedTable\":false,\"lakeFormationTags\":{\"sample-key1\":\"sample-value1\",\"sample-key2\":\"sample-value2\"}},\"primaryKeys\":[\"sample-Key1\",\"sample-Key2\"],\"region\":\"us-east-1\",\"sortKeys\":[\"sample-sortKey1\"],\"sourceClassification\":\"sample-sourceClassification\",\"sourceLocation\":\"sample-sourceLocation\",\"tableArn\":\"sample-tableArn\",\"tableDescription\":\"sample-tableDescription\",\"tableName\":\"sample-tableName\"}"
            }
        ],
        glossaryTerms = ["<glossaryTermId:>"]
    )
      

您可以使用下列範例指令碼來發佈資產：

def publish_asset(domainId, assetId):
    return dzclient.create_listing_change_set(
        domainIdentifier = domainId,
        entityIdentifier = assetId,
        entityType = "ASSET",
        action = "PUBLISH",
    )
      

搜尋資料目錄並訂閱資料

您可以使用下列範例指令碼來搜尋資料目錄並訂閱資料：

def search_asset(domainId, projectId, text):
    return dzclient.search(
        domainIdentifier = domainId,
        owningProjectIdentifier = projectId,
        searchScope = "ASSET",
        searchText = text,
    )
      

您可以使用下列範例指令碼來取得資產的清單 ID：

def search_listings(domainId, assetName, assetId):
    listings = dzclient.search_listings(
        domainIdentifier=domainId,
        searchText=assetName,
        additionalAttributes=["FORMS"]
    )
    
    assetListing = None
    for listing in listings['items']:
        if listing['assetListing']['entityId'] == assetId: 
            assetListing = listing
    
    return listing['assetListing']['listingId'] 
      

您可以使用下列範例指令碼，使用清單 ID 建立訂閱請求：

create_subscription_response = def create_subscription_request(domainId, projectId, listingId):
    return dzclient.create_subscription_request(
        subscribedPrincipals=[{
            "project": {
                "identifier": projectId
            }
        }],
        subscribedListings=[{
            "identifier": listingId
        }],
        requestReason="Give request reason here."
    )
      

使用create_subscription_response上述 取得 subscription_request_id，然後使用下列範例指令碼接受/核准訂閱：

subscription_request_id = create_subscription_response["id"]

def accept_subscription_request(domainId, subscriptionRequestId): 
    return dzclient.accept_subscription_request(
        domainIdentifier=domainId,
        identifier=subscriptionRequestId
    )
      

在資料目錄中搜尋資產

您可以使用下列範例指令碼，利用自由文字搜尋來查詢 Amazon DataZone 目錄中已發佈的資料資產 （清單）。

您也可以使用篩選條件搜尋清單。中的 filters 參數 SearchListings API可讓您從網域擷取篩選結果。API 支援多個預設篩選條件，您也可以合併兩個或多個篩選條件，並對其執行 AND/OR 操作。篩選條件子句包含兩個參數：屬性和值。預設支援的篩選條件屬性為 typeName、 owningProjectId和 glossaryTerms。

其他有用的範例指令碼

當您在 Amazon 中使用資料時，您可以使用下列範例指令碼來完成各種任務 DataZone。

使用下列範例指令碼列出現有的 Amazon DataZone 網域：

def list_domains():
    datazone = boto3.client('datazone')
    response = datazone.list_domains(status='AVAILABLE')
    [print("%12s | %16s | %12s | %52s" % (item['id'], item['name'], item['managedAccountId'], item['portalUrl'])) for item in response['items']]
    return
      

使用下列範例指令碼來列出現有的 Amazon DataZone 專案：

def list_projects(domain_id):
    datazone = boto3.client('datazone')
    response = datazone.list_projects(domainIdentifier=domain_id)
    [print("%12s | %16s " % (item['id'], item['name'])) for item in response['items']]
    return
      

使用下列範例指令碼列出現有的 Amazon DataZone 中繼資料表單：

def list_metadata_forms(domain_id):
    datazone = boto3.client('datazone')
    response = datazone.search_types(domainIdentifier=domain_id, 
        managed=False,
        searchScope='FORM_TYPE')
    [print("%16s | %16s | %3s | %8s" % (item['formTypeItem']['name'], item['formTypeItem']['owningProjectId'],item['formTypeItem']['revision'], item['formTypeItem']['status'])) for item in response['items']]
    return