使用 DynamoDB 專用 PartiQL 執行交易

本節說明如何搭配 DynamoDB 專用 PartiQL 使用交易。PartiQL 交易限制為全部 100 個陳述式 (動作)。

如需 DynamoDB 交易的詳細資訊，請參閱管理包含 DynamoDB 交易的複雜工作流程。

注意

整個交易必須由讀取陳述式或寫入陳述式所組成。你不能在一個交易中混合兩者。EXISTS 函數為例外。您可以使用它來檢查項目的特定屬性的條件，以類似的方式ConditionCheck在TransactWrite項目 API 操作。

語法

[
   {
      "Statement":" statement ",
      "Parameters":[
         {
            " parametertype " : " parametervalue "
         }, ...]
   } , ...
]

參數

陳述式

(必要) DynamoDB 專用 PartiQL 支援的陳述式。

注意

整個交易必須由讀取陳述式或寫入陳述式所組成。你不能在一個交易中混合兩者。

parametertype

(選用) DynamoDB 類型，如果在指定 PartiQL 陳述式時使用了參數。

parametervalue

(選用) 參數值，如果在指定 PartiQL 陳述式時使用了參數。

傳回值

此陳述式不會傳回寫入作業 (插入、更新或刪除) 的任何值。但是，它根據 WHERE 子句中指定的條件回傳讀取操作 (SELECT) 的不同值。

注意

如果任何單一 INSERT、UPDATE 或 DELETE 操作傳回錯誤，交易會取消並出現 TransactionCanceledException 例外狀況，且取消原因程式碼會包含個別單一操作的錯誤。

範例

以下範例會以交易形式執行多個陳述式。

AWS CLI

1. 將以下 JSON 程式碼儲存至名為 partiql.json 的檔案

   [
       {
           "Statement": "EXISTS(SELECT * FROM \"Music\" where Artist='No One You Know' and SongTitle='Call Me Today' and Awards is  MISSING)"
       },
       {
           "Statement": "INSERT INTO Music value {'Artist':?,'SongTitle':'?'}",
           "Parameters": [{\"S\": \"Acme Band\"}, {\"S\": \"Best Song\"}]
       },
       {
           "Statement": "UPDATE \"Music\" SET AwardsWon=1 SET AwardDetail={'Grammys':[2020, 2018]}  where Artist='Acme Band' and SongTitle='PartiQL Rocks'"
       }
   ]

2. 在命令提示中執行下列命令。

   aws dynamodb execute-transaction --transact-statements  file://partiql.json

Java

public class DynamoDBPartiqlTransaction {

    public static void main(String[] args) {
        // Create the DynamoDB Client with the region you want
        AmazonDynamoDB dynamoDB = createDynamoDbClient("us-west-2");
        
        try {
            // Create ExecuteTransactionRequest
            ExecuteTransactionRequest executeTransactionRequest = createExecuteTransactionRequest();
            ExecuteTransactionResult executeTransactionResult = dynamoDB.executeTransaction(executeTransactionRequest);
            System.out.println("ExecuteTransaction successful.");
            // Handle executeTransactionResult

        } catch (Exception e) {
            handleExecuteTransactionErrors(e);
        }
    }

    private static AmazonDynamoDB createDynamoDbClient(String region) {
        return AmazonDynamoDBClientBuilder.standard().withRegion(region).build();
    }

    private static ExecuteTransactionRequest createExecuteTransactionRequest() {
        ExecuteTransactionRequest request = new ExecuteTransactionRequest();
        
        // Create statements
        List<ParameterizedStatement> statements = getPartiQLTransactionStatements();

        request.setTransactStatements(statements);
        return request;
    }

    private static List<ParameterizedStatement> getPartiQLTransactionStatements() {
        List<ParameterizedStatement> statements = new ArrayList<ParameterizedStatement>();

        statements.add(new ParameterizedStatement()
                               .withStatement("EXISTS(SELECT * FROM "Music" where Artist='No One You Know' and SongTitle='Call Me Today' and Awards is  MISSING)"));

        statements.add(new ParameterizedStatement()
                               .withStatement("INSERT INTO "Music" value {'Artist':'?','SongTitle':'?'}")
                               .withParameters(new AttributeValue("Acme Band"),new AttributeValue("Best Song")));

        statements.add(new ParameterizedStatement()
                               .withStatement("UPDATE "Music" SET AwardsWon=1 SET AwardDetail={'Grammys':[2020, 2018]}  where Artist='Acme Band' and SongTitle='PartiQL Rocks'"));

        return statements;
    }

    // Handles errors during ExecuteTransaction execution. Use recommendations in error messages below to add error handling specific to 
    // your application use-case.
    private static void handleExecuteTransactionErrors(Exception exception) {
        try {
            throw exception;
        } catch (TransactionCanceledException tce) {
            System.out.println("Transaction Cancelled, implies a client issue, fix before retrying. Error: " + tce.getErrorMessage());
        } catch (TransactionInProgressException tipe) {
            System.out.println("The transaction with the given request token is already in progress, consider changing " +
                "retry strategy for this type of error. Error: " + tipe.getErrorMessage());
        } catch (IdempotentParameterMismatchException ipme) {
            System.out.println("Request rejected because it was retried with a different payload but with a request token that was already used, " +
                "change request token for this payload to be accepted. Error: " + ipme.getErrorMessage());
        } catch (Exception e) {
            handleCommonErrors(e);
        }
    }

    private static void handleCommonErrors(Exception exception) {
        try {
            throw exception;
        } catch (InternalServerErrorException isee) {
            System.out.println("Internal Server Error, generally safe to retry with exponential back-off. Error: " + isee.getErrorMessage());
        } catch (RequestLimitExceededException rlee) {
            System.out.println("Throughput exceeds the current throughput limit for your account, increase account level throughput before " + 
                "retrying. Error: " + rlee.getErrorMessage());
        } catch (ProvisionedThroughputExceededException ptee) {
            System.out.println("Request rate is too high. If you're using a custom retry strategy make sure to retry with exponential back-off. " +
                "Otherwise consider reducing frequency of requests or increasing provisioned capacity for your table or secondary index. Error: " + 
                ptee.getErrorMessage());
        } catch (ResourceNotFoundException rnfe) {
            System.out.println("One of the tables was not found, verify table exists before retrying. Error: " + rnfe.getErrorMessage());
        } catch (AmazonServiceException ase) {
            System.out.println("An AmazonServiceException occurred, indicates that the request was correctly transmitted to the DynamoDB " + 
                "service, but for some reason, the service was not able to process it, and returned an error response instead. Investigate and " +
                "configure retry strategy. Error type: " + ase.getErrorType() + ". Error message: " + ase.getErrorMessage());
        } catch (AmazonClientException ace) {
            System.out.println("An AmazonClientException occurred, indicates that the client was unable to get a response from DynamoDB " +
                "service, or the client was unable to parse the response from the service. Investigate and configure retry strategy. "+
                "Error: " + ace.getMessage());
        } catch (Exception e) {
            System.out.println("An exception occurred, investigate and configure retry strategy. Error: " + e.getMessage());
        }
    }

}

下列範例顯示當 DynamoDB 讀取 WHERE 子句中指定之不同條件的項目時，不同的傳回值。

AWS CLI

1. 將以下 JSON 程式碼儲存至名為 partiql.json 的檔案

   [
       // Item exists and projected attribute exists
       {
           "Statement": "SELECT * FROM "Music" WHERE Artist='No One You Know' and SongTitle='Call Me Today'"
       },
       // Item exists but projected attributes do not exist
       {
           "Statement": "SELECT non_existent_projected_attribute FROM "Music" WHERE Artist='No One You Know' and SongTitle='Call Me Today'"
       },
       // Item does not exist
       {
           "Statement": "SELECT * FROM "Music" WHERE Artist='No One I Know' and SongTitle='Call You Today'"
       }
   ]

2. 在命令提示中執行命令。

   aws dynamodb execute-transaction --transact-statements  file://partiql.json

3. 會傳回下列回應：

   {
       "Responses": [
           // Item exists and projected attribute exists
           {
               "Item": {
                   "Artist":{
                       "S": "No One You Know"
                   },
                   "SongTitle":{
                       "S": "Call Me Today"
                   }    
               }
           },
           // Item exists but projected attributes do not exist
           {
               "Item": {}
           },
           // Item does not exist
           {}
       ]
   }