如何使用 AWS Java SDK 按降序从 DynamoDb 获取项目？

Question

我在 AWS DynamoDB 中有一个名为 accountantHistoric 的 table。它有一个主分区键 (id) 和一个主排序键 (tenant_id)。我还创建了另一个索引，称为 "ExpirationDate"。该索引的排序键是 "expirationDate"(String) 属性。我想要的是：获取基于 storeId 的 itens（此字段不是键）和基于 "expirationDate"（从最新到最旧）降序排列的 tenant_id。我也在做分页。

我正在尝试使用下面的代码，但它没有按降序返回。

我需要更改什么？

谢谢大家！

public AccountantEmailHistoricInfo getHistoric(String storeId, int page, int countReg, int fixedPage) {
            Condition conditionStoreId = new Condition().withComparisonOperator(ComparisonOperator.EQ)
                    .withAttributeValueList(new AttributeValue().withS(storeId));

            Condition conditionTenantId = new Condition().withComparisonOperator(ComparisonOperator.EQ)
                    .withAttributeValueList(new AttributeValue().withS(TenantIdentifierLocalThread.getTenantIdentifier()));

            final DynamoDBScanExpression scanPageExpression = new DynamoDBScanExpression()
                    .withFilterConditionEntry("tenant_id", conditionTenantId)
                    .withFilterConditionEntry("storeId", conditionStoreId)
                    .withIndexName("ExpirationDate")
                    .withLimit(countReg);

            List<AccountantEmailHistoric> accountantEmailHistoricList = new ArrayList<>();
            int i = 0;

            do {
                ScanResultPage<AccountantEmailHistoric> scanPage = mapper.scanPage(AccountantEmailHistoric.class, scanPageExpression);
                scanPageExpression.setExclusiveStartKey(scanPage.getLastEvaluatedKey());

                if (i == page) {
                    accountantEmailHistoricList.addAll(scanPage.getResults());
                }
                i++;
            } while (scanPageExpression.getExclusiveStartKey() != null);

            final DynamoDBScanExpression scan = new DynamoDBScanExpression()
                    .withFilterConditionEntry("tenant_id", conditionTenantId)
                    .withFilterConditionEntry("storeId", conditionStoreId);

            AccountantEmailHistoricInfo info = new AccountantEmailHistoricInfo(accountantEmailHistoricList);
            info.setPage(fixedPage);
            info.setTotalRecords(mapper.scan(AccountantEmailHistoric.class, scan).size());
            return info;
        }

Answer 1

您需要query，而不是扫描。然后你可以把ScanIndexForward设为false。