本文属于机器翻译版本。若本译文内容与英语原文存在差异，则一律以英文原文为准。

# AWS Marketplace 中的卖家传输数据源
<a name="data-feed-service"></a>

AWS Marketplace 将数据源作为一种机制提供，用于将最新的结构化产品和客户信息从 AWS Marketplace 系统发送到卖家 Amazon S3 存储桶，以便在卖家所有的商业智能工具之间进行 ETL（提取、转换和加载）。数据源收集逗号分隔值 (CSV) 文件并将其传送到您提供的已加密 Amazon S3 存储桶。数据源在一天内生成，包含前一天 24 小时的数据。以下各节概述了数据源，并说明了如何访问和使用它们。随后的部分将介绍每个数据源。

交易数据以双时态结构传输和附加，因此卖家可以根据两个时间线存储和查询数据，且它们都具有时间戳
+ 有效时间：某个事实实际发生的时间（“您所知道的”）
+ 系统时间：将该事实记录到数据库的时间（“当您知道时”）。

数据源将在世界标准时间每天午夜传输，此前一天的更新包含前一天的 24 小时数据。更新可以通过客户订阅、向客户开具发票或 AWS 付款来定义。

**Topics**
+ [AWS Marketplace 数据源的存储和结构](data-feed-details.md)
+ [访问数据源](data-feed-accessing.md)
+ [使用数据源收集和分析数据](data-feed-using.md)
+ [数据源表概述](data-feed-joining.md)
+ [数据源查询示例](data-feed-full-examples.md)
+ [数据源](data-feeds.md)

# AWS Marketplace 数据源的存储和结构
<a name="data-feed-details"></a>

AWS Marketplace 将数据源作为一种机制提供，用于将最新的结构化产品和客户信息从 AWS Marketplace 系统发送到卖家 Amazon S3 存储桶，以便在卖家所有的商业智能工具之间进行 ETL（提取、转换和加载）。本主题将提供有关数据源的结构和存储的更多信息。

数据源收集逗号分隔值 (CSV) 文件并将其传送到您提供的已加密 Amazon S3 存储桶。CSV 文件具有以下特征：
+ 它们遵循 [4180 标准](https://tools.ietf.org/html/rfc4180)。
+ 字符编码是无 BOM 的 UTF-8。
+ 使用逗号作为值之间的分隔符。
+ 字段由双引号进行转义。
+ `\n` 是换行符。
+ 日期按 UTC 时区进行报告，采用 ISO 8601 日期和时间格式，并精确到 1 秒。
+ 所有 `*_period_start_date` 和 `*_period_end_date` 值都包括在内，这意味着 `23:59:59` 是任何一天的最后一个可能的时间戳。
+ 所有货币字段前面都有币种字段。
+ 货币字段使用句点 (`.`) 字符作为小数分隔符，不使用逗号 (,) 作为千位分隔符。

数据源的生成和存储方式如下：
+ 数据源在一天内生成，包含前一天 24 小时的数据。
+ 在 Amazon S3 存储桶中，数据源使用以下格式按月进行组织：

  `bucket-name/data-feed-name_version/year=YYYY/month=MM/data.csv`
+ 生成每个每日数据源时，它将附加到该月的现有 CSV 文件中。当新月份开始时，将为每个数据源生成一个新的 CSV 文件。
+ 数据源中的信息从 2010/01/01 到 2020/04/30（含）进行回填，并且可以在 `year=2010/month=01` 子文件夹的 [CSV 文件](#data-feed-details)中找到。

  您可能会注意到本月的给定数据源文件仅包含列标题，而不包含数据的情况。这意味着该月该源没有新条目。对于更新频率较低的数据源（如产品源），可能会发生这种情况。在这些情况下，在回填文件夹中提供数据。
+ 在 Amazon S3 中，您可以创建 [Amazon S3 生命周期策略](https://docs.aws.amazon.com/AmazonS3/latest/userguide/create-lifecycle.html)来管理文件在存储桶中的保留时间。
+ 您可以配置 Amazon SNS 以在数据传送到已加密的 Amazon S3 存储桶时向您发送通知。有关配置通知的更多信息，请参阅《Amazon Simple Notification Service 开发人员指南》**中的 [Amazon SNS 入门](https://docs.aws.amazon.com/sns/latest/dg/sns-getting-started.html)。

## 数据历史记录
<a name="data-feed-historization"></a>

每个数据源都包含记录数据历史记录的列。除了 `valid_to` 之外，这些列对于所有数据源都是通用的。它们作为通用历史记录架构包含在内，并且在查询数据时非常有用。


| 列名称  | 描述  | 
| --- | --- | 
| valid\$1from | 主键值相对于其他字段的值有效的第一个日期。 | 
| valid\$1to | 此列仅显示在[地址](data-feed-address.md)数据源中，且始终为空。 | 
| insert\$1date | 将记录插入数据源中的日期。 | 
| update\$1date | 上次更新记录的日期。 | 
| delete\$1date | 该列始终为空。 | 

下面显示了这些列的示例。


|  valid\$1from  |  valid\$1to  |  insert\$1date  |  update\$1date  |  delete\$1date  | 
| --- | --- | --- | --- | --- | 
|  2018-12-12T02:00:00Z  |   |  2018-12-12T02:00:00Z  |  2018-12-12T02:00:00Z  |   | 
|  2019-03-29T03:00:00Z  |   |  2019-03-29T03:00:00Z  |  2019-03-29T03:00:00Z  |   | 
|  2019-03-29T03:00:00Z  |   |  2019-03-29T03:00:00Z  |  2019-04-28T03:00:00Z  |   | 

`valid_from` 和 `update_date` 字段共同构成了*双时态数据模型*。`valid_from` 字段的名字会表明该项目何时有效。如果项目已被编辑，源中可以有多条记录，每条记录都有不同 `update_date` 和相同的 `valid_from` 日期。例如，要查找某项的当前值，可以从具有最新 `valid_from` 日期的记录列表中查找具有最新 `update_date` 的记录。

在上面的示例中，该记录最初创建于 2018-12-12。然后在 2019-03-29 进行了更改（例如，记录中的地址发生了变更）。后来，在 2019-04-28，地址变更得到了更正（所以 `valid_from` 没有改变，但 `update_date` 变了）。更正地址（罕见事件）会追溯更改原始 `valid_from` 日期的记录，因此该字段不会更改。查找最新 `valid_from` 的查询将返回两条记录，具有最新 `update_date` 的记录会给出实际的当前记录。

# 访问数据源
<a name="data-feed-accessing"></a>

借助 AWS Marketplace，您可以将数据源用作一种机制，将最新的结构化产品和客户信息从 AWS Marketplace 系统发送到 Amazon S3 存储桶，以便在卖家所有的商业智能工具之间进行 ETL（提取、转换和加载）。您需要对环境进行配置，以将数据源接收到加密的 Amazon S3 存储桶中。本主题将向您介绍如何访问和取消订阅数据源。

**Topics**
+ [访问数据源](#data-feed-accessing-procedure)
+ [数据源策略](#data-feed-policies)
+ [取消订阅数据源](#data-feed-unsubscribing)

## 访问数据源
<a name="data-feed-accessing-procedure"></a>

1. 分配具有 SQL 和 ETL（提取、转换、加载）经验的商业智能或数据工程师。此人还需要有设置 API 的经验。

1. 设置 Amazon Simple Storage Service 存储桶并订阅数据源。使用与您的 Marketplace 产品列表关联的 AWS 卖家账户 ID。为此，您可以[观看此 YouTube 视频](https://www.youtube.com/watch?v=heCsZdOT-hw)或按照以下步骤操作。

   该视频和步骤说明了如何使用有助于简化配置的 [CloudFormation 模板](https://s3.amazonaws.com/aws-marketplace-reports-resources/DataFeedsResources.yaml)。

   1. 打开 Web 浏览器并登录 [AWS Marketplace 管理门户](https://aws.amazon.com/marketplace/management/)，然后转到[设置客户数据存储](https://aws.amazon.com/marketplace/management/reports/data-feed-configuration)。

   1. 选择**使用 CloudFormation 模板创建资源**，在 CloudFormation 控制台的另一个窗口中打开模板。

   1. 在模板中，指定以下内容，然后选择**下一步**：
      + 堆栈名称 – 您将创建的允许访问数据源的资源集合。
      + Amazon S3 存储桶名称 – 用于存储数据源的存储桶。
      + （可选）Amazon SNS 主题名称 – 在 Amazon Simple Storage Service 存储桶发生相关事件时接收通知的主题。

   1. 在**审核**页面上，确认您的条目，然后选择**创建堆栈**。这将打开一个包含 CloudFormation 状态和详细信息的新页面。

   1. 从**资源**选项卡中，将 CloudFormation 页面上的以下资源的 Amazon 资源名称 (ARN) 复制到 AWS Marketplace [设置客户数据存储](https://aws.amazon.com/marketplace/management/reports/data-feed-configuration)页面上的字段中：
      + 用于存储数据源的 Amazon S3 存储桶
      + 用于加密 Amazon S3 存储桶的 AWS KMS 密钥
      + （可选）AWS 将新数据传送到 Amazon S3 存储桶时接收通知的 Amazon SNS 主题

   1. 在**设置客户数据存储**页面上，选择**提交**。

   1. （可选）编辑由 CloudFormation 模板创建的策略。有关更多信息，请参阅[数据源策略](#data-feed-policies)。

      您现在已订阅数据源。下次生成数据源时，您可以访问数据。

1. 使用 ETL（提取、转换、加载）操作将数据源连接到您的数据仓库或关系数据库。
**注意**  
数据工具功能各异。您必须让商业智能工程师或数据工程师参与设置集成，以便与您工具的功能相匹配。

1. 要运行或创建 SQL 查询，请将数据源配置为在数据工具中强制使用主键和外键。每个数据源代表一个唯一的表，您必须使用实体关系在数据架构中设置所有数据源。有关表和实体关系的更多信息，请参阅本指南中的[数据源表概述](data-feed-joining.md)。

1. 设置 Amazon Simple Notification Service 以自动刷新您的数据仓库或关系数据库。您可以配置 Amazon SNS 通知，在来自每个唯一源的数据传送到 Amazon S3 存储桶时发送提醒。通过数据源接收新数据时，可以利用这些通知来自动刷新卖家数据仓库（如果卖家数据工具支持此功能）。请参阅《Amazon Simple Notification Service 开发人员指南》**中的 [Amazon SNS 入门](https://docs.aws.amazon.com/sns/latest/dg/sns-getting-started.html)。

   示例通知：

   ```
   {
           "mainExecutionId": "1bc08b11-ab4b-47e1-866a-9c8f38423a98",
           "executionId": "52e862a9-42d2-41e0-8010-810af84d39b1",
           "subscriptionId": "27ae3961-b13a-44bc-a1a7-365b2dc181fd",
           "processedFiles": [],
           "executionStatus": "SKIPPED",
           "errors": [],
           "feedType": "[data feed name]"
           }
   ```

   通知可能具有以下 `executionStatus` 状态：
   + `SKIPPED` – 卖家没有当天的新数据。
   + `COMPLETED` – 我们传输了包含新数据的源。
   + `FAILED` – 源传输存在问题。

1. 通过运行 SQL 查询来验证设置。您可以使用[本指南中的示例查询](https://docs.aws.amazon.com/marketplace/latest/userguide/data-feed-full-examples.html)，也可以使用 GitHub 上的查询，网址为 [https://github.com/aws-samples/aws-marketplace-api-samples/tree/main/seller-data-feeds/queries](https://github.com/aws-samples/aws-marketplace-api-samples/tree/main/seller-data-feeds/queries)。
**注意**  
本指南中的示例查询是为 AWS Athena 编写的。您可能需要修改查询才能与您的工具配合使用。

1. 确定业务用户想要在哪里使用数据。例如，您可以：
   + 从您的数据仓库或 SQL 数据库中导出 .csv 数据。
   + 将您的数据连接到 PowerBI 或 Tableau 等可视化工具。
   + 将数据映射到您的 CRM、ERP 或财务工具，例如 Salesforce、Infor 或 Netsuite。

有关使用 CloudFormation 模板的更多信息，请参阅《AWS CloudFormation 用户指南》**中的[使用 CloudFormation 模板](https://docs.aws.amazon.com/AWSCloudFormation/latest/UserGuide/template-guide.html)。

## 数据源策略
<a name="data-feed-policies"></a>

当使用 CloudFormation 模板创建 Amazon S3 存储桶时，它将为附加到该存储桶、AWS KMS 密钥和 Amazon SNS 主题的访问权限创建策略。这些策略允许 AWS Marketplace 报告服务使用数据源信息写入您的存储桶和 SNS 主题。每个策略都将包含如下所示的部分（此示例来自 Amazon S3 存储桶）。

```
        {
            "Sid": "AwsMarketplaceDataFeedsAccess",
            "Effect": "Allow",
            "Principal": {
                "Service": "reports.marketplace.amazonaws.com"
            },
            "Action": [
                "s3:ListBucket",
                "s3:GetObject",
                "s3:PutObject",
                "s3:GetEncryptionConfiguration",
                "s3:GetBucketAcl",
                "s3:PutObjectAcl"
            ],
            "Resource": [
                "arn:aws:s3:::amzn-s3-demo-bucket",
                "arn:aws:s3:::amzn-s3-demo-bucket/*"
            ]
        },
```

在本策略中，AWS Marketplace 使用 `reports.marketplace.amazonaws.com` 服务主体将数据推送到 Amazon S3 存储桶。您在 CloudFormation 模板中指定了 *amzn-s3-demo-bucket*。

当 AWS Marketplace 报告服务调用 Amazon S3、AWS KMS 或 Amazon SNS 时，它将提供意图写入存储桶的数据的 ARN。为确保写入存储桶的唯一数据是以您的名义写入的数据，您可以在策略的条件中指定 `aws:SourceArn`。在以下示例中，您必须将 *account-id* 替换为自己的 AWS 账户 ID。

```
        {
           "Sid": "AwsMarketplaceDataFeedsAccess",
           "Effect": "Allow",
           "Principal": {
                "Service": "reports.marketplace.amazonaws.com"
            },
            "Action": [
                "s3:ListBucket",
                "s3:GetObject",
                "s3:PutObject",
                "s3:GetEncryptionConfiguration",
                "s3:GetBucketAcl",
                "s3:PutObjectAcl"
            ],
            "Resource": [
                "arn:aws:s3:::amzn-s3-demo-bucket",
                "arn:aws:s3:::amzn-s3-demo-bucket/*"
            ,
            "Condition": {
                "StringEquals": {
                        "aws:SourceAccount": "account-id",
                        "aws:SourceArn": ["arn:aws:marketplace::account-id:AWSMarketplace/SellerDataSubscription/DataFeeds_V1",
                        "arn:aws:marketplace::account-id:AWSMarketplace/SellerDataSubscription/Example-Report"]
                        }
                }
        },
```

## 取消订阅数据源
<a name="data-feed-unsubscribing"></a>

打开 Web 浏览器并登录 [AWS Marketplace Management Portal](https://aws.amazon.com/marketplace/management/)。然后，转到[联系我们页面](https://aws.amazon.com/marketplace/management/contact-us/)，向 AWS Marketplace 卖家运营团队提交取消订阅请求。取消订阅请求可能需要 10 个工作日来处理。

# 使用数据源收集和分析数据
<a name="data-feed-using"></a>

AWS Marketplace 提供数据馈送作为一种机制，用于将结构化、 up-to-date产品和客户信息从 AWS Marketplace 系统发送到卖家 Amazon S3 存储桶，以便在卖家拥有的商业智能工具之间进行 ETL（提取、转换和加载）。当 Amazon S3 存储桶中存在数据时，您可以通过以下方式使用数据源：
+ 从您在[访问数据源](data-feed-accessing.md)中创建的 Amazon S3 存储桶下载 .CSV 文件，以便您可以在电子表格中查看数据。
+ 使用 ETL（提取、转换和加载）、SQL 查询、业务分析工具来收集和分析数据。

  您可以使用 AWS 服务来收集和分析数据，也可以使用任何可以对基于.csv 的数据集进行分析的第三方工具。

有关使用数据源来收集和分析数据的更多信息，请参阅以下示例。

## 示例：使用 AWS 服务收集和分析数据
<a name="data-feed-using-example"></a>

以下过程假定您已配置环境以将数据源接收到 Amazon S3 存储桶，并且该存储桶包含数据源。

**从数据源收集数据并进行分析**

1. 在 [AWS Glue 控制台](https://console.aws.amazon.com/glue) 中，[创建一个爬网程序](https://docs.aws.amazon.com/glue/latest/dg/add-crawler.html)以连接到存储数据源的 Amazon S3 存储桶、提取所需数据并在 AWS Glue Data Catalog中创建元数据表。

   有关的更多信息 AWS Glue，请参阅《[https://docs.aws.amazon.com/glue/latest/dg/what-is-glue.html](https://docs.aws.amazon.com/glue/latest/dg/what-is-glue.html)。

1. 在 [Athena 控制台](https://console.aws.amazon.com/athena) 中，[对 AWS Glue Data Catalog中的数据运行 SQL 查询](https://docs.aws.amazon.com/athena/latest/ug/querying-athena-tables.html)。

   有关 Athena 的更多信息，请参阅 [《Amazon Athena 用户指南》**](https://docs.aws.amazon.com/athena/latest/ug/what-is.html)。

1. 在 [Quick 控制台](http://quicksight.aws.amazon.com)中，[创建分析](https://docs.aws.amazon.com/quick/latest/userguide/creating-an-analysis.html)，然后[创建数据的可视化视](https://docs.aws.amazon.com/quick/latest/userguide/creating-a-visual.html)图。

   有关 Quick 的更多信息，请参阅 [https://docs.aws.amazon.com/quick/latest/userguide/welcome.html](https://docs.aws.amazon.com/quick/latest/userguide/welcome.html)。

有关使用 AWS 服务收集和分析数据上传数据的一种方式的详细示例，请参阅博客上的[使用卖家数据源配送服务、Amazon Athena 和快速创建卖家](https://aws.amazon.com/blogs/awsmarketplace/using-seller-data-feed-delivery-service-amazon-athena-and-amazon-quicksight-to-create-seller-reports/)报告。 AWS Marketplace 

# 数据源表概述
<a name="data-feed-joining"></a>

AWS Marketplace 提供的数据源以一组表格的形式呈现，您可以将这些表格连接在一起，为查询提供更多上下文。

AWS Marketplace 会在您的数据源中包含以下通用领域或相关类别：
+ **目录** - 包括有关您账户中产品和优惠的信息。
+ **账户** - 包括有关在 AWS Marketplace 中提供或购买产品的账户（您自己的账户或与您合作的各方账户，例如渠道合作伙伴或买家）的信息。
+ **收入** - 包括有关账单、支付和税费的信息。
+ **采购** - 包括您作为登记卖家创建的产品优惠的协议相关信息。

此图显示了目录、账户和收入领域下的相关表格。

![\[Entity relationship diagram showing how data feeds relate to each other.\]](http://docs.aws.amazon.com/zh_cn/marketplace/latest/userguide/images/datafeeds-overview.png) 

## 与目录相关的表
<a name="data-feed-catalog-domain"></a>

下图显示了目录域中各表之间的关系以及表中各字段的关系。

![\[目录域中“Product”、“Offer_Product”、“Offer”、“Offer_Target”和“Legacy_id_mapping”表之间的关系。\]](http://docs.aws.amazon.com/zh_cn/marketplace/latest/userguide/images/datafeeds-catalog-details.png)


`Product`、`Offer_Product`、`Offer`、`Offer_Target` 和 `Legacy_id_mapping` 表位于目录域中。

`Offer_Target` 表包含目标的 `account_id` 的值字段，但仅当 `target_type` 的值为 `account` 时。

`Legacy_id_mapping` 表不用于当前数据。

**注意**  
有关这些表的更多信息，包括对表中每个字段的描述以及可以创建的连接，请参阅以下主题：  
[产品数据源](data-feed-product.md)
[优惠产品数据源](data-feed-offer-product.md)
[优惠数据源](data-feed-offer.md)
[优惠目标数据源](data-feed-offer-target.md)
[传统映射数据源](data-feed-legacy-mapping.md)

## 与账户相关的表
<a name="data-feed-accounts-domain"></a>

下图显示了账户域中 `Account` 和 `Address` 表之间的关系以及表中各字段的关系。

![\[账户域中的“Account”表和“Address”表与每个表中的字段之间的关系。\]](http://docs.aws.amazon.com/zh_cn/marketplace/latest/userguide/images/datafeeds-accounts-details.png)


**注意**  
有关这些表的更多信息，包括对表中每个字段的描述以及您可以创建的连接，请参阅以下主题：  
[账户数据源](data-feed-account.md)
[地址数据源](data-feed-address.md)

## 与收入相关的表
<a name="data-feed-revenue-domain"></a>

下图显示了收入域中 `Billing_Event` 和 `Tax_Item` 表之间的关系以及表中各字段的关系。`Billing_Event` 表包括有关支付和账单事件的信息。

![\[收入域中 Billing_Event 和 Tax_Item 表之间的关系以及表中各字段的关系。\]](http://docs.aws.amazon.com/zh_cn/marketplace/latest/userguide/images/datafeeds-revenue-details.png)


**注意**  
有关这些表的更多信息，包括对表中每个字段的描述以及您可以创建的连接，请参阅以下主题：  
[账单事件数据源](data-feed-billing-event.md)
[税项数据源](data-feed-tax-item.md)

### 与采购相关的表
<a name="procurement-tables"></a>

下图显示了采购域中“Agreement”表中的字段。

**注意**  
有关这些表的更多信息，包括对表中每个字段的描述以及可以创建的连接，请参阅本指南中的[协议数据源](data-feed-agreements.md)。

以下各部分提供了每个域的*实体关系* (ER) 图。每个 ER 图都显示表和每个表中的字段，以及可用于联接表的字段。

**注意**  
本部分中的 ER 图不包括所有数据源的公共字段。有关字段的更多信息，请参阅[AWS Marketplace 数据源的存储和结构](data-feed-details.md)。

下表介绍了 ER 图中显示的符号。


| 符号 | 描述 | 
| --- | --- | 
|  ![\[字母“PK”作为符号的图像。\]](http://docs.aws.amazon.com/zh_cn/marketplace/latest/userguide/images/datafeeds-primary-key.png)  |  **主键** - 表的主键。与 `valid_from` 和 `update_date` 字段一起使用时，它是唯一的。有关将这些字段一起使用的更多详细信息，请参阅[数据历史记录](data-feed-details.md#data-feed-historization)。如果多个字段被标记为主键，则这些字段共同构成主键。  | 
|  ![\[字母“FK”作为符号的图像。\]](http://docs.aws.amazon.com/zh_cn/marketplace/latest/userguide/images/datafeeds-foreign-key.png)  |   **外键** - 表示不同表中主键的字段。表中不一定是唯一的。   在某些情况下，如果当前表中的记录在外表中没有相应的记录，则外键可能为空。   | 
|  ![\[字母“AK”作为符号的图像。\]](http://docs.aws.amazon.com/zh_cn/marketplace/latest/userguide/images/datafeeds-alternate-key.png)  |   **备用密钥** - 可在表中用作密钥的密钥。遵循与主键相同的唯一性规则。  | 
|  ![\[一端是十字架，另一端是圆圈和叉子的一条线的图像。\]](http://docs.aws.amazon.com/zh_cn/marketplace/latest/userguide/images/datafeeds-one-to-many.png)  |   **连接器** - 字段之间的线代表一个连接，即两个可用于连接表的字段。线的末端代表连接的类型。此示例表示一对多连接。  | 

**连接器类型**

下表显示每个连接器可以具有的端部类型。


| 连接器类型 | 描述 | 
| --- | --- | 
|  ![\[一端是十字架的一条线的图像。\]](http://docs.aws.amazon.com/zh_cn/marketplace/latest/userguide/images/datafeeds-one-to-n.png)  |   **一对 n** – 具有此端的连接器表示在此侧只有一个值的连接。  | 
|  ![\[一端是十字架，一端是圆圈的一条线的图像。\]](http://docs.aws.amazon.com/zh_cn/marketplace/latest/userguide/images/datafeeds-zero-or-one-to-n.png)  |   **零或一到 n** - 具有此端的连接器表示在此侧具有零或一个值的连接。  | 
|  ![\[一端是圆圈和叉子的一条线的图像。\]](http://docs.aws.amazon.com/zh_cn/marketplace/latest/userguide/images/datafeeds-optional-many-to-n.png)  |   **零或多对 n** - 具有此端的连接器表示此侧具有零、一个或多个值的连接。  | 
|  ![\[一端是十字架和叉子的一条线的图像。\]](http://docs.aws.amazon.com/zh_cn/marketplace/latest/userguide/images/datafeeds-one-or-more-to-n.png)  |   **一或多对 n** - 具有此端的连接器表示在此侧具有一个或多个值的连接。  | 

# 数据源查询示例
<a name="data-feed-full-examples"></a>

本节举例说明了使用提供的数据馈送进行复杂查询 AWS Marketplace。这些示例与您从 AWS Marketplace 管理门户中获得的[卖家控制面板](dashboards.md)类似。您可以自定义这些查询以创建所需的其他报告。

**Topics**
+ [协议和续订](#data-feed-example-agreements)
+ [计费收入](#data-feed-example-billed-revenue)
+ [未收取或已支付的发票](#data-feed-example-collections)
+ [已纳税发票](#data-feed-example-tax)
+ [按产品列出的款项](#data-feed-example-disbursement-by-product)
+ [销售补偿报告](#data-feed-example-sales-compensation)

## 协议和续订
<a name="data-feed-example-agreements"></a>

要查找协议和续订数据，您可以运行如以下示例所示的一组查询。这些查询相互依存，可创建**协议和续订**控制面板，即精细数据部分。您可以使用下方所示的示例，也可以根据您的数据和用例进行自定义。

查询中的注释说明了查询的作用以及如何修改查询。

```
      Query currently under development.
```

## 计费收入
<a name="data-feed-example-billed-revenue"></a>

要查找发票数据，您可以运行如以下示例所示的一组查询。这些查询相互依存，可创建**计费收入**报告。您可以使用下方所示的示例，也可以根据您的数据和用例进行自定义。

查询中的注释说明了查询的作用以及如何修改查询。

```
-- Billed revenue report

-- General note: When executing this query we are assuming that the data ingested in the database uses 
-- two time axes (the valid_from column and the update_date column).
-- See documentation for more details: https://docs.aws.amazon.com/marketplace/latest/userguide/data-feed.html#data-feed-details

-- An account_id has several valid_from dates (each representing a separate revision of the data)
-- but because of bi-temporality, an account_id + valid_from tuple can appear multiple times with a different update_date.
-- We are only interested in the most recent tuple (ie, uni-temporal model)
with accounts_with_uni_temporal_data as (
  select
    account_id,
    aws_account_id,
    encrypted_account_id,
    mailing_address_id,
    tax_address_id,
    tax_legal_name,
    from_iso8601_timestamp(valid_from) as valid_from,
    tax_registration_number
  from
    (
      select
        account_id,
        aws_account_id,
        encrypted_account_id,
        mailing_address_id,
        tax_address_id,
        tax_legal_name,
        valid_from,
        delete_date,
        tax_registration_number,
        row_number() over (partition by account_id, valid_from order by from_iso8601_timestamp(update_date) desc) as row_num
      from
        accountfeed_v1
    )
    where
      -- keep latest ...
      row_num = 1
      -- ... and remove the soft-deleted one.
      and (delete_date is null or delete_date = '')
  ),

accounts_with_history as (
  with accounts_with_history_with_extended_valid_from as (
    select
      account_id,
      -- sometimes, this columns gets imported as a "bigint" and loses heading 0s -> casting to a char and re-adding heading 0s (if need be)
      substring('000000000000'||cast(aws_account_id as varchar),-12) as aws_account_id,
      encrypted_account_id,
      mailing_address_id,
      tax_address_id,
      tax_legal_name,
      -- The start time of account valid_from is extended to '1970-01-01 00:00:00', because:
      -- ... in tax report transformations, some tax line items with invoice_date cannot
      -- ... fall into the default valid time range of the associated account
      case
        when lag(valid_from) over (partition by account_id order by valid_from asc) is null
          then cast('1970-01-01 00:00:00' as timestamp)
        else valid_from
      end as valid_from
    from accounts_with_uni_temporal_data
    )
  select
    account_id,
    aws_account_id,
    encrypted_account_id,
    mailing_address_id,
    tax_address_id,
    tax_legal_name,
    valid_from,
    coalesce(
      lead(valid_from) over (partition by account_id order by valid_from asc),
      cast('2999-01-01 00:00:00' as timestamp)
    ) as valid_to
  from
    accounts_with_history_with_extended_valid_from
),

-- An address_id has several valid_from dates (each representing a separate revision of the data)
-- but because of bi-temporality, an account_id + valid_from tuple can appear multiple times with a different update_date.
-- We are only interested in the most recent tuple (ie, uni-temporal model)
address_with_uni_temporal_data as (
  select
    from_iso8601_timestamp(valid_from) as valid_from,
    address_id,
    company_name,
    email_domain,
    country_code,
    state_or_region,
    city,
    postal_code,
    row_num
  from
  (
    select
      valid_from,
      update_date,
      delete_date,
      address_id,
      company_name,
      email_domain,
      country_code,
      state_or_region,
      city,
      postal_code,
      row_number() over (partition by address_id, valid_from order by from_iso8601_timestamp(update_date) desc) as row_num
    from
      addressfeed_v1
  )
  where
    -- keep latest ...
    row_num = 1
    -- ... and remove the soft-deleted one.
    and (delete_date is null or delete_date = '')
),

-- We are only interested in the most recent tuple (BTW: a given address is not supposed to change over time but when bugs ;-) so this query mainly does nothing)
address_with_latest_revision as (
  select
    valid_from,
    address_id,
    company_name,
    email_domain,
    country_code,
    state_or_region,
    city,
    postal_code,
    row_num_latest_revision
  from
  (
    select
      valid_from,
      address_id,
      company_name,
      email_domain,
      country_code,
      state_or_region,
      city,
      postal_code,
      row_number() over (partition by address_id order by valid_from desc) as row_num_latest_revision
    from
      address_with_uni_temporal_data
  )
  where
    row_num_latest_revision = 1
),

accounts_with_history_with_company_name as (
  select
    awh.account_id,
    awh.aws_account_id,
    awh.encrypted_account_id,
    awh.mailing_address_id,
    awh.tax_address_id,
    coalesce(
      --empty value in Athena shows as '', change all '' value to null
      case when address.company_name = '' then null else address.company_name end,
      awh.tax_legal_name) as mailing_company_name,
    address.email_domain,
    awh.valid_from,
    -- For BYOL, the agreement might be accepted (using some external non-AWS system or manual process) days before
    -- that BYOL agreement is entered into AWS Marketplace by the buyer. Therefore, the buyer is permitted to manually
    -- enter a backdated acceptance date, which might predate the point in time when the account was created.
    -- To work around this, we need to adjust the valid_from of the account to be
    -- earlier than the earliest possible backdated BYOL agreement acceptance date.
    case
      when lag(awh.valid_from) over (partition by aws_account_id order by awh.valid_from asc) is null
      then date_add('Day', -212, awh.valid_from)
      -- 212 is the longest delay between acceptance_date of the agreement and the account start_Date
      else awh.valid_from
    end as valid_from_adjusted,
    awh.valid_to
  from accounts_with_history as awh
  left join address_with_latest_revision as address on
    awh.mailing_address_id = address.address_id and awh.mailing_address_id is not null
),

-- An agreement_id has several valid_from dates (each representing an agreement revision)
-- but because of bi-temporality, an agreement_id + valid_from tuple can appear multiple times with a different update_date.
-- We are only interested in the most recent tuple (ie, uni-temporal model)
agreements_with_uni_temporal_data as (
  select
    agreement_id,
    origin_offer_id,
    proposer_account_id,
    acceptor_account_id,
    agreement_revision,
    from_iso8601_timestamp(valid_from) as valid_from,
    from_iso8601_timestamp(start_date) as start_date,
    from_iso8601_timestamp(end_date) as end_date,
    from_iso8601_timestamp(acceptance_date) as acceptance_date,
    agreement_type,
    previous_agreement_id,
    agreement_intent
  from
  (
    select
      --empty value in Athena shows as '', change all '' value to null
      case when agreement_id = '' then null else agreement_id end as agreement_id,
      origin_offer_id,
      proposer_account_id,
      acceptor_account_id,
      agreement_revision,
      valid_from,
      delete_date,
      start_date,
      end_date,
      acceptance_date,
      agreement_type,
      previous_agreement_id,
      agreement_intent,
      row_number() over (partition by agreement_id, valid_from order by from_iso8601_timestamp(update_date) desc) as row_num
    from
      -- TODO change to agreementfeed_v1 when Agreement Feed is GA'ed
      agreementfeed
  )
  where
    -- keep latest ...
    row_num = 1
    -- ... and remove the soft-deleted one.
    and (delete_date is null or delete_date = '')
),

agreements_with_history as (
  with agreements_with_window_functions as (
    select
      agreement_id,
      origin_offer_id as offer_id,
      proposer_account_id,
      acceptor_account_id,
      agreement_revision,
      start_date,
      end_date,
      acceptance_date,
      -- The start time of agreement valid_from is extended to '1970-01-01 00:00:00', because:
      -- ... in usage report transformations, some usage line items with usage_date cannot
      -- ... fall into the default valid time range of the associated agreement
      case
          when lag(valid_from) over (PARTITION BY agreement_id order by valid_from asc) is null
          then timestamp '1970-01-01 00:00:00'
          else valid_from
      end as valid_from,
      coalesce(
          lead(valid_from) over (partition by agreement_id order by valid_from asc),
          timestamp '2999-01-01 00:00:00'
      ) as valid_to,
      rank() over (partition by agreement_id order by valid_from asc) version,
      agreement_type,
      previous_agreement_id,
      agreement_intent
    from
      agreements_with_uni_temporal_data
  )
  select
    agreement_id,
    offer_id,
    proposer_account_id,
    acceptor_account_id,
    agreement_revision,
    start_date,
    end_date,
    acceptance_date,
    valid_from,
    case
        when version=1 and valid_from<timestamp '2023-03-03 06:16:08.743' then timestamp '1970-01-01'
        -- The following 60 minute adjustment is to handle special case where When Renewal happens for a contract
        when version=1 then date_add('minute',-60,valid_from)
        else valid_from
    end as valid_from_adjusted,
    valid_to,
    agreement_type,
    previous_agreement_id,
    agreement_intent
  from
    agreements_with_window_functions
),

-- An offer_id has several valid_from dates (each representing an offer revision)
-- but because of bi-temporality, an offer_id + valid_from tuple can appear multiple times with a different update_date.
-- We are only interested in the most recent tuple (ie, uni-temporal model)
offers_with_uni_temporal_data as (
  select
    from_iso8601_timestamp(valid_from) as valid_from,
    from_iso8601_timestamp(update_date) as update_date,
    from_iso8601_timestamp(delete_date) as delete_date,
    offer_id,
    offer_revision,
    name,
    expiration_date,
    opportunity_id,
    opportunity_name,
    opportunity_description,
    seller_account_id
  from
  (
    select
      valid_from,
      update_date,
      delete_date,
      offer_id,
      offer_revision,
      name,
      expiration_date,
      opportunity_id,
      opportunity_name,
      opportunity_description,
      seller_account_id,
      row_number() over (partition by offer_id, valid_from order by from_iso8601_timestamp(update_date) desc) as row_num
    from
      offerfeed_v1
  )
  where
    -- keep latest ...
    row_num = 1
    -- ... and remove the soft-deleted one.
    and (delete_date is null or delete_date = '')
),

-- Here, we build the validity time range (adding valid_to on top of valid_from) of each offer revision.
-- We will use it to get Offer name at invoice time.
-- NB: If you'd rather get "current" offer name, un-comment "offers_with_latest_revision"
offers_with_history as (
  select
    offer_id,
    offer_revision,
    name,
    opportunity_id,
    opportunity_name,
    opportunity_description,
    valid_from,
    -- When we try to look up an offer revision as at the acceptance date of a BYOL agreement, we run into a problem.
    -- For BYOL, the agreement might be accepted (using some external non-AWS system or manual process) days before
    -- that BYOL agreement is entered into AWS Marketplace by the buyer. Therefore, the buyer is permitted to manually
    -- enter a backdated acceptance date, which might predate the point in time when the first revision of the offer
    -- was created. To work around this, we need to adjust the valid_from on the first revision of the offer to be
    -- earlier than the earliest possible backdated BYOL agreement acceptance date.
    case
      when lag(valid_from) over (partition by offer_id order by valid_from asc) is null and valid_from<cast('2021-04-01' as timestamp)
      then date_add('Day', -3857, valid_from)
      -- 3857 is the longest delay between acceptance_date of an agreement and the first revision of the offer
      when lag(valid_from) over (partition by offer_id order by valid_from asc) is null and valid_from >= cast('2021-04-01' as timestamp)
      then date_add('Day', -1460, valid_from)
      --after 2021 for the two offers we need to adjust for 2 more years
      else valid_from
    end as valid_from_adjusted,
    coalesce(
      lead(valid_from) over (partition by offer_id order by valid_from asc),
      cast('2999-01-01 00:00:00' as timestamp))
    as valid_to
  from offers_with_uni_temporal_data
),
-- provided for reference only if you are interested into get "current" offer name
-- (ie. not used afterwards)
offers_with_latest_revision as (
  select
    offer_id,
    offer_revision,
    name,
    opportunity_name,
    opportunity_description,
    valid_from,
    null valid_to
  from
  (
    select
      offer_id,
      offer_revision,
      name,
      opportunity_name,
      opportunity_description,
      valid_from,
      null valid_to,
      row_number() over (partition by offer_id order by valid_from desc) as row_num_latest_revision
    from
      offers_with_uni_temporal_data
  )
  where
    row_num_latest_revision = 1
),

-- An offer_target_id has several valid_from dates (each representing an offer revision)
-- but because of bi-temporality, an offer_target_id + valid_from tuple can appear multiple times with a different update_date.
-- We are only interested in the most recent tuple (ie, uni-temporal model)
offer_targets_with_uni_temporal_data as (
  select
    from_iso8601_timestamp(valid_from) as valid_from,
    from_iso8601_timestamp(update_date) as update_date,
    from_iso8601_timestamp(delete_date) as delete_date,
    offer_target_id,
    offer_id,
    offer_revision,
    target_type,
    polarity,
    value
  from
  (
    select
      valid_from,
      update_date,
      delete_date,
      offer_target_id,
      offer_id,
      offer_revision,
      target_type,
      polarity,
      value,
      row_number() over (partition by offer_target_id, valid_from order by from_iso8601_timestamp(update_date) desc) as row_num
    from
      offertargetfeed_v1
  )
  where
    -- keep latest ...
    row_num = 1
    -- ... and remove the soft-deleted one.
    and (delete_date is null or delete_date = '')
),

offer_target_type as (
  select
    offer_id,
    offer_revision,
    substring(
      -- The first character indicates the priority (lower value means higher precedence):
      min(
        case
          when offer_target.target_type='BuyerAccounts' then '1Private'
          when offer_target.target_type='ParticipatingPrograms' then '2Program:'||cast(offer_target.value as varchar)
          when offer_target.target_type='CountryCodes' then '3GeoTargeted'
          -- well, there is no other case today, but rather be safe...
          else '4Other Targeting'
        end
      ),
      -- Remove the first character that was only used for th priority in the "min" aggregate function:
      2
    ) as offer_target
  from
    offer_targets_with_uni_temporal_data as offer_target
  group by
    offer_id,
    offer_revision
),

offers_with_history_with_target_type as (
  select
    offer.offer_id,
    offer.offer_revision,
    -- even though today it is not possible to combine several types of targeting in a single offer, let's ensure the query is still predictable if this gets possible in the future
    max(
      case
        when off_tgt.offer_target is null then 'Public'
        else off_tgt.offer_target
      end
    ) as offer_target,
    min(offer.name) as name,
    min(offer.opportunity_name) as opportunity_name,
    min(offer.opportunity_description) as opportunity_description,
    offer.valid_from,
    offer.valid_from_adjusted,
    offer.valid_to,
    offer.opportunity_id
  from
    offers_with_history as offer
  left join offer_target_type as off_tgt on
    offer.offer_id = off_tgt.offer_id
    and offer.offer_revision = off_tgt.offer_revision
  group by
    offer.offer_id,
    offer.offer_revision,
    offer.valid_from,
    offer.valid_from_adjusted,
    offer.valid_to,
    offer.opportunity_id
),

-- provided for reference only if you are interested into get "current" offer targets
-- (ie. not used afterwards)
offers_with_latest_revision_with_target_type as (
  select
    offer.offer_id,
    offer.offer_revision,
    -- even though today it is not possible to combine several types of targeting in a single offer, let's ensure the query is still predictable if this gets possible in the future
    max(
      distinct
      case
        when off_tgt.target_type is null then 'Public'
        when off_tgt.target_type='BuyerAccounts' then 'Private'
        when off_tgt.target_type='ParticipatingPrograms' then 'Program:'||cast(off_tgt.value as varchar)
        when off_tgt.target_type='CountryCodes' then 'GeoTargeted'
        -- well, there is no other case today, but rather be safe...
        else 'Other Targeting'
      end
    ) as offer_target,
    min(offer.name) as name,
    min(offer.opportunity_name) as opportunity_name,
    min(offer.opportunity_description) as opportunity_description,
    offer.valid_from,
    offer.valid_to
  from
    offers_with_latest_revision offer
    -- left joining because public offers don't have targets
    left join offer_targets_with_uni_temporal_data off_tgt on
      offer.offer_id=off_tgt.offer_id and offer.offer_revision=off_tgt.offer_revision
  group by
    offer.offer_id,
    offer.offer_revision,
    -- redundant with offer_revision, as each revision has a dedicated valid_from (but cleaner in the group by)
    offer.valid_from,
    offer.valid_to
),

-- A product_id has several valid_from dates (each representing a product revision),
-- but because of bi-temporality, each product_id + valid_from tuple can appear multiple times with a different update_date.
-- We are only interested in the most recent tuple (ie, uni-temporal model)
products_with_uni_temporal_data as (
  select
    from_iso8601_timestamp(valid_from) as valid_from,
    from_iso8601_timestamp(update_date) as update_date,
    from_iso8601_timestamp(delete_date) as delete_date,
    product_id,
    manufacturer_account_id,
    product_code,
    title
  from
  (
    select
      valid_from,
      update_date,
      delete_date,
      product_id,
      manufacturer_account_id,
      product_code,
      title,
      row_number() over (partition by product_id, valid_from order by from_iso8601_timestamp(update_date) desc) as row_num
    from
      productfeed_v1
  )
  where
    -- keep latest ...
    row_num = 1
    -- ... and remove the soft-deleted one.
    and (delete_date is null or delete_date = '')
),

products_with_history as (
  select
    product_id,
    title,
    valid_from,
    -- Offerv2 can have upto 50 years and Offerv3 is upto 5 years of past date
    case
      when lag(valid_from) over (partition by product_id order by valid_from asc) is null and valid_from<cast('2021-04-01' as timestamp)
        then date_add('Day', -3857, valid_from)
      -- 3827 is the longest delay between acceptance_date of an agreement and the product
      -- we are keeping 3857 as a consistency between the offers and products
      when lag(valid_from) over (partition by product_id order by valid_from asc) is null and valid_from >= cast('2021-04-01' as timestamp)
        then date_add('Day', -2190, valid_from)
      --after 2021 for the two offers we need to adjust for 2 more years
      else valid_from 
    end as valid_from_adjusted,
    coalesce(
      lead(valid_from) over (partition by product_id order by valid_from asc),
      cast('2999-01-01 00:00:00' as timestamp)
    ) as valid_to,
    product_code,
    manufacturer_account_id
  from
    products_with_uni_temporal_data
),

legacy_products as (
  select
    legacy_id,
    new_id
  from
    legacyidmappingfeed_v1
  where
    mapping_type='PRODUCT'
  group by
    legacy_id,
    new_id
),

-- A given billing_event_id represents an accounting event and thus has only one valid_from date,
-- but because of bi-temporality, a billing_event_id (+ its valid_from) can appear multiple times with a different update_date.
-- We are only interested in the most recent tuple (ie, uni-temporal model)
billing_events_with_uni_temporal_data as (
  select
    billing_event_id,
    valid_from,
    update_date,
    delete_date,
    invoice_date,
    transaction_type,
    transaction_reference_id,
    parent_billing_event_id,
    bank_trace_id,
    broker_id,
    product_id,
    disbursement_billing_event_id,
    action,
    from_account_id,
    to_account_id,
    end_user_account_id,
    billing_address_id,
    amount,
    currency,
    balance_impacting,
    --empty value in Athena shows as '', change all '' value to null
    case when agreement_id = '' then null else agreement_id end as agreement_id,
    invoice_id,
    payment_due_date,
    usage_period_start_date,
    usage_period_end_date,
    buyer_transaction_reference_id,
    row_num
  from
  (
    select
      billing_event_id,
      from_iso8601_timestamp(valid_from) as valid_from,
      from_iso8601_timestamp(update_date) as update_date,
      delete_date,
      from_iso8601_timestamp(invoice_date) as invoice_date,
      transaction_type,
      transaction_reference_id,
      parent_billing_event_id,
      -- casting in case data was imported as number
      cast(bank_trace_id as varchar) as bank_trace_id,
      broker_id,
      product_id,
      disbursement_billing_event_id,
      action,
      from_account_id,
      to_account_id,
      end_user_account_id,
      billing_address_id,
      -- casting in case data was imported as varchar
      cast(amount as decimal(38,6)) as amount,
      currency,
      balance_impacting,
      agreement_id,
      invoice_id,
      case when payment_due_date is null or payment_due_date = '' then null else from_iso8601_timestamp(payment_due_date) end as payment_due_date,
      from_iso8601_timestamp(usage_period_start_date) as usage_period_start_date,
      from_iso8601_timestamp(usage_period_end_date) as usage_period_end_date,
      buyer_transaction_reference_id,
      row_number() over (partition by billing_event_id, valid_from order by from_iso8601_timestamp(update_date) desc) as row_num
    from
      billingeventfeed_v1
  )
  where
    -- keep latest ...
    row_num = 1
    -- ... and remove the soft-deleted one.
    and (delete_date is null or delete_date = '')
),

-- Here we select the account_id of the current seller (We identify this by looking for the to_account_id related to revenue transactions).
-- We will use it later to distinguish own agreements from agreements generated by channel partners.
seller_account as (
  select
    from_account_id as seller_account_id
  from
    billing_events_with_uni_temporal_data bill
  where
    -- Assumption here is only seller will pay listing fee. As of 12/21/2021, there are cases that Channel partner have 0 listing fee for CPPO, so the amount could be 0.
    bill.transaction_type like 'AWS_REV_SHARE' and amount <= 0 and action = 'INVOICED'
  group by
    -- from_account_id is always the same for all those "listing fee" transactions == the seller of record himself.
    -- If this view returns more than 1 record, the overall query will fail (on purpose). Please contact AWS Marketplace if this happens.
    from_account_id
),

billing_event_with_business_flags as (
  select
    bl.billing_event_id,
    bl.end_user_account_id,
    bl.agreement_id,
    aggrement.proposer_account_id,
    aggrement.offer_id,
    aggrement.acceptor_account_id,
    case
      -- For AWS and BALANCE_ADJUSTMENT, the billing event feed will show the "AWS Marketplace" account as the
      -- receiver of the funds and the seller as the payer. We are not interested in this information here.
      -- Null values will be ignored by the `max` aggregation function.
      when bl.transaction_type like 'AWS%' then null
      -- For BALANCE_ADJUSTMENT, payer is seller themselves
      when bl.invoice_id is null then bl.to_account_id
      -- We get the payer of the invoice from *any* transaction type that is not AWS and not BALANCE_ADJUSTMENT (because they are the same for a given end user + agreement + product).
      else bl.from_account_id
    end as payer_account_id,
    bl.product_id,
    bl.action,
    bl.transaction_type,
    bl.parent_billing_event_id,
    bl.disbursement_billing_event_id,
    bl.amount,
    bl.currency,
    bl.balance_impacting,
    bl.invoice_date,
    bl.payment_due_date,
    bl.usage_period_start_date,
    bl.usage_period_end_date,
    bl.invoice_id,
    bl.billing_address_id,
    bl.transaction_reference_id,
    bl.buyer_transaction_reference_id,
    case when disbursement.bank_trace_id = 'EMEA_MP_TEST_TRACE_ID' then null else disbursement.bank_trace_id end as bank_trace_id,
    case when disbursement.bank_trace_id = 'EMEA_MP_TEST_TRACE_ID' then null else disbursement.invoice_date end as disbursement_date,
    disbursement.billing_event_id as disbursement_id,
    -- We will use disbursement_id_or_invoiced as part of the PK, so it cannot be null:
    coalesce(
      --empty value in Athena shows as '', change all '' value to null
      case when disbursement.billing_event_id = '' then null else disbursement.billing_event_id end,
      '<invoiced>') as disbursement_id_or_invoiced,
    bl.broker_id,
    case
      when bl.invoice_id is null /* transaction_type = 'BALANCE_ADJUSTMENT' */
        then (select seller_account_id from seller_account) ||':'|| cast(bl.invoice_date as varchar)
      else bl.buyer_transaction_reference_id
        ||'-'|| case when bl.agreement_id is null or bl.agreement_id = '' then ' ' else bl.agreement_id end
        ||'-'|| case when bl.end_user_account_id is null or bl.end_user_account_id = '' then ' ' else bl.end_user_account_id end
        ||'-'|| coalesce(cast(bl.usage_period_start_date as varchar),' ')
        ||'-'|| coalesce(cast(bl.usage_period_end_date as varchar),' ')
    end as internal_buyer_line_item_id,
    bl.buyer_transaction_reference_id <> bl.transaction_reference_id as is_seller_invoice,
    case when bl.transaction_type = 'SELLER_REV_SHARE' and (select seller_account_id from seller_account) <> bl.to_account_id then true else false end as is_cog,
    case when bl.transaction_type in('SELLER_REV_SHARE_CREDIT', 'SELLER_REV_SHARE_REFUND') and (select seller_account_id from seller_account) <> bl.to_account_id then true else false end as is_cog_refund,
    --TODO: replace below logic once we can create a logic the identify reseller/manufacturer without depending on agreement feed
    case when aggrement.proposer_account_id <> (select seller_account_id from seller_account) then true else false end as is_manufacturer_view_of_reseller
  from
    billing_events_with_uni_temporal_data as bl
    left join billing_events_with_uni_temporal_data as disbursement on
      disbursement.transaction_type like 'DISBURSEMENT%'
        and disbursement.action = 'DISBURSED'
        and disbursement.transaction_type IN ('DISBURSEMENT', 'DISBURSEMENT_FAILURE')
        and bl.disbursement_billing_event_id = disbursement.billing_event_id
    left join agreements_with_history as aggrement on
      bl.agreement_id = aggrement.agreement_id
        and bl.invoice_date >= aggrement.valid_from_adjusted
        and bl.invoice_date<aggrement.valid_to
    left join accounts_with_history awh on
      bl.to_account_id = awh.account_id
        and bl.invoice_date >= awh.valid_from
        and bl.invoice_date<awh.valid_to
  where
    bl.transaction_type not like 'DISBURSEMENT%' and
      (bl.agreement_id is null or bl.agreement_id = ''
      or aggrement.agreement_id is not null)
),

-- listagg function in athena does not support partitioning, grouping here and then joining to the main query
seller_invoice_list as (
  select
    internal_buyer_line_item_id,
    listagg(case when not is_seller_invoice then null else invoice_id end,',') within group (order by case when not is_seller_invoice then null else cast(invoice_date as varchar) end) as seller_invoice_id_or_null,
    listagg(case when not is_seller_invoice then null else cast(invoice_date as varchar) end,',') within group (order by case when not is_seller_invoice then null else cast(invoice_date as varchar) end) as seller_invoice_date_or_null
  from
    (
    -- listagg function in athena does not support ordering by another field when distinct is used,
    -- here we first select distinct invoices and then do the listagg order by invoice_date
    select distinct internal_buyer_line_item_id, is_seller_invoice, invoice_id, invoice_date
    from billing_event_with_business_flags) distinct_invoices
  group by internal_buyer_line_item_id
  order by internal_buyer_line_item_id
),

billing_event_with_categorized_transaction as (
-- Use the flags that were created in the previous transformation in more calculated columns:
-- NOTE: This transformation has no joins and no window functions
  select
    billing_event_id,
    end_user_account_id,
    agreement_id,
    proposer_account_id,
    offer_id,
    acceptor_account_id,
    case when is_cog or is_cog_refund then null else payer_account_id end as payer_account_id,
    product_id,
    action,
    transaction_type,
    parent_billing_event_id,
    disbursement_billing_event_id,
    amount,
    currency,
    balance_impacting,
    invoice_date,
    payment_due_date,
    usage_period_start_date,
    usage_period_end_date,
    invoice_id,
    billing_address_id,
    transaction_reference_id,
    buyer_transaction_reference_id,
    bank_trace_id,
    disbursement_date,
    disbursement_id,
    disbursement_id_or_invoiced,
    broker_id,
    bl.internal_buyer_line_item_id,
    is_seller_invoice,
    is_cog,
    is_cog_refund,
    is_manufacturer_view_of_reseller,

    -- Buyer/seller columns:
    case when is_seller_invoice then null else invoice_id end as buyer_invoice_id_or_null,
    seller_invoices.seller_invoice_id_or_null,
    case when is_seller_invoice then null else invoice_date end as buyer_invoice_date_or_null,
    seller_invoices.seller_invoice_date_or_null,

    -- Categorized amounts by transaction type:
    case when transaction_type =   'SELLER_REV_SHARE' and not is_cog then amount else 0 end as gross_revenue,
    case when transaction_type in ('SELLER_REV_SHARE_REFUND','SELLER_REV_SHARE_CREDIT') and not is_cog_refund then amount else 0 end as gross_refund,
    case when transaction_type =   'SELLER_REV_SHARE' and     is_cog then amount else 0 end as cogs,
    case when transaction_type in ('SELLER_REV_SHARE_REFUND','SELLER_REV_SHARE_CREDIT') and     is_cog_refund then amount else 0 end as cogs_refund,
    case when transaction_type =   'AWS_REV_SHARE' then amount else 0 end as aws_rev_share,
    case when transaction_type in ('AWS_REV_SHARE_REFUND','AWS_REV_SHARE_CREDIT') then amount else 0 end as aws_refund_share,
    case when transaction_type =   'AWS_TAX_SHARE' and not is_seller_invoice then amount else 0 end as aws_tax_share,             -- AWS tax share from _buyer_  invoice
    case when transaction_type =   'AWS_TAX_SHARE' and     is_seller_invoice then amount else 0 end as aws_tax_share_listing_fee, -- AWS tax share from _seller_ invoice
    case when transaction_type =   'AWS_TAX_SHARE_REFUND' and not is_seller_invoice then amount else 0 end as aws_tax_share_refund,
    case when transaction_type =   'AWS_TAX_SHARE_REFUND' and     is_seller_invoice then amount else 0 end as aws_tax_share_refund_listing_fee,
    case when transaction_type =   'SELLER_TAX_SHARE' then amount else 0 end as seller_tax_share,
    case when transaction_type =   'SELLER_TAX_SHARE_REFUND' then amount else 0 end as seller_tax_share_refund,
    case when transaction_type =   'BALANCE_ADJUSTMENT' then amount else 0 end as balance_adjustment,
    case when transaction_type =   'SELLER_REV_SHARE_CREDIT' then amount else 0 end as seller_rev_credit,
    case when transaction_type =   'AWS_REV_SHARE_CREDIT' then amount else 0 end as aws_ref_fee_credit
  from
    billing_event_with_business_flags as bl
    left join seller_invoice_list as seller_invoices
      on bl.internal_buyer_line_item_id = seller_invoices.internal_buyer_line_item_id
    ),

line_items_aggregated as (
-- This transformation has the only "group by" in all of these transformations.
-- NOTE: This transformation has no joins and no window functions
  select
    internal_buyer_line_item_id,
    disbursement_id,
    disbursement_id_or_invoiced,
    product_id,
    broker_id,
    currency,
    agreement_id,
    proposer_account_id,
    acceptor_account_id,
    max(payer_account_id) as payer_account_id,
    offer_id,
    end_user_account_id,
    usage_period_start_date,
    usage_period_end_date,
    max(payment_due_date) payment_due_date,
    buyer_transaction_reference_id,
    bank_trace_id,
    disbursement_date,
    max(billing_address_id) as billing_address_id,

    -- Buyer/seller columns:
    max(buyer_invoice_id_or_null) as buyer_invoice_id,
    max(seller_invoice_id_or_null) as seller_invoice_id,
    max(buyer_invoice_date_or_null) as buyer_invoice_date,
    max(seller_invoice_date_or_null) as seller_invoice_date,
  
    -- Categorized amounts by transaction type:
    -- When disbursement_id_or_invoiced = '<invoiced>',    these are invoiced amounts
    -- When disbursement_id_or_invoiced <> '<invoiced>' these are disbursed amounts for _this_ specific disbursement_id
    sum(gross_revenue) as gross_revenue_this_disbursement_id_or_invoiced,
    sum(gross_refund) as gross_refund_this_disbursement_id_or_invoiced,
    sum(cogs) as cogs_this_disbursement_id_or_invoiced,
    sum(cogs_refund) as cogs_refund_this_disbursement_id_or_invoiced,
    sum(aws_rev_share) as aws_rev_share_this_disbursement_id_or_invoiced,
    sum(aws_refund_share) as aws_refund_share_this_disbursement_id_or_invoiced,
    sum(aws_tax_share) as aws_tax_share_this_disbursement_id_or_invoiced,
    sum(aws_tax_share_listing_fee) as aws_tax_share_listing_fee_this_disbursement_id_or_invoiced,
    sum(aws_tax_share_refund) as aws_tax_share_refund_this_disbursement_id_or_invoiced,
    sum(aws_tax_share_refund_listing_fee) as aws_tax_share_refund_listing_fee_this_disbursement_id_or_invoiced,
    sum(seller_tax_share) as seller_tax_share_this_disbursement_id_or_invoiced,
    sum(seller_tax_share_refund) as seller_tax_share_refund_this_disbursement_id_or_invoiced,
    sum(balance_adjustment) as balance_adjustment_this_disbursement_id_or_invoiced,
    sum(seller_rev_credit) as seller_rev_credit_this_disbursement_id_or_invoiced,
    sum(aws_ref_fee_credit) as aws_ref_fee_credit_this_disbursement_id_or_invoiced
  from
    billing_event_with_categorized_transaction as billing_categorized
  group by
    internal_buyer_line_item_id,
    disbursement_id,
    disbursement_id_or_invoiced,
    broker_id,
    -- The following columns are included the in group by but they are intentionally omitted from the PK.
    -- These columns should have the _same_ values for each record in the PK.
    product_id,
    currency,
    agreement_id,
    proposer_account_id,
    acceptor_account_id,
    offer_id,
    end_user_account_id,
    usage_period_start_date,
    usage_period_end_date,
    buyer_transaction_reference_id,
    bank_trace_id,
    disbursement_date
),

-- listagg function in athena does not support partitioning, grouping here and then joining to the main query
disbursement_list as (
  select
    internal_buyer_line_item_id,
    listagg(cast(disbursement_date as varchar),',') within group (order by cast(disbursement_date as varchar)) as disbursement_date_list,
    listagg(bank_trace_id,',') within group (order by cast(disbursement_date as varchar)) as disburse_bank_trace_id_list
    from (
    -- listagg function in athena does not support ordering by another field when distinct is used,
    -- here we first select distinct bank_trace_ids and then do the listagg order by disbursement_date
    select distinct internal_buyer_line_item_id, disbursement_date, bank_trace_id
  from billing_event_with_business_flags) distinct_disbursements
  group by internal_buyer_line_item_id
  order by internal_buyer_line_item_id
),

line_items_with_window_functions as (
--add flag next step compare gross_revenue and gross_revenue_disbursed or gross_refund and gross_refund_disbursed
  select
    line_item.internal_buyer_line_item_id,
    disbursement_id,
    disbursement_id_or_invoiced,
    product_id,
    broker_id,
    currency,
    agreement_id,
    proposer_account_id,
    acceptor_account_id,
    -- when there's aws_rev_Share adjustment/refund to a seller_rev_share invoice, it can happen that for the same aws_rev_share invoice_id, there are multiple disbursement events,
    -- using windows function to map payer_account_id of seller_rev_share to all corresponding aws_rev_Share
    max(payer_account_id) over (partition by line_item.internal_buyer_line_item_id) as payer_account_id,
    offer_id,
    end_user_account_id,
    usage_period_start_date,
    usage_period_end_date,
    payment_due_date,
    bank_trace_id,
    disbursement_date,
    billing_address_id,

    -- Buyer/seller columns:
    max(buyer_invoice_id) over (partition by line_item.internal_buyer_line_item_id) as buyer_invoice_id,
    seller_invoice_id,
    max(buyer_invoice_date) over (partition by line_item.internal_buyer_line_item_id) as buyer_invoice_date,
    seller_invoice_date,

    -- When disbursement_id_or_invoiced = '<invoiced>', these are actually invoiced amounts
    -- When disbursement_id_or_invoiced <> '<invoiced>' these are disbursed amounts for _this_ specific disbursement_id
    gross_revenue_this_disbursement_id_or_invoiced,
    gross_refund_this_disbursement_id_or_invoiced,
    cogs_this_disbursement_id_or_invoiced,
    cogs_refund_this_disbursement_id_or_invoiced,
    aws_rev_share_this_disbursement_id_or_invoiced,
    aws_refund_share_this_disbursement_id_or_invoiced,
    aws_tax_share_this_disbursement_id_or_invoiced,
    aws_tax_share_listing_fee_this_disbursement_id_or_invoiced,
    aws_tax_share_refund_this_disbursement_id_or_invoiced,
    aws_tax_share_refund_listing_fee_this_disbursement_id_or_invoiced,
    seller_tax_share_this_disbursement_id_or_invoiced,
    seller_tax_share_refund_this_disbursement_id_or_invoiced,
    balance_adjustment_this_disbursement_id_or_invoiced,
    seller_rev_credit_this_disbursement_id_or_invoiced,
    aws_ref_fee_credit_this_disbursement_id_or_invoiced,
    -- IMPORTANT: All window functions partitioned by internal_buyer_line_item_id:

    -- Invoiced amounts, categorized by transaction type:
    sum(case when disbursement_id_or_invoiced = '<invoiced>' then gross_revenue_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end)over (partition by line_item.internal_buyer_line_item_id) as gross_revenue_invoiced,
    sum(case when disbursement_id_or_invoiced = '<invoiced>' then gross_refund_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as gross_refund_invoiced,
    sum(case when disbursement_id_or_invoiced = '<invoiced>' then cogs_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as cogs_invoiced,
    sum(case when disbursement_id_or_invoiced = '<invoiced>' then cogs_refund_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as cogs_refund_invoiced,
    sum(case when disbursement_id_or_invoiced = '<invoiced>' then aws_rev_share_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as aws_rev_share_invoiced,
    sum(case when disbursement_id_or_invoiced = '<invoiced>' then aws_refund_share_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as aws_refund_share_invoiced,
    sum(case when disbursement_id_or_invoiced = '<invoiced>' then aws_tax_share_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as aws_tax_share_invoiced,
    sum(case when disbursement_id_or_invoiced = '<invoiced>' then aws_tax_share_listing_fee_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as aws_tax_share_listing_fee_invoiced,
    sum(case when disbursement_id_or_invoiced = '<invoiced>' then aws_tax_share_refund_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as aws_tax_share_refund_invoiced,
    sum(case when disbursement_id_or_invoiced = '<invoiced>' then aws_tax_share_refund_listing_fee_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as aws_tax_share_refund_listing_fee_invoiced,
    sum(case when disbursement_id_or_invoiced = '<invoiced>' then seller_tax_share_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as seller_tax_share_invoiced,
    sum(case when disbursement_id_or_invoiced = '<invoiced>' then seller_tax_share_refund_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as seller_tax_share_refund_invoiced,
    sum(case when disbursement_id_or_invoiced = '<invoiced>' then balance_adjustment_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as balance_adjustment_invoiced,
    sum(case when disbursement_id_or_invoiced = '<invoiced>' then seller_rev_credit_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as seller_rev_credit_invoiced,
    sum(case when disbursement_id_or_invoiced = '<invoiced>' then aws_ref_fee_credit_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as aws_ref_fee_credit_invoiced,

    -- Total disbursed amounts (for all disbursement_id values), categorized by transaction type:
    sum(case when disbursement_id_or_invoiced <> '<invoiced>' then gross_revenue_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as gross_revenue_disbursed,
    sum(case when disbursement_id_or_invoiced <> '<invoiced>' then gross_refund_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as gross_refund_disbursed,
    sum(case when disbursement_id_or_invoiced <> '<invoiced>' then cogs_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as cogs_disbursed,
    sum(case when disbursement_id_or_invoiced <> '<invoiced>' then cogs_refund_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as cogs_refund_disbursed,
    sum(case when disbursement_id_or_invoiced <> '<invoiced>' then aws_rev_share_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as aws_rev_share_disbursed,
    sum(case when disbursement_id_or_invoiced <> '<invoiced>' then aws_refund_share_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as aws_refund_share_disbursed,
    sum(case when disbursement_id_or_invoiced <> '<invoiced>' then aws_tax_share_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as aws_tax_share_disbursed,
    sum(case when disbursement_id_or_invoiced <> '<invoiced>' then aws_tax_share_listing_fee_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as aws_tax_share_listing_fee_disbursed,
    sum(case when disbursement_id_or_invoiced <> '<invoiced>' then aws_tax_share_refund_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as aws_tax_share_refund_disbursed,
    sum(case when disbursement_id_or_invoiced <> '<invoiced>' then aws_tax_share_refund_listing_fee_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as aws_tax_share_refund_listing_fee_disbursed,
    sum(case when disbursement_id_or_invoiced <> '<invoiced>' then seller_tax_share_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as seller_tax_share_disbursed,
    sum(case when disbursement_id_or_invoiced <> '<invoiced>' then seller_tax_share_refund_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as seller_tax_share_refund_disbursed,
    sum(case when disbursement_id_or_invoiced <> '<invoiced>' then balance_adjustment_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as balance_adjustment_disbursed,
    sum(case when disbursement_id_or_invoiced <> '<invoiced>' then seller_rev_credit_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as seller_rev_credit_disbursed,
    sum(case when disbursement_id_or_invoiced <> '<invoiced>' then aws_ref_fee_credit_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as aws_ref_fee_credit_disbursed,

    -- aggregate multiple disbursement
    max(disbursement_date) over (partition by line_item.internal_buyer_line_item_id) as last_disbursement_date,
    first_value(case when disbursement_id_or_invoiced = '<invoiced>' then null else disbursement_id_or_invoiced end) over(partition by line_item.internal_buyer_line_item_id order by coalesce(disbursement_date,cast('1900-01-01' as timestamp)) desc rows between unbounded preceding and unbounded following) as last_disbursement_id,
    first_value(bank_trace_id) over (partition by line_item.internal_buyer_line_item_id order by coalesce(disbursement_date,cast('1900-01-01' as timestamp)) desc rows between unbounded preceding and unbounded following) as last_disburse_bank_trace_id,
    disb_list.disbursement_date_list,
    disb_list.disburse_bank_trace_id_list
  from
    line_items_aggregated as line_item
    left join disbursement_list disb_list
      on line_item.internal_buyer_line_item_id = disb_list.internal_buyer_line_item_id
),

cppo_offer_id as (
  select
    -- Channel partner offers do not exist in offertargetfeed_v1 table (as per legal requirement), causing cppo offer be defined as 'Public' in previous step, we will convert them back to 'Private' in next step
    offer_id
  from
    offers_with_uni_temporal_data
  where
    -- seller_account_id is null means the ISV owns the offer
    seller_account_id is not null
    and seller_account_id <>  (select seller_account_id from seller_account)
  group by
    offer_id
),

line_items_with_window_functions_enrich_offer_product_address as (
  select
    internal_buyer_line_item_id,
    disbursement_id,
    disbursement_id_or_invoiced,
    line.product_id,
    legacy_product.legacy_id as legacy_product_id,
    products.title as product_title,
    line.broker_id,
    line.currency,
    line.end_user_account_id,
    acc_enduser.encrypted_account_id as end_user_encrypted_account_id,
    acc_enduser.aws_account_id as end_user_aws_account_id,
    acc_payer.aws_account_id as payer_aws_account_id,
    acc_payer.encrypted_account_id payer_encrypted_account_id,
    line.agreement_id,
    agreement.agreement_revision,
    line.proposer_account_id,
    case when offer.offer_id like 'aiqoffer-%' then null else agreement.start_date end as Agreement_Start_Date,
    case when offer.offer_id like 'aiqoffer-%' then null else agreement.end_date end as Agreement_End_Date,
    case when offer.offer_id like 'aiqoffer-%' then null else agreement.acceptance_date end as Agreement_Acceptance_Date,
    case when offer.offer_id like 'aiqoffer-%' then null else agreement.valid_from end as agreement_updated_date,
    case when offer.offer_id like 'aiqoffer-%' then null else line.usage_period_start_date end as Usage_Period_Start_Date,
    case when offer.offer_id like 'aiqoffer-%' then null else line.usage_period_end_date end as Usage_Period_End_Date,

    line.acceptor_account_id,
    acc_subscriber.aws_account_id as subscriber_aws_account_id,
    acc_subscriber.encrypted_account_id as subscriber_encrypted_account_id,
    offer.offer_id,
    case
      when offer.offer_id in (
        select distinct offer_id
        from cppo_offer_id)
        then 'Private'
      else offer.offer_target
    end as offer_target,
    offer.name offer_name,
    offer.opportunity_name offer_opportunity_name,
    offer.opportunity_description offer_opportunity_description,
    offer.opportunity_id,
    payment_due_date,
    line.bank_trace_id,
    disbursement_date,
    billing_address_id,
    buyer_invoice_id,
    seller_invoice_id,
    buyer_invoice_date,
    seller_invoice_date,
    gross_revenue_this_disbursement_id_or_invoiced,
    gross_refund_this_disbursement_id_or_invoiced,
    cogs_this_disbursement_id_or_invoiced,
    cogs_refund_this_disbursement_id_or_invoiced,
    aws_rev_share_this_disbursement_id_or_invoiced,
    aws_refund_share_this_disbursement_id_or_invoiced,
    aws_tax_share_this_disbursement_id_or_invoiced,
    aws_tax_share_listing_fee_this_disbursement_id_or_invoiced,
    aws_tax_share_refund_this_disbursement_id_or_invoiced,
    aws_tax_share_refund_listing_fee_this_disbursement_id_or_invoiced,
    seller_tax_share_this_disbursement_id_or_invoiced,
    seller_tax_share_refund_this_disbursement_id_or_invoiced,
    balance_adjustment_this_disbursement_id_or_invoiced,
    seller_rev_credit_this_disbursement_id_or_invoiced,
    aws_ref_fee_credit_this_disbursement_id_or_invoiced,
    gross_revenue_invoiced,
    gross_refund_invoiced,
    cogs_invoiced,
    cogs_refund_invoiced,
    aws_rev_share_invoiced,
    aws_refund_share_invoiced,
    aws_tax_share_invoiced,
    aws_tax_share_listing_fee_invoiced,
    aws_tax_share_refund_invoiced,
    aws_tax_share_refund_listing_fee_invoiced,
    seller_tax_share_invoiced,
    seller_tax_share_refund_invoiced,
    balance_adjustment_invoiced,
    seller_rev_credit_invoiced,
    aws_ref_fee_credit_invoiced,
    gross_revenue_disbursed,
    gross_refund_disbursed,
    cogs_disbursed,
    cogs_refund_disbursed,
    aws_rev_share_disbursed,
    aws_refund_share_disbursed,
    aws_tax_share_disbursed,
    aws_tax_share_listing_fee_disbursed,
    aws_tax_share_refund_disbursed,
    aws_tax_share_refund_listing_fee_disbursed,
    seller_tax_share_disbursed,
    seller_tax_share_refund_disbursed,
    balance_adjustment_disbursed,
    seller_rev_credit_disbursed,
    aws_ref_fee_credit_disbursed,
    last_disbursement_date,
    last_disbursement_id,
    last_disburse_bank_trace_id,
    disbursement_date_list,
    disburse_bank_trace_id_list,
    products.product_code,
    acc_products.aws_account_id as manufacturer_aws_account_id,
    products.manufacturer_account_id,
    --add subscriber and payer addressID, payer address preference order: tax address > billing address > mailing address,  subscriber address preference order: tax address >  mailing address
    coalesce (
      --empty value in Athena shows as '', change all '' value to null in order to follow the preference order logic above
      case when acc_subscriber.tax_address_id ='' then null else acc_subscriber.tax_address_id end,
      case when acc_subscriber.mailing_address_id = '' then null else acc_subscriber.mailing_address_id end) as subscriber_address_id,
    coalesce (
      case when acc_payer.tax_address_id = '' then null else acc_payer.tax_address_id end,
      case when line.billing_address_id = '' then null else line.billing_address_id end,
      case when acc_payer.mailing_address_id = '' then null else acc_payer.mailing_address_id end) as payer_address_id,
    coalesce (
      case when acc_enduser.tax_address_id = '' then null else acc_enduser.tax_address_id end,
      case when line.billing_address_id = '' then null else line.billing_address_id end,
      case when acc_enduser.mailing_address_id = '' then null else acc_enduser.mailing_address_id end) as end_user_address_id
  from
    line_items_with_window_functions as line
  left join agreements_with_history as agreement on
      line.agreement_id = agreement.agreement_id and line.buyer_invoice_date >= agreement.valid_from_adjusted and line.buyer_invoice_date<agreement.valid_to
  left join offers_with_history_with_target_type as offer on
        line.offer_id = offer.offer_id and line.buyer_invoice_date >= offer.valid_from and line.buyer_invoice_date<offer.valid_to
  left join products_with_history as products on
        line.product_id = products.product_id and line.buyer_invoice_date >= products.valid_from_adjusted and line.buyer_invoice_date<products.valid_to
  left join legacy_products as legacy_product on
        line.product_id = legacy_product.new_id
  left join accounts_with_history_with_company_name as acc_payer on
        line.payer_account_id = acc_payer.account_id and line.buyer_invoice_date >= acc_payer.valid_from and line.buyer_invoice_date<acc_payer.valid_to
  left join accounts_with_history_with_company_name as acc_enduser on
        line.end_user_account_id = acc_enduser.account_id and line.buyer_invoice_date >= acc_enduser.valid_from and line.buyer_invoice_date<acc_enduser.valid_to
  left join accounts_with_history_with_company_name as acc_subscriber on
        line.acceptor_account_id = acc_subscriber.account_id and line.buyer_invoice_date >= acc_subscriber.valid_from and line.buyer_invoice_date<acc_subscriber.valid_to
  left join accounts_with_history_with_company_name as acc_products on
        products.manufacturer_account_id = acc_products.account_id and line.buyer_invoice_date >= acc_products.valid_from and line.buyer_invoice_date<acc_products.valid_to

),

line_items_with_window_functions_enrich_offer_product_address_name as (
  select
    line.internal_buyer_line_item_id,
    disbursement_id,
    disbursement_id_or_invoiced,
    product_id,
    legacy_product_id,
    product_title,
    broker_id,
    currency,
    end_user_address_id,
    end_user_account_id,
    end_user_encrypted_account_id,
    end_user_aws_account_id,
    add_enduser.company_name end_user_company_name,
    add_enduser.email_domain end_user_email_domain,
    add_enduser.city end_user_city,
    add_enduser.state_or_region end_user_state,
    add_enduser.country_code end_user_country,
    add_enduser.postal_code end_user_postal_code,
    payer_aws_account_id,
    payer_encrypted_account_id,
    payer_address_id,
    add_payer.company_name payer_company_name,
    add_payer.email_domain payer_email_domain,
    add_payer.city payer_city,
    add_payer.state_or_region payer_state,
    add_payer.country_code payer_country,
    add_payer.postal_code payer_postal_code,
    agreement_id,
    agreement_revision,
    agreement_start_date,
    agreement_end_date,
    agreement_acceptance_date,
    agreement_updated_date,
    case when proposer_account_id = (select seller_account_id from seller_account) then null else acc_proposer.aws_account_id end as reseller_aws_account_id,
    case when proposer_account_id = (select seller_account_id from seller_account) then null else acc_proposer.mailing_company_name end as reseller_company_name,
    usage_period_start_date,
    usage_period_end_date,
    proposer_account_id,
    acc_proposer.aws_account_id as proposer_aws_account_id,
    acceptor_account_id,
    subscriber_aws_account_id,
    subscriber_encrypted_account_id,
    subscriber_address_id,
    add_subscriber.company_name subscriber_company_name,
    add_subscriber.email_domain subscriber_email_domain,
    add_subscriber.city subscriber_city,
    add_subscriber.state_or_region subscriber_state,
    add_subscriber.country_code subscriber_country,
    add_subscriber.postal_code subscriber_postal_code,
    offer_id,
    offer_target,
    offer_name,
    offer_opportunity_name,
    offer_opportunity_description,
    opportunity_id,
    payment_due_date,
    bank_trace_id,
    disbursement_date,
    billing_address_id,
    max(buyer_invoice_id)as buyer_invoice_id,
    max(seller_invoice_id)as seller_invoice_id,
    max(buyer_invoice_date)as buyer_invoice_date,
    max(seller_invoice_date)as seller_invoice_date,
    gross_revenue_this_disbursement_id_or_invoiced,
    gross_refund_this_disbursement_id_or_invoiced,
    cogs_this_disbursement_id_or_invoiced,
    cogs_refund_this_disbursement_id_or_invoiced,
    aws_rev_share_this_disbursement_id_or_invoiced,
    aws_refund_share_this_disbursement_id_or_invoiced,
    aws_tax_share_this_disbursement_id_or_invoiced,
    aws_tax_share_listing_fee_this_disbursement_id_or_invoiced,
    aws_tax_share_refund_this_disbursement_id_or_invoiced,
    aws_tax_share_refund_listing_fee_this_disbursement_id_or_invoiced,
    seller_tax_share_this_disbursement_id_or_invoiced,
    seller_tax_share_refund_this_disbursement_id_or_invoiced,
    balance_adjustment_this_disbursement_id_or_invoiced,
    seller_rev_credit_this_disbursement_id_or_invoiced,
    aws_ref_fee_credit_this_disbursement_id_or_invoiced,
    (gross_revenue_this_disbursement_id_or_invoiced + gross_refund_this_disbursement_id_or_invoiced + aws_rev_share_this_disbursement_id_or_invoiced + aws_refund_share_this_disbursement_id_or_invoiced + seller_tax_share_this_disbursement_id_or_invoiced + seller_tax_share_refund_this_disbursement_id_or_invoiced
      + cogs_this_disbursement_id_or_invoiced + cogs_refund_this_disbursement_id_or_invoiced + aws_tax_share_listing_fee_this_disbursement_id_or_invoiced + aws_tax_share_refund_listing_fee_this_disbursement_id_or_invoiced) as seller_net_revenue_this_disbursement_id_or_invoiced,
    gross_revenue_invoiced,
    gross_refund_invoiced,
    cogs_invoiced,
    cogs_refund_invoiced,
    aws_rev_share_invoiced,
    aws_refund_share_invoiced,
    aws_tax_share_invoiced,
    aws_tax_share_listing_fee_invoiced,
    aws_tax_share_refund_invoiced,
    aws_tax_share_refund_listing_fee_invoiced,
    seller_tax_share_invoiced,
    seller_tax_share_refund_invoiced,
    balance_adjustment_invoiced,
    seller_rev_credit_invoiced,
    aws_ref_fee_credit_invoiced,
    gross_revenue_disbursed,
    gross_refund_disbursed,
    cogs_disbursed,
    cogs_refund_disbursed,
    aws_rev_share_disbursed,
    aws_refund_share_disbursed,
    aws_tax_share_disbursed,
    aws_tax_share_listing_fee_disbursed,
    aws_tax_share_refund_disbursed,
    aws_tax_share_refund_listing_fee_disbursed,
    seller_tax_share_disbursed,
    seller_tax_share_refund_disbursed,
    balance_adjustment_disbursed,
    seller_rev_credit_disbursed,
    aws_ref_fee_credit_disbursed,
    (gross_revenue_invoiced + gross_revenue_disbursed) as uncollected_gross_revenue,
    -- net revenue = gross revenue - listing fee - tax - cogs
    (gross_revenue_invoiced + gross_refund_invoiced + aws_rev_share_invoiced + aws_refund_share_invoiced + seller_tax_share_invoiced + seller_tax_share_refund_invoiced + cogs_invoiced + cogs_refund_invoiced + aws_tax_share_listing_fee_invoiced + aws_tax_share_refund_listing_fee_invoiced) as seller_net_revenue,
    (gross_revenue_invoiced + gross_refund_invoiced + aws_rev_share_invoiced + aws_refund_share_invoiced + seller_tax_share_invoiced + seller_tax_share_refund_invoiced + cogs_invoiced + cogs_refund_invoiced + aws_tax_share_listing_fee_invoiced + aws_tax_share_refund_listing_fee_invoiced
      + gross_revenue_disbursed + gross_refund_disbursed + aws_rev_share_disbursed + aws_refund_share_disbursed + seller_tax_share_disbursed + seller_tax_share_refund_disbursed + cogs_disbursed + cogs_refund_disbursed + aws_tax_share_listing_fee_disbursed + aws_tax_share_refund_listing_fee_disbursed) as uncollected_seller_net_revenue,
    last_disbursement_date,
    last_disbursement_id,
    last_disburse_bank_trace_id,
    disbursement_date_list,
    disburse_bank_trace_id_list,
    product_code,
    manufacturer_aws_account_id,
    manufacturer_account_id,
    acc_manu.mailing_company_name as manufacturer_company_name,
    cast(null as varchar) as AR_Period,
    case
      when (
        (gross_revenue_invoiced <>0 and gross_revenue_invoiced = -1 * gross_revenue_disbursed)
        or (gross_refund_invoiced <> 0 and gross_refund_invoiced = -1 * gross_refund_disbursed)
        or (balance_adjustment_invoiced <> 0 and balance_adjustment_invoiced = -1 * balance_adjustment_disbursed)
        or (seller_tax_share_refund_invoiced <> 0 and seller_tax_share_refund_invoiced = -1 * seller_tax_share_refund_disbursed)
        or (gross_revenue_invoiced = 0 and gross_refund_invoiced = 0 and balance_adjustment_invoiced = 0 and seller_tax_share_refund_invoiced = 0 and last_disbursement_id is not null)) then 'Yes'
      when gross_revenue_disbursed = 0 and gross_refund_disbursed = 0 and balance_adjustment_disbursed = 0 and seller_tax_share_disbursed = 0 and seller_tax_share_refund_disbursed = 0 then 'No'
      else 'Partial'
    end as Disbursement_Flag
  from line_items_with_window_functions_enrich_offer_product_address as line
  left join accounts_with_history_with_company_name as acc_manu on
    line.manufacturer_account_id = acc_manu.account_id and line.buyer_invoice_date >= acc_manu.valid_from_adjusted and line.buyer_invoice_date <= acc_manu.valid_to
  left join accounts_with_history_with_company_name as acc_proposer on
    line.proposer_account_id = acc_proposer.account_id and line.buyer_invoice_date >= acc_proposer.valid_from and line.buyer_invoice_date<acc_proposer.valid_to
  left join address_with_latest_revision as add_payer on
    line.payer_address_id = add_payer.address_id
  left join address_with_latest_revision as add_subscriber on
    line.subscriber_address_id = add_subscriber.address_id
  left join address_with_latest_revision as add_enduser on
    line.end_user_address_id = add_enduser.address_id
  group by
    line.internal_buyer_line_item_id,
    disbursement_id,
    disbursement_id_or_invoiced,
    product_id,
    legacy_product_id,
    product_title,
    broker_id,
    currency,
    end_user_address_id,
    end_user_account_id,
    end_user_encrypted_account_id,
    end_user_aws_account_id,
    add_enduser.company_name,
    add_enduser.email_domain,
    add_enduser.city,
    add_enduser.state_or_region,
    add_enduser.country_code,
    add_enduser.postal_code,
    payer_aws_account_id,
    payer_encrypted_account_id,
    payer_address_id,
    add_payer.company_name,
    add_payer.email_domain,
    add_payer.city,
    add_payer.state_or_region,
    add_payer.country_code,
    add_payer.postal_code,
    agreement_id,
    agreement_revision,
    case when proposer_account_id = (select seller_account_id from seller_account) then null else acc_proposer.aws_account_id end,
    case when proposer_account_id = (select seller_account_id from seller_account) then null else acc_proposer.mailing_company_name end,
    agreement_start_date,
    agreement_end_date,
    agreement_acceptance_date,
    agreement_updated_date,
    usage_period_start_date,
    usage_period_end_date,
    acceptor_account_id,
    subscriber_aws_account_id,
    subscriber_encrypted_account_id,
    subscriber_address_id,
    add_subscriber.company_name,
    add_subscriber.email_domain,
    add_subscriber.city,
    add_subscriber.state_or_region,
    add_subscriber.country_code,
    add_subscriber.postal_code,
    offer_id,
    offer_target,
    offer_name,
    offer_opportunity_name,
    offer_opportunity_description,
    opportunity_id,
    payment_due_date,
    bank_trace_id,
    disbursement_date,
    billing_address_id,
    gross_revenue_this_disbursement_id_or_invoiced,
    gross_refund_this_disbursement_id_or_invoiced,
    cogs_this_disbursement_id_or_invoiced,
    cogs_refund_this_disbursement_id_or_invoiced,
    aws_rev_share_this_disbursement_id_or_invoiced,
    aws_refund_share_this_disbursement_id_or_invoiced,
    aws_tax_share_this_disbursement_id_or_invoiced,
    aws_tax_share_listing_fee_this_disbursement_id_or_invoiced,
    aws_tax_share_refund_this_disbursement_id_or_invoiced,
    aws_tax_share_refund_listing_fee_this_disbursement_id_or_invoiced,
    seller_tax_share_this_disbursement_id_or_invoiced,
    seller_tax_share_refund_this_disbursement_id_or_invoiced,
    balance_adjustment_this_disbursement_id_or_invoiced,
    seller_rev_credit_this_disbursement_id_or_invoiced,
    aws_ref_fee_credit_this_disbursement_id_or_invoiced,
    gross_revenue_invoiced,
    gross_refund_invoiced,
    cogs_invoiced,
    cogs_refund_invoiced,
    aws_rev_share_invoiced,
    aws_refund_share_invoiced,
    aws_tax_share_invoiced,
    aws_tax_share_listing_fee_invoiced,
    aws_tax_share_refund_invoiced,
    aws_tax_share_refund_listing_fee_invoiced,
    seller_tax_share_invoiced,
    seller_tax_share_refund_invoiced,
    balance_adjustment_invoiced,
    seller_rev_credit_invoiced,
    aws_ref_fee_credit_invoiced,
    gross_revenue_disbursed,
    gross_refund_disbursed,
    cogs_disbursed,
    cogs_refund_disbursed,
    aws_rev_share_disbursed,
    aws_refund_share_disbursed,
    aws_tax_share_disbursed,
    aws_tax_share_listing_fee_disbursed,
    aws_tax_share_refund_disbursed,
    aws_tax_share_refund_listing_fee_disbursed,
    seller_tax_share_disbursed,
    seller_tax_share_refund_disbursed,
    balance_adjustment_disbursed,
    seller_rev_credit_disbursed,
    aws_ref_fee_credit_disbursed,
    last_disbursement_date,
    last_disbursement_id,
    last_disburse_bank_trace_id,
    disbursement_date_list,
    disburse_bank_trace_id_list,
    product_code,
    manufacturer_aws_account_id,
    manufacturer_account_id,
    acc_manu.mailing_company_name,
    proposer_account_id,
    acc_proposer.aws_account_id
),

billed_revenue as (
  select
    ------------------
    -- Invoice Info --
    ------------------
    buyer_invoice_date as Invoice_Date,
    Payment_Due_Date as Payment_Due_Date,
    concat(
      'Net ',
      case
        when abs(date_diff('Day', Payment_due_date, buyer_invoice_date))>180 then '180+'
        else cast(abs(date_diff('Day', Payment_due_date, buyer_invoice_date)) as varchar)
        end,
      ' days'
    ) as payment_terms,
    buyer_invoice_id as Invoice_ID,
    coalesce(
      --empty value in Athena shows as '', change all '' value to null
      case when seller_invoice_id = '' then null else seller_invoice_id end,
      'Not applicable') as Listing_Fee_Invoice_ID,

    ---------------------------
    --End user Information --
    ---------------------------
    coalesce(
      --empty value in Athena shows as '', change all '' value to null
      case when End_User_Company_Name = '' then null else End_User_Company_Name end,
      'Not available') as End_User_Company_Name,
    End_User_AWS_Account_ID,
    End_User_Encrypted_Account_ID,
    End_User_Email_Domain,
    End_User_City,
    End_User_State as End_User_State_or_Region,
    End_User_Country,
    End_User_Postal_Code,
    End_User_Address_ID,

    ---------------------------
    --Subscriber Information --
    ---------------------------
    case
      when Agreement_Id is null or Agreement_ID = '' then 'Not available'
      when Subscriber_Company_Name is null or Subscriber_Company_Name = '' then 'Not provided'
      else Subscriber_Company_Name
      end as Subscriber_Company_Name,
    case
      when Agreement_Id is null or Agreement_ID = '' then 'Not available'
      else Subscriber_AWS_Account_ID
      end as Subscriber_AWS_Account_ID,
    case
      when Agreement_Id is null or Agreement_ID = '' then 'Not available'
      else Subscriber_Encrypted_Account_ID
      end as Subscriber_Encrypted_Account_ID,
    case
      when Agreement_Id is null or Agreement_ID = '' then 'Not available'
      when Subscriber_Email_Domain is null or Subscriber_Email_Domain = '' then 'Not provided'
      else Subscriber_Email_Domain
      end as Subscriber_Email_Domain,
    case
      when Agreement_id is null or Agreement_ID = '' then 'Not available'
      when Subscriber_City is null or Subscriber_City = '' then 'Not provided'
      else Subscriber_City
      end as Subscriber_City,
    case
      when Agreement_Id is null or Agreement_ID = '' then 'Not available'
      when Subscriber_State is null or Subscriber_State = '' then 'Not provided'
      else Subscriber_State
      end as Subscriber_State_or_Region,
    case
      when Agreement_Id is null or Agreement_ID = '' then 'Not available'
      when Subscriber_Country is null or Subscriber_Country = '' then 'Not provided'
      else Subscriber_Country
      end as Subscriber_Country,
    case
      when Agreement_Id is null or Agreement_ID = '' then 'Not available'
      when Subscriber_Postal_Code is null or Subscriber_Postal_Code = '' then 'Not provided'
      else Subscriber_Postal_Code
      end as Subscriber_Postal_Code,
    case
      when Agreement_ID is null or Agreement_ID = '' then 'Not available'
      when Subscriber_Address_ID is null or Subscriber_Address_ID = '' then 'Not provided'
      else Subscriber_Address_ID
      end as Subscriber_Address_ID,

    ----------------------
    -- Procurement Info --
    ----------------------
     -- product title at time of invoice. It is possible that the title changes over time and therefore there may be multiple product titles mapped to a single product id.
    coalesce(
      --empty value in Athena shows as '', change all '' value to null
      case when Product_Title = '' then null else Product_Title end,
      'Not provided') as Product_Title,
    -- offer name at time of invoice. It is possible that the name changes over time therefore there may be multiple offer names mapped to a single offer id.
    case
      when Agreement_Id is null or Agreement_ID = '' then 'Not available'
      when (Offer_Name is null or Offer_Name = '') and Offer_Target = 'Public' then 'Not applicable'
      else Offer_Name
      end as Offer_Name,
    case
      when Agreement_Id is null or Agreement_ID = ''
      then 'Not available'
      else Offer_ID
      end as Offer_ID,
    -- offer target at time of invoice.,
    case
      when Agreement_Id is null or Agreement_ID = '' then 'Not available'
      else Offer_Target
      end as Offer_Visibility,
    coalesce(
      --empty value in Athena shows as '', change all '' value to null
      case when Agreement_ID = '' then null else Agreement_ID end,
      'Not available') as Agreement_ID,
    Agreement_Start_Date,
    Agreement_Acceptance_Date,
    Agreement_End_Date,
    Usage_Period_Start_Date,
    Usage_Period_End_Date,

    -----------------------
    -- Disbursement Info --
    -----------------------
    case
      when Disbursement_Flag = 'Yes' then 'Disbursed'
      when Disbursement_Flag = 'No' then 'Not Disbursed'
      else 'Other'
      end as Disbursement_Status,
    last_disbursement_date as disbursement_date,
    case
      when Disbursement_Flag = 'No' then 'Not applicable'
      when disburse_bank_trace_id_list is null or disburse_bank_trace_id_list = '' then 'Not available'
      else disburse_bank_trace_id_list
    end as disburse_bank_trace_id,

    --------------
    -- Revenues --
    --------------
    -- We are rounding the sums using 2 decimal precision
    -- Note that the rounding method might differ between SQL implementations.
    -- The monthly revenue report is using RoundingMode.HALF_UP. This might create tiny discrepancies between this SQL output
    -- and the legacy report
    round(gross_revenue_invoiced,2) as Gross_Revenue,
    round(gross_refund_invoiced,2) as Gross_Refund,
    round(aws_rev_share_invoiced,2) as Listing_Fee,
    round(aws_refund_share_invoiced,2) as Listing_Fee_Refund,
    truncate(
      case
        when gross_revenue_invoiced != 0 then abs(aws_rev_share_invoiced/gross_revenue_invoiced)
        when gross_refund_invoiced != 0 then abs(aws_refund_share_invoiced/gross_refund_invoiced)
        else 0
      end
      ,4) as Listing_Fee_Percentage,
    round(seller_tax_share_invoiced,2) as Seller_Tax_Share,
    round(seller_tax_share_refund_invoiced,2) as Seller_Tax_Share_Refund,
    round(aws_tax_share_invoiced,2) as AWS_Tax_Share,
    round(aws_tax_share_refund_invoiced,2) as AWS_Tax_Share_Refund,
    round(aws_tax_share_listing_fee_invoiced,2) as AWS_Tax_Share_Listing_Fee,
    round(aws_tax_share_refund_listing_fee_invoiced,2) as AWS_Tax_Share_Refund_Listing_Fee,
    round(cogs_invoiced,2) as Wholesale_cost,
    round(cogs_refund_invoiced,2) as Wholesale_cost_Refund,
    round(seller_net_revenue,2) as Seller_Net_Revenue,
    currency as Currency,

    substring(internal_buyer_line_item_id,1,strpos(internal_buyer_line_item_id,'-')-1) as Transaction_Reference_ID,
    broker_id as AWS_seller_of_record,

    -----------------
    -- Resale info --
    -----------------
    case
      when Opportunity_Id is null or Opportunity_Id = '' then
        case
          when Offer_Target = 'Public' then 'Not applicable'
          when (Offer_Target is null or Offer_Target = '') and (Agreement_Id is not null and Agreement_Id != '') then 'Not applicable'
          else null
        end
      else Opportunity_Id
    end as Resale_authorization_ID,
    case
      when Offer_Opportunity_Name is null or Offer_Opportunity_Name = '' then
        case
          when Offer_Target = 'Public' then 'Not applicable'
          when (Offer_Target is null or Offer_Target = '') and (Agreement_Id is not null and Agreement_Id != '') then 'Not applicable'
          else null
        end
      else Offer_Opportunity_Name
    end as Resale_authorization_name,
    case
      when Offer_Opportunity_Description is null or Offer_Opportunity_Description = '' then
        case
         when Offer_Target = 'Public' then 'Not applicable'
         when (Offer_Target is null or Offer_Target = '') and (Agreement_Id is not null and Agreement_Id != '') then 'Not applicable'
         else null
        end
      else Offer_Opportunity_Name
    end as Resale_authorization_description,
    case
      when (Reseller_AWS_Account_ID is not null and Reseller_AWS_Account_ID != '')
        and (Reseller_Company_Name is null or Reseller_Company_Name = '') then 'Not available'
      when (Reseller_AWS_Account_ID is null or Reseller_AWS_Account_ID = '')
        and (opportunity_id is null or opportunity_id = '') then 'Not applicable'
      when (select seller_account_id from seller_account) <> manufacturer_aws_account_id
        and (Reseller_AWS_Account_ID is null or Reseller_AWS_Account_ID = '') then 'Not applicable'
      else Reseller_Company_Name
    end as Reseller_Company_Name,
    case
      when (Reseller_AWS_Account_ID is null or Reseller_AWS_Account_ID = '')
        and (Opportunity_Id is null or Opportunity_Id = '') then 'Not applicable'
      when (select seller_account_id from seller_account) <> manufacturer_aws_account_id
        and (Reseller_AWS_Account_ID is null or Reseller_AWS_Account_ID = '') then 'Not applicable'
      else Reseller_AWS_Account_ID
    end as Reseller_AWS_Account_ID,

    -----------------------
    -- Payer Information --
    -----------------------
    coalesce(
      --empty value in Athena shows as '', change all '' value to null
      case when Payer_Company_Name = '' then null else Payer_Company_Name end,
      'Not available') as Payer_Company_Name,
    Payer_AWS_Account_ID,
    Payer_Encrypted_Account_ID,
    Payer_Email_Domain,
    Payer_City,
    Payer_State as Payer_State_or_Region,
    Payer_Country,
    Payer_Postal_Code,
    Payer_Address_ID,

    ---------------------
    -- ISV Information --
    ---------------------
    manufacturer_aws_account_id as ISV_Account_ID,
    coalesce(
      --empty value in Athena shows as '', change all '' value to null
      case when Manufacturer_Company_Name = '' then null else Manufacturer_Company_Name end,
      'Not available') as ISV_Company_Name,

    ---------------------
    -- Products info --
    ---------------------
    Legacy_Product_ID,
    coalesce(
      --empty value in Athena shows as '', change all '' value to null
      case when Product_ID = '' then null else Product_ID end,
      'Not provided') as Product_ID,
    Product_Code
  from
    line_items_with_window_functions_enrich_offer_product_address_name as line
  where disbursement_id_or_invoiced = '<invoiced>'

)

select *
from billed_revenue
where invoice_date >= date_add('DAY', -90, current_date)
--where invoice_date between cast('2023-01-01' as timestamp) and cast('2024-03-01' as timestamp)
```

## 未收取或已支付的发票
<a name="data-feed-example-collections"></a>

要查找未收取或已支付的发票，您可以运行如以下示例所示的一组查询。这些查询相互依存，可创建**收款和支出**报告。您可以使用下方所示的示例，也可以根据您的数据和用例进行自定义。

查询中的注释说明了查询的作用以及如何修改查询。

```
-- Collections and disbursements report

-- General note: When running this query, we assume that the data ingested in the database uses
-- two time axes (the valid_from column and the update_date column).
-- See documentation for more details: https://docs.aws.amazon.com/marketplace/latest/userguide/data-feed.html#data-feed-details

-- An account_id has several valid_from dates (each representing a separate revision of the data)
-- but because of bi-temporality, an account_id + valid_from tuple can appear multiple times with a different update_date.
-- We are only interested in the most recent tuple (ie, uni-temporal model)
with accounts_with_uni_temporal_data as (
  select
    account_id,
    aws_account_id,
    encrypted_account_id,
    mailing_address_id,
    tax_address_id,
    tax_legal_name,
    from_iso8601_timestamp(valid_from) as valid_from,
    tax_registration_number
  from
    (
      select
        account_id,
        aws_account_id,
        encrypted_account_id,
        mailing_address_id,
        tax_address_id,
        tax_legal_name,
        valid_from,
        delete_date,
        tax_registration_number,
        row_number() over (partition by account_id, valid_from order by from_iso8601_timestamp(update_date) desc) as row_num
      from
        accountfeed_v1
    )
    where
      -- keep latest ...
      row_num = 1
      -- ... and remove the soft-deleted one.
      and (delete_date is null or delete_date = '')
  ),

accounts_with_history as (
  with accounts_with_history_with_extended_valid_from as (
    select
      account_id,
      -- sometimes, this columns gets imported as a "bigint" and loses heading 0s -> casting to a char and re-adding heading 0s (if need be)
      substring('000000000000'||cast(aws_account_id as varchar),-12) as aws_account_id,
      encrypted_account_id,
      mailing_address_id,
      tax_address_id,
      tax_legal_name,
      -- The start time of account valid_from is extended to '1970-01-01 00:00:00', because:
      -- ... in tax report transformations, some tax line items with invoice_date cannot
      -- ... fall into the default valid time range of the associated account
      case
        when lag(valid_from) over (partition by account_id order by valid_from asc) is null
          then cast('1970-01-01 00:00:00' as timestamp)
        else valid_from
      end as valid_from
    from accounts_with_uni_temporal_data
    )
  select
    account_id,
    aws_account_id,
    encrypted_account_id,
    mailing_address_id,
    tax_address_id,
    tax_legal_name,
    valid_from,
    coalesce(
      lead(valid_from) over (partition by account_id order by valid_from asc),
      cast('2999-01-01 00:00:00' as timestamp)
    ) as valid_to
  from
    accounts_with_history_with_extended_valid_from
),

-- An address_id has several valid_from dates (each representing a separate revision of the data)
-- but because of bi-temporality, an account_id + valid_from tuple can appear multiple times with a different update_date.
-- We are only interested in the most recent tuple (ie, uni-temporal model)
address_with_uni_temporal_data as (
  select
    from_iso8601_timestamp(valid_from) as valid_from,
    address_id,
    company_name,
    email_domain,
    country_code,
    state_or_region,
    city,
    postal_code,
    row_num
  from
  (
    select
      valid_from,
      update_date,
      delete_date,
      address_id,
      company_name,
      email_domain,
      country_code,
      state_or_region,
      city,
      postal_code,
      row_number() over (partition by address_id, valid_from order by from_iso8601_timestamp(update_date) desc) as row_num
    from
      addressfeed_v1
  )
  where
    -- keep latest ...
    row_num = 1
    -- ... and remove the soft-deleted one.
    and (delete_date is null or delete_date = '')
),

-- We are only interested in the most recent tuple (BTW: a given address is not supposed to change over time but when bugs ;-) so this query mainly does nothing)
address_with_latest_revision as (
  select
    valid_from,
    address_id,
    company_name,
    email_domain,
    country_code,
    state_or_region,
    city,
    postal_code,
    row_num_latest_revision
  from
  (
    select
      valid_from,
      address_id,
      company_name,
      email_domain,
      country_code,
      state_or_region,
      city,
      postal_code,
      row_number() over (partition by address_id order by valid_from desc) as row_num_latest_revision
    from
      address_with_uni_temporal_data
  )
  where
    row_num_latest_revision = 1
),

accounts_with_history_with_company_name as (
  select
    awh.account_id,
    awh.aws_account_id,
    awh.encrypted_account_id,
    awh.mailing_address_id,
    awh.tax_address_id,
    coalesce(
      --empty value in Athena shows as '', change all '' value to null
      case when address.company_name = '' then null else address.company_name end,
      awh.tax_legal_name) as mailing_company_name,
    address.email_domain,
    awh.valid_from,
    -- For BYOL, the agreement might be accepted (using some external non-AWS system or manual process) days before
    -- that BYOL agreement is entered into AWS Marketplace by the buyer. Therefore, the buyer is permitted to manually
    -- enter a backdated acceptance date, which might predate the point in time when the account was created.
    -- To work around this, we need to adjust the valid_from of the account to be
    -- earlier than the earliest possible backdated BYOL agreement acceptance date.
    case
      when lag(awh.valid_from) over (partition by aws_account_id order by awh.valid_from asc) is null
      then date_add('Day', -212, awh.valid_from)
      -- 212 is the longest delay between acceptance_date of the agreement and the account start_Date
      else awh.valid_from
    end as valid_from_adjusted,
    awh.valid_to
  from accounts_with_history as awh
  left join address_with_latest_revision as address on
    awh.mailing_address_id = address.address_id and awh.mailing_address_id is not null
),

-- An agreement_id has several valid_from dates (each representing an agreement revision)
-- but because of bi-temporality, an agreement_id + valid_from tuple can appear multiple times with a different update_date.
-- We are only interested in the most recent tuple (ie, uni-temporal model)
agreements_with_uni_temporal_data as (
  select
    agreement_id,
    origin_offer_id,
    proposer_account_id,
    acceptor_account_id,
    agreement_revision,
    from_iso8601_timestamp(valid_from) as valid_from,
    from_iso8601_timestamp(start_date) as start_date,
    from_iso8601_timestamp(end_date) as end_date,
    from_iso8601_timestamp(acceptance_date) as acceptance_date,
    agreement_type,
    previous_agreement_id,
    agreement_intent
  from
  (
    select
      --empty value in Athena shows as '', change all '' value to null
      case when agreement_id = '' then null else agreement_id end as agreement_id,
      origin_offer_id,
      proposer_account_id,
      acceptor_account_id,
      agreement_revision,
      valid_from,
      delete_date,
      start_date,
      end_date,
      acceptance_date,
      agreement_type,
      previous_agreement_id,
      agreement_intent,
      row_number() over (partition by agreement_id, valid_from order by from_iso8601_timestamp(update_date) desc) as row_num
    from
      -- TODO change to agreementfeed_v1 when Agreement Feed is GA'ed
      agreementfeed
  )
  where
    -- keep latest ...
    row_num = 1
    -- ... and remove the soft-deleted one.
    and (delete_date is null or delete_date = '')
),

agreements_with_history as (
  with agreements_with_window_functions as (
    select
      agreement_id,
      origin_offer_id as offer_id,
      proposer_account_id,
      acceptor_account_id,
      agreement_revision,
      start_date,
      end_date,
      acceptance_date,
      -- The start time of agreement valid_from is extended to '1970-01-01 00:00:00', because:
      -- ... in usage report transformations, some usage line items with usage_date cannot
      -- ... fall into the default valid time range of the associated agreement
      case
          when lag(valid_from) over (PARTITION BY agreement_id order by valid_from asc) is null
          then timestamp '1970-01-01 00:00:00'
          else valid_from
      end as valid_from,
      coalesce(
          lead(valid_from) over (partition by agreement_id order by valid_from asc),
          timestamp '2999-01-01 00:00:00'
      ) as valid_to,
      rank() over (partition by agreement_id order by valid_from asc) version,
      agreement_type,
      previous_agreement_id,
      agreement_intent
    from
      agreements_with_uni_temporal_data
  )
  select
    agreement_id,
    offer_id,
    proposer_account_id,
    acceptor_account_id,
    agreement_revision,
    start_date,
    end_date,
    acceptance_date,
    valid_from,
    case
        when version=1 and valid_from < timestamp '2023-03-03 06:16:08.743' then timestamp '1970-01-01'
        -- The following 60 minute adjustment is to handle special case where When Renewal happens for a contract
        when version=1 then date_add('minute',-60,valid_from)
        else valid_from
    end as valid_from_adjusted,
    valid_to,
    agreement_type,
    previous_agreement_id,
    agreement_intent
  from
    agreements_with_window_functions
),

-- An offer_id has several valid_from dates (each representing an offer revision)
-- but because of bi-temporality, an offer_id + valid_from tuple can appear multiple times with a different update_date.
-- We are only interested in the most recent tuple (ie, uni-temporal model)
offers_with_uni_temporal_data as (
  select
    from_iso8601_timestamp(valid_from) as valid_from,
    from_iso8601_timestamp(update_date) as update_date,
    from_iso8601_timestamp(delete_date) as delete_date,
    offer_id,
    offer_revision,
    name,
    expiration_date,
    opportunity_id,
    opportunity_name,
    opportunity_description,
    seller_account_id
  from
  (
    select
      valid_from,
      update_date,
      delete_date,
      offer_id,
      offer_revision,
      name,
      expiration_date,
      opportunity_id,
      opportunity_name,
      opportunity_description,
      seller_account_id,
      row_number() over (partition by offer_id, valid_from order by from_iso8601_timestamp(update_date) desc) as row_num
    from
      offerfeed_v1
  )
  where
    -- keep latest ...
    row_num = 1
    -- ... and remove the soft-deleted one.
    and (delete_date is null or delete_date = '')
),

-- Here, we build the validity time range (adding valid_to on top of valid_from) of each offer revision.
-- We will use it to get Offer name at invoice time.
-- NB: If you'd rather get "current" offer name, un-comment "offers_with_latest_revision"
offers_with_history as (
  select
    offer_id,
    offer_revision,
    name,
    opportunity_id,
    opportunity_name,
    opportunity_description,
    valid_from,
    -- When we try to look up an offer revision as at the acceptance date of a BYOL agreement, we run into a problem.
    -- For BYOL, the agreement might be accepted (using some external non-AWS system or manual process) days before
    -- that BYOL agreement is entered into AWS Marketplace by the buyer. Therefore, the buyer is permitted to manually
    -- enter a backdated acceptance date, which might predate the point in time when the first revision of the offer
    -- was created. To work around this, we need to adjust the valid_from on the first revision of the offer to be
    -- earlier than the earliest possible backdated BYOL agreement acceptance date.
    case
      when lag(valid_from) over (partition by offer_id order by valid_from asc) is null and valid_from < cast('2021-04-01' as timestamp)
      then date_add('Day', -3857, valid_from)
      -- 3857 is the longest delay between acceptance_date of an agreement and the first revision of the offer
      when lag(valid_from) over (partition by offer_id order by valid_from asc) is null and valid_from >= cast('2021-04-01' as timestamp)
      then date_add('Day', -1460, valid_from)
      --after 2021 for the two offers we need to adjust for 2 more years
      else valid_from
    end as valid_from_adjusted,
    coalesce(
      lead(valid_from) over (partition by offer_id order by valid_from asc),
      cast('2999-01-01 00:00:00' as timestamp))
    as valid_to
  from offers_with_uni_temporal_data
),
-- provided for reference only if you are interested into get "current" offer name
-- (ie. not used afterwards)
offers_with_latest_revision as (
  select
    offer_id,
    offer_revision,
    name,
    opportunity_name,
    opportunity_description,
    valid_from,
    null valid_to
  from
  (
    select
      offer_id,
      offer_revision,
      name,
      opportunity_name,
      opportunity_description,
      valid_from,
      null valid_to,
      row_number() over (partition by offer_id order by valid_from desc) as row_num_latest_revision
    from
      offers_with_uni_temporal_data
  )
  where
    row_num_latest_revision = 1
),

-- An offer_target_id has several valid_from dates (each representing an offer revision)
-- but because of bi-temporality, an offer_target_id + valid_from tuple can appear multiple times with a different update_date.
-- We are only interested in the most recent tuple (ie, uni-temporal model)
offer_targets_with_uni_temporal_data as (
  select
    from_iso8601_timestamp(valid_from) as valid_from,
    from_iso8601_timestamp(update_date) as update_date,
    from_iso8601_timestamp(delete_date) as delete_date,
    offer_target_id,
    offer_id,
    offer_revision,
    target_type,
    polarity,
    value
  from
  (
    select
      valid_from,
      update_date,
      delete_date,
      offer_target_id,
      offer_id,
      offer_revision,
      target_type,
      polarity,
      value,
      row_number() over (partition by offer_target_id, valid_from order by from_iso8601_timestamp(update_date) desc) as row_num
    from
      offertargetfeed_v1
  )
  where
    -- keep latest ...
    row_num = 1
    -- ... and remove the soft-deleted one.
    and (delete_date is null or delete_date = '')
),

offer_target_type as (
  select
    offer_id,
    offer_revision,
    substring(
      -- The first character indicates the priority (lower value means higher precedence):
      min(
        case
          when offer_target.target_type='BuyerAccounts' then '1Private'
          when offer_target.target_type='ParticipatingPrograms' then '2Program:'||cast(offer_target.value as varchar)
          when offer_target.target_type='CountryCodes' then '3GeoTargeted'
          -- well, there is no other case today, but rather be safe...
          else '4Other Targeting'
        end
      ),
      -- Remove the first character that was only used for th priority in the "min" aggregate function:
      2
    ) as offer_target
  from
    offer_targets_with_uni_temporal_data as offer_target
  group by
    offer_id,
    offer_revision
),

offers_with_history_with_target_type as (
  select
    offer.offer_id,
    offer.offer_revision,
    -- even though today it is not possible to combine several types of targeting in a single offer, let's ensure the query is still predictable if this gets possible in the future
    max(
      case
        when off_tgt.offer_target is null then 'Public'
        else off_tgt.offer_target
      end
    ) as offer_target,
    min(offer.name) as name,
    min(offer.opportunity_name) as opportunity_name,
    min(offer.opportunity_description) as opportunity_description,
    offer.valid_from,
    offer.valid_from_adjusted,
    offer.valid_to,
    offer.opportunity_id
  from
    offers_with_history as offer
  left join offer_target_type as off_tgt on
    offer.offer_id = off_tgt.offer_id
    and offer.offer_revision = off_tgt.offer_revision
  group by
    offer.offer_id,
    offer.offer_revision,
    offer.valid_from,
    offer.valid_from_adjusted,
    offer.valid_to,
    offer.opportunity_id
),

-- provided for reference only if you are interested into get "current" offer targets
-- (ie. not used afterwards)
offers_with_latest_revision_with_target_type as (
  select
    offer.offer_id,
    offer.offer_revision,
    -- even though today it is not possible to combine several types of targeting in a single offer, let's ensure the query is still predictable if this gets possible in the future
    max(
      distinct
      case
        when off_tgt.target_type is null then 'Public'
        when off_tgt.target_type='BuyerAccounts' then 'Private'
        when off_tgt.target_type='ParticipatingPrograms' then 'Program:'||cast(off_tgt.value as varchar)
        when off_tgt.target_type='CountryCodes' then 'GeoTargeted'
        -- well, there is no other case today, but rather be safe...
        else 'Other Targeting'
      end
    ) as offer_target,
    min(offer.name) as name,
    min(offer.opportunity_name) as opportunity_name,
    min(offer.opportunity_description) as opportunity_description,
    offer.valid_from,
    offer.valid_to
  from
    offers_with_latest_revision offer
    -- left joining because public offers don't have targets
    left join offer_targets_with_uni_temporal_data off_tgt on
      offer.offer_id=off_tgt.offer_id and offer.offer_revision=off_tgt.offer_revision
  group by
    offer.offer_id,
    offer.offer_revision,
    -- redundant with offer_revision, as each revision has a dedicated valid_from (but cleaner in the group by)
    offer.valid_from,
    offer.valid_to
),

-- A product_id has several valid_from dates (each representing a product revision),
-- but because of bi-temporality, each product_id + valid_from tuple can appear multiple times with a different update_date.
-- We are only interested in the most recent tuple (ie, uni-temporal model)
products_with_uni_temporal_data as (
  select
    from_iso8601_timestamp(valid_from) as valid_from,
    from_iso8601_timestamp(update_date) as update_date,
    from_iso8601_timestamp(delete_date) as delete_date,
    product_id,
    manufacturer_account_id,
    product_code,
    title
  from
  (
    select
      valid_from,
      update_date,
      delete_date,
      product_id,
      manufacturer_account_id,
      product_code,
      title,
      row_number() over (partition by product_id, valid_from order by from_iso8601_timestamp(update_date) desc) as row_num
    from
      productfeed_v1
  )
  where
    -- keep latest ...
    row_num = 1
    -- ... and remove the soft-deleted one.
    and (delete_date is null or delete_date = '')
),

products_with_history as (
  select
    product_id,
    title,
    valid_from,
    -- Offerv2 can have upto 50 years and Offerv3 is upto 5 years of past date
    case
      when lag(valid_from) over (partition by product_id order by valid_from asc) is null and valid_from < cast('2021-04-01' as timestamp)
        then date_add('Day', -3857, valid_from)
      -- 3827 is the longest delay between acceptance_date of an agreement and the product
      -- we are keeping 3857 as a consistency between the offers and products
      when lag(valid_from) over (partition by product_id order by valid_from asc) is null and valid_from >= cast('2021-04-01' as timestamp)
        then date_add('Day', -2190, valid_from)
      --after 2021 for the two offers we need to adjust for 2 more years
      else valid_from 
    end as valid_from_adjusted,
    coalesce(
      lead(valid_from) over (partition by product_id order by valid_from asc),
      cast('2999-01-01 00:00:00' as timestamp)
    ) as valid_to,
    product_code,
    manufacturer_account_id
  from
    products_with_uni_temporal_data
),

legacy_products as (
  select
    legacy_id,
    new_id
  from
    legacyidmappingfeed_v1
  where
    mapping_type='PRODUCT'
  group by
    legacy_id,
    new_id
),

-- A given billing_event_id represents an accounting event and thus has only one valid_from date,
-- but because of bi-temporality, a billing_event_id (+ its valid_from) can appear multiple times with a different update_date.
-- We are only interested in the most recent tuple (ie, uni-temporal model)
billing_events_with_uni_temporal_data as (
  select
    billing_event_id,
    valid_from,
    update_date,
    delete_date,
    invoice_date,
    transaction_type,
    transaction_reference_id,
    parent_billing_event_id,
    bank_trace_id,
    broker_id,
    product_id,
    disbursement_billing_event_id,
    action,
    from_account_id,
    to_account_id,
    end_user_account_id,
    billing_address_id,
    amount,
    currency,
    balance_impacting,
    --empty value in Athena shows as '', change all '' value to null
    case when agreement_id = '' then null else agreement_id end as agreement_id,
    invoice_id,
    payment_due_date,
    usage_period_start_date,
    usage_period_end_date,
    buyer_transaction_reference_id,
    row_num
  from
  (
    select
      billing_event_id,
      from_iso8601_timestamp(valid_from) as valid_from,
      from_iso8601_timestamp(update_date) as update_date,
      delete_date,
      from_iso8601_timestamp(invoice_date) as invoice_date,
      transaction_type,
      transaction_reference_id,
      parent_billing_event_id,
      -- casting in case data was imported as number
      cast(bank_trace_id as varchar) as bank_trace_id,
      broker_id,
      product_id,
      disbursement_billing_event_id,
      action,
      from_account_id,
      to_account_id,
      end_user_account_id,
      billing_address_id,
      -- casting in case data was imported as varchar
      cast(amount as decimal(38,6)) as amount,
      currency,
      balance_impacting,
      agreement_id,
      invoice_id,
      case when payment_due_date is null or payment_due_date = '' then null else from_iso8601_timestamp(payment_due_date) end as payment_due_date,
      from_iso8601_timestamp(usage_period_start_date) as usage_period_start_date,
      from_iso8601_timestamp(usage_period_end_date) as usage_period_end_date,
      buyer_transaction_reference_id,
      row_number() over (partition by billing_event_id, valid_from order by from_iso8601_timestamp(update_date) desc) as row_num
    from
      billingeventfeed_v1
  )
  where
    -- keep latest ...
    row_num = 1
    -- ... and remove the soft-deleted one.
    and (delete_date is null or delete_date = '')
),

-- Here we select the account_id of the current seller (We identify this by looking for the to_account_id related to revenue transactions).
-- We will use it later to distinguish own agreements from agreements generated by channel partners.
seller_account as (
  select
    from_account_id as seller_account_id
  from
    billing_events_with_uni_temporal_data bill
  where
    -- Assumption here is only seller will pay listing fee. As of 12/21/2021, there are cases that Channel partner have 0 listing fee for CPPO, so the amount could be 0.
    bill.transaction_type like 'AWS_REV_SHARE' and amount <= 0 and action = 'INVOICED'
  group by
    -- from_account_id is always the same for all those "listing fee" transactions == the seller of record himself.
    -- If this view returns more than 1 record, the overall query will fail (on purpose). Please contact AWS Marketplace if this happens.
    from_account_id
),

billing_event_with_business_flags as (
  select
    bl.billing_event_id,
    bl.end_user_account_id,
    bl.agreement_id,
    aggrement.proposer_account_id,
    aggrement.offer_id,
    aggrement.acceptor_account_id,
    case
      -- For AWS and BALANCE_ADJUSTMENT, the billing event feed will show the "AWS Marketplace" account as the
      -- receiver of the funds and the seller as the payer. We are not interested in this information here.
      -- Null values will be ignored by the `max` aggregation function.
      when bl.transaction_type like 'AWS%' then null
      -- For BALANCE_ADJUSTMENT, payer is seller themselves
      when bl.invoice_id is null then bl.to_account_id
      -- We get the payer of the invoice from *any* transaction type that is not AWS and not BALANCE_ADJUSTMENT (because they are the same for a given end user + agreement + product).
      else bl.from_account_id
    end as payer_account_id,
    bl.product_id,
    bl.action,
    bl.transaction_type,
    bl.parent_billing_event_id,
    bl.disbursement_billing_event_id,
    bl.amount,
    bl.currency,
    bl.balance_impacting,
    bl.invoice_date,
    bl.payment_due_date,
    bl.usage_period_start_date,
    bl.usage_period_end_date,
    bl.invoice_id,
    bl.billing_address_id,
    bl.transaction_reference_id,
    bl.buyer_transaction_reference_id,
    case when disbursement.bank_trace_id = 'EMEA_MP_TEST_TRACE_ID' then null else disbursement.bank_trace_id end as bank_trace_id,
    case when disbursement.bank_trace_id = 'EMEA_MP_TEST_TRACE_ID' then null else disbursement.invoice_date end as disbursement_date,
    disbursement.billing_event_id as disbursement_id,
    -- We will use disbursement_id_or_invoiced as part of the PK, so it cannot be null:
    coalesce(
      --empty value in Athena shows as '', change all '' value to null
      case when disbursement.billing_event_id = '' then null else disbursement.billing_event_id end,
      '<invoiced>') as disbursement_id_or_invoiced,
    bl.broker_id,
    case
      when bl.invoice_id is null /* transaction_type = 'BALANCE_ADJUSTMENT' */
        then (select seller_account_id from seller_account) ||':'|| cast(bl.invoice_date as varchar)
      else bl.buyer_transaction_reference_id
        ||'-'|| case when bl.agreement_id is null or bl.agreement_id = '' then ' ' else bl.agreement_id end
        ||'-'|| case when bl.end_user_account_id is null or bl.end_user_account_id = '' then ' ' else bl.end_user_account_id end
        ||'-'|| coalesce(cast(bl.usage_period_start_date as varchar),' ')
        ||'-'|| coalesce(cast(bl.usage_period_end_date as varchar),' ')
    end as internal_buyer_line_item_id,
    bl.buyer_transaction_reference_id <> bl.transaction_reference_id as is_seller_invoice,
    case when bl.transaction_type = 'SELLER_REV_SHARE' and (select seller_account_id from seller_account) <> bl.to_account_id then true else false end as is_cog,
    case when bl.transaction_type in('SELLER_REV_SHARE_CREDIT', 'SELLER_REV_SHARE_REFUND') and (select seller_account_id from seller_account) <> bl.to_account_id then true else false end as is_cog_refund,
    --TODO: replace below logic once we can create a logic the identify reseller/manufacturer without depending on agreement feed
    case when aggrement.proposer_account_id <> (select seller_account_id from seller_account) then true else false end as is_manufacturer_view_of_reseller
  from
    billing_events_with_uni_temporal_data as bl
    left join billing_events_with_uni_temporal_data as disbursement on
      disbursement.transaction_type like 'DISBURSEMENT%'
        and disbursement.action = 'DISBURSED'
        and disbursement.transaction_type IN ('DISBURSEMENT', 'DISBURSEMENT_FAILURE')
        and bl.disbursement_billing_event_id = disbursement.billing_event_id
    left join agreements_with_history as aggrement on
      bl.agreement_id = aggrement.agreement_id
        and bl.invoice_date >= aggrement.valid_from_adjusted
        and bl.invoice_date < aggrement.valid_to
    left join accounts_with_history awh on
      bl.to_account_id = awh.account_id
        and bl.invoice_date >= awh.valid_from
        and bl.invoice_date < awh.valid_to
  where
    bl.transaction_type not like 'DISBURSEMENT%' and
      (bl.agreement_id is null or bl.agreement_id = ''
      or aggrement.agreement_id is not null)
),

-- listagg function in athena does not support partitioning, grouping here and then joining to the main query
seller_invoice_list as (
  select
    internal_buyer_line_item_id,
    listagg(case when not is_seller_invoice then null else invoice_id end,',') within group (order by case when not is_seller_invoice then null else cast(invoice_date as varchar) end) as seller_invoice_id_or_null,
    listagg(case when not is_seller_invoice then null else cast(invoice_date as varchar) end,',') within group (order by case when not is_seller_invoice then null else cast(invoice_date as varchar) end) as seller_invoice_date_or_null
  from
    (
    -- listagg function in athena does not support ordering by another field when distinct is used,
    -- here we first select distinct invoices and then do the listagg order by invoice_date
    select distinct internal_buyer_line_item_id, is_seller_invoice, invoice_id, invoice_date
    from billing_event_with_business_flags) distinct_invoices
  group by internal_buyer_line_item_id
  order by internal_buyer_line_item_id
),

billing_event_with_categorized_transaction as (
-- Use the flags that were created in the previous transformation in more calculated columns:
-- NOTE: This transformation has no joins and no window functions
  select
    billing_event_id,
    end_user_account_id,
    agreement_id,
    proposer_account_id,
    offer_id,
    acceptor_account_id,
    case when is_cog or is_cog_refund then null else payer_account_id end as payer_account_id,
    product_id,
    action,
    transaction_type,
    parent_billing_event_id,
    disbursement_billing_event_id,
    amount,
    currency,
    balance_impacting,
    invoice_date,
    payment_due_date,
    usage_period_start_date,
    usage_period_end_date,
    invoice_id,
    billing_address_id,
    transaction_reference_id,
    buyer_transaction_reference_id,
    bank_trace_id,
    disbursement_date,
    disbursement_id,
    disbursement_id_or_invoiced,
    broker_id,
    bl.internal_buyer_line_item_id,
    is_seller_invoice,
    is_cog,
    is_cog_refund,
    is_manufacturer_view_of_reseller,

    -- Buyer/seller columns:
    case when is_seller_invoice then null else invoice_id end as buyer_invoice_id_or_null,
    seller_invoices.seller_invoice_id_or_null,
    case when is_seller_invoice then null else invoice_date end as buyer_invoice_date_or_null,
    seller_invoices.seller_invoice_date_or_null,

    -- Categorized amounts by transaction type:
    case when transaction_type =   'SELLER_REV_SHARE' and not is_cog then amount else 0 end as gross_revenue,
    case when transaction_type in ('SELLER_REV_SHARE_REFUND','SELLER_REV_SHARE_CREDIT') and not is_cog_refund then amount else 0 end as gross_refund,
    case when transaction_type =   'SELLER_REV_SHARE' and     is_cog then amount else 0 end as cogs,
    case when transaction_type in ('SELLER_REV_SHARE_REFUND','SELLER_REV_SHARE_CREDIT') and     is_cog_refund then amount else 0 end as cogs_refund,
    case when transaction_type =   'AWS_REV_SHARE' then amount else 0 end as aws_rev_share,
    case when transaction_type in ('AWS_REV_SHARE_REFUND','AWS_REV_SHARE_CREDIT') then amount else 0 end as aws_refund_share,
    case when transaction_type =   'AWS_TAX_SHARE' and not is_seller_invoice then amount else 0 end as aws_tax_share,             -- AWS tax share from _buyer_  invoice
    case when transaction_type =   'AWS_TAX_SHARE' and     is_seller_invoice then amount else 0 end as aws_tax_share_listing_fee, -- AWS tax share from _seller_ invoice
    case when transaction_type =   'AWS_TAX_SHARE_REFUND' and not is_seller_invoice then amount else 0 end as aws_tax_share_refund,
    case when transaction_type =   'AWS_TAX_SHARE_REFUND' and     is_seller_invoice then amount else 0 end as aws_tax_share_refund_listing_fee,
    case when transaction_type =   'SELLER_TAX_SHARE' then amount else 0 end as seller_tax_share,
    case when transaction_type =   'SELLER_TAX_SHARE_REFUND' then amount else 0 end as seller_tax_share_refund,
    case when transaction_type =   'BALANCE_ADJUSTMENT' then amount else 0 end as balance_adjustment,
    case when transaction_type =   'SELLER_REV_SHARE_CREDIT' then amount else 0 end as seller_rev_credit,
    case when transaction_type =   'AWS_REV_SHARE_CREDIT' then amount else 0 end as aws_ref_fee_credit
  from
    billing_event_with_business_flags as bl
    left join seller_invoice_list as seller_invoices
      on bl.internal_buyer_line_item_id = seller_invoices.internal_buyer_line_item_id
    ),

line_items_aggregated as (
-- This transformation has the only "group by" in all of these transformations.
-- NOTE: This transformation has no joins and no window functions
  select
    internal_buyer_line_item_id,
    disbursement_id,
    disbursement_id_or_invoiced,
    product_id,
    broker_id,
    currency,
    agreement_id,
    proposer_account_id,
    acceptor_account_id,
    max(payer_account_id) as payer_account_id,
    offer_id,
    end_user_account_id,
    usage_period_start_date,
    usage_period_end_date,
    max(payment_due_date) payment_due_date,
    buyer_transaction_reference_id,
    bank_trace_id,
    disbursement_date,
    max(billing_address_id) as billing_address_id,

    -- Buyer/seller columns:
    max(buyer_invoice_id_or_null) as buyer_invoice_id,
    max(seller_invoice_id_or_null) as seller_invoice_id,
    max(buyer_invoice_date_or_null) as buyer_invoice_date,
    max(seller_invoice_date_or_null) as seller_invoice_date,
  
    -- Categorized amounts by transaction type:
    -- When disbursement_id_or_invoiced = '<invoiced>',    these are invoiced amounts
    -- When disbursement_id_or_invoiced <> ''<invoiced>' these are disbursed amounts for _this_ specific disbursement_id
    sum(gross_revenue) as gross_revenue_this_disbursement_id_or_invoiced,
    sum(gross_refund) as gross_refund_this_disbursement_id_or_invoiced,
    sum(cogs) as cogs_this_disbursement_id_or_invoiced,
    sum(cogs_refund) as cogs_refund_this_disbursement_id_or_invoiced,
    sum(aws_rev_share) as aws_rev_share_this_disbursement_id_or_invoiced,
    sum(aws_refund_share) as aws_refund_share_this_disbursement_id_or_invoiced,
    sum(aws_tax_share) as aws_tax_share_this_disbursement_id_or_invoiced,
    sum(aws_tax_share_listing_fee) as aws_tax_share_listing_fee_this_disbursement_id_or_invoiced,
    sum(aws_tax_share_refund) as aws_tax_share_refund_this_disbursement_id_or_invoiced,
    sum(aws_tax_share_refund_listing_fee) as aws_tax_share_refund_listing_fee_this_disbursement_id_or_invoiced,
    sum(seller_tax_share) as seller_tax_share_this_disbursement_id_or_invoiced,
    sum(seller_tax_share_refund) as seller_tax_share_refund_this_disbursement_id_or_invoiced,
    sum(balance_adjustment) as balance_adjustment_this_disbursement_id_or_invoiced,
    sum(seller_rev_credit) as seller_rev_credit_this_disbursement_id_or_invoiced,
    sum(aws_ref_fee_credit) as aws_ref_fee_credit_this_disbursement_id_or_invoiced
  from
    billing_event_with_categorized_transaction as billing_categorized
  group by
    internal_buyer_line_item_id,
    disbursement_id,
    disbursement_id_or_invoiced,
    broker_id,
    -- The following columns are included the in group by but they are intentionally omitted from the PK.
    -- These columns should have the _same_ values for each record in the PK.
    product_id,
    currency,
    agreement_id,
    proposer_account_id,
    acceptor_account_id,
    offer_id,
    end_user_account_id,
    usage_period_start_date,
    usage_period_end_date,
    buyer_transaction_reference_id,
    bank_trace_id,
    disbursement_date
),

-- listagg function in athena does not support partitioning, grouping here and then joining to the main query
disbursement_list as (
  select
    internal_buyer_line_item_id,
    listagg(cast(disbursement_date as varchar),',') within group (order by cast(disbursement_date as varchar)) as disbursement_date_list,
    listagg(bank_trace_id,',') within group (order by cast(disbursement_date as varchar)) as disburse_bank_trace_id_list
    from (
    -- listagg function in athena does not support ordering by another field when distinct is used,
    -- here we first select distinct bank_trace_ids and then do the listagg order by disbursement_date
    select distinct internal_buyer_line_item_id, disbursement_date, bank_trace_id
  from billing_event_with_business_flags) distinct_disbursements
  group by internal_buyer_line_item_id
  order by internal_buyer_line_item_id
),

line_items_with_window_functions as (
--add flag next step compare gross_revenue and gross_revenue_disbursed or gross_refund and gross_refund_disbursed
  select
    line_item.internal_buyer_line_item_id,
    disbursement_id,
    disbursement_id_or_invoiced,
    product_id,
    broker_id,
    currency,
    agreement_id,
    proposer_account_id,
    acceptor_account_id,
    -- when there's aws_rev_Share adjustment/refund to a seller_rev_share invoice, it can happen that for the same aws_rev_share invoice_id, there are multiple disbursement events,
    -- using windows function to map payer_account_id of seller_rev_share to all corresponding aws_rev_Share
    max(payer_account_id) over (partition by line_item.internal_buyer_line_item_id) as payer_account_id,
    offer_id,
    end_user_account_id,
    usage_period_start_date,
    usage_period_end_date,
    payment_due_date,
    bank_trace_id,
    disbursement_date,
    billing_address_id,

    -- Buyer/seller columns:
    max(buyer_invoice_id) over (partition by line_item.internal_buyer_line_item_id) as buyer_invoice_id,
    seller_invoice_id,
    max(buyer_invoice_date) over (partition by line_item.internal_buyer_line_item_id) as buyer_invoice_date,
    seller_invoice_date,

    -- When disbursement_id_or_invoiced = ''<invoiced>', these are actually invoiced amounts
    -- When disbursement_id_or_invoiced <> ''<invoiced>' these are disbursed amounts for _this_ specific disbursement_id
    gross_revenue_this_disbursement_id_or_invoiced,
    gross_refund_this_disbursement_id_or_invoiced,
    cogs_this_disbursement_id_or_invoiced,
    cogs_refund_this_disbursement_id_or_invoiced,
    aws_rev_share_this_disbursement_id_or_invoiced,
    aws_refund_share_this_disbursement_id_or_invoiced,
    aws_tax_share_this_disbursement_id_or_invoiced,
    aws_tax_share_listing_fee_this_disbursement_id_or_invoiced,
    aws_tax_share_refund_this_disbursement_id_or_invoiced,
    aws_tax_share_refund_listing_fee_this_disbursement_id_or_invoiced,
    seller_tax_share_this_disbursement_id_or_invoiced,
    seller_tax_share_refund_this_disbursement_id_or_invoiced,
    balance_adjustment_this_disbursement_id_or_invoiced,
    seller_rev_credit_this_disbursement_id_or_invoiced,
    aws_ref_fee_credit_this_disbursement_id_or_invoiced,
    -- IMPORTANT: All window functions partitioned by internal_buyer_line_item_id:

    -- Invoiced amounts, categorized by transaction type:
    sum(case when disbursement_id_or_invoiced = ''<invoiced>' then gross_revenue_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end)over (partition by line_item.internal_buyer_line_item_id) as gross_revenue_invoiced,
    sum(case when disbursement_id_or_invoiced = ''<invoiced>' then gross_refund_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as gross_refund_invoiced,
    sum(case when disbursement_id_or_invoiced = ''<invoiced>' then cogs_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as cogs_invoiced,
    sum(case when disbursement_id_or_invoiced = ''<invoiced>' then cogs_refund_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as cogs_refund_invoiced,
    sum(case when disbursement_id_or_invoiced = ''<invoiced>' then aws_rev_share_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as aws_rev_share_invoiced,
    sum(case when disbursement_id_or_invoiced = ''<invoiced>' then aws_refund_share_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as aws_refund_share_invoiced,
    sum(case when disbursement_id_or_invoiced = ''<invoiced>' then aws_tax_share_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as aws_tax_share_invoiced,
    sum(case when disbursement_id_or_invoiced = ''<invoiced>' then aws_tax_share_listing_fee_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as aws_tax_share_listing_fee_invoiced,
    sum(case when disbursement_id_or_invoiced = ''<invoiced>' then aws_tax_share_refund_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as aws_tax_share_refund_invoiced,
    sum(case when disbursement_id_or_invoiced = ''<invoiced>' then aws_tax_share_refund_listing_fee_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as aws_tax_share_refund_listing_fee_invoiced,
    sum(case when disbursement_id_or_invoiced = ''<invoiced>' then seller_tax_share_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as seller_tax_share_invoiced,
    sum(case when disbursement_id_or_invoiced = ''<invoiced>' then seller_tax_share_refund_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as seller_tax_share_refund_invoiced,
    sum(case when disbursement_id_or_invoiced = ''<invoiced>' then balance_adjustment_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as balance_adjustment_invoiced,
    sum(case when disbursement_id_or_invoiced = ''<invoiced>' then seller_rev_credit_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as seller_rev_credit_invoiced,
    sum(case when disbursement_id_or_invoiced = ''<invoiced>' then aws_ref_fee_credit_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as aws_ref_fee_credit_invoiced,

    -- Total disbursed amounts (for all disbursement_id values), categorized by transaction type:
    sum(case when disbursement_id_or_invoiced '<> ''<invoiced>' then gross_revenue_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as gross_revenue_disbursed,
    sum(case when disbursement_id_or_invoiced '<> ''<invoiced>' then gross_refund_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as gross_refund_disbursed,
    sum(case when disbursement_id_or_invoiced '<> ''<invoiced>' then cogs_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as cogs_disbursed,
    sum(case when disbursement_id_or_invoiced '<> ''<invoiced>' then cogs_refund_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as cogs_refund_disbursed,
    sum(case when disbursement_id_or_invoiced '<> ''<invoiced>' then aws_rev_share_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as aws_rev_share_disbursed,
    sum(case when disbursement_id_or_invoiced '<> ''<invoiced>' then aws_refund_share_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as aws_refund_share_disbursed,
    sum(case when disbursement_id_or_invoiced '<> ''<invoiced>' then aws_tax_share_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as aws_tax_share_disbursed,
    sum(case when disbursement_id_or_invoiced '<> ''<invoiced>' then aws_tax_share_listing_fee_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as aws_tax_share_listing_fee_disbursed,
    sum(case when disbursement_id_or_invoiced '<> ''<invoiced>' then aws_tax_share_refund_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as aws_tax_share_refund_disbursed,
    sum(case when disbursement_id_or_invoiced '<> ''<invoiced>' then aws_tax_share_refund_listing_fee_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as aws_tax_share_refund_listing_fee_disbursed,
    sum(case when disbursement_id_or_invoiced '<> ''<invoiced>' then seller_tax_share_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as seller_tax_share_disbursed,
    sum(case when disbursement_id_or_invoiced '<> ''<invoiced>' then seller_tax_share_refund_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as seller_tax_share_refund_disbursed,
    sum(case when disbursement_id_or_invoiced '<> ''<invoiced>' then balance_adjustment_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as balance_adjustment_disbursed,
    sum(case when disbursement_id_or_invoiced '<> ''<invoiced>' then seller_rev_credit_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as seller_rev_credit_disbursed,
    sum(case when disbursement_id_or_invoiced '<> ''<invoiced>' then aws_ref_fee_credit_this_disbursement_id_or_invoiced else cast(0 as decimal(38,6)) end) over (partition by line_item.internal_buyer_line_item_id) as aws_ref_fee_credit_disbursed,

    -- aggregate multiple disbursement
    max(disbursement_date) over (partition by line_item.internal_buyer_line_item_id) as last_disbursement_date,
    first_value(case when disbursement_id_or_invoiced = ''<invoiced>' then null else disbursement_id_or_invoiced end) over(partition by line_item.internal_buyer_line_item_id order by coalesce(disbursement_date,cast('1900-01-01' as timestamp)) desc rows between unbounded preceding and unbounded following) as last_disbursement_id,
    first_value(bank_trace_id) over (partition by line_item.internal_buyer_line_item_id order by coalesce(disbursement_date,cast('1900-01-01' as timestamp)) desc rows between unbounded preceding and unbounded following) as last_disburse_bank_trace_id,
    disb_list.disbursement_date_list,
    disb_list.disburse_bank_trace_id_list
  from
    line_items_aggregated as line_item
    left join disbursement_list disb_list
      on line_item.internal_buyer_line_item_id = disb_list.internal_buyer_line_item_id
),

cppo_offer_id as (
  select
    -- Channel partner offers do not exist in offertargetfeed_v1 table (as per legal requirement), causing cppo offer be defined as 'Public' in previous step, we will convert them back to 'Private' in next step
    offer_id
  from
    offers_with_uni_temporal_data
  where
    -- seller_account_id is null means the ISV owns the offer
    seller_account_id is not null
    and seller_account_id '<>  (select seller_account_id from seller_account)
  group by
    offer_id
),

line_items_with_window_functions_enrich_offer_product_address as (
  select
    internal_buyer_line_item_id,
    disbursement_id,
    disbursement_id_or_invoiced,
    line.product_id,
    legacy_product.legacy_id as legacy_product_id,
    products.title as product_title,
    line.broker_id,
    line.currency,
    line.end_user_account_id,
    acc_enduser.encrypted_account_id as end_user_encrypted_account_id,
    acc_enduser.aws_account_id as end_user_aws_account_id,
    acc_payer.aws_account_id as payer_aws_account_id,
    acc_payer.encrypted_account_id payer_encrypted_account_id,
    line.agreement_id,
    agreement.agreement_revision,
    line.proposer_account_id,
    case when offer.offer_id like 'aiqoffer-%' then null else agreement.start_date end as Agreement_Start_Date,
    case when offer.offer_id like 'aiqoffer-%' then null else agreement.end_date end as Agreement_End_Date,
    case when offer.offer_id like 'aiqoffer-%' then null else agreement.acceptance_date end as Agreement_Acceptance_Date,
    case when offer.offer_id like 'aiqoffer-%' then null else agreement.valid_from end as agreement_updated_date,
    case when offer.offer_id like 'aiqoffer-%' then null else line.usage_period_start_date end as Usage_Period_Start_Date,
    case when offer.offer_id like 'aiqoffer-%' then null else line.usage_period_end_date end as Usage_Period_End_Date,

    line.acceptor_account_id,
    acc_subscriber.aws_account_id as subscriber_aws_account_id,
    acc_subscriber.encrypted_account_id as subscriber_encrypted_account_id,
    offer.offer_id,
    case
      when offer.offer_id in (
        select distinct offer_id
        from cppo_offer_id)
        then 'Private'
      else offer.offer_target
    end as offer_target,
    offer.name offer_name,
    offer.opportunity_name offer_opportunity_name,
    offer.opportunity_description offer_opportunity_description,
    offer.opportunity_id,
    payment_due_date,
    line.bank_trace_id,
    disbursement_date,
    billing_address_id,
    buyer_invoice_id,
    seller_invoice_id,
    buyer_invoice_date,
    seller_invoice_date,
    gross_revenue_this_disbursement_id_or_invoiced,
    gross_refund_this_disbursement_id_or_invoiced,
    cogs_this_disbursement_id_or_invoiced,
    cogs_refund_this_disbursement_id_or_invoiced,
    aws_rev_share_this_disbursement_id_or_invoiced,
    aws_refund_share_this_disbursement_id_or_invoiced,
    aws_tax_share_this_disbursement_id_or_invoiced,
    aws_tax_share_listing_fee_this_disbursement_id_or_invoiced,
    aws_tax_share_refund_this_disbursement_id_or_invoiced,
    aws_tax_share_refund_listing_fee_this_disbursement_id_or_invoiced,
    seller_tax_share_this_disbursement_id_or_invoiced,
    seller_tax_share_refund_this_disbursement_id_or_invoiced,
    balance_adjustment_this_disbursement_id_or_invoiced,
    seller_rev_credit_this_disbursement_id_or_invoiced,
    aws_ref_fee_credit_this_disbursement_id_or_invoiced,
    gross_revenue_invoiced,
    gross_refund_invoiced,
    cogs_invoiced,
    cogs_refund_invoiced,
    aws_rev_share_invoiced,
    aws_refund_share_invoiced,
    aws_tax_share_invoiced,
    aws_tax_share_listing_fee_invoiced,
    aws_tax_share_refund_invoiced,
    aws_tax_share_refund_listing_fee_invoiced,
    seller_tax_share_invoiced,
    seller_tax_share_refund_invoiced,
    balance_adjustment_invoiced,
    seller_rev_credit_invoiced,
    aws_ref_fee_credit_invoiced,
    gross_revenue_disbursed,
    gross_refund_disbursed,
    cogs_disbursed,
    cogs_refund_disbursed,
    aws_rev_share_disbursed,
    aws_refund_share_disbursed,
    aws_tax_share_disbursed,
    aws_tax_share_listing_fee_disbursed,
    aws_tax_share_refund_disbursed,
    aws_tax_share_refund_listing_fee_disbursed,
    seller_tax_share_disbursed,
    seller_tax_share_refund_disbursed,
    balance_adjustment_disbursed,
    seller_rev_credit_disbursed,
    aws_ref_fee_credit_disbursed,
    last_disbursement_date,
    last_disbursement_id,
    last_disburse_bank_trace_id,
    disbursement_date_list,
    disburse_bank_trace_id_list,
    products.product_code,
    acc_products.aws_account_id as manufacturer_aws_account_id,
    products.manufacturer_account_id,
    --add subscriber and payer addressID, payer address preference order: tax address>billing address>mailing address,  subscriber address preference order: tax address> mailing address
    coalesce (
      --empty value in Athena shows as '', change all '' value to null in order to follow the preference order logic above
      case when acc_subscriber.tax_address_id ='' then null else acc_subscriber.tax_address_id end,
      case when acc_subscriber.mailing_address_id = '' then null else acc_subscriber.mailing_address_id end) as subscriber_address_id,
    coalesce (
      case when acc_payer.tax_address_id = '' then null else acc_payer.tax_address_id end,
      case when line.billing_address_id = '' then null else line.billing_address_id end,
      case when acc_payer.mailing_address_id = '' then null else acc_payer.mailing_address_id end) as payer_address_id,
    coalesce (
      case when acc_enduser.tax_address_id = '' then null else acc_enduser.tax_address_id end,
      case when line.billing_address_id = '' then null else line.billing_address_id end,
      case when acc_enduser.mailing_address_id = '' then null else acc_enduser.mailing_address_id end) as end_user_address_id
  from
    line_items_with_window_functions as line
  left join agreements_with_history as agreement on
      line.agreement_id = agreement.agreement_id and line.buyer_invoice_date >= agreement.valid_from_adjusted and line.buyer_invoice_date < agreement.valid_to
  left join offers_with_history_with_target_type as offer on
        line.offer_id = offer.offer_id and line.buyer_invoice_date >= offer.valid_from and line.buyer_invoice_date < offer.valid_to
  left join products_with_history as products on
        line.product_id = products.product_id and line.buyer_invoice_date >= products.valid_from_adjusted and line.buyer_invoice_date < products.valid_to
  left join legacy_products as legacy_product on
        line.product_id = legacy_product.new_id
  left join accounts_with_history_with_company_name as acc_payer on
        line.payer_account_id = acc_payer.account_id and line.buyer_invoice_date >= acc_payer.valid_from and line.buyer_invoice_date < acc_payer.valid_to
  left join accounts_with_history_with_company_name as acc_enduser on
        line.end_user_account_id = acc_enduser.account_id and line.buyer_invoice_date >= acc_enduser.valid_from and line.buyer_invoice_date < acc_enduser.valid_to
  left join accounts_with_history_with_company_name as acc_subscriber on
        line.acceptor_account_id = acc_subscriber.account_id and line.buyer_invoice_date >= acc_subscriber.valid_from and line.buyer_invoice_date < acc_subscriber.valid_to
  left join accounts_with_history_with_company_name as acc_products on
        products.manufacturer_account_id = acc_products.account_id and line.buyer_invoice_date >= acc_products.valid_from and line.buyer_invoice_date < acc_products.valid_to

),

line_items_with_window_functions_enrich_offer_product_address_name as (
  select
    line.internal_buyer_line_item_id,
    disbursement_id,
    disbursement_id_or_invoiced,
    product_id,
    legacy_product_id,
    product_title,
    broker_id,
    currency,
    end_user_address_id,
    end_user_account_id,
    end_user_encrypted_account_id,
    end_user_aws_account_id,
    add_enduser.company_name end_user_company_name,
    add_enduser.email_domain end_user_email_domain,
    add_enduser.city end_user_city,
    add_enduser.state_or_region end_user_state,
    add_enduser.country_code end_user_country,
    add_enduser.postal_code end_user_postal_code,
    payer_aws_account_id,
    payer_encrypted_account_id,
    payer_address_id,
    add_payer.company_name payer_company_name,
    add_payer.email_domain payer_email_domain,
    add_payer.city payer_city,
    add_payer.state_or_region payer_state,
    add_payer.country_code payer_country,
    add_payer.postal_code payer_postal_code,
    agreement_id,
    agreement_revision,
    agreement_start_date,
    agreement_end_date,
    agreement_acceptance_date,
    agreement_updated_date,
    case when proposer_account_id = (select seller_account_id from seller_account) then null else acc_proposer.aws_account_id end as reseller_aws_account_id,
    case when proposer_account_id = (select seller_account_id from seller_account) then null else acc_proposer.mailing_company_name end as reseller_company_name,
    usage_period_start_date,
    usage_period_end_date,
    proposer_account_id,
    acc_proposer.aws_account_id as proposer_aws_account_id,
    acceptor_account_id,
    subscriber_aws_account_id,
    subscriber_encrypted_account_id,
    subscriber_address_id,
    add_subscriber.company_name subscriber_company_name,
    add_subscriber.email_domain subscriber_email_domain,
    add_subscriber.city subscriber_city,
    add_subscriber.state_or_region subscriber_state,
    add_subscriber.country_code subscriber_country,
    add_subscriber.postal_code subscriber_postal_code,
    offer_id,
    offer_target,
    offer_name,
    offer_opportunity_name,
    offer_opportunity_description,
    opportunity_id,
    payment_due_date,
    bank_trace_id,
    disbursement_date,
    billing_address_id,
    max(buyer_invoice_id)as buyer_invoice_id,
    max(seller_invoice_id)as seller_invoice_id,
    max(buyer_invoice_date)as buyer_invoice_date,
    max(seller_invoice_date)as seller_invoice_date,
    gross_revenue_this_disbursement_id_or_invoiced,
    gross_refund_this_disbursement_id_or_invoiced,
    cogs_this_disbursement_id_or_invoiced,
    cogs_refund_this_disbursement_id_or_invoiced,
    aws_rev_share_this_disbursement_id_or_invoiced,
    aws_refund_share_this_disbursement_id_or_invoiced,
    aws_tax_share_this_disbursement_id_or_invoiced,
    aws_tax_share_listing_fee_this_disbursement_id_or_invoiced,
    aws_tax_share_refund_this_disbursement_id_or_invoiced,
    aws_tax_share_refund_listing_fee_this_disbursement_id_or_invoiced,
    seller_tax_share_this_disbursement_id_or_invoiced,
    seller_tax_share_refund_this_disbursement_id_or_invoiced,
    balance_adjustment_this_disbursement_id_or_invoiced,
    seller_rev_credit_this_disbursement_id_or_invoiced,
    aws_ref_fee_credit_this_disbursement_id_or_invoiced,
    (gross_revenue_this_disbursement_id_or_invoiced + gross_refund_this_disbursement_id_or_invoiced + aws_rev_share_this_disbursement_id_or_invoiced + aws_refund_share_this_disbursement_id_or_invoiced + seller_tax_share_this_disbursement_id_or_invoiced + seller_tax_share_refund_this_disbursement_id_or_invoiced
      + cogs_this_disbursement_id_or_invoiced + cogs_refund_this_disbursement_id_or_invoiced + aws_tax_share_listing_fee_this_disbursement_id_or_invoiced + aws_tax_share_refund_listing_fee_this_disbursement_id_or_invoiced) as seller_net_revenue_this_disbursement_id_or_invoiced,
    gross_revenue_invoiced,
    gross_refund_invoiced,
    cogs_invoiced,
    cogs_refund_invoiced,
    aws_rev_share_invoiced,
    aws_refund_share_invoiced,
    aws_tax_share_invoiced,
    aws_tax_share_listing_fee_invoiced,
    aws_tax_share_refund_invoiced,
    aws_tax_share_refund_listing_fee_invoiced,
    seller_tax_share_invoiced,
    seller_tax_share_refund_invoiced,
    balance_adjustment_invoiced,
    seller_rev_credit_invoiced,
    aws_ref_fee_credit_invoiced,
    gross_revenue_disbursed,
    gross_refund_disbursed,
    cogs_disbursed,
    cogs_refund_disbursed,
    aws_rev_share_disbursed,
    aws_refund_share_disbursed,
    aws_tax_share_disbursed,
    aws_tax_share_listing_fee_disbursed,
    aws_tax_share_refund_disbursed,
    aws_tax_share_refund_listing_fee_disbursed,
    seller_tax_share_disbursed,
    seller_tax_share_refund_disbursed,
    balance_adjustment_disbursed,
    seller_rev_credit_disbursed,
    aws_ref_fee_credit_disbursed,
    (gross_revenue_invoiced + gross_revenue_disbursed) as uncollected_gross_revenue,
    -- net revenue = gross revenue - listing fee - tax - cogs
    (gross_revenue_invoiced + gross_refund_invoiced + aws_rev_share_invoiced + aws_refund_share_invoiced + seller_tax_share_invoiced + seller_tax_share_refund_invoiced + cogs_invoiced + cogs_refund_invoiced + aws_tax_share_listing_fee_invoiced + aws_tax_share_refund_listing_fee_invoiced) as seller_net_revenue,
    (gross_revenue_invoiced + gross_refund_invoiced + aws_rev_share_invoiced + aws_refund_share_invoiced + seller_tax_share_invoiced + seller_tax_share_refund_invoiced + cogs_invoiced + cogs_refund_invoiced + aws_tax_share_listing_fee_invoiced + aws_tax_share_refund_listing_fee_invoiced
      + gross_revenue_disbursed + gross_refund_disbursed + aws_rev_share_disbursed + aws_refund_share_disbursed + seller_tax_share_disbursed + seller_tax_share_refund_disbursed + cogs_disbursed + cogs_refund_disbursed + aws_tax_share_listing_fee_disbursed + aws_tax_share_refund_listing_fee_disbursed) as uncollected_seller_net_revenue,
    last_disbursement_date,
    last_disbursement_id,
    last_disburse_bank_trace_id,
    disbursement_date_list,
    disburse_bank_trace_id_list,
    product_code,
    manufacturer_aws_account_id,
    manufacturer_account_id,
    acc_manu.mailing_company_name as manufacturer_company_name,
    cast(null as varchar) as AR_Period,
    case
      when (
        (gross_revenue_invoiced '<>0 and gross_revenue_invoiced = -1 * gross_revenue_disbursed)
        or (gross_refund_invoiced '<> 0 and gross_refund_invoiced = -1 * gross_refund_disbursed)
        or (balance_adjustment_invoiced '<> 0 and balance_adjustment_invoiced = -1 * balance_adjustment_disbursed)
        or (seller_tax_share_refund_invoiced '<> 0 and seller_tax_share_refund_invoiced = -1 * seller_tax_share_refund_disbursed)
        or (gross_revenue_invoiced = 0 and gross_refund_invoiced = 0 and balance_adjustment_invoiced = 0 and seller_tax_share_refund_invoiced = 0 and last_disbursement_id is not null)) then 'Yes'
      when gross_revenue_disbursed = 0 and gross_refund_disbursed = 0 and balance_adjustment_disbursed = 0 and seller_tax_share_disbursed = 0 and seller_tax_share_refund_disbursed = 0 then 'No'
      else 'Partial'
    end as Disbursement_Flag
  from line_items_with_window_functions_enrich_offer_product_address as line
  left join accounts_with_history_with_company_name as acc_manu on
    line.manufacturer_account_id = acc_manu.account_id and line.buyer_invoice_date >= acc_manu.valid_from_adjusted and line.buyer_invoice_date <= acc_manu.valid_to
  left join accounts_with_history_with_company_name as acc_proposer on
    line.proposer_account_id = acc_proposer.account_id and line.buyer_invoice_date >= acc_proposer.valid_from and line.buyer_invoice_date < acc_proposer.valid_to
  left join address_with_latest_revision as add_payer on
    line.payer_address_id = add_payer.address_id
  left join address_with_latest_revision as add_subscriber on
    line.subscriber_address_id = add_subscriber.address_id
  left join address_with_latest_revision as add_enduser on
    line.end_user_address_id = add_enduser.address_id
  group by
    line.internal_buyer_line_item_id,
    disbursement_id,
    disbursement_id_or_invoiced,
    product_id,
    legacy_product_id,
    product_title,
    broker_id,
    currency,
    end_user_address_id,
    end_user_account_id,
    end_user_encrypted_account_id,
    end_user_aws_account_id,
    add_enduser.company_name,
    add_enduser.email_domain,
    add_enduser.city,
    add_enduser.state_or_region,
    add_enduser.country_code,
    add_enduser.postal_code,
    payer_aws_account_id,
    payer_encrypted_account_id,
    payer_address_id,
    add_payer.company_name,
    add_payer.email_domain,
    add_payer.city,
    add_payer.state_or_region,
    add_payer.country_code,
    add_payer.postal_code,
    agreement_id,
    agreement_revision,
    case when proposer_account_id = (select seller_account_id from seller_account) then null else acc_proposer.aws_account_id end,
    case when proposer_account_id = (select seller_account_id from seller_account) then null else acc_proposer.mailing_company_name end,
    agreement_start_date,
    agreement_end_date,
    agreement_acceptance_date,
    agreement_updated_date,
    usage_period_start_date,
    usage_period_end_date,
    acceptor_account_id,
    subscriber_aws_account_id,
    subscriber_encrypted_account_id,
    subscriber_address_id,
    add_subscriber.company_name,
    add_subscriber.email_domain,
    add_subscriber.city,
    add_subscriber.state_or_region,
    add_subscriber.country_code,
    add_subscriber.postal_code,
    offer_id,
    offer_target,
    offer_name,
    offer_opportunity_name,
    offer_opportunity_description,
    opportunity_id,
    payment_due_date,
    bank_trace_id,
    disbursement_date,
    billing_address_id,
    gross_revenue_this_disbursement_id_or_invoiced,
    gross_refund_this_disbursement_id_or_invoiced,
    cogs_this_disbursement_id_or_invoiced,
    cogs_refund_this_disbursement_id_or_invoiced,
    aws_rev_share_this_disbursement_id_or_invoiced,
    aws_refund_share_this_disbursement_id_or_invoiced,
    aws_tax_share_this_disbursement_id_or_invoiced,
    aws_tax_share_listing_fee_this_disbursement_id_or_invoiced,
    aws_tax_share_refund_this_disbursement_id_or_invoiced,
    aws_tax_share_refund_listing_fee_this_disbursement_id_or_invoiced,
    seller_tax_share_this_disbursement_id_or_invoiced,
    seller_tax_share_refund_this_disbursement_id_or_invoiced,
    balance_adjustment_this_disbursement_id_or_invoiced,
    seller_rev_credit_this_disbursement_id_or_invoiced,
    aws_ref_fee_credit_this_disbursement_id_or_invoiced,
    gross_revenue_invoiced,
    gross_refund_invoiced,
    cogs_invoiced,
    cogs_refund_invoiced,
    aws_rev_share_invoiced,
    aws_refund_share_invoiced,
    aws_tax_share_invoiced,
    aws_tax_share_listing_fee_invoiced,
    aws_tax_share_refund_invoiced,
    aws_tax_share_refund_listing_fee_invoiced,
    seller_tax_share_invoiced,
    seller_tax_share_refund_invoiced,
    balance_adjustment_invoiced,
    seller_rev_credit_invoiced,
    aws_ref_fee_credit_invoiced,
    gross_revenue_disbursed,
    gross_refund_disbursed,
    cogs_disbursed,
    cogs_refund_disbursed,
    aws_rev_share_disbursed,
    aws_refund_share_disbursed,
    aws_tax_share_disbursed,
    aws_tax_share_listing_fee_disbursed,
    aws_tax_share_refund_disbursed,
    aws_tax_share_refund_listing_fee_disbursed,
    seller_tax_share_disbursed,
    seller_tax_share_refund_disbursed,
    balance_adjustment_disbursed,
    seller_rev_credit_disbursed,
    aws_ref_fee_credit_disbursed,
    last_disbursement_date,
    last_disbursement_id,
    last_disburse_bank_trace_id,
    disbursement_date_list,
    disburse_bank_trace_id_list,
    product_code,
    manufacturer_aws_account_id,
    manufacturer_account_id,
    acc_manu.mailing_company_name,
    proposer_account_id,
    acc_proposer.aws_account_id
),
invoiced_not_disbursed as(
  select
    --we will filter on rownum =1 in next step,
    -- means internal_buyer_line_item_id, there's only '<invoiced> record, no disbursement_id linked
    *,
    max(case when disbursement_id_or_invoiced = ''<invoiced>' then 1 else 2 end)
      over (partition by internal_buyer_line_item_id) rownum
  from line_items_with_window_functions_enrich_offer_product_address_name as line_items

),
collections_and_disbursements as (
  select
    ------------------
    -- Invoice Info --
    ------------------
    buyer_invoice_date as Invoice_Date,
    Payment_Due_Date as Payment_Due_Date,
    concat(
      'Net ',
      case
        when abs(date_diff('Day', Payment_due_date, buyer_invoice_date))>180 then '180+'
        else cast(abs(date_diff('Day', Payment_due_date, buyer_invoice_date)) as varchar)
        end,
      ' days'
    ) as payment_terms,
    buyer_invoice_id as Invoice_ID,
    coalesce(
      --empty value in Athena shows as '', change all '' value to null
      case when seller_invoice_id = '' then null else seller_invoice_id end,
      'Not applicable') as Listing_Fee_Invoice_ID,

    ---------------------------
    --End user Information --
    ---------------------------
    coalesce(
      --empty value in Athena shows as '', change all '' value to null
      case when End_User_Company_Name = '' then null else End_User_Company_Name end,
      'Not available') as End_User_Company_Name,
    End_User_AWS_Account_ID,
    End_User_Encrypted_Account_ID,
    End_User_Email_Domain,
    End_User_City,
    End_User_State as End_User_State_or_Region,
    End_User_Country,
    End_User_Postal_Code,
    End_User_Address_ID,

    ---------------------------
    --Subscriber Information --
    ---------------------------
    case
      when Agreement_Id is null or Agreement_ID = '' then 'Not available'
      when Subscriber_Company_Name is null or Subscriber_Company_Name = '' then 'Not provided'
      else Subscriber_Company_Name
    end as Subscriber_Company_Name,
    case
      when Agreement_Id is null or Agreement_ID = '' then 'Not available'
      else Subscriber_AWS_Account_ID
    end as Subscriber_AWS_Account_ID,
    case
      when Agreement_Id is null or Agreement_ID = '' then 'Not available'
      else Subscriber_Encrypted_Account_ID
    end as Subscriber_Encrypted_Account_ID,
    case
      when Agreement_Id is null or Agreement_ID = '' then 'Not available'
      when Subscriber_Email_Domain is null or Subscriber_Email_Domain = '' then 'Not provided'
      else Subscriber_Email_Domain
    end as Subscriber_Email_Domain,
    case
      when Agreement_id is null or Agreement_ID = '' then 'Not available'
      when Subscriber_City is null or Subscriber_City = '' then 'Not provided'
      else Subscriber_City
    end as Subscriber_City,
    case
      when Agreement_Id is null or Agreement_ID = '' then 'Not available'
      when Subscriber_State is null or Subscriber_State = '' then 'Not provided'
      else Subscriber_State
    end as Subscriber_State_or_Region,
    case
      when Agreement_Id is null or Agreement_ID = '' then 'Not available'
      when Subscriber_Country is null or Subscriber_Country = '' then 'Not provided'
      else Subscriber_Country
    end as Subscriber_Country,
    case
      when Agreement_Id is null or Agreement_ID = '' then 'Not available'
      when Subscriber_Postal_Code is null or Subscriber_Postal_Code = '' then 'Not provided'
      else Subscriber_Postal_Code
    end as Subscriber_Postal_Code,
    case
      when Agreement_ID is null or Agreement_ID = '' then 'Not available'
      when Subscriber_Address_ID is null or Subscriber_Address_ID = '' then 'Not provided'
      else Subscriber_Address_ID
    end as Subscriber_Address_ID,

    ----------------------
    -- Procurement Info --
    ----------------------
    -- product title at time of invoice. It is possible that the title changes over time and therefore there may be multiple product titles mapped to a single product id.
    coalesce(
      --empty value in Athena shows as '', change all '' value to null
      case when Product_Title = '' then null else Product_Title end,
      'Not provided') as Product_Title,
    -- offer name at time of invoice. It is possible that the name changes over time therefore there may be multiple offer names mapped to a single offer id.
    case
      when Agreement_Id is null or Agreement_ID = '' then 'Not available'
      when (Offer_Name is null or Offer_Name = '') and Offer_Target = 'Public' then 'Not applicable'
      else Offer_Name
    end as Offer_Name,
    case
      when Agreement_Id is null or Agreement_ID = ''
      then 'Not available'
      else Offer_ID
    end as Offer_ID,
    -- offer visibility at time of invoice.,
    case
      when Agreement_Id is null or Agreement_ID = '' then 'Not available'
      else Offer_Target
    end as Offer_Visibility,
    coalesce(
      --empty value in Athena shows as '', change all '' value to null
      case when Agreement_ID = '' then null else Agreement_ID end,
      'Not available') as Agreement_ID,
    Agreement_Start_Date,
    Agreement_Acceptance_Date,
    Agreement_End_Date,

    Usage_Period_Start_Date,
    Usage_Period_End_Date,

    -----------------------
    -- Disbursement Info --
    -----------------------
    case
      when Disbursement_Flag = 'Yes' then 'Disbursed'
      when Disbursement_Flag = 'No' then 'Not Disbursed'
      else 'Other'
    end as Disbursement_Status,
    last_disbursement_date as disbursement_date,
    case
      when Disbursement_Flag = 'Yes' then date_diff('DAY', date_trunc('DAY',payment_due_date), date_trunc('DAY',last_disbursement_date))
      else null
    end as Disbursement_Time,
    case
      when Disbursement_Flag = 'No' then 'Not applicable'
      when bank_trace_id is null or bank_trace_id = '' then 'Not available'
      else bank_trace_id
    end as disburse_bank_trace_id,

    --------------
    -- Revenues --
    --------------
    -- We are rounding the sums using 2 decimal precision
    -- Note that the rounding method might differ between SQL implementations.
    -- The monthly revenue report is using RoundingMode.HALF_UP. This might create tiny discrepancies between this SQL output
    -- and the legacy report
    round(-1 * gross_revenue_this_disbursement_id_or_invoiced,2) as Gross_Revenue,
    round(-1 * gross_refund_this_disbursement_id_or_invoiced,2) as Gross_Refund,
    round(-1 * aws_rev_share_this_disbursement_id_or_invoiced,2) as Listing_Fee,
    round(-1 * aws_refund_share_this_disbursement_id_or_invoiced,2) as Listing_Fee_Refund,
    truncate(
      case
        when gross_revenue_this_disbursement_id_or_invoiced != 0 then abs(aws_rev_share_this_disbursement_id_or_invoiced/gross_revenue_this_disbursement_id_or_invoiced)
        when gross_refund_this_disbursement_id_or_invoiced != 0 then abs(aws_refund_share_this_disbursement_id_or_invoiced/gross_refund_this_disbursement_id_or_invoiced)
        else 0
      end
      ,4) as Listing_Fee_Percentage,
    round(-1 * seller_tax_share_this_disbursement_id_or_invoiced,2) as Seller_Tax_Share,
    round(-1 * seller_tax_share_refund_this_disbursement_id_or_invoiced,2) as Seller_Tax_Share_Refund,
    round(-1 * aws_tax_share_listing_fee_this_disbursement_id_or_invoiced,2) as AWS_Tax_Share_Listing_Fee,
    round(-1 * aws_tax_share_refund_listing_fee_this_disbursement_id_or_invoiced,2) as AWS_Tax_Share_Refund_Listing_Fee,
    round(-1 * cogs_this_disbursement_id_or_invoiced,2) as Wholesale_cost,
    round(-1 * cogs_refund_this_disbursement_id_or_invoiced,2) as Wholesale_cost_Refund,
    round(-1 * seller_net_revenue_this_disbursement_id_or_invoiced,2) as Seller_Net_Revenue,
    currency as Currency,

    substring(internal_buyer_line_item_id,1,strpos(internal_buyer_line_item_id,'-')-1) as Transaction_Reference_ID,
    broker_id as AWS_seller_of_record,

    -----------------
    -- Resale info --
    -----------------
    case
      when Opportunity_Id is null or Opportunity_Id = '' then
        case
          when Offer_Target = 'Public' then 'Not applicable'
          when (Offer_Target is null or Offer_Target = '') and (Agreement_Id is not null and Agreement_Id != '') then 'Not applicable'
          else null
        end
      else Opportunity_Id
    end as Resale_authorization_ID,
    case
      when Offer_Opportunity_Name is null or Offer_Opportunity_Name = '' then
        case
          when Offer_Target = 'Public' then 'Not applicable'
          when (Offer_Target is null or Offer_Target = '') and (Agreement_Id is not null and Agreement_Id != '') then 'Not applicable'
          else null
        end
      else Offer_Opportunity_Name
    end as Resale_authorization_name,
    case
      when Offer_Opportunity_Description is null or Offer_Opportunity_Description = '' then
        case
          when Offer_Target = 'Public' then 'Not applicable'
          when (Offer_Target is null or Offer_Target = '') and (Agreement_Id is not null and Agreement_Id != '') then 'Not applicable'
          else null
        end
      else Offer_Opportunity_Description
    end as Resale_authorization_description,
    case
      when (Reseller_AWS_Account_ID is not null and Reseller_AWS_Account_ID != '')
        and (Reseller_Company_Name is null or Reseller_Company_Name = '') then 'Not available'
      when (Reseller_AWS_Account_ID is null or Reseller_AWS_Account_ID = '')
        and (opportunity_id is null or opportunity_id = '') then 'Not applicable'
      when (select seller_account_id from seller_account) '<> manufacturer_aws_account_id
        and (Reseller_AWS_Account_ID is null or Reseller_AWS_Account_ID = '') then 'Not applicable'
      else Reseller_Company_Name
    end as Reseller_Company_Name,
    case
      when (Reseller_AWS_Account_ID is null or Reseller_AWS_Account_ID = '')
        and (Opportunity_Id is null or Opportunity_Id = '') then 'Not applicable'
      when (select seller_account_id from seller_account) '<> manufacturer_aws_account_id
        and (Reseller_AWS_Account_ID is null or Reseller_AWS_Account_ID = '') then 'Not applicable'
      else Reseller_AWS_Account_ID
    end as Reseller_AWS_Account_ID,

    -----------------------
    -- Payer Information --
    -----------------------
    coalesce(
      --empty value in Athena shows as '', change all '' value to null
      case when Payer_Company_Name = '' then null else Payer_Company_Name end,
      'Not available') as Payer_Company_Name,
    Payer_AWS_Account_ID, -- "Customer AWS Account Number" in legacy report
    Payer_Encrypted_Account_ID,
    Payer_Email_Domain,
    Payer_City,
    Payer_State as Payer_State_or_Region,
    Payer_Country,
    Payer_Postal_Code,
    Payer_Address_ID,

    ---------------------
    -- ISV Information --
    ---------------------
    manufacturer_aws_account_id as ISV_Account_ID,
    coalesce(
      --empty value in Athena shows as '', change all '' value to null
      case when Manufacturer_Company_Name = '' then null else Manufacturer_Company_Name end,
      'Not available') as ISV_Company_Name,

    ---------------------
    -- Products info --
    ---------------------
    Legacy_Product_ID,
    coalesce(
      --empty value in Athena shows as '', change all '' value to null
      case when Product_ID = '' then null else Product_ID end,
      'Not provided') as Product_ID,
    -- this is to get the legacy product id https://sim.amazon.com/issues/MP-INSIGHTS-2561
    Product_Code,

    case when Disbursement_Flag = 'Yes' then round(-1 * seller_net_revenue_this_disbursement_id_or_invoiced,2) else 0 end as Disbursed_Net_Revenue,
    case when Disbursement_Flag = 'No' then round(-1 * seller_net_revenue_this_disbursement_id_or_invoiced,2) else 0 end as Undisbursed_Net_Revenue,
    case
      when case when Disbursement_Flag = 'Yes' then date_diff('DAY', date_trunc('DAY',payment_due_date), date_trunc('DAY',last_disbursement_date)) else null end <= 0 then 'Not due'
      when case when Disbursement_Flag = 'Yes' then date_diff('DAY', date_trunc('DAY',payment_due_date), date_trunc('DAY',last_disbursement_date)) else null end <=30 then '1-30 days late'
      when case when Disbursement_Flag = 'Yes' then date_diff('DAY', date_trunc('DAY',payment_due_date), date_trunc('DAY',last_disbursement_date)) else null end <=60 then '31-60 days late'
      when case when Disbursement_Flag = 'Yes' then date_diff('DAY', date_trunc('DAY',payment_due_date), date_trunc('DAY',last_disbursement_date)) else null end <=90 then '61-90 days late'
      when case when Disbursement_Flag = 'Yes' then date_diff('DAY', date_trunc('DAY',payment_due_date), date_trunc('DAY',last_disbursement_date)) else null end <=120 then '91-120 days late'
      when case when Disbursement_Flag = 'Yes' then date_diff('DAY', date_trunc('DAY',payment_due_date), date_trunc('DAY',last_disbursement_date)) else null end >=121 then '121+ days late'
      else null
    end as Disbursement_Period
  from
    line_items_with_window_functions_enrich_offer_product_address_name as line
  where disbursement_id_or_invoiced != ''<invoiced>'

  union

  select
    ------------------
    -- Invoice Info --
    ------------------
    buyer_invoice_date as Invoice_Date,
    Payment_Due_Date as Payment_Due_Date,
    concat(
      'Net ',
      case
        when abs(date_diff('Day', Payment_due_date, buyer_invoice_date)) >180 then '180+'
        else cast(abs(date_diff('Day', Payment_due_date, buyer_invoice_date)) as varchar)
      end,
      ' days'
    ) as payment_terms,
    buyer_invoice_id as Invoice_ID,
    coalesce(
      --empty value in Athena shows as '', change all '' value to null
      case when seller_invoice_id = '' then null else seller_invoice_id end,
      'Not applicable') as Listing_Fee_Invoice_ID,

    ---------------------------
    --End user Information --
    ---------------------------
    coalesce(
      --empty value in Athena shows as '', change all '' value to null
      case when End_User_Company_Name = '' then null else End_User_Company_Name end,
      'Not available') as End_User_Company_Name,
    End_User_AWS_Account_ID,
    End_User_Encrypted_Account_ID,
    End_User_Email_Domain,
    End_User_City,
    End_User_State as End_User_State_or_Region,
    End_User_Country,
    End_User_Postal_Code,
    End_User_Address_ID,

    ---------------------------
    --Subscriber Information --
    ---------------------------
    case
      when Agreement_Id is null or Agreement_ID = '' then 'Not available'
      when Subscriber_Company_Name is null or Subscriber_Company_Name = '' then 'Not provided'
      else Subscriber_Company_Name
    end as Subscriber_Company_Name,
    case
      when Agreement_Id is null or Agreement_ID = '' then 'Not available'
      else Subscriber_AWS_Account_ID
    end as Subscriber_AWS_Account_ID,
    case
      when Agreement_Id is null or Agreement_ID = '' then 'Not available'
      else Subscriber_Encrypted_Account_ID
    end as Subscriber_Encrypted_Account_ID,
    case
      when Agreement_Id is null or Agreement_ID = '' then 'Not available'
      when Subscriber_Email_Domain is null or Subscriber_Email_Domain = '' then 'Not provided'
      else Subscriber_Email_Domain
    end as Subscriber_Email_Domain,
    case
      when Agreement_id is null or Agreement_ID = '' then 'Not available'
      when Subscriber_City is null or Subscriber_City = '' then 'Not provided'
      else Subscriber_City
    end as Subscriber_City,
    case
      when Agreement_Id is null or Agreement_ID = '' then 'Not available'
      when Subscriber_State is null or Subscriber_State = '' then 'Not provided'
      else Subscriber_State
    end as Subscriber_State_or_Region,
    case
      when Agreement_Id is null or Agreement_ID = '' then 'Not available'
      when Subscriber_Country is null or Subscriber_Country = '' then 'Not provided'
      else Subscriber_Country
    end as Subscriber_Country,
    case
      when Agreement_Id is null or Agreement_ID = '' then 'Not available'
      when Subscriber_Postal_Code is null or Subscriber_Postal_Code = '' then 'Not provided'
      else Subscriber_Postal_Code
    end as Subscriber_Postal_Code,
    case
      when Agreement_ID is null or Agreement_ID = '' then 'Not available'
      when Subscriber_Address_ID is null or Subscriber_Address_ID = '' then 'Not provided'
      else Subscriber_Address_ID
    end as Subscriber_Address_ID,

    ----------------------
    -- Procurement Info --
    ----------------------
    coalesce(
      --empty value in Athena shows as '', change all '' value to null
      case when Product_Title = '' then null else Product_Title end,
      'Not provided') as Product_Title,
    -- offer name at time of invoice. It is possible that the name changes over time therefore there may be multiple offer names mapped to a single offer id.
    case
      when Agreement_Id is null or Agreement_ID = '' then 'Not available'
      when (Offer_Name is null or Offer_Name = '') and Offer_Target = 'Public' then 'Not applicable'
      else Offer_Name
    end as Offer_Name,
    case
      when Agreement_Id is null or Agreement_ID = ''
      then 'Not available'
      else Offer_ID
    end as Offer_ID,
    -- offer visibility at time of invoice.,
    case
      when Agreement_Id is null or Agreement_ID = '' then 'Not available'
      else Offer_Target
    end as Offer_Visibility,
    coalesce(
      --empty value in Athena shows as '', change all '' value to null
      case when Agreement_ID = '' then null else Agreement_ID end,
      'Not available') as Agreement_ID,
    --case when Agreement_Id is null or Agreement_Id = '' then cast(null as timestamp) else Agreement_Start_Date end as Agreement_Start_Date,
    --case when Agreement_Id is null or Agreement_Id = '' then cast(null as timestamp) else Agreement_End_Date end as Agreement_End_Date,
    --case when Agreement_Id is null or Agreement_Id = '' then cast(null as timestamp) else Agreement_Acceptance_Date end as Agreement_Acceptance_Date,
    Agreement_Start_Date,
    Agreement_Acceptance_Date,
    Agreement_End_Date,

    Usage_Period_Start_Date,
    Usage_Period_End_Date,

    -----------------------
    -- Disbursement Info --
    -----------------------
    case
      when Disbursement_Flag = 'Yes' then 'Disbursed'
      when Disbursement_Flag = 'No' then 'Not Disbursed'
      else 'Other'
    end as Disbursement_Status,
    last_disbursement_date as disbursement_date,
    case
      when Disbursement_Flag = 'Yes' then date_diff('DAY', date_trunc('DAY',payment_due_date), date_trunc('DAY',last_disbursement_date))
      else null
    end as Disbursement_Time,
    case
      when Disbursement_Flag = 'No' then 'Not applicable'
      when bank_trace_id is null or bank_trace_id = '' then 'Not available'
      else bank_trace_id
    end as disburse_bank_trace_id,

    --------------
    -- Revenues --
    --------------
    -- We are rounding the sums using 2 decimal precision
    -- Note that the rounding method might differ between SQL implementations.
    -- The monthly revenue report is using RoundingMode.HALF_UP. This might create tiny discrepancies between this SQL output
    -- and the legacy report
    round(gross_revenue_this_disbursement_id_or_invoiced,2) as Gross_Revenue,
    round(gross_refund_this_disbursement_id_or_invoiced,2) as Gross_Refund,
    round(aws_rev_share_this_disbursement_id_or_invoiced,2) as Listing_Fee,
    round(aws_refund_share_this_disbursement_id_or_invoiced,2) as Listing_Fee_Refund,
    truncate(
      case
        when gross_revenue_this_disbursement_id_or_invoiced != 0 then abs(aws_rev_share_this_disbursement_id_or_invoiced/gross_revenue_this_disbursement_id_or_invoiced)
        when gross_refund_this_disbursement_id_or_invoiced != 0 then abs(aws_refund_share_this_disbursement_id_or_invoiced/gross_refund_this_disbursement_id_or_invoiced)
        else 0
      end
      ,4) as Listing_Fee_Percentage,
    round(seller_tax_share_this_disbursement_id_or_invoiced,2) as Seller_Tax_Share,
    round(seller_tax_share_refund_this_disbursement_id_or_invoiced,2) as Seller_Tax_Share_Refund,
    round(aws_tax_share_listing_fee_this_disbursement_id_or_invoiced,2) as AWS_Tax_Share_Listing_Fee,
    round(aws_tax_share_refund_listing_fee_this_disbursement_id_or_invoiced,2) as AWS_Tax_Share_Refund_Listing_Fee,
    round(cogs_this_disbursement_id_or_invoiced,2) as Wholesale_cost,
    round(cogs_refund_this_disbursement_id_or_invoiced,2) as Wholesale_cost_Refund,
    round(seller_net_revenue_this_disbursement_id_or_invoiced,2) as Seller_Net_Revenue,
    currency as Currency,

    substring(internal_buyer_line_item_id,1,strpos(internal_buyer_line_item_id,'-')-1) as Transaction_Reference_ID,
    broker_id as AWS_seller_of_record,

    -----------------
    -- Resale info --
    -----------------
    case
      when Opportunity_Id is null or Opportunity_Id = '' then
        case
          when Offer_Target = 'Public' then 'Not applicable'
          when (Offer_Target is null or Offer_Target = '') and (Agreement_Id is not null and Agreement_Id != '') then 'Not applicable'
          else null
        end
      else Opportunity_Id
    end as Resale_authorization_ID,
    case
      when Offer_Opportunity_Name is null or Offer_Opportunity_Name = '' then
        case
          when Offer_Target = 'Public' then 'Not applicable'
          when (Offer_Target is null or Offer_Target = '') and (Agreement_Id is not null and Agreement_Id != '') then 'Not applicable'
          else null
        end
      else Offer_Opportunity_Name
    end as Resale_authorization_name,
    case
      when Offer_Opportunity_Description is null or Offer_Opportunity_Description = '' then
        case
          when Offer_Target = 'Public' then 'Not applicable'
          when (Offer_Target is null or Offer_Target = '') and (Agreement_Id is not null and Agreement_Id != '') then 'Not applicable'
          else null
        end
      else Offer_Opportunity_Description
    end as Resale_authorization_description,
    case
      when (Reseller_AWS_Account_ID is not null and Reseller_AWS_Account_ID != '')
        and (Reseller_Company_Name is null or Reseller_Company_Name = '') then 'Not available'
      when (Reseller_AWS_Account_ID is null or Reseller_AWS_Account_ID = '')
        and (opportunity_id is null or opportunity_id = '') then 'Not applicable'
      when (select seller_account_id from seller_account) '<> manufacturer_aws_account_id
        and (Reseller_AWS_Account_ID is null or Reseller_AWS_Account_ID = '') then 'Not applicable'
      else Reseller_Company_Name
    end as Reseller_Company_Name,
    case
      when (Reseller_AWS_Account_ID is null or Reseller_AWS_Account_ID = '')
        and (Opportunity_Id is null or Opportunity_Id = '') then 'Not applicable'
      when (select seller_account_id from seller_account) '<> manufacturer_aws_account_id
        and (Reseller_AWS_Account_ID is null or Reseller_AWS_Account_ID = '') then 'Not applicable'
      else Reseller_AWS_Account_ID
    end as Reseller_AWS_Account_ID,

    -----------------------
    -- Payer Information --
    -----------------------
    coalesce(
      --empty value in Athena shows as '', change all '' value to null
      case when Payer_Company_Name = '' then null else Payer_Company_Name end,
      'Not available') as Payer_Company_Name,
    Payer_AWS_Account_ID, -- "Customer AWS Account Number" in legacy report
    Payer_Encrypted_Account_ID,
    Payer_Email_Domain,
    Payer_City,
    Payer_State as Payer_State_or_Region,
    Payer_Country,
    Payer_Postal_Code,
    Payer_Address_ID,

    ---------------------
    -- ISV Information --
    ---------------------
    manufacturer_aws_account_id as ISV_Account_ID,
    coalesce(
      --empty value in Athena shows as '', change all '' value to null
      case when Manufacturer_Company_Name = '' then null else Manufacturer_Company_Name end,
      'Not available') as ISV_Company_Name,

    ---------------------
    -- Products info --
    ---------------------
    -- product title at time of invoice. It is possible that the title changes over time and therefore there may be multiple product titles mapped to a single product id.
    Legacy_Product_ID,
    coalesce(
      --empty value in Athena shows as '', change all '' value to null
      case when Product_ID = '' then null else Product_ID end,
      'Not provided') as Product_ID,
    -- this is to get the legacy product id https://sim.amazon.com/issues/MP-INSIGHTS-2561
    Product_Code,

    case when Disbursement_Flag = 'Yes' then round(seller_net_revenue_this_disbursement_id_or_invoiced,2) else 0 end as Disbursed_Net_Revenue,
    case when Disbursement_Flag = 'No' then round(seller_net_revenue_this_disbursement_id_or_invoiced,2) else 0 end as Undisbursed_Net_Revenue,
    case
      when case when Disbursement_Flag = 'Yes' then date_diff('DAY', date_trunc('DAY',payment_due_date), date_trunc('DAY',last_disbursement_date)) else null end <= 0 then 'Not due'
      when case when Disbursement_Flag = 'Yes' then date_diff('DAY', date_trunc('DAY',payment_due_date), date_trunc('DAY',last_disbursement_date)) else null end <=30 then '1-30 days late'
      when case when Disbursement_Flag = 'Yes' then date_diff('DAY', date_trunc('DAY',payment_due_date), date_trunc('DAY',last_disbursement_date)) else null end <=60 then '31-60 days late'
      when case when Disbursement_Flag = 'Yes' then date_diff('DAY', date_trunc('DAY',payment_due_date), date_trunc('DAY',last_disbursement_date)) else null end <=90 then '61-90 days late'
      when case when Disbursement_Flag = 'Yes' then date_diff('DAY', date_trunc('DAY',payment_due_date), date_trunc('DAY',last_disbursement_date)) else null end <=120 then '91-120 days late'
      when case when Disbursement_Flag = 'Yes' then date_diff('DAY', date_trunc('DAY',payment_due_date), date_trunc('DAY',last_disbursement_date)) else null end >=121 then '121+ days late'
      else null
    end as Disbursement_Period
  from
    invoiced_not_disbursed
  where rownum = 1

)

select *
from collections_and_disbursements
where payment_due_date >= date_add('DAY', -90, current_date)
--where payment_due_date between cast('2023-01-01' as timestamp) and cast('2024-12-31' as timestamp)

--where disbursement_date >= date_add('DAY', -90, current_date)
--where disbursement_date between cast('2023-01-01' as timestamp) and cast('2024-12-31' as timestamp)
```

## 已纳税发票
<a name="data-feed-example-tax"></a>

要查找已纳税发票，您可以运行如以下示例所示的一组查询。这些查询相互依存，可创建**税务**报告。您可以使用下方所示的示例，也可以根据您的数据和用例进行自定义。

查询中的注释说明了查询的作用以及如何修改查询。

```
-- Taxation report

-- General note: When executing this query we are assuming that the data ingested in the database is using
-- two time axes (the valid_from column and the update_date column).
-- See documentation for more details: https://docs.aws.amazon.com/marketplace/latest/userguide/data-feed.html#data-feed-details

-- An account_id has several valid_from dates (each representing a separate revision of the data)
-- but because of bi-temporality, an account_id + valid_from tuple can appear multiple times with a different update_date.
-- We are only interested in the most recent tuple (ie, uni-temporal model)
with accounts_with_uni_temporal_data as (
  select
    account_id,
    aws_account_id,
    encrypted_account_id,
    mailing_address_id,
    tax_address_id,
    tax_legal_name,
    from_iso8601_timestamp(valid_from) as valid_from,
    tax_registration_number
  from
    (
      select
        account_id,
        aws_account_id,
        encrypted_account_id,
        mailing_address_id,
        tax_address_id,
        tax_legal_name,
        valid_from,
        delete_date,
        tax_registration_number,
        row_number() over (partition by account_id, valid_from order by from_iso8601_timestamp(update_date) desc) as row_num
      from
        accountfeed_v1
    )
  where
    -- keep latest ...
    row_num = 1
    -- ... and remove the soft-deleted one.
    and (delete_date is null or delete_date = '')
),

accounts_with_history as (
  with accounts_with_history_with_extended_valid_from as (
    select
      account_id,
      -- sometimes, this columns gets imported as a "bigint" and loses heading 0s -> casting to a char and re-adding heading 0s (if need be)
      substring('000000000000'||cast(aws_account_id as varchar),-12) as aws_account_id,
      encrypted_account_id,
      mailing_address_id,
      tax_address_id,
      tax_legal_name tax_legal_name,
      -- The start time of account valid_from is extended to '1970-01-01 00:00:00', because:
      -- ... in tax report transformations, some tax line items with invoice_date cannot
      -- ... fall into the default valid time range of the associated account
      CASE
        WHEN LAG(valid_from) OVER (PARTITION BY account_id ORDER BY valid_from ASC) IS NULL
            THEN CAST('1970-01-01 00:00:00' as timestamp)
        ELSE valid_from
      END AS valid_from
    from
      (select * from accounts_with_uni_temporal_data ) as account
  )
  select
    account_id,
    aws_account_id,
    encrypted_account_id,
    mailing_address_id,
    tax_address_id,
    tax_legal_name,
    valid_from,
    coalesce(
      lead(valid_from) over (partition by account_id order by valid_from asc),
      cast('2999-01-01 00:00:00' as timestamp)
    ) as valid_to
  from
    accounts_with_history_with_extended_valid_from
),

-- A product_id has several valid_from dates (each representing a product revision),
-- but because of bi-temporality, each product_id + valid_from tuple can appear multiple times with a different update_date.
-- We are only interested in the most recent tuple (ie, uni-temporal model)
products_with_uni_temporal_data as (
  select
    from_iso8601_timestamp(valid_from) as valid_from,
    from_iso8601_timestamp(update_date) as update_date,
    from_iso8601_timestamp(delete_date) as delete_date,
    product_id,
    manufacturer_account_id,
    product_code,
    title
  from
    (
      select
        valid_from,
        update_date,
        delete_date,
        product_id,
        manufacturer_account_id,
        product_code,
        title,
        row_number() over (partition by product_id, valid_from order by from_iso8601_timestamp(update_date) desc) as row_num
      from
       productfeed_v1
      )
  where
    -- keep latest ...
    row_num = 1
    -- ... and remove the soft-deleted one.
    and (delete_date is null or delete_date = '')
),

products_with_history as (
  select
    product_id,
    title,
    valid_from,    
    case
      when lag(valid_from) over (partition by product_id order by valid_from asc) is null and valid_from < cast('2021-04-01' as timestamp)
        then date_add('Day', -3857, valid_from)
      -- 3827 is the longest delay between acceptance_date of an agreement and the product
      -- we are keeping 3857 as a consistency between the offers and products
      when lag(valid_from) over (partition by product_id order by valid_from asc) is null and valid_from >= cast('2021-04-01' as timestamp)
        then date_add('Day', -2190, valid_from)
      --after 2021 for the two offers we need to adjust for 2 more years
      else valid_from end as valid_from_adjusted,
    coalesce(
      lead(valid_from) over (partition by product_id order by valid_from asc),
      cast('2999-01-01 00:00:00' as timestamp)
    ) as valid_to,
    product_code,
    manufacturer_account_id
  from
    products_with_uni_temporal_data
),

-- A tax_item_id has several valid_from dates (each representing a product revision),
-- but because of bi-temporality, each tax_item_id + valid_from tuple can appear multiple times with a different update_date.
-- We are only interested in the most recent tuple (ie, uni-temporal model)
tax_items_with_uni_temporal_data as (
  select
    from_iso8601_timestamp(valid_from) as valid_from,
    from_iso8601_timestamp(update_date) as update_date,
    delete_date,
    cast(tax_item_id as varchar) as tax_item_id,
    cast(invoice_id as varchar) as invoice_id,
    cast(line_item_id as varchar) as line_item_id,
    cast(customer_bill_id as varchar) as customer_bill_id,
    tax_liable_party,
    transaction_type_code,
    product_id,
    product_tax_code,
    from_iso8601_timestamp(invoice_date) as invoice_date,
    taxed_customer_account_id,
    taxed_customer_country,
    taxed_customer_state_or_region,
    taxed_customer_city,
    taxed_customer_postal_code,
    tax_location_code_taxed_jurisdiction,
    tax_type_code,
    jurisdiction_level,
    taxed_jurisdiction,
    display_price_taxability_type,
    tax_jurisdiction_rate,
    tax_amount,
    tax_currency,
    tax_calculation_reason_code,
    date_used_for_tax_calculation,
    customer_exemption_certificate_id,
    customer_exemption_certificate_id_domain,
    customer_exemption_certificate_level,
    customer_exemption_code,
    customer_exemption_domain,
    transaction_reference_id
  from
    (
      select
        valid_from,
        update_date,
        delete_date,
        tax_item_id,
        invoice_id,
        line_item_id,
        customer_bill_id,
        tax_liable_party,
        transaction_type_code,
        product_id,
        product_tax_code,
        invoice_date,
        taxed_customer_account_id,
        taxed_customer_country,
        taxed_customer_state_or_region,
        taxed_customer_city,
        taxed_customer_postal_code,
        tax_location_code_taxed_jurisdiction,
        tax_type_code,
        jurisdiction_level,
        taxed_jurisdiction,
        display_price_taxability_type,
        tax_jurisdiction_rate,
        tax_amount,
        tax_currency,
        tax_calculation_reason_code,
        date_used_for_tax_calculation,
        customer_exemption_certificate_id,
        customer_exemption_certificate_id_domain,
        customer_exemption_certificate_level,
        customer_exemption_code,
        customer_exemption_domain,
        transaction_reference_id,
        row_number() over (partition by tax_item_id, valid_from order by from_iso8601_timestamp(update_date) desc) as row_num
      from
        taxitemfeed_v1
    )
  where
    -- keep latest ...
    row_num = 1
    -- ... and remove the soft-deleted one.
    and (delete_date is null or delete_date = '')
),

taxation as (
  select
    tax_items.invoice_id,
    tax_items.line_item_id,
    tax_items.customer_bill_id,
    tax_items.tax_liable_party,
    tax_items.transaction_type_code,
    tax_items.product_id,
    product_tax_item.title as product_title,
    tax_items.product_tax_code,
    tax_items.invoice_date,
    accounts_with_history.aws_account_id as taxed_customer_account_id,
    tax_items.taxed_customer_country,
    tax_items.taxed_customer_state_or_region,
    tax_items.taxed_customer_city,
    tax_items.taxed_customer_postal_code,
    tax_items.tax_type_code as tax_type,
    tax_items.jurisdiction_level,
    tax_items.taxed_jurisdiction,
    tax_items.display_price_taxability_type,
    tax_items.tax_jurisdiction_rate,
    tax_items.tax_amount,
    tax_items.tax_currency,
    tax_items.tax_calculation_reason_code,
    tax_items.date_used_for_tax_calculation,
    coalesce(
      --empty value in Athena shows as '', change all '' value to null
      case when tax_items.customer_exemption_certificate_id = '' then null else tax_items.customer_exemption_certificate_id end,
      'Not exempt') customer_exemption_certificate_id,
    coalesce(--empty value in Athena shows as '', change all '' value to null
      case when tax_items.customer_exemption_certificate_id_domain = '' then null else tax_items.customer_exemption_certificate_id_domain end,
      'Not exempt') customer_exemption_certificate_id_domain,
    coalesce(--empty value in Athena shows as '', change all '' value to null
      case when tax_items.customer_exemption_certificate_level = '' then null else tax_items.customer_exemption_certificate_level end,
      'Not exempt') customer_exemption_certificate_level,
    coalesce(--empty value in Athena shows as '', change all '' value to null
      case when tax_items.customer_exemption_code = '' then null else tax_items.customer_exemption_code end,
      'Not exempt') customer_exemption_code,
    tax_items.transaction_reference_id
  from
    tax_items_with_uni_temporal_data as tax_items
    left join products_with_history as product_tax_item on
      tax_items.product_id = product_tax_item.product_id and tax_items.invoice_date >= product_tax_item.valid_from_adjusted and tax_items.invoice_date < product_tax_item.valid_to
    left join accounts_with_history as accounts_with_history on
      tax_items.taxed_customer_account_id = accounts_with_history.account_id and tax_items.invoice_date >= accounts_with_history.valid_from and tax_items.invoice_date < accounts_with_history.valid_to

)

select *
from taxation
where invoice_date >= date_add('DAY', -90, current_date)
--where invoice_date between cast('2023-01-01' as timestamp) and cast('2024-12-31' as timestamp)
```

## 按产品列出的款项
<a name="data-feed-example-disbursement-by-product"></a>

要查找按产品列出的已支付金额，您可以运行如以下示例所示的一组查询。

这些示例查询相互依存，可创建包含付款的最终产品详细信息列表。它还显示了如何在特定时间点获取产品信息。您可以使用下方所示的示例，也可以根据您的数据和用例进行自定义。

查询中的注释说明了查询的作用以及如何修改查询。

**注意**  
运行此查询时，我们假设提取的数据使用两个时间轴，分别为 `valid_from` 和 `update` 列。有关这些轴的更多信息，请参阅[AWS Marketplace 数据源的存储和结构](data-feed-details.md)。

```
    -- Get all the products and keep the latest product_id, valid_from tuple
    with products_with_uni_temporal_data as (
      select
       *
      from
      (
        select
         *,
         ROW_NUMBER() OVER (PARTITION BY product_id, valid_from 
             ORDER BY from_iso8601_timestamp(update_date) desc) 
             as row_num
        from
         productfeed_v1
      )
      where
        -- A product_id can appear multiple times with the same 
        -- valid_from date but with a different update_date column,
        -- making it effectively bi-temporal. By only taking the most
        -- recent tuple, we are converting to a uni-temporal model.
        row_num = 1
    ),

    -- Gets the latest revision of a product
    -- A product can have multiple revisions where some of the 
    -- columns, like the title, can change.
    -- For the purpose of the disbursement report, we want 
    -- to get the latest revision of a product
    products_with_latest_version as (
     select
      *
     from
     (
      select
       *,
       ROW_NUMBER() OVER (PARTITION BY product_id 
           ORDER BY from_iso8601_timestamp(valid_from) desc) 
           as row_num_latest_version
      from
       products_with_uni_temporal_data
     )
     where
      row_num_latest_version = 1
   ),

    -- Get all the accounts and keep the latest account_id, valid_from tuple
    accounts_with_uni_temporal_data as (
      select
       *
      from
      (
        select
         *,
         ROW_NUMBER() OVER (PARTITION BY account_id, valid_from ORDER BY from_iso8601_timestamp(update_date) desc) as row_num
        from
         accountfeed_v1
      )
      where
        -- An account_id can appear multiple times with the same 
        -- valid_from date but with a different update_date column,
        -- making it effectively bi-temporal. By only taking the most
        -- recent tuple, we are converting to a uni-temporal model.
        row_num = 1
    ),

    -- Gets the latest revision of an account
    -- An account can have multiple revisions where some of the 
    -- columns, like the mailing_address_id, can change.
    -- For the purpose of the disbursement report, we want 
    -- to get the latest revision of a product
    accounts_with_latest_version as (
     select
      *
     from
     (
      select
       *,
       ROW_NUMBER() OVER (PARTITION BY account_id 
           ORDER BY from_iso8601_timestamp(valid_from) desc) 
           as row_num_latest_version
      from
       accounts_with_uni_temporal_data
     )
     where
      row_num_latest_version = 1
   ),

    -- Get all the billing events and keep the 
    -- latest billing_event_id, valid_from tuple:
    billing_events_with_uni_temporal_data as (
      select
       *
      from (
        select
          billing_event_id,
          from_iso8601_timestamp(valid_from) as valid_from,
          from_iso8601_timestamp(update_date) as update_date,
          from_iso8601_timestamp(invoice_date) as invoice_date,
          transaction_type,
          transaction_reference_id,
          product_id,
          disbursement_billing_event_id,
          action,
          from_account_id,
          to_account_id,
          end_user_account_id,
          CAST(amount as decimal(20, 10)) invoice_amount,
          bank_trace_id,
          ROW_NUMBER() OVER (PARTITION BY billing_event_id, valid_from 
              ORDER BY from_iso8601_timestamp(update_date) desc) 
              as row_num
        from
          billingeventfeed_v1
        )
      where row_num = 1
    ),

    -- Get all the disbursements
    -- The billing events data is immutable.
    -- It is not required to use time windows based on the 
    -- valid_from column to get the most recent billing event
    disbursement_events as (
      select
        billing_events_raw.billing_event_id as disbursement_id,
        billing_events_raw.invoice_date as disbursement_date,
        billing_events_raw.bank_trace_id
      from
        billing_events_with_uni_temporal_data billing_events_raw
      where
        -- Only interested in disbursements, so filter out
        -- non-disbursements by selecting transaction type 
        -- to be DISBURSEMENT:
        billing_events_raw.transaction_type = 'DISBURSEMENT'
        -- Select a time period, you can adjust the dates 
        -- below if need be. For billing events use the 
        -- invoice date as the point in time of the 
        -- disbursement being initiated:
        and billing_events_raw.invoice_date >= 
            from_iso8601_timestamp('2020-10-01T00:00:00Z')
        and billing_events_raw.invoice_date < 
            from_iso8601_timestamp('2020-11-01T00:00:00Z')
    ),

    -- Get the invoices along with the line items that 
    -- are part of the above filtered disbursements
    disbursed_line_items as (
      select
        line_items.transaction_reference_id,
        line_items.product_id,
        line_items.transaction_type,
        (case
           -- Get the payer of the invoice from any 
           -- transaction type that is not AWS and 
           -- not BALANCE_ADJUSTMENT.
           -- For AWS and BALANCE_ADJUSTMENT, the billing 
           -- event feed will show the "AWS Marketplace" 
           -- account as the receiver of the funds and the 
           -- seller as the payer. Filter those out.
           when line_items.transaction_type 
               not like '%AWS%' and transaction_type 
               not like 'BALANCE_ADJUSTMENT' 
               then line_items.from_account_id
        end) as payer_account_id,
        line_items.end_user_account_id,
        invoice_amount,
        disbursements.disbursement_date,
        disbursements.disbursement_id,
        disbursements.bank_trace_id
      from
        billing_events_with_uni_temporal_data line_items
        -- Each disbursed line item is linked to the parent 
        -- disbursement via the disbursement_billing_event_id
        join disbursement_events disbursements 
          on disbursements.disbursement_id 
          = line_items.disbursement_billing_event_id
      where
        -- we are interested only in the invoice line 
        -- items that are DISBURSED
        line_items.action = 'DISBURSED'
    ),

  -- An invoice can contain multiple line items
  -- Create a pivot table to calculate the different 
  -- amounts that are part of an invoice.
  -- The new row is aggregated at 
  -- transaction_reference_id - end_user_account_id level
  invoice_amounts_aggregated as (
    select
      transaction_reference_id,
      product_id,
      -- a given disbursement id should have the 
      -- same disbursement_date
      max(disbursement_date) as disbursement_date,
      -- Build a pivot table in order to provide all the
      -- data related to a transaction in a single row.
      -- Note that the amounts are negated. This is because 
      -- when an invoice is generated, we give you the 
      -- positive amounts and the disbursement event 
      -- negates the amounts
      sum(case when transaction_type = 'SELLER_REV_SHARE' 
          then -invoice_amount else 0 end) as seller_rev_share,
      sum(case when transaction_type = 'AWS_REV_SHARE'  
          then -invoice_amount else 0 end) as aws_rev_share,
      sum(case when transaction_type = 'SELLER_REV_SHARE_REFUND'  
          then -invoice_amount else 0 end) as seller_rev_refund,
      sum(case when transaction_type = 'AWS_REV_SHARE_REFUND'  
          then -invoice_amount else 0 end) as aws_rev_refund,
      sum(case when transaction_type = 'SELLER_REV_SHARE_CREDIT'  
          then -invoice_amount else 0 end) as seller_rev_credit,
      sum(case when transaction_type = 'AWS_REV_SHARE_CREDIT'  
          then -invoice_amount else 0 end) as aws_rev_credit,
      sum(case when transaction_type = 'SELLER_TAX_SHARE'  
          then -invoice_amount else 0 end) as seller_tax_share,
      sum(case when transaction_type = 'SELLER_TAX_SHARE_REFUND'  
          then -invoice_amount else 0 end) as seller_tax_refund,
      -- This is the account that pays the invoice:
      max(payer_account_id) as payer_account_id,
      -- This is the account that subscribed to the product:
      end_user_account_id as customer_account_id,
      bank_trace_id
    from
      disbursed_line_items
    group by
      transaction_reference_id,
      product_id,
      disbursement_id,
      -- There might be a different end-user for the same 
      -- transaction reference id. Distributed licenses 
      -- is an example
      end_user_account_id,
      bank_trace_id
),

disbursed_amount_by_product as (
  select
    products.title as ProductTitle,
    products.product_code as ProductCode,
    -- We are rounding the sums using 2 decimal precision
    -- Note that the rounding method might differ 
    -- between SQL implementations.
    -- The disbursement seller report is using 
    -- RoundingMode.HALF_UP. This might create 
    -- discrepancies between this SQL output
    -- and the disbursement seller report
    round(invoice_amounts.seller_rev_share, 2) as SellerRev,
    round(invoice_amounts.aws_rev_share, 2) as AWSRefFee,
    round(invoice_amounts.seller_rev_refund, 2) as SellerRevRefund,
    round(invoice_amounts.aws_rev_refund, 2) as AWSRefFeeRefund,
    round(invoice_amounts.seller_rev_credit, 2) as SellerRevCredit,
    round(invoice_amounts.aws_rev_credit, 2) as AWSRefFeeCredit,
    (
        round(invoice_amounts.seller_rev_share, 2) +
        round(invoice_amounts.aws_rev_share, 2) +
        round(invoice_amounts.seller_rev_refund, 2) +
        round(invoice_amounts.aws_rev_refund, 2) +
        round(invoice_amounts.seller_rev_credit, 2) +
        round(invoice_amounts.aws_rev_credit, 2)
    ) as NetAmount,
    invoice_amounts.transaction_reference_id  
          as TransactionReferenceID,
    round(invoice_amounts.seller_tax_share, 2)  
          as SellerSalesTax,
    round(invoice_amounts.seller_tax_refund, 2)  
          as SellerSalesTaxRefund,
    payer_info.aws_account_id  
          as PayerAwsAccountId,
    customer_info.aws_account_id  
          as EndCustomerAwsAccountId,
    invoice_amounts.disbursement_date  
          as DisbursementDate,
    invoice_amounts.bank_trace_id  
          as BankTraceId
  from
    invoice_amounts_aggregated invoice_amounts
    join products_with_latest_version products  
      on products.product_id = invoice_amounts.product_id
    left join accounts_with_latest_version payer_info  
      on payer_info.account_id = invoice_amounts.payer_account_id
    left join accounts_with_latest_version customer_info  
      on customer_info.account_id = invoice_amounts.customer_account_id
)

select * from disbursed_amount_by_product;
```

## 销售补偿报告
<a name="data-feed-example-sales-compensation"></a>

要按客户查找计费收入，您可以运行如下所示的查询。这些示例查询相互依存，可创建最终的客户详细信息列表，其中包含向每位客户收取的软件使用费总额。您可以使用下方所示的查询，也可以根据您的数据和用例进行自定义。

 查询中的注释说明了查询的作用以及如何修改查询。

**注意**  
运行此查询时，我们假设提取的数据使用两个时间轴（ `valid_from` 和 `update` 列）。有关更多信息，请参阅[AWS Marketplace 数据源的存储和结构](data-feed-details.md)。

```
    -- Gets all the products and keeps the latest product_id, 
    -- valid_from tuple.
    with products_with_uni_temporal_data as (
      select
       *
      from
      (
        select
         *,
         ROW_NUMBER() OVER (PARTITION BY product_id, valid_from 
                  ORDER BY from_iso8601_timestamp(update_date) desc) 
                  as row_num
        from
         productfeed_v1
      )
      where
        -- A product_id can appear multiple times with the same 
        -- valid_from date but with a different update_date column,
        -- making it effectively bi-temporal. By only taking the most
        -- recent tuple, we are converting to a uni-temporal model.
        row_num = 1
    ),

    -- Gets the latest revision of a product
    -- A product can have multiple revisions where some of the 
    -- columns, like the title, can change.
    -- For the purpose of the sales compensation report, we want 
    -- to get the latest revision of a product
    products_with_latest_revision as (
     select
      *
     from
     (
      select
       *,
       ROW_NUMBER() OVER (PARTITION BY product_id ORDER BY from_iso8601_timestamp(valid_from) desc) as row_num_latest_revision
      from
       products_with_uni_temporal_data
     )
     where
      row_num_latest_revision = 1
   ),

     -- Gets all the addresses and keeps the latest address_id, 
     -- aws_account_id, and valid_from combination.
     -- We're transitioning from a bi-temporal data model to an 
     -- uni-temporal data_model
     piifeed_with_uni_temporal_data as (
       select
        *
       from
       (
         select
          *,
          ROW_NUMBER() OVER (
             PARTITION BY address_id, aws_account_id, valid_from 
             ORDER BY from_iso8601_timestamp(update_date) desc) 
             as row_num
         from
          piifeed
       )
       where
         -- An address_id can appear multiple times with the same
         -- valid_from date but with a different update_date column.
         -- We are only interested in the most recent.
         row_num = 1
     ),

    -- Gets the latest revision of an address.
    -- An address_id can have multiple revisions where some of 
    -- the columns can change.
    -- For the purpose of the sales compensation report, we want to
    -- get the latest revision of an address + account_id pair.
    pii_with_latest_revision as (
      select
       *
      from
      (
       select
        *,
        ROW_NUMBER() OVER (PARTITION BY address_id, aws_account_id 
              ORDER BY from_iso8601_timestamp(valid_from) desc) 
              as row_num_latest_revision
       from
        piifeed_with_uni_temporal_data
      )
      where
       row_num_latest_revision = 1
    ),

    -- Gets all the accounts and keeps the latest 
    -- account_id, valid_from tuple.
    -- We're transitioning from a bi-temporal data 
    -- model to an uni-temporal data_model.
    accounts_with_uni_temporal_data as (
      select
       *
      from
      (
        select
         *,
         ROW_NUMBER() OVER (PARTITION BY account_id, valid_from 
             ORDER BY from_iso8601_timestamp(update_date) desc) 
             as row_num
        from
         accountfeed_v1
      )
      where
        -- An account_id can appear multiple times with the same 
        -- valid_from date but with a different update_date column.
        -- We are only interested in the most recent tuple.
        row_num = 1
    ),

    -- Gets all the historical dates for an account
    -- An account can have multiple revisions where some of the 
    -- columns like the mailing_address_id can change.
    accounts_with_history as (
     select
      *,
      -- This interval's begin_date
      case
        when
        -- First record for a given account_id
          lag(valid_from, 1) over (partition by account_id 
             order by from_iso8601_timestamp(valid_from) asc) is null
        then
          -- 'force' begin_date a bit earlier because of different 
          -- data propagation times. We'll subtract one day as one
          -- hour is not sufficient
          from_iso8601_timestamp(valid_from) - INTERVAL '1' DAY
        else
          -- not the first line -> return the real date
          from_iso8601_timestamp(valid_from)
      end as begin_date,
      -- This interval's end date.
      COALESCE(
           LEAD(from_iso8601_timestamp(valid_from), 1) 
                OVER (partition by account_id 
                ORDER BY from_iso8601_timestamp(valid_from)),
           from_iso8601_timestamp('9999-01-01T00:00:00Z')
      ) as end_date
     from
       accounts_with_uni_temporal_data
   ),

    -- Gets all the billing events and keeps the latest 
    -- billing_event_id, valid_from tuple.
    -- We're transitioning from a bi-temporal data 
    -- model to an uni-temporal data_model.
    billing_events_with_uni_temporal_data as (
      select
       *
      from (
        select
          billing_event_id,
          from_iso8601_timestamp(valid_from) as valid_from,
          from_iso8601_timestamp(update_date) as update_date,
          from_iso8601_timestamp(invoice_date) as invoice_date,
          transaction_type,
          transaction_reference_id,
          product_id,
          disbursement_billing_event_id,
          action,
          currency,
          from_account_id,
          to_account_id,
          end_user_account_id,
          -- convert an empty billing address to null. This will 
          -- later be used in a COALESCE call
          case
           when billing_address_id <> '' then billing_address_id else null
          end as billing_address_id,
          CAST(amount as decimal(20, 10)) invoice_amount,
          ROW_NUMBER() OVER (PARTITION BY billing_event_id, valid_from 
              ORDER BY from_iso8601_timestamp(update_date) desc) 
              as row_num
        from
          billingeventfeed_v1
        where
          -- The Sales Compensation Report does not contain BALANCE 
          -- ADJUSTMENTS, so we filter them out here
          transaction_type <> 'BALANCE_ADJUSTMENT'
          -- Keep only the transactions that will affect any 
          -- future disbursed amounts.
          and balance_impacting = '1'
        )
      where row_num = 1
    ),

    -- Gets the billing address for all DISBURSED invoices. This 
    -- will be the address of the payer when the invoice was paid.
    -- NOTE: For legal reasons, for CPPO transactions, the 
    -- manufacturer will not see the payer's billing address id
    billing_addresses_for_disbursed_invoices as (
      select
        billing_events_raw.transaction_reference_id,
        billing_events_raw.billing_address_id,
        billing_events_raw.from_account_id
      from
        billing_events_with_uni_temporal_data billing_events_raw
      where
        -- the disbursed items will contain the billing address id
        billing_events_raw.action = 'DISBURSED'
        -- we only want to get the billing address id for the 
        -- transaction line items where the seller is the receiver
        -- of the amount
        and billing_events_raw.transaction_type like 'SELLER_%'
      group by
        billing_events_raw.transaction_reference_id,
        billing_events_raw.billing_address_id,
        billing_events_raw.from_account_id
    ),

  -- An invoice can contain multiple line items.
  -- We create a pivot table to calculate the different amounts 
  -- that are part of an invoice.
  -- The new row is aggregated at 
  -- transaction_reference_id - end_user_account_id level
  invoiced_and_forgiven_transactions as (
    select
      transaction_reference_id,
      product_id,
      -- A transaction will have the same invoice date for all 
      -- of its line items (transaction types)
      max(invoice_date) as invoice_date,
      -- A transaction will have the same billing_address_id 
      -- for all of its line items. Remember that the billing event
      -- is uni temporal and we retrieved only the latest valid_from item
      max(billing_address_id) as billing_address_id,
      --  A transaction will have the same currency for all 
      -- of its line items
      max(currency) as currency,
      -- We're building a pivot table in order to provide all the 
      -- data related to a transaction in a single row
      sum(case when transaction_type = 'SELLER_REV_SHARE' 
            then invoice_amount else 0 end) as seller_rev_share,
      sum(case when transaction_type = 'AWS_REV_SHARE' 
            then invoice_amount else 0 end) as aws_rev_share,
      sum(case when transaction_type = 'SELLER_REV_SHARE_REFUND' 
            then invoice_amount else 0 end) as seller_rev_refund,
      sum(case when transaction_type = 'AWS_REV_SHARE_REFUND' 
            then invoice_amount else 0 end) as aws_rev_refund,
      sum(case when transaction_type = 'SELLER_REV_SHARE_CREDIT' 
            then invoice_amount else 0 end) as seller_rev_credit,
      sum(case when transaction_type = 'AWS_REV_SHARE_CREDIT' 
            then invoice_amount else 0 end) as aws_rev_credit,
      sum(case when transaction_type = 'SELLER_TAX_SHARE' 
            then invoice_amount else 0 end) as seller_tax_share,
      sum(case when transaction_type = 'SELLER_TAX_SHARE_REFUND' 
            then invoice_amount else 0 end) as seller_tax_refund,
      -- this is the account that pays the invoice.
      max(case
        -- Get the payer of the invoice from any transaction type 
        -- that is not AWS and not BALANCE_ADJUSTMENT.
        -- For AWS and BALANCE_ADJUSTMENT, the billing event feed 
        -- will show the "AWS Marketplace" account as the
        -- receiver of the funds and the seller as the payer. We 
        -- are not interested in this information here.
        when
         transaction_type not like '%AWS%' 
           and transaction_type not like 'BALANCE_ADJUSTMENT' 
         then from_account_id
       end) as payer_account_id,
      -- this is the account that subscribed to your product
      end_user_account_id as customer_account_id
    from
      billing_events_with_uni_temporal_data
    where
      -- Get invoiced or forgiven items. Disbursements are 
      -- not part of the sales compensation report
      action in ('INVOICED', 'FORGIVEN')
    group by
      transaction_reference_id,
      product_id,
      -- There might be a different end-user for the same 
      -- transaction reference id. Distributed licenses 
      -- is an example.
      end_user_account_id
),

invoiced_items_with_product_and_billing_address as (
  select
    invoice_amounts.*,
    products.product_code,
    products.title,
    payer_info.aws_account_id as payer_aws_account_id,
    payer_info.account_id as payer_reference_id,
    customer_info.aws_account_id as end_user_aws_account_id,
    (
        invoice_amounts.seller_rev_share +
        invoice_amounts.aws_rev_share +
        invoice_amounts.seller_rev_refund +
        invoice_amounts.aws_rev_refund +
        invoice_amounts.seller_rev_credit +
        invoice_amounts.aws_rev_credit +
        invoice_amounts.seller_tax_share +
        invoice_amounts.seller_tax_refund
    ) as seller_net_revenue,
    -- Try to get the billing address from the DISBURSED event 
    -- (if any). If there is no DISBURSEMENT, get the billing 
    -- address from the INVOICED item. If still no billing address, 
    -- then default to getting the mailing address of the payer.
    coalesce(billing_add.billing_address_id, 
             invoice_amounts.billing_address_id, 
             payer_info.mailing_address_id) 
          as final_billing_address_id
  from
    invoiced_and_forgiven_transactions invoice_amounts
    join products_with_latest_revision products 
        on products.product_id = invoice_amounts.product_id
    left join accounts_with_history payer_info 
        on payer_info.account_id = invoice_amounts.payer_account_id
          -- Get the Payer Information at the time of invoice creation
          and payer_info.begin_date <= invoice_amounts.invoice_date 
          and invoice_amounts.invoice_date < payer_info.end_date
    left join accounts_with_history customer_info 
        on customer_info.account_id = invoice_amounts.customer_account_id
          -- Get the End User Information at the time of invoice creation
          and customer_info.begin_date <= invoice_amounts.invoice_date 
          and invoice_amounts.invoice_date < customer_info.end_date
    left join billing_addresses_for_disbursed_invoices billing_add 
        on billing_add.transaction_reference_id = 
           invoice_amounts.transaction_reference_id
        and billing_add.from_account_id = 
            invoice_amounts.payer_account_id
),

invoices_with_full_address as (
  select
    payer_aws_account_id as "Customer AWS Account Number",
    pii_data.country as "Country",
    pii_data.state_or_region as "State",
    pii_data.city as "City",
    pii_data.postal_code as "Zip Code",
    pii_data.email_domain as "Email Domain",
    product_code as "Product Code",
    title as "Product Title",
    seller_rev_share as "Gross Revenue",
    aws_rev_share as "AWS Revenue Share",
    seller_rev_refund as "Gross Refunds",
    aws_rev_refund as "AWS Refunds Share",
    seller_net_revenue as "Net Revenue",
    currency as "Currency",
    date_format(invoice_date, '%Y-%m')as "AR Period",
    transaction_reference_id as "Transaction Reference ID",
    payer_reference_id as "Payer Reference ID",
    end_user_aws_account_id as "End Customer AWS Account ID"
  from
    invoiced_items_with_product_and_billing_address invoice_amounts
    left join pii_with_latest_revision pii_data 
        on pii_data.aws_account_id = invoice_amounts.payer_aws_account_id
        and pii_data.address_id = invoice_amounts.final_billing_address_id
    -- Filter out FORGIVEN and Field Demonstration Pricing transactions
    where seller_net_revenue <> 0
)

select * from invoices_with_full_address;
```

# 数据源
<a name="data-feeds"></a>

AWS Marketplace 提供大量数据馈送，以帮助卖家收集和分析有关您的商品销售的信息。所有注册 AWS Marketplace 卖家均可使用数据源。由于数据源是在一天内生成的，因此其中包含最新的可用数据。

以下各个部分介绍了每种数据源。

**Topics**
+ [账户数据源](data-feed-account.md)
+ [地址数据源](data-feed-address.md)
+ [协议数据源](data-feed-agreements.md)
+ [协议费用数据馈送](data-feed-agreement-charge.md)
+ [协议期限数据馈送](data-feed-agreement-term.md)
+ [账单事件数据源](data-feed-billing-event.md)
+ [每日使用量数据源](data-feed-daily-usage.md)
+ [传统映射数据源](data-feed-legacy-mapping.md)
+ [优惠数据源](data-feed-offer.md)
+ [优惠产品数据源](data-feed-offer-product.md)
+ [优惠目标数据源](data-feed-offer-target.md)
+ [产品数据源](data-feed-product.md)
+ [税项数据源](data-feed-tax-item.md)

# 账户数据源
<a name="data-feed-account"></a>

此数据源提供有关您与之交互的所有账户的信息：您自己的账户、与您合作的任何渠道合作伙伴、买家、付款人和所有纳税账户。

账户数据是不可变的，它不与版本号相关联。将附加对字段的更改，因此此数据源可能包含多个具有相同 `account_id` 和不同 `valid_from` 值的行。有关数据历史记录字段的信息，请参阅[数据历史记录](data-feed-details.md#data-feed-historization)。

账户数据源每 24 小时刷新一次，因此每天提供的都是新数据。

下表说明了数据源列的名称和说明。


| 列名称  | 描述  | 
| --- | --- | 
| account\$1id  | 账户的全局唯一标识符 (GUID)。可用于连接 `Product`、`Offer_Target`、`Billing_Event` 和 `Tax_Item` 数据源的字段。有关可用于连接的字段的信息，请参阅这些数据源。 | 
| aws\$1account\$1id  | 卖家的 AWS 账户的 AWS 账号，就 AWS 分区而言是唯一的。 | 
| encrypted\$1account\$1id | 您的应用程序的个人买家的唯一已加密 ID。例如，AWS Marketplace Metering Service 将 encrypted\$1account\$1id 的值用作 [https://docs.aws.amazon.com/marketplacemetering/latest/APIReference/API_ResolveCustomer.html](https://docs.aws.amazon.com/marketplacemetering/latest/APIReference/API_ResolveCustomer.html) 操作返回 CustomerIdentifier 的值。 | 
| mailing\$1address\$1id | 此账户的邮寄地址参考。 | 
| tax\$1address\$1id | 此账户的纳税地址参考。 | 
| tax\$1registration\$1number | 对于非美国账户，为此账户的税务登记号。 | 
| tax\$1legal\$1name | 对于非美国账户，为法定公司名称。这是税务发票上使用的名称。 | 

## 账户数据源示例
<a name="data-feed-account-sample-data"></a>

下面显示了账户数据源的示例。为了便于阅读，未显示数据历史记录列。有关数据历史记录字段的信息，请参阅[数据历史记录](data-feed-details.md#data-feed-historization)。


| account\$1id  | aws\$1account\$1id  | encrypted\$1account\$1id | mailing\$1address\$1id | tax\$1address\$1id | tax\$1registration\$1number | tax\$1legal\$1name | 
| --- | --- | --- | --- | --- | --- | --- | 
| xk0CSmiAm6PQ4QqEog9iiaochIzuPlkMfba7a1oDlZ | 444456660000 | Zf7oMzheGWpH | 25o3k46eN6eViOfFiiqtxwX8e3kaOiPalUiofjyFa3 |  |  |  | 
| 7nyo5jwTRoPlyX81vx9ji04eEwTurO1Ff8biQi88W8 | 555567679999 | 373vuQUqmQ8v | 5oJ6vTjSzMrrF2gvh2Vj9HfqiM800MuLEHmyFY5Lr42s8 | 5oJ6vTjSzMrrF2gvh2Vj9HfqiM800MuLEHmyFY5Lr42s8 | SE823935083345 |  | 
| VIeGa2t9j3MuxioH9wc8lsndXXCgGCGUreeXriocM5 | 73739998888 | 8SPxAYmi8MwX | NLUc5UeiMlGFTrDWCoftDPhDUF1oaSd8xgl5QM8Db7 | V5NhBYBiYogwy0WMhndGU4AfMggmuoTC2j7Pm8ZKKNNyT | DE469558025 |  | 

# 地址数据源
<a name="data-feed-address"></a>

**重要**  
AWS Marketplace 将于 2023 年 12 月停止使用[卖家传输数据源服务](https://docs.aws.amazon.com/marketplace/latest/userguide/data-feed-service.html) PIIFeed。使用 AddressFeed\$1v1 数据源来满足您的地址数据需求。如果您有任何疑问或需要支持，请联系 [AWS Marketplace 卖家运营](https://aws.amazon.com/marketplace/management/contact-us/)团队。

此数据源提供了您与之交互的所有账户的联系信息：您自己的账户、与您合作的任何渠道合作伙伴、买家、付款人和所有纳税账户。每次进行新交易时，都会扫描交易的客户地址；如果在数据源中不包含该地址，则会在数据源文件中添加新条目。

地址数据是不可变的。

地址数据源每 24 小时刷新一次，因此每天提供的都是新数据。

下表说明了数据源列的名称和说明。


| 列名称  | 描述  | 
| --- | --- | 
| address\$1id  | 地址的唯一键。可用于连接 `Billing_Event` 数据源的 `billing_address_id` 字段，或 `Account` 数据源的 `mailing_address_id` 或 `tax_address_id` 字段。 | 
| email\$1domain  | 为该账户登记的电子邮件地址域。 | 
| company\$1name  | 为该账户登记的公司名称。 | 
| country\$1code | 为该地址登记的 ISO 3166 alpha-2 国家/地区代码。 | 
| state\$1or\$1region  | 为该地址登记的州/省或地区。 | 
| city  | 为该地址登记的城市。 | 
| postal\$1code  | 为该地址登记的邮政编码。 | 
| address\$1line\$11  | 为该地址登记的地址第一行。 | 
| address\$1line\$12  | 为该地址登记的地址第二行。 | 
| address\$1line\$13  | 为该地址登记的地址第三行。 | 

## 地址数据源示例
<a name="data-feed-address-sample-data"></a>

下面显示了地址数据源的示例。在数据源中，此信息以单个表格形式显示。为了便于阅读，数据显示在两个表格中，并且未显示数据历史记录列。有关数据历史记录字段的信息，请参阅[数据历史记录](data-feed-details.md#data-feed-historization)。


| address\$1id  | email\$1domain  | company\$1name  | country\$1code  | state\$1or\$1region  | city  | postal\$1code  | 
| --- | --- | --- | --- | --- | --- | --- | 
| V5NhBYBiYogwy0WMhndGU4AfMggmuoTC2j7Pm8ZKKNNyT | a.com | Mateo Jackson's Company | DE |  | Hamburg | 67568 | 
| G68xdbkZQDVVHzfBGw6yf5yos0A6NiSVWHmH5ViLjf | b.com | Mary Major's Company | 美国 | OH | Dayton | 57684 | 
| NLUc5UeiMlGFTrDWCoftDPhDUF1oaSd8xgl5QM8Db7 | c.com | Our Seller | 美国 | NY | New York | 89475 | 




| address\$1line\$11  | address\$1line\$12  | address\$1line\$13  | 
| --- | --- | --- | 
|   |   |  | 
|  |   |  | 
|  | 19th Floor |  | 



# 协议数据源
<a name="data-feed-agreements"></a>

此数据源提供有关协议的信息，协议是提议者（记录在案的卖方）和接受者（AWS 买方）之间为开始使用产品而签订的合同。此数据源提供了您作为登记卖家创建的产品优惠的协议相关信息。

协议数据源每 24 小时刷新一次。

下表列出并描述了数据源中的项目。


| 列 | 说明 | 
| --- | --- | 
| valid\$1from | 主键值相对于其他字段的值有效的第一个日期。 | 
| insert\$1date | 将记录插入数据源中的日期。 | 
| update\$1date | 上次更新记录的日期。 | 
| delete\$1date | 该列始终为空。 | 
| Status |  协议的当前状态。支持的状态包括： [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/zh_cn/marketplace/latest/userguide/data-feed-agreements.html)  | 
| estimated\$1agreement\$1value | 客户在协议的整个生命周期内可能支付的已知总金额。 | 
| currency\$1code | 客户在协议的整个生命周期内可能支付的已知总金额的货币。 | 
| agreement\$1id | 协议的唯一标识符。 | 
| license\$1ids | 与协议关联的许可证标识符，以数组表示。 | 
| proposer\$1account\$1id | 提出此建议的卖家 PurchaseAgreement，由卖家账户的全球唯一标识符 (GUID) 表示。可用于连接“账户”数据源。 | 
| acceptor\$1account\$1id | 接受这一点的买家 PurchaseAgreement，由买家账户的全球唯一标识符 (GUID) 表示。可用于连接“账户”数据源。 | 
| offer\$1revision\$1at\$1acceptance | 与此协议对应的优惠的友好 ID。可用于连接“优惠”和“优惠目标”数据源。 | 
| 报价集\$1ID | 与该优惠关联的优惠套餐的标识符。 | 
| start\$1time | 协议的开始日期和时间。 | 
| end\$1time | 协议的结束日期和时间。对于没有结束日期的 pay-as-you-go协议，该字段为空。 | 
| acceptance\$1time | 买家接受优惠或协议创建的日期和时间。 [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/zh_cn/marketplace/latest/userguide/data-feed-agreements.html)  | 
| intent | 上次修改协议时买家的意图。 | 
| preceding\$1agreement\$1id | 之前的协议的协议 ID。 | 
| status\$1reason\$1code | 协议状态更改原因。 | 
| 收件人\$1账号\$1ID | 在上传数据中接收数据的卖家账户。可用于连接 `account_id` 数据源的 `Account` 字段。 | 

## 协议数据源示例
<a name="agreements-feed-example"></a>


| valid\$1from | agreement\$1id | proposer\$1account\$1id | acceptor\$1account\$1id | offer\$1id | offer\$1revision\$1at\$1acceptance | 报价集\$1ID | start\$1time | end\$1time | acceptancet\$1time | intent | preceding\$1agreement\$1id | status | status\$1reason\$1code | estimated\$1agreement\$1value | currency\$1code | 
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | 
| 2024-06-21 20:58:00.0 | agmt-34g544dfgsd5678adsrgwe5t | 88a3xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx | 88a3xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx | offer-krxxxxxxxxxxx | 1 | offerset-yricpu74oqox2 | 2024-06-21 20:58:00.0 | 2025-06-21 20:58:00.0 | 2024-06-21 20:58:00.0 | 新 |  | ACTIVE |  | 1000 | USD | 

# 协议费用数据馈送
<a name="data-feed-agreement-charge"></a>

此数据源概述了与您作为登记卖家创建的协议相关的费用。该数据源目前仅提供与付款时间表条款相关的协议的费用信息。

协议费用数据源每 24 小时刷新一次。

下表列出并描述了数据源中的项目。


| 列 | 说明 | 
| --- | --- | 
| valid\$1from | 主键值相对于其他字段的值有效的第一个日期。 | 
| insert\$1date | 将记录插入数据源中的日期。 | 
| update\$1date | 上次更新记录的日期。 | 
| delete\$1date | 该列始终为空。 | 
| 量 | 要收取的金额。 | 
| currency\$1code | 付款的定价货币。 | 
| 时间 | 计划分期付款的预定收费日期。 | 
| agreement\$1id | 协议的唯一标识符。 | 
| 术语\$1id | 术语的唯一标识符。 | 

## 协议费用数据馈送示例
<a name="agreement-charge-feed-example"></a>


| valid\$1from | insert\$1date | update\$1date | delete\$1date | 量 | currency\$1code | 时间 | agreement\$1id | 术语\$1id | 
| --- | --- | --- | --- | --- | --- | --- | --- | --- | 
| 2024-12-24 11:31:47.0 | 2025-10-01 01:03:41.0 | 2025-10-01 01:03:41.0 | null | 28440.00 | USD | 2025-11-06 00:00:00.000 000 | agmt-3kk39tbw3j6id2vakbp0xxxx | term-3986e2c7f73768ed4ff7cd8a97b41ac0ae2a02ada6b68deb9349c8604cxxxx | 

# 协议期限数据馈送
<a name="data-feed-agreement-term"></a>

此数据源概述了与您作为登记卖家创建的协议相关的协议期限。

协议条款数据源每 24 小时刷新一次。

下表列出并描述了数据源中的项目。


| 列 | 说明 | 
| --- | --- | 
| valid\$1from | 主键值相对于其他字段的值有效的第一个日期。 | 
| insert\$1date | 将记录插入数据源中的日期。 | 
| update\$1date | 上次更新记录的日期。 | 
| delete\$1date | 删除记录的日期。 | 
| agreement\$1id | 协议的唯一标识符。 | 
| 术语\$1id | 术语的唯一标识符。 | 
| 术语类型 | 与协议相关的条款类型。有关可用术语的完整列表，请参阅 https://docs.aws.amazon.com/marketplace/ latest/APIReference/API \$1Types\$1 AWS\$1M arketplace\$1Agreement\$1Service.html。 | 
| 术语配置 | 接受者在接受术语时指定的其他参数。这适用于 ConfigurableUpfrontPricingTerms 和 RenewalTerms。 | 

## 协议条款数据馈送示例
<a name="agreement-term-feed-example"></a>


| valid\$1from | insert\$1date | update\$1date | delete\$1date | agreement\$1id | 术语\$1id | 术语类型 | 术语配置 | 
| --- | --- | --- | --- | --- | --- | --- | --- | 
| 2024-12-24 11:31:47 .0 | 2025-10-01 01:03:41.0 | 2025-10-01 01:03:41.0 | null | agmt-3kk39tbw3j6id2vakbp0xxxx | term-3986e2c7f73768ed4ff7cd8a97b41ac0ae2a02ada6b68deb9349c8604cxxxx | ConfigurableUpfrontPricingTerm | \$1“selectorValue”: “P36M”，“dimensionKey”：[\$1“dimensionKey”: “应用程序”，“dimensionValue”: 1\$1]]\$1 | 

# 账单事件数据源
<a name="data-feed-billing-event"></a>

此数据源提供有关账单事件的信息，包括开发票和付款。

例如，您可以使用此数据源来了解为买家开发票的时间和内容。您还可以使用[示例 SQL 查询](#data-feeds-billing-event-query-examples)来分析来自此数据源的数据。

此数据源包含与您作为其记录卖家的账单事件相关的信息。对于通过渠道合作伙伴达成的协议，此数据源包含有关制造商和记录卖家之间的账单事件的信息。

账单事件数据源每 24 小时刷新一次，因此每天提供的都是新数据。

账单事件数据是不可变的。

下表说明了数据源列的名称和说明。


| 列名称  | 说明  | 
| --- | --- | 
| billing\$1event\$1id | 账单事件的标识符。此 ID 在卖家的环境中是唯一的。 | 
| from\$1account\$1id |  启动账单事件的账户。如果 `transaction_type` 为 `SELLER_REV_SHARE`，则它是买家的付款人账户。这是[账户](data-feed-account.md)数据源的外键。 可用于连接 `account_id` 数据源的 `Account` 字段。  | 
| to\$1account\$1id | 接收产品的交易金额的账户。这是账户数据源的外键。可用于连接 `account_id` 数据源的 `Account` 字段。 | 
| end\$1user\$1account\$1id | 使用产品的账户。此账户可能与买家账户和付款人账户不同。可用于连接 `account_id` 数据源的 `Account` 字段。 | 
| product\$1id | 产品的标识符。这是[产品](data-feed-product.md)数据源的外键。可用于连接 `product_id` 数据源的 `Product` 字段。 | 
| action |  此事件的操作类型。可能值如下所示： [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/zh_cn/marketplace/latest/userguide/data-feed-billing-event.html)  | 
| transaction\$1type |  交易的类型。有关示例，请参阅 [征税方案](#data-feeds-billing-event-tax-examples)。可能值如下所示： [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/zh_cn/marketplace/latest/userguide/data-feed-billing-event.html)  | 
| parent\$1billing\$1event\$1id |  如果 `broker_id` 的值为 `AWS_INC`，`action` 的值为 `DISBURSED` 或 `FORGIVEN`，`transaction_type` 的值为 `DISBURSEMENT`，则 `parent_billing_event_id` 表示发起此账单事件的原始 `billing_event_id`。如果 `action` 具有其他值，则此字段为空。 当 `broker_id` 的值为 `AWS_EUROPE` 时，`parent_billing_event_id` 表示在以下情况下启动此账单事件的原始 `billing_event_id`： [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/zh_cn/marketplace/latest/userguide/data-feed-billing-event.html) 当 `broker_id` 的值为 `AWS_EUROPE`，则 `parent_billing_event_id` 表示在以下情景中上一次成功的支付账单事件的原始 `billing_event_id`： [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/zh_cn/marketplace/latest/userguide/data-feed-billing-event.html) 当 `broker_id` 的值为 `AWS_EUROPE` 时，对于所有剩余情景，此字段均为空。  | 
| disbursement\$1billing\$1event\$1id |  `action` 的值为 `DISBURSED` 且以下参数之一为 True 时的相关支付： [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/zh_cn/marketplace/latest/userguide/data-feed-billing-event.html) 其他所有情景下，此值为 Null。  | 
| 量 | 账单事件金额。 | 
| currency | ISO 639 币种代码。 | 
| disbursement\$1amount | 付款金额。对于双重开票场景，此字段显示付款金额。对于统一开票场景，此字段为空。 | 
| disbursement\$1currency | 付款金额对应的 ISO 639 货币代码。对于双重开票场景，此字段显示付款货币代码。对于统一开票场景，此字段为空。 | 
| balance\$1impacting | 在计算卖家付款时是否考虑该金额。值为 0 指示显示该金额以供参考，对余额没有影响。值为 1 指示在确定卖家付款时考虑该金额。 | 
| invoice\$1date | 创建发票的日期。 | 
| payment\$1due\$1date |  如果 `action` 的值为 `INVOICED`，则为发票的到期日期。  | 
| usage\$1period\$1start\$1date | 记录中期间的开始日期。 | 
| usage\$1period\$1end\$1date | 记录中期间的结束日期。 | 
| invoice\$1id |  发 AWS 票编号。  | 
| billing\$1address\$1id | 地址数据源中的付款人账单地址参考。可用于连接 `address_id` 数据源的 `Address` 字段。 | 
| transaction\$1reference\$1id |  允许您交叉引用以下报告中的数据的标识符： [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/zh_cn/marketplace/latest/userguide/data-feed-billing-event.html)  | 
| bank\$1trace\$1id |  对于支付交易（`transaction_type` = `DISBURSEMENT` 且操作为`DISBURSED`），是银行分配的追踪 ID。追踪 ID 可用于关联卖家银行提供的银行报告。  | 
| broker\$1id |  为交易提供便利的商业实体的标识符。可能值如下所示： [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/zh_cn/marketplace/latest/userguide/data-feed-billing-event.html)  | 
| buyer\$1transaction\$1reference\$1id |  一种标识符，它使用 `GROUP BY` 或在 SQL 中构造的窗口函数对账单源中的所有相关记录进行分组。这些相关记录可以包含买家发票、卖家发票和上市费的增值税 (VATs)。  | 
| 操作日期 |  交易事件发生的日期。具体日期取决于操作类型： [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/zh_cn/marketplace/latest/userguide/data-feed-billing-event.html)  | 
| 收件人\$1账号\$1id | 在上传数据中接收数据的卖家账户。可用于连接 `account_id` 数据源的 `Account` 字段。 | 
| offer\$1id | 与买家发票单项关联的报价的友好编号。可用于连接 `offer_id` 数据源的 `Offer` 字段。 | 
| 买家发票日期 | 关联的买家发票明细项目的发票日期。 | 
| line\$1item\$1id | 按支出粒度分列的发票明细项目编号。 | 
| buyer\$1line\$1item\$1id | 按支付粒度划分的买家发票单项编号。 | 
| 冲锋变体 | 表示正在买入或卖出什么。 | 
| charge\$1side | 指明卖家是买入还是卖出费用变体。 | 

## 征税方案
<a name="data-feeds-billing-event-tax-examples"></a>

买家和卖家所在国家/地区和州/省实施的征税模式决定了税款的收取和汇款方式。以下是可能的方案：
+ 税款由收缴和汇出。 AWS在这些情况下，`transaction_type` 为 `AWS_TAX_SHARE`。
+ 税款由卖家收取 AWS，支付给卖家，然后由卖家汇给税务机关。在这些情况下，`transaction_type` 为 `SELLER_TAX_SHARE`。
+ 税款不是由收取的 AWS。卖家必须计算税款并将其汇款给税务机关。在这些情况下， AWS Marketplace 不进行税务计算或接收税务信息。卖家从收入份额中支付税款。

## 账单事件数据源示例
<a name="data-feed-billing-event-sample-scenario"></a>

本部分显示开发票时和一个月之后的账单事件数据期间的示例。对于本部分中的所有表，请注意以下各项：
+ 在数据源中，`billing_event_id` 值为 40 个字符的字母数字字符串。为了便于阅读，它们在这里显示为两个字符的字符串。
+ 在数据源中，此信息以单个表格形式显示。为了便于阅读，数据显示在此处的多个表中，并且不显示所有列。

对于本部分中的示例，假定以下内容：
+ Arnav 是买家。
  + 他的账户 ID 是 `737399998888`。
  + 他位于法国，需要遵守电商平台促进法的规定。有关更多信息，请参阅 [Amazon Web Service 税务帮助](https://aws.amazon.com/tax-help/)。
  + 他购买了 `prod-o4grxfafcxxxx` 并因他每月使用该产品而获得了 120.60 美元的发票。
  + 他在当月内支付了发票。
+ Jane 是制造商。
  + 她的账户 ID 是 `111122223333`。
+ Paulo 是记录卖家。
  + 他的账户 ID 是 `777788889999`。
  + 他住在堪萨斯州，这里不受电商平台促进法约束。

### 记录卖家的账单事件数据源
<a name="billing-event-example-seller-of-record"></a>

作为记录卖家，Paulo 向买家 Arnav 开具发票。

下表显示了在 Paulo 向 Arnav 开发票时其数据源中的相关信息。


| billing\$1event\$1id  | from\$1account\$1id  | to\$1account\$1id  | end\$1user\$1account\$1id | product\$1id | action | transaction\$1type | 
| --- | --- | --- | --- | --- | --- | --- | 
| I0 | 737399998888 | 777788889999 | 737399998888 | prod-o4grxfafcxxxx | INVOICED | SELLER\$1REV\$1SHARE | 
| I1 | 737399998888 | 采买 | 737399998888 | prod-o4grxfafcxxxx | INVOICED | AWS\$1TAX\$1分享 | 
| I2 | 777788889999 | 111122223333 | 737399998888 | prod-o4grxfafcxxxx | INVOICED | SELLER\$1REV\$1SHARE | 
| I3 | 777788889999 | 采买 | 737399998888 | prod-o4grxfafcxxxx | INVOICED | AWS\$1REV\$1分享 | 




| parent\$1billing\$1event\$1id | disbursement\$1billing\$1event\$1id | 量 | currency | invoice\$1date | invoice\$1id | 
| --- | --- | --- | --- | --- | --- | 
|  |  | 100 | USD | 2018-12-31T00:00:00Z | 781216640 | 
|  |  | 20.6 | USD | 2018-12-31T00:00:00Z | 781216640 | 
|  |  | -80 | USD | 2018-12-31T00:04:07Z | 788576665 | 
|  |  | -0.2 | USD | 2018-12-31T00:04:07Z | 788576665 | 

下表显示了在 Arnav 支付发票后，月底时 Paulo 数据源中的相关信息。


| billing\$1event\$1id  | from\$1account\$1id  | to\$1account\$1id  | end\$1user\$1account\$1id | product\$1id | action | transaction\$1type | 
| --- | --- | --- | --- | --- | --- | --- | 
| I10 | 737399998888 | 777788889999 | 737399998888 |  | DISBURSED | SELLER\$1REV\$1SHARE | 
| I12 | 777788889999 | 111122223333 | 737399998888 |  | DISBURSED | SELLER\$1REV\$1SHARE | 
| I13 | 777788889999 | 采买 | 737399998888 | prod-o4grxfafcxxxx | DISBURSED | AWS\$1REV\$1分享 | 
| I14 | 采买 | 777788889999 |  |  | DISBURSED | DISBURSEMENT | 




| parent\$1billing\$1event\$1id | disbursement\$1billing\$1event\$1id | 量 | currency | invoice\$1date | invoice\$1id | 
| --- | --- | --- | --- | --- | --- | 
| I0 | I14 | -100 | USD | 2018-12-31T00:00:00Z | 781216640 | 
| I2 | I14 | 80 | USD | 2018-12-31T00:04:07Z | 788576665 | 
| I3 | I14 | 0.2 | USD | 2018-12-31T00:04:07Z | 788576665 | 
|  |  | 19.8 | USD |  |  | 

### 制造商的账单事件数据源
<a name="billing-event-example-manufacturer"></a>

下表显示了在 Paulo 向 Arnav 开发票时 Jane 数据源中的相关信息。


| billing\$1event\$1id  | from\$1account\$1id  | to\$1account\$1id  | end\$1user\$1account\$1id | product\$1id | action | transaction\$1type | 
| --- | --- | --- | --- | --- | --- | --- | 
| I5 | 777788889999 | 111122223333 |  | prod-o4grxfafcxxxx | INVOICED | SELLER\$1REV\$1SHARE | 
| I6 | 777788889999 | 111122223333 |  | prod-o4grxfafcxxxx | INVOICED | SELLER\$1TAX\$1SHARE | 
| I7 | 111122223333 | 采买 |  | prod-o4grxfafcxxxx | INVOICED | AWS\$1REV\$1分享 | 




| parent\$1billing\$1event\$1id | disbursement\$1billing\$1event\$1id | 量 | currency | invoice\$1date | invoice\$1id | 
| --- | --- | --- | --- | --- | --- | 
|  |  | 73.5 |  | 2018-12-31T00:04:07Z | 788576665 | 
|  |  | 6.5 |  | 2018-12-31T00:04:07Z | 788576665 | 
|  |  | -7.35 |  | 2018-12-31T00:04:07Z | 788576665 | 

下表显示了发票支付发票后，月底时 Jane 数据源中的相关信息。


| billing\$1event\$1id  | from\$1account\$1id  | to\$1account\$1id  | end\$1user\$1account\$1id | product\$1id | action | transaction\$1type | 
| --- | --- | --- | --- | --- | --- | --- | 
| I30 | 777788889999 | 111122223333 |  | prod-o4grxfafcxxxx | DISBURSED | SELLER\$1REV\$1SHARE | 
| I31 | 777788889999 | 111122223333 |  | prod-o4grxfafcxxxx | DISBURSED | SELLER\$1TAX\$1SHARE | 
| I32 | 111122223333 | 采买 |  | prod-o4grxfafcxxxx | DISBURSED | AWS\$1REV\$1分享 | 
| I33 | 采买 | 111122223333 |  |  | DISBURSED | DISBURSEMENT | 




| parent\$1billing\$1event\$1id | disbursement\$1billing\$1event\$1id | 量 | currency | invoice\$1date | invoice\$1id | 
| --- | --- | --- | --- | --- | --- | 
| I5 | I33 | -73.5 | USD |  |  | 
| I6 | I33 | -6.5 | USD |  |  | 
| I7 | I33 | 7.35 | USD |  |  | 
|  |  | 72.65 | USD |  |  | 

## 示例查询
<a name="data-feeds-billing-event-query-examples"></a>

如[使用数据源收集和分析数据](data-feed-using.md)中所述，您可以使用 [Athena](https://docs.aws.amazon.com/athena/latest/ug/what-is.html) 对收集并作为数据源存储在托管的 Amazon 存储桶中的数据运行查询。本部分提供了您执行此操作可能采取的常见方法的一些示例。所有示例都假定使用单一币种。

## 示例 1：开具发票的金额，含税
<a name="data-feed-example-query-tax-invoice"></a>

要了解向买家开具发票的金额（含税），您可以运行查询，如以下示例所示。

```
SELECT sum(amount) FROM billing_event 
WHERE 
  action = 'INVOICED'
  AND
  (
    (transaction_type in ('SELLER_REV_SHARE', 'SELLER_TAX_SHARE')
      -- to discard SELLER_REV_SHARE from Manufacturer to Channel Partner, aka cost of goods
      AND to_account_id='seller-account-id'
    )
  OR transaction_type= 'AWS_TAX_SHARE'
  );
```

## 示例 2：代表卖家向买家开具发票的金额
<a name="data-feed-example-query-invoice-for-seller"></a>

要了解代表卖家向买家开具发票的金额，您可以运行查询，如以下示例所示。

```
SELECT sum(amount) FROM billing_event 
WHERE
  action = 'INVOICED'
  AND transaction_type in ('SELLER_REV_SHARE', 'SELLER_TAX_SHARE')
  AND to_account_id='seller-account-id'
;
```

## 示例 3： AWS 可以代表卖家收取的金额
<a name="data-feed-example-query-aws-collect"></a>

要了解 AWS 可以代表卖家收取的金额（减去任何退款、积分和免除的账户），您可以运行查询，如以下示例所示。

```
SELECT sum(amount) FROM billing_event 
WHERE
  -- what is invoiced on behalf of SELLER, incl. refunds/ credits and cost of goods
  transaction_type like 'SELLER_%' 
  -- FORGIVEN action records will "negate" related INVOICED
  and action in ('INVOICED','FORGIVEN') 
;
```

## 示例 4：卖家可以收取的金额
<a name="data-feed-example-query-seller-collect"></a>

要了解卖家可以收取的金额，您可以运行查询，如以下示例所示。此示例删除了房源费用和收取的 AWS 税费，并添加了任何特殊余额调整。

```
SELECT sum(amount) FROM billing_event
WHERE
  (transaction_type like 'SELLER_%' -- what is invoiced on behalf of SELLER
  or transaction_type like 'AWS_REV_%' -- what is owed to AWS
  or transaction_type = 'BALANCE_ADJUSTMENT' -- exceptionnal case
  )
  and action in ('INVOICED','FORGIVEN')
;
```

您也可以使用以下查询来收集相同的信息，如以下示例所示。

```
SELECT sum(amount) FROM billing_event
WHERE
  balance_impacting = 1
  and action in ('INVOICED','FORGIVEN')
;
```

以下示例显示相同信息，但仅限于 2018 年交易，并假定所有买家都按照发票支付了款项。

```
SELECT sum(amount) FROM billing_event
WHERE
  invoice_date between '2018-01-01' and '2018-12-31'
  and balance_impacting = 1
  and action in ('INVOICED','FORGIVEN')
;
```

## 示例 5：支付金额
<a name="data-feed-example-query-disbursements"></a>

要了解已支付的金额，您可以运行查询，如以下示例所示。

```
select sum(amount) FROM billing_event
WHERE
  action ='DISBURSED'
  and transaction_type like 'DISBURSEMENT%'
;
```

## 示例 6：待支付金额
<a name="data-feed-example-query-pending-disbursement"></a>

要了解待支付的金额，您可以运行查询，如以下示例所示。此查询删除了已付款的金额。

```
SELECT sum(amount) FROM billing_event targeted 
WHERE
   (transaction_type like 'SELLER_%'  -- what is invoiced on behalf of SELLER
    or transaction_type like 'AWS_REV_%'  -- what is owed to AWS
    or transaction_type = 'BALANCE_ADJUSTMENT' -- exceptionnal case
   ) 
  -- DISBURSEMENT action records will "negate" 'INVOICED'
  -- but do not take into account failed disbursements
   AND 
    (not exists
      (select 1 
        from billing_event disbursement
          join billing_event failed_disbursement
           on disbursement.billing_event_id=failed_disbursement.parent_billing_event_id
        where
         disbursement.transaction_type='DISBURSEMENT'
         and failed_disbursement.transaction_type='DISBURSEMENT_FAILURE'
         and targeted.disbursement_billing_event_id=disbursement.billing_event_id
      )
    ) 
;
```

获取相同信息的另一种方法是运行查询来获取卖家余额，如以下示例所示。

```
SELECT sum(amount) FROM billing_event
WHERE
 balance_impacting = 1
;
```

以下查询扩展了我们的示例。它将结果限制为 2018 年交易，并返回有关交易的其他详细信息。

```
select sum(residual_amount_per_transaction)
from
 (SELECT
    max(billed_invoices.amount) invoiced_amount,
    sum(nvl(disbursed_invoices.amount,0)) disbursed_amount,
    -- Exercise left to the reader:
    -- use transaction_type to distinguish listing fee vs seller-owed money
    -- still pending collection
    max(transaction_type) transaction_type,
    max(billed_invoices.amount) 
      + sum(nvl(disbursed_invoices.amount,0)) residual_amount_per_transaction
  FROM billing_event billed_invoices
    -- find related disbursements
    left join billing_event disbursed_invoices
      on disbursed_invoices.action='DISBURSED'
      and disbursed_invoices.parent_billing_event_id=billed_invoices.billing_event_id
  WHERE
    billed_invoices.invoice_date between '2018-01-01' and '2018-12-31'
    and billed_invoices.transaction_type like 'SELLER_%' -- invoiced on behalf of SELLER
    and billed_invoices.action in ('INVOICED','FORGIVEN')
    -- do not take into account failed disbursements
    AND not exists
      (select 1 from billing_event failed_disbursement
       where disbursed_invoices.disbursement_billing_event_id = failed_disbursement.parent_billing_event_id
      )
   GROUP BY billed_invoices.billing_event_id
);
```

## 示例 7：一组发票的余额
<a name="data-feed-example-query-balance-invoice-set"></a>

要了解一组发票的总额，您可以运行查询，如以下示例所示。

```
SELECT invoice_id, sum(amount) FROM billing_event targeted
WHERE
  -- invoice_id is only not null for invoiced records AND disbursed records 
  -- linking them to related disbursement -> no need to filter more precisely
  invoice_id in ('XXX','YYY') 
  -- filter out failed disbursements 
  AND not exists
      (select 1 
        from billing_event disbursement
          join billing_event failed_disbursement
           on disbursement.billing_event_id=failed_disbursement.parent_billing_event_id
        where
         disbursement.transaction_type='DISBURSEMENT'
         and failed_disbursement.transaction_type='DISBURSEMENT_FAILURE'
         and targeted.disbursement_billing_event_id=disbursement.billing_event_id
      ) 
  group by invoice_id;
```

# 每日使用量数据源
<a name="data-feed-daily-usage"></a>

此数据源提供您产品的详细每日使用信息，包括买家使用量指标、预计收入和定价详情。这些数据可帮助您跟踪客户如何使用您的产品，并根据使用模式计算估计收入。

每日使用量数据源每 24 小时刷新一次。

下表列出并描述了数据源中的项目。


| 列 | 说明 | 
| --- | --- | 
| valid\$1from | 主键值相对于其他字段的值有效的第一个日期。 | 
| insert\$1date | 将记录插入数据源中的日期。 | 
| update\$1date | 上次更新记录的日期。 | 
| delete\$1date | 软删除记录的日期。 | 
| usage\$1feed\$1id | 使用记录的唯一标识符。这是针对每个卖家的加盐哈希值，可确保数据隐私，同时保持记录的唯一性。 | 
| usage\$1date | 客户使用日期，不含时间部分。由于每天的使用量是汇总的，因此省略了该时间。 | 
| product\$1id | 产品的友好 ID。可用于加入账户、Billing\$1Event 和 Offer\$1Product 数据馈送的`product_id`字段。 | 
| agreement\$1id | 协议的唯一标识符。如果存在，则始终以`agmnt-`。此字段偶尔可能为空或与相应的发票行项目不匹配。 | 
| end\$1user\$1account\$1id | 实际使用产品的账户，由最终用户账户的全局唯一标识符 (GUID) 表示。可用于连接“账户”数据源。这是每个卖家的加盐哈希值，而不是原始的 AWS 账户 ID。 | 
| payer\$1account\$1id | 应支付产品使用费用的账户，由付款人账户的全球唯一标识符 (GUID) 表示。可用于连接“账户”数据源。这是每个卖家的加盐哈希值，而不是原始的 AWS 账户 ID。 | 
| region | 买家使用发生的地 AWS 区。 | 
| 维度键 | 卖家在发布报价时配置的维度密钥。 | 
| 用量\$1单位 | 使用单位的分类，描述使用量测量的类型。 | 
| usage\$1quantity | 与账单记录关联的 AWS 服务或使用类型的使用值。该值最多可提供 2 位小数。 | 
| pricing\$1currency | 预计收入的货币。 | 
| 以定价货币计算的估计收入 | 使用`usage_rate_per_unit`和计算的估计收入`usage_quantity`。该值最多可提供 2 位小数。要获取 ISV 对经销商产品使用情况的看法，则需要将其乘以收入分成百分比。 | 
| 收件人\$1账号\$1ID | 在上传数据中接收数据的卖家账户。 | 
| offer\$1id | 优惠的友好 ID。这将与购买协议的 offer\$1id（如果有）匹配。可用于连接“优惠”和“优惠目标”数据源。为了保持一致性，此字段始终为非空。 | 
| 每单位用量费率定价货币 | 每单位的使用率，可以乘以`usage_quantity`来验证。`estimated_revenue_in_pricing_currency`该值最多可提供 6 位小数。 | 
| 充电商品描述 | 完整的费用项目描述，通常格式为 offer\$1term\$1description\$1region\$1region\$1dimension\$1description，表示基于使用量的 | 

## 每日使用量数据 Feed 示例
<a name="daily-usage-feed-example"></a>

以下是带有关键列的 “每日使用量” 数据馈送的示例。为了便于阅读，某些列未显示。


| usage\$1date | product\$1id | agreement\$1id | region | 维度键 | 用量\$1单位 | usage\$1quantity | pricing\$1currency | 以定价货币计算的估计收入 | offer\$1id | 每单位用量费率定价货币 | 充电商品描述 | 
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | 
| 2025-01-15 | prod-abcd1234efgh5678 | agmnt-wxyz9876abcd5432 | us-east-1 | USE1\$1InputTokenCount | 单位 | 24.00 | USD | 12.00 | offer-mnop5432qrst7890 | 0.500000 | AWS Marketplace 软件使用情况\$1美国东部-100万个输入代币 | 
| 2025-01-15 | prod-ijkl9876mnop1234 | agmnt-stuv5432wxyz9876 | us-west-2 | USE1\$1InputTokenCount | 单位 | 1000.00 | USD | 5.50 | offer-abcd9876efgh5432 | 0.005500 | AWS Marketplace 软件使用情况\$1us-west-2\$1API 调用 | 

# 传统映射数据源
<a name="data-feed-legacy-mapping"></a>

此数据源列出了产品 IDs 和报价如何 IDs 映射到传统的全球唯一标识符 (GUIDs)。旧版用于 GUIDs 较旧的报表，而新的 IDs 则用于数据馈送和中 AWS Marketplace APIs。

此数据源提供了有关您作为记录卖家创建的所有产品以及您有权转售的所有产品的信息。

传统映射数据源每 24 小时刷新一次，因此每天提供的都是新数据。

下表说明了数据源列的名称和说明。


| 列名称  | 说明  | 
| --- | --- | 
| mapping\$1type | 这是产品 ID 还是优惠 ID。 | 
| legacy\$1id | 此产品或优惠的传统 ID。 | 
| new\$1id | 此产品或优惠的友好 ID。此 ID 用作主键，并与所有当前 API 操作一起使用。 | 

## 传统映射数据源示例
<a name="data-feed-legacy-mapping-sample-data"></a>

下面显示了传统映射数据源的示例。为了便于阅读，未显示数据历史记录列。有关数据历史记录字段的信息，请参阅[数据历史记录](data-feed-details.md#data-feed-historization)。


| mapping\$1type | legacy\$1id  | new\$1id | 
| --- | --- | --- | 
| 优惠 | 8a806c74-dbd6-403e-9362-bb08f417ff37 | offer-dacpxznflfwin | 
| 产品 | 1368541d-890b-4b6c-9bb9-4a55306ab642 | prod-o4grxfafcxxxy | 
| 优惠 | 558d8382-6b3a-4c75-8345-a627b552f5f1 | offer-gszhmle5npzip | 



# 优惠数据源
<a name="data-feed-offer"></a>

优惠数据源提供了有关您作为记录卖家创建的所有优惠的信息。如果单个优惠具有多个修订版，则所有修订版都包含在数据源中。

当您创建优惠修订版且公开字段中的数据发生更改时，会在数据源中为同一主键（`offer_id` 和 `offer_revision`）创建一条新记录。但 `valid_from` 字段具有不同的值。有关数据源历史记录列的更多信息，请参阅[数据历史记录](data-feed-details.md#data-feed-historization)。

每 24 小时刷新一次优惠数据源，因此每天都提供新数据。

下表提供了数据源列的名称和说明。


| 列名称  | 说明  | 
| --- | --- | 
| offer\$1id | 优惠的友好标识符。可用于连接 `Offer_Product` 数据源的 `offer_id` 字段。 | 
| offer\$1revision | 优惠修订版。此字段与 offer\$1id 字段组合使用来构成主键。`offer_id` 可用于连接 `Target_Offer` 数据源的 `offer_id` 和 `offer_revision` 字段。 | 
| name | 卖家定义的优惠名称。 | 
| expiration\$1date | 优惠到期的日期和时间。 | 
| opportunity\$1name | 与此优惠相关的任何机会数据。如果优惠绑定到 AWS 机会，则会填充此字段。 | 
| opportunity\$1description | 与此优惠相关的任何描述性信息。如果优惠绑定到 AWS 机会，则会填充此字段。 | 
| seller\$1account\$1id | 卖家账户的全局唯一标识符 (GUID)。可用于连接账户数据源的 account\$1id 字段。 | 
| opportunity\$1id | 只有当经销商销售您的产品时，才会填写机会的标识符。如果产品相同，则由不同渠道合作伙伴（或卖家）创建的所有优惠具有相同的 opportunity\$1id。 | 
| 收件人\$1账号\$1ID | 在上传数据中接收数据的卖家账户。可用于连接 `account_id` 数据源的 `Account` 字段。 | 

## 优惠数据源示例
<a name="data-feed-offer-sample-data"></a>

下面显示了优惠数据源的示例。为了便于阅读，未显示数据历史记录列。有关数据历史记录字段的信息，请参阅[数据历史记录](data-feed-details.md#data-feed-historization)。


| offer\$1id  | offer\$1revision | name | expiration\$1date | opportunity\$1name | opportunity\$1description | seller\$1account\$1id | opportunity\$1id | 
| --- | --- | --- | --- | --- | --- | --- | --- | 
| offer-dacpxznflfwin | 1 | 企业合同计划优惠 | 9999-01-01T00:00:00Z |  |  |  |  | 
| offer-gszhmle5npzip | 1 | 卖家创建的专属优惠 | 2020-10-31T00:00:00Z |  |  |  |  | 
| offer-hmzhyle8nphlp | 1 | 企业合同计划优惠 | 9999-01-01T00:00:00Z |  |  |  |  | 

# 优惠产品数据源
<a name="data-feed-offer-product"></a>

一种优惠可以包含多种产品，一种产品可以包含在不同优惠中。此数据源列出了有关优惠与产品之间的关系的信息。

此数据源提供了有关您作为记录卖家创建的所有产品优惠的信息。

当您在优惠中添加或删除产品时，您将创建优惠修订版。

每 24 小时刷新一次优惠产品数据源，因此每天都提供新数据。

下表说明了数据源列的名称和说明。有关数据源历史记录列的信息，请参阅[数据历史记录](data-feed-details.md#data-feed-historization)。


| 列名称  | 描述  | 
| --- | --- | 
| offer\$1id | 此优惠的友好标识符。可用于连接 `Offer` 数据源的 `offer_id` 字段。 | 
| offer\$1revision | 与 offer\$1id 字段组合使用来构成优惠修订版的外键。 | 
| product\$1id | 产品的友好标识符，这是此优惠公开的产品的外键。可用于连接 `Product` 数据源的 `product_id` 字段。 | 

## 优惠产品数据源示例
<a name="data-feed-offer-product-sample-data"></a>

下面显示了优惠产品数据源的示例。


| offer\$1id  | offer\$1revision | product\$1id | 
| --- | --- | --- | 
| offer-dacpxznflfwin | 10 | prod-o4grxfafcxxxx | 
| offer-gszhmle5npzip | 24 | prod-o4grxfafcxxxy | 

# 优惠目标数据源
<a name="data-feed-offer-target"></a>

此数据源为您作为记录卖家创建的所有优惠列出了优惠修订版的目标。如果单个优惠具有多个修订版，则所有修订版都包含在数据源中。

当您创建优惠修订版且公开字段中的数据发生更改时，会在数据源中为同一主键（`offer_id` 和 `offer_revision`）创建一条新记录，但 `valid_from` 字段的值不同。

优惠目标数据源每 24 小时刷新一次，因此每天提供的都是新数据。

下表说明了数据源列的名称和说明。


| 列名称  | 描述  | 
| --- | --- | 
| offer\$1target\$1id | 源的主键。 | 
| offer\$1id\$1offer\$1revision | 优惠的标识符和修订版。这两列引用了与此目标相关的优惠。可用于连接 `Target` 数据源的 `offer_id` 和 `offer_revision`字段。 | 
| target\$1type | 指示优惠接收者是 BuyerAccounts（这指示专属优惠）还是 ParticipatingPrograms。 | 
| polarity | 指示优惠是否打算面向 `target_type` 提供。可接受的值如下所示： [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/zh_cn/marketplace/latest/userguide/data-feed-offer-target.html)  | 
| 值 | 表示目标的字符串：AWS 账户 ID 或可与优惠一起使用的计划。例如，[AWS Marketplace 的标准合同 (SCMP)](standardized-license-terms.md#standard-contracts)、[AWS Marketplace 现场演示计划 (FDP)](field-demonstration-program.md)。 | 

## 优惠目标数据源示例
<a name="data-feed-offer-target-sample-data"></a>

下面显示了优惠目标数据源的示例。为了便于阅读，未显示数据历史记录列。有关数据历史记录字段的信息，请参阅[数据历史记录](data-feed-details.md#data-feed-historization)。


| offer\$1target\$1id  | offer\$1id  | offer\$1revision | target\$1type | polarity | value | 
| --- | --- | --- | --- | --- | --- | 
| 925ddc73f6a373b7d5544ea3210610803b600 | offer-dacpxznflfwin | 1 | ParticipatingPrograms | PositiveTargeting | EnterpriseContract | 
| 471ff22ae3165278f1fb960d3e14517bcd601 | offer-gszhmle5npzip | 1 | ParticipatingPrograms | PositiveTargeting | FieldDemonstration | 
| 511ff22adfj65278f1fb960d3e14517bcd6e602 | offer-gszhmle5npzip | 1 | ParticipatingPrograms  | PositiveTargeting | EnterpriseContract | 

# 产品数据源
<a name="data-feed-product"></a>

此数据源提供了有关您作为记录卖家创建的所有产品以及您有权转售的所有产品的信息。

产品数据是可变的。这意味着，当您更改以下某个字段的值时，将在数据源中创建一条新记录，其中包含不同的 `valid_from` 字段值。有关数据源历史记录列的更多信息，请参阅[数据历史记录](data-feed-details.md#data-feed-historization)。

每 24 小时刷新一次产品数据源，因此每天都提供新数据。

下表说明了数据源列的名称和说明。


| 列名称  | 描述  | 
| --- | --- | 
| product\$1id | 产品的友好标识符。可用于连接 `Account`、`Billing_Event` 和 `Offer_Product` 数据源的 `product_id` 字段。 | 
| manufacturer\$1account\$1id | 产品拥有者的标识符。这是[账户](data-feed-account.md)数据源的外键。可用于连接 `Account` 数据源的 `account_id` 字段。 | 
| product\$1code | 用于计量产品的现有授权产品代码。此值还用于将数据与报告联接，或引用 AWS Marketplace Metering Service 中提供的内容。 | 
| title | 产品的标题。 | 

## 产品数据源示例
<a name="data-feed-product-sample-data"></a>

下面显示了优惠目标数据源的示例。为了便于阅读，未显示数据历史记录列。有关数据历史记录字段的信息，请参阅[数据历史记录](data-feed-details.md#data-feed-historization)。


| product\$1id  | manufacturer\$1account\$1id  | product\$1code | title | 
| --- | --- | --- | --- | 
| prod-o4grxfafcxxxx | 555568000000 | product\$1code\$11 | Product1 | 
| prod-t3grxfafcxxxy | 444457000000 | product\$1code\$12 | Product2 | 
| prod-x8faxxfafcxxy | 666678000000 | product\$1code\$13 | Product3 | 

# 税项数据源
<a name="data-feed-tax-item"></a>

此数据源提供有关客户发票的税务计算的信息。

对于给定客户发票 (`invoice_id`) 的给定产品 (`product_id`)，可能有多个行项目 (`line_item_id`)，每个税务管辖区一个或多个。例如，对于使用不同 AWS 实体（例如美国和爱尔兰）使用不同 AWS 区域规则的客户而言，会发生这种情况。要详细了解在哪里对您的销售征 AWS 收销售税、增值税或商品及服务税，并以 AWS, Inc. 的名义将此类税款汇给当地税务机关，请参阅[亚马逊网络服务税务帮助](https://aws.amazon.com/tax-help/)。

税项数据源每 24 小时刷新一次，因此每天提供的都是新数据。

税项数据是不可变的。

下表说明了数据源列的名称和说明。有关数据历史记录列的信息，请参阅[数据历史记录](data-feed-details.md#data-feed-historization)。


| 列名称  | 说明  | 
| --- | --- | 
| tax\$1item\$1id | 税项记录的唯一标识符。 | 
| invoice\$1id | 发 AWS 票编号。您可以将此值与 product\$1id 值结合使用，以查找相关的税务记账事件。 | 
| line\$1item\$1id | 客户账单行项目的唯一标识符。退款交易具有与远期税务交易相同的行项目 ID。 | 
| customer\$1bill\$1id | 客户账单的唯一标识符。买家可与卖家共享此标识符，以帮助识别和解决税务计算问题。 | 
| tax\$1liable\$1party | `AWS` 或 `Seller`。如果卖家是税务责任方，则会征收税款。如果 AWS 是纳税方，则由 AWS其征收和汇出销售税。有关更多信息，请参阅 [AWS Marketplace 卖家和税务征收](https://aws.amazon.com/tax-help/marketplace)。 如果未征收税款，则此处不显示值。卖家需要确定是否针对每张发票征收某些税务，因为卖家有征税责任。 | 
| transaction\$1type\$1code | 交易的类型。可能值如下所示： [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/zh_cn/marketplace/latest/userguide/data-feed-tax-item.html) 退款交易具有与原始远期交易相同的行项目 ID。 | 
| product\$1id | 产品的外键。可用于连接 `product_id` 数据源的 `Product` 字段。 | 
| product\$1tax\$1code | 用于标识产品税费属性的标准代码。卖家在创建或修改产品时选择属性。 | 
| invoice\$1date | 创建发票的日期。 | 
| taxed\$1customer\$1account\$1id | 纳税的账户实体的外键。可用于连接 `account_id` 数据源的 `Account` 字段。 | 
| taxed\$1customer\$1country | 用于税务计算的地址的 ISO 3166 Alpha 2 国家/地区代码。 | 
| taxed\$1customer\$1state\$1or\$1region | 用于税务计算的州、地区或省。 | 
| taxed\$1customer\$1city | 用于税务计算的城市。 | 
| taxed\$1customer\$1postal\$1code | 用于税务计算的邮政编码。 | 
| tax\$1location\$1code\$1taxed\$1jurisdiction | 与纳税位置关联的 vertex 地理编码。 | 
| tax\$1type\$1code | 应用于交易的税费类型。可能的值为 None、Sales 和 SellerUse。 | 
| jurisdiction\$1level | 用于税务位置的地址的管辖区级别。可能的值为 State、County、City 和 District。 | 
| taxed\$1jurisdiction | 税务管辖区的名称。 | 
| display\$1price\$1taxability\$1type | 买家看到的价格是含税还是不含税。所有 AWS Marketplace 优惠均不含税。 | 
| taxable\$1amount | 在此管辖区级别应纳税的交易金额。 | 
| nontaxable\$1amount | 在此管辖区级别可免税的交易金额。 | 
| tax\$1jurisdiction\$1rate | 在此管辖区级别应用的税率。 | 
| tax\$1amount | 在此管辖区级别收取的税额。 | 
| tax\$1currency | 以上金额的 ISO 4217 Alpha 3 货币代码。 | 
| tax\$1calculation\$1reason\$1code | 交易是应税、非应税、免税还是零税率，按管辖区级别划分。 | 
| date\$1used\$1for\$1tax\$1calculation | 用于计算交易税费的日期。 | 
| customer\$1exemption\$1certificate\$1id | 免税证明的证明 ID。 | 
| customer\$1exemption\$1certificate\$1id\$1domain | 证书在 Amazon 系统中的存储位置。 | 
| customer\$1exemption\$1certificate\$1level | 提供免税的管辖区级别。 | 
| customer\$1exemption\$1code | 指定免税的代码，例如 RESALE。 | 
| customer\$1exemption\$1domain | 用于捕获客户免税信息（如果有）的 Amazon 系统。 | 
| transaction\$1reference\$1id | 允许您交叉引用以下报告中的数据的标识符： [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/zh_cn/marketplace/latest/userguide/data-feed-tax-item.html)  | 
| legal\$1amount | 需缴纳税款对应的法定货币金额。 | 
| legal\$1currency | 需缴纳税款对应的法定货币。 | 
| tax\$1invoice\$1id | 在双重开票的情况下，税目对应的税务发票标识符。对于统一开票场景，此字段将为空。 | 
| tax\$1invoice\$1date | 税务发票的开具日期。对于统一开票场景，此字段将为空。 | 

**注意**  
从 2021 年 8 月 5 日起，国际 Marketplace 服务商 AWS Marketplace 销售税将在纳税项目数据源中输入内容。这意味着，从 2021 年 8 月 5 日起，账单事件数据源中的每条 AWS\$1TAX\$1SHARE 和 SELLER\$1TAX\$1SHARE 记录都应在纳税项目数据源中都有相应的记录。

## 税项数据源示例
<a name="data-feed-tax-item-sample-data"></a>

下面显示了税项数据源的示例。在数据源中，此信息以单个表格形式显示。为了便于阅读，数据显示在此处的多个表中，并且不显示所有列。


| tax\$1item\$1id | invoice\$1id | line\$1item\$1id | customer\$1bill\$1id | 
| --- | --- | --- | --- | 
| 6p2ni6tu041xagvhbyanbgxl3xameha16txjoav\$10001 | 781216640 | 71000000000000000000 | 2210000000000000000 | 
| 6p2ni6tu041xagvhbyanbgxl3xameha16txjoav\$10002 | 781216640 | 53000000000000000000 | 2210000000000000000 | 
| flr4jobxjzww8czdsrq4noue2uxd56j39wxw0k7\$10001 | 250816266 | 76400000000000000000 | 5720000000000000000 | 
| gfkjjobxjzw56jgkrsrqgjtk52uxd56j39wgj567d\$10002 | 280336288 | 76400000000000000000 | 5724390000000000000 | 
| wwk1qpvb8ran3geiw8e3mp6dgs2qj7wpkuwhgk1\$10001 | 451431024 | 99300000000000000000 | 1230000000000000000 | 
| wwk1qpvb8ran3geiw8e3mp6dgs2qj7wpkuwhgk1\$10002 | 451431024 | 99300000000000000000 | 3120000000000000000 | 
| fnohdid8kwgqq9lvii2k30spn3ftgwihbe8h75x\$10001 | 229987654 | 92100000000000000000 | 6390000000000000000 | 




| tax\$1liable\$1party | transaction\$1type\$1code | product\$1id | product\$1tax\$1code | invoice\$1date | 
| --- | --- | --- | --- | --- | 
| Seller | AWS | prod-o4grxfafcxxxx | AWSMP\$1SOFTWARE\$1RA | 2018-12-31T00:00:00Z | 
| Seller | AWS | prod-o4grxfafcxxxx | AWSMP\$1SOFTWARE\$1RA | 2018-12-31T00:00:00Z | 
| Seller | AWS | prod-t3grxfafcxxxy | AWS\$1REMOTE\$1访问软件 | 2018-08-31T00:00:00Z | 
| Seller | REFUND | prod-t3grxfafcxxxy | AWS\$1REMOTE\$1访问软件 | 2018-08-31T00:00:00Z | 
| Seller | AWS | prod-x8faxxfafcxxy | AWS\$1REMOTE\$1访问软件 | 2018-08-31T00:00:00Z | 
| Seller | TAXONLYREFUND | prod-x8faxxfafcxxy | AWS\$1REMOTE\$1访问软件 | 2018-05-31T00:00:00Z | 
| AWS | AWS | prod-wghj8xfafrhgj | AWS\$1REMOTE\$1访问软件 | 2019-07-31T00:00:00Z | 




| taxed\$1customer\$1account\$1id | taxed\$1customer\$1country | taxed\$1customer\$1state\$1or\$1region | taxed\$1customer\$1city | taxed\$1customer\$1postal\$1code | 
| --- | --- | --- | --- | --- | 
| VIega2 XXCg GCGUree t9j3muxioh9wc8lsnd xriocm5 | 美国 | GA | MILTON | 48573-4839 | 
| VIega2 XXCg GCGUree t9j3muxioh9wc8lsnd xriocm5 | 美国 | GA | MILTON | 48573-4839 | 
| 7nyo5jw plyx81vx9ji04 o TRo 1ff8biqi88w8 eEwTur | 美国 | NC | DURHAM | 27517-4834 | 
| 7nyo5jw plyx81vx9ji04 o TRo 1ff8biqi88w8 eEwTur | 美国 | NC | DURHAM | 27517-4834 | 
| 7nyo5jw plyx81vx9ji04 o TRo 1ff8biqi88w8 eEwTur | 美国 | TX | NOT APPLICABLE | 75844-1235 | 
| 7nyo5jw plyx81vx9ji04 o TRo 1ff8biqi88w8 eEwTur | 美国 | TX | HOUSTON | 75844-1235 | 
| 192a0421313e41f069b52962ed7babf716291b688 | 美国 | CT | NEW HAVEN | 06002-2948 | 




| tax\$1location\$1code\$1taxed\$1jurisdiction | tax\$1type\$1code | jurisdiction\$1level | taxed\$1jurisdiction | display\$1price\$1taxability\$1type | taxable\$1amount | nontaxable\$1amount | 
| --- | --- | --- | --- | --- | --- | --- | 
| 460473664 | Sales | State | GA | 排除 | 100 | 0 | 
| 66301164 | Sales | County | FULTON | 排除 | 0 | 100 | 
| 692938178 | SellerUse | 州 | NC | 排除 | 58.1 | 523.8 | 
| 692938178 | SellerUse | 州 | NC | 排除 | -58.1 | 523.8 | 
| 356794387 | Sales | 州 | TX | Exclusive | 1105.14 | 0 | 
| 528887443 | Sales | City | HOUSTON | 排除 | -36 | 0 | 
| 171248162 | Sales | 州 | CT | 排除 | 0 | 114.55 | 




| tax\$1jurisdication\$1rate | tax\$1amount | tax\$1currency | tax\$1calculation\$1reason\$1code | date\$1used\$1for\$1tax\$1calculation | 
| --- | --- | --- | --- | --- | 
| 0.206 | 20.6 | USD | 应税 | 2018-10-31T00:00:00Z | 
| 0 | 0 | USD | NonTaxable | 2018-10-31T00:00:00Z | 
| 0.1 | 5.8 | USD | 应税 | 2018-07-31T00:00:00Z | 
| 0.1 | -5.8 | USD | 应税 | 2018-07-31T00:00:00Z | 
| 0.06 | 66.3 | USD | 应税 | 2018-07-31T00:00:00Z | 
| 0.01 | -0.36 | USD | NonTaxable | 2018-07-31T00:00:00Z | 
| 0 | 0 | USD | Exempt | 2019-06-30T00:00:00Z |