AWS AppSync GraphQL - 使用联合作为突变返回导致错误

Question

我正在使用 AWS AppSync 创建一个 GraphQL 架构，我想Union用作突变返回。我想这样写突变：

mutation addUpdateTariff($tariff: TariffInput!, $seasonalTimeTariff: [SeasonalTimeTariffInput!]) {
  addUpdateTariff(tariff: $tariff, seasonalTimeTariff: $seasonalTimeTariff) {
    id
    type
    values {
      ... on SteppedTariff {
        endDate
      }
      ... on SeasonalTimeTariff {
        endDate
        peakConsumption
      }
    }
  }
}

但我收到此错误：

Request failed with status code 400

该字段values可以是SteppedTariff或类型SeasonalTimeTariff，取决于addUpdateTariff突变的输入。据我搜索，Union仅用于查询，我没有找到一些说明它不能以不同方式使用的文档。

我错过了什么还是我真的不能用Union这种方式？

架构：

type Tariff {
  id: ID!
  type: TariffType!
  values: [TariffValue!]
}

type SteppedTariff {
  endDate: AWSDate
}

type SeasonalTimeTariff {
  endDate: AWSDate
  peakConsumption: Float
}

union TariffValue = SeasonalTimeTariff | SteppedTariff

测试关税值的不同模式：

1. values: [String]

询问：

mutation addUpdateTariff($tariff: TariffInput!, $seasonalTimeTariff: [SeasonalTimeTariffInput!]) {
  addUpdateTariff(tariff: $tariff, seasonalTimeTariff: $seasonalTimeTariff) {
    id
    type
    values
  }
}

回复：

{
  "data": {
    "addUpdateTariff": {
      "id": "843eadcf-48bd-4d58-93ec-8bbe96db3635",
      "type": "SeasonalTime",
      "values": [
        "{endDate=2019-02-02}"
      ]
    }
  }
}

2. values: [SeasonalTimeTariff]

询问：

mutation addUpdateTariff($tariff: TariffInput!, $seasonalTimeTariff: [SeasonalTimeTariffInput!]) {
  addUpdateTariff(tariff: $tariff, seasonalTimeTariff: $seasonalTimeTariff) {
    id
    type
    values {
      endDate
    }
  }
}

回复：

{
  "data": {
    "addUpdateTariff": {
      "id": "7f77c5d9-2b06-4bb8-a678-10996addc4e1",
      "type": "SeasonalTime",
      "values": [
        {
          "endDate": "2019-02-02"
        }
      ]
    }
  }
}

Answer 1

您绝对可以在突变的返回类型上使用联合。确保您__typename在值中包含该字段。您的解析器模板应返回如下内容：

{
  "id": 1,
  "type": "?",
  "values": [
    {
      "__typename": "StepTariff",
      "endDate": "2019-01-01"
    },
    {
      "__typename": "SeasonalTimeTariff",
      "endDate": "2019-01-01",
      "peakConsumption": 1.0
    }
  ]
}