Support for Batching

[michaelstaib]

[
  {
    query: `
      mutation ($input: TokenizeCreditCardInput!) {
        tokenizeCreditCard(input: $input) {
          paymentMethod {
            id @export(as: "id")
          }
        }
      }
    `,
    variables: { input: "..." }
  },
  {
    query: `
      mutation ($id: ID, $transaction: TransactionInput!) {
        chargePaymentMethod(input: { id: $id, transaction: $transaction }) {
          transaction {
            id
            status
          }
        }
      }
    `,
    variables: { transaction: "..." }
  },
]

https://github.com/apollographql/apollo-link/tree/master/packages/apollo-link-batch-http
https://blog.apollographql.com/batching-client-graphql-queries-a685f5bcd41b
https://sangria-graphql.org/learn/#batch-executor

[michaelstaib]

graphql/graphql-spec#377

[michaelstaib]

In order to support this we should introduce a new IBatchQueryExecutor that yields always a response stream on execute.

public interface IBatchQueryExecutor
{
    Task<IResponseStream> ExecuteAsync(IReadOnlyBatchQueryRequest request, CancellationToken cancellationToken);
}

[michaelstaib]

@export(as: "foo")

The export directive can export a field result as variable under the same name as the field name or under a name specified by the as argument.

The exported variables are pushed to a list.

When a query is executed the batch executer will analyze if the query has all variables declared, if there are variables collected from a previous query that are not declared then the variable will be inferred from the argument/field in which they are used.