Dapper.FSharp.Pipeline

Experiments with Dapper's F# idiomatic API

What does the code look like?

F# Async style

Use CommandDef pipeline API

[<CLIMutable>]
type Customer = { Id: int; Name: string }

use cts = new CancellationTokenSource()

let coldTask =
    async {
        use connection = // ...
        do! connection |> DbConnection.openAsync

        return!
            "select Id, Name from Customer where ID = @ID"
            |> CommandDef.create
            |> CommandDef.withParameters {| ID = 42 |}
            |> CommandDef.querySingleAsync<Customer> connection
    }

Async.RunSynchronously(coldTask, cancellationToken = cts.Token)

.NET Task style

Use ordinary (but overloaded) IDbConnection extension methods

[<CLIMutable>]
type Customer = { Id: int; Name: string }

use cts = new CancellationTokenSource()

let hotTask =
    task {
        use connection = // ...
        do! connection.OpenAsync(cts.Token)

        return!
            "select Id, Name from Customer where ID = @ID"
            |> CommandDef.create
            |> CommandDef.withParameters {| ID = 42 |}
            |> CommandDef.withCancellationToken cts.Token
            |> connection.QuerySingleAsync<Customer>
    }

hotTask.Result

Naming convention

• Extension methods of IDbConnection return Task<'T>
  • Overload with the same name as the extension method of Dapper
• Pipeline API functions return Async<'T>
  • Add Async to the function name
  • Naming in camelCase to avoid confusion with APIs that return Task<'T>