using System;

using System.Diagnostics;

using System.IO;

using System.Threading.Tasks;

using Microsoft.Data.Sqlite;

namespace AsyncSample

{

class Program

{

static async Task Main()

{

var connection = new SqliteConnection("Data Source=AsyncSample.db");

connection.Open();

var createCommand = connection.CreateCommand();

createCommand.CommandText =

@"

CREATE TABLE data (

value BLOB

)

";

createCommand.ExecuteNonQuery();

// SQLite doesn't support asynchronous I/O. Instead, they recommend using a

// write -ahead log (WAL) which improves write performance. This sample

// demonstrates the anti-pattern of using ADO.NET's async methods with

// Microsoft.Data.Sqlite.

var insertCommand = connection.CreateCommand();

insertCommand.CommandText =

@"

INSERT INTO data

VALUES ($value)

";

Console.WriteLine("Generating 100 MB of data...");

var value = new byte[100_000_000];

var random = new Random();

random.NextBytes(value);

insertCommand.Parameters.AddWithValue("$value", value);

Console.WriteLine("Inserting data...");

var stopwatch = Stopwatch.StartNew();

var task = insertCommand.ExecuteNonQueryAsync();

Console.WriteLine($"Blocked for {stopwatch.ElapsedMilliseconds} ms");

stopwatch.Restart();

await task;

Console.WriteLine($"Yielded for {stopwatch.ElapsedMilliseconds} ms");

// Clean up

connection.Close();

File.Delete("AsyncSample.db");

}

}

}