I'm trying to use StreamReader.ReadAsync(Memory, CancellationToken) overload that accepts a Memory<T>
.
Here's the sample code: (.NET Core 2.1)
static async Task Main(string[] args)
{
string code = "Hello, World";
var memoryStream = Encoding.UTF8.GetBytes(code.ToCharArray(), 0, code.Length);
using (var stream = new MemoryStream(memoryStream))
using (var reader = new StreamReader(stream))
{
// var text = await reader.ReadLineAsync();
// Console.WriteLine(text); -> Outputs: Hello, World
Memory<char> memory = new Memory<char>();
await reader.ReadAsync(memory);
Console.WriteLine("Memory<char> length:" + memory.Length);
}
}
Code tries to read string content through StreamReader
and populate Memory<char>
. However, code keeps outputting 0 for Memory<char>
length.
What's the proper way to use this method to populate Memory<char>
?
My .csproj
file content:
<Project Sdk="Microsoft.NET.Sdk">
<PropertyGroup>
<OutputType>Exe</OutputType>
<TargetFramework>netcoreapp2.1</TargetFramework>
<LangVersion>7.2</LangVersion>
</PropertyGroup>
</Project>
The working solution is to initialize memory with a buffer and to care the return value of ReadAsync so you know how many characters (not bytes) you did read.
You did run into the issue that Memory is a struct which always allows you to create an instance with no arguments although from an API usage perspective this is not allowed.
using System;
using System.IO;
using System.Text;
using System.Threading.Tasks;
namespace ReadMemory
{
class Program
{
public static async Task Main(string[] args)
{
string code = "Hello, World";
var memoryStream = Encoding.UTF8.GetBytes(code.ToCharArray(), 0, code.Length);
using (var stream = new MemoryStream(memoryStream))
using (var reader = new StreamReader(stream))
{
Memory<char> memory = new Memory<char>(new char[1024]); // Init with backing buffer. Otherwise you are trying to read 0 bytes into a zero sized buffer.
int charsRead = await reader.ReadAsync(memory);
Console.WriteLine($"Memory<char> len: {memory.Length}, bytes read: {charsRead}");
}
}
}
}
To answer you second question. Yes you can use the ReadAsync(char[] ...) instead of memory. They are equivalent since under the hood it is passed further as System.Memory anyway which is then when it comes to the actual reading converted to a Span anyway. See
public virtual System.Threading.Tasks.Task<int> ReadAsync(char[] buffer, int index, int count)
{
return this.ReadAsyncInternal(new Memory<char>(buffer, index, count), default(System.Threading.CancellationToken)).AsTask();
}
If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With