Soenneker.Blazor.ApiClient 11.0.3306

Soenneker.Blazor.ApiClient

A scoped Blazor API client for JSON requests, authenticated session tokens, optional browser-console logging, and multipart uploads.

Installation and registration

dotnet add package Soenneker.Blazor.ApiClient

using Soenneker.Blazor.ApiClient.Registrars;

builder.Services.AddApiClientAsScoped();

The registrar also adds the session, JSON logging, and shared HttpClient cache dependencies.

Initialize the scoped client

Initialize() must run before any request:

using Soenneker.Blazor.ApiClient.Abstract;

public sealed class WeatherApi
{
    private readonly IApiClient _api;

    public WeatherApi(IApiClient api, IConfiguration configuration)
    {
        _api = api;
        _api.Initialize(
            configuration["Api:BaseUrl"]!,
            requestResponseLogging: false);
    }
}

The base address must be absolute and HTTPS. Plain HTTP is accepted only for loopback development addresses.

Authenticated JSON request

public async Task<WeatherForecast?> GetForecast(
    CancellationToken cancellationToken)
{
    using HttpResponseMessage response = await _api.Get(
        "weather/forecast",
        cancellationToken: cancellationToken);

    response.EnsureSuccessStatusCode();
    return await response.Content.ReadFromJsonAsync<WeatherForecast>(
        cancellationToken);
}

Authenticated wrapper methods request the current access token from ISessionUtil and attach it as a bearer header to that request. Token headers are reused only while the token string remains unchanged.

Use allowAnonymous: true for a public endpoint:

using HttpResponseMessage response = await _api.Get(
    "health",
    allowAnonymous: true,
    cancellationToken: cancellationToken);

Authenticated requests cannot use an absolute URL on a different scheme, host, or port than the configured base address. This prevents forwarding a session token to another origin. Anonymous requests may use absolute URLs.

Request options

using Soenneker.Blazor.ApiClient.Dtos;

using HttpResponseMessage response = await _api.Post(
    new RequestOptions
    {
        Uri = "orders",
        Object = new { productId, quantity = 1 },
        LogRequest = true,
        LogResponse = false
    },
    cancellationToken);

LogRequest and LogResponse take effect only when requestResponseLogging was enabled during initialization. Logging may expose request or response data in the browser console; keep it disabled for sensitive production traffic.

File upload

await using FileStream stream = File.OpenRead(path);

using HttpResponseMessage response = await _api.Upload(
    new RequestUploadOptions
    {
        Uri = "documents",
        Stream = stream,
        FileName = Path.GetFileName(path),
        Object = new { category = "invoice" },
        LogRequest = true
    },
    cancellationToken);

Uploads are always authenticated. The multipart names are file for the binary stream and json for optional serialized metadata. The upload disposes its multipart content, which also disposes the supplied stream; do not reuse it afterward.

All methods return the raw HttpResponseMessage. The caller owns that response and is responsible for disposal, status handling, and response deserialization. GetClient() returns the cached transport client but does not attach authorization automatically; prefer the wrapper methods for authenticated calls.