MassiveDotNet.Rest 0.4.1

MassiveDotNet.Rest

REST client for the Massive market data platform (formerly Polygon.io), for .NET 10.

dotnet add package MassiveDotNet.Rest

Quick start

using MassiveDotNet;
using MassiveDotNet.Rest;
using MassiveDotNet.Rest.Models;

// One client per application. It is thread-safe, owns a pooled HttpClient, and is meant to be
// long-lived: creating one per request exhausts sockets under load.
using MassiveRestClient client = new(Environment.GetEnvironmentVariable("MASSIVE_API_KEY")!);

await foreach (TickerSummary ticker in client.Reference.EnumerateTickersAsync(
    market: MarketType.Stocks,
    active: true,
    limit: 100))
{
    Console.WriteLine($"{ticker.Ticker,-8} {ticker.Name}");
}

Every snippet above is lifted from samples/MassiveDotNet.Quickstart, which the build compiles, so a signature change breaks that project rather than silently rotting this page.

The shape of it

Endpoints hang off groups that mirror the platform's own taxonomy, and groups are readonly struct over the shared transport, so navigating to one allocates nothing:

client.Stocks.ListAggregatesAsync(...)
client.Reference.ListTickersAsync(...)

Every paginated endpoint gets two methods, following the BCL's own Directory.EnumerateFiles/GetFiles distinction:

• ListXxxAsync returns one page and reports whether more exist.
• EnumerateXxxAsync returns IAsyncEnumerable<T> and walks every page, one in flight at a time, retaining no memory proportional to the pages traversed.

Comparator parameters collapse to one filter-typed argument per field rather than six flat ones, so an unsupported comparator is a compile error instead of a silently dropped query string.

Why it is built this way

The whole REST surface is generated from Massive's own OpenAPI description, and the generated code is committed — so the surface tracks the platform rather than drifting from it, and every change to it arrives as a reviewable diff. Deprecated operations ship marked [Obsolete] with the replacement named; unreleased ones ship marked [Experimental]. Nothing is silently omitted.

Native AOT clean, source-generated JSON only, and deliberately allocation-conscious: a 50,000-row page deserializes into one array rather than 50,000 objects.

Licensed MIT. Issues and source at github.com/jerbersoft/massivedotnet.