Zaps Loader

The Zaps Loader fetches zap receipt events for inspecting payments and senders. It wraps the Tag Value Loader and returns finite, partial results rather than a live or complete total.

How to use it

Create the loader once and subscribe with a target event:

import { createZapsLoader } from "applesauce-loaders/loaders";

const loader = createZapsLoader(pool, { eventStore });
const subscription = loader(target, ["wss://relay.example.com"]).subscribe((event) =>
  console.log(event.pubkey, event.content),
);

subscription.unsubscribe();

The loader queries kind 9735 using e for regular events and a for replaceable or addressable events. Relay hints include the target's seen relays by default; set useSeenRelays: false to use only explicit hints and configured extra relays.

It accepts the Tag Value Loader options except kinds, which is fixed. Each wrapper has separate internal e and a loaders, so concurrency budgets apply independently to each. Pointer force is available through the generic tag value API, not this wrapper's call signature.

Integration

Pass the application's EventStore to update its models and casts. An optional cacheRequest runs in parallel with relay requests and does not prevent network loading.

Use the zap helpers to parse and validate receipts before displaying payment details. The loader requests kind 9735 events; it does not itself validate invoices or prove payment authenticity.

Use an abort signal to stop both internal loaders:

const controller = new AbortController();
const loader = createZapsLoader(pool, {
  eventStore,
  signal: controller.signal,
  onError: (error, source) => console.warn("Partial zaps load", source, error),
});

controller.abort();

The count service handles supported shared feed metrics, but has no built-in zap count or amount total. Use this loader when receipt details are required, and treat aggregates over its results as partial.

Best Practices

• Reuse one wrapper for related views to preserve batching and compatible request sharing.
• Display “No events loaded” rather than interpreting an empty result as proof of absence.
• Report partial source failures with onError while keeping received events visible.
• Use signal for deterministic teardown; the wrapper does not expose stop().
• Keep the default deadlines or choose finite values appropriate to the view.