How does trpc know which invocations are on the same batch?

[ylc395]

According to doc:

After that, you can make use of batching by setting all your procedures in a Promise.all. The code below will produce exactly one HTTP request and on the server exactly one database query:

This feature is cool, I wonder how trpc knows some invocation are on the same batch? Does trpc rewrite the Promise.all?

[KATT]

We have a dataloder that waits one tick before doing any calls to the server (when you use the httpBatchLink) and groups everything into 1 req

[ylc395]

so all invocations in one tick will be batched?

[mnpenner]

The part I want to know about is this:

and on the server exactly one database query:

Can we see the implementation of byId? I'm assuming byId takes a single ID and it would have been typed as such in the .input method on the server, so how are these being combined into a single database query?

[theoludwig]

Yes. I don't think this statement is true.

It indeed produce exactly one HTTP request, but still nonetheless produce multiple database query.

[KATT]

With Prisma you get automatic batching for findUnique-queries that solves the so called N+1 problem - using dataloaders work for this too