Using Bun s3 client with different buckets

[EvHaus]

I'm looking at migrating to Bun's S3 API, but one thing is bothering me. Looks like Bun's S3 API requires that you specify the Bucket as part of the client configuration, whereas S3's official API has you specify it for each operation. In my app I work across several various buckets. Is the expectation that I create a new Bun S3Client for each operation? That feels very inefficient and verbose to me.

With @aws-sdk/client-s3

import { DeleteObjectCommand, S3Client } from '@aws-sdk/client-s3';

// Setup the client once
const s3 = new S3Client({
	credentials: {
		accessKeyId: ACCESS_KEY_ID,
		secretAccessKey: SECRET_KEY,
	},
	region: AWS_REGION,
});

// Now I can delete from various different buckets
s3.send(new DeleteObjectCommand({
	Bucket: 'Bucket A',
	Key: 'SomeKey',
}));
s3.send(new DeleteObjectCommand({
	Bucket: 'Bucket B',
	Key: 'SomeOtherKey',
}));

With Bun's S3Client

import { S3Client } from "bun";

// Have to create a separate S3Client instance for each bucket I'm operating on
const s3ClientA = new Bun.S3Client({
	accessKeyId: ACCESS_KEY_ID,
	secretAccessKey: SECRET_KEY,
	region: AWS_REGION,
	bucket: 'Bucket A',
});

s3ClientA.delete('SomeKey')

const s3ClientB = new Bun.S3Client({
	accessKeyId: ACCESS_KEY_ID,
	secretAccessKey: SECRET_KEY,
	region: AWS_REGION,
	bucket: 'Bucket B',
});

s3ClientB.delete('SomeOtherKey')

Just confirming -- is this expected behaviour and a deliberate API choice?

[kravetsone]

Mostly u don't need dynamic buckets

But there https://bun.sh/docs/api/s3#s3client-static-methods
Static options like u wants