Cache | blurkit

When to use

Use cache integration when your job may encode the same source multiple times.

Example

import { createFilesystemCache, encode } from 'blurkit/node'

const cache = createFilesystemCache({
  dir: './.cache/blurkit',
  ttlMs: 120_000,
})

const result = await encode('./public/hero.jpg', { cache })

Inputs / Options / Behavior

interface BlurKitCache {
  get(key: string): Promise<BlurResult | undefined> | BlurResult | undefined
  set(key: string, value: BlurResult): Promise<void> | void
}

Built-in helpers:

createMemoryCache({ max? }) from blurkit/node and blurkit/deno
createFilesystemCache({ dir, ttlMs? }) from blurkit/node and blurkit/deno
createCloudflareCache({ name?, ttlSeconds? }) from blurkit/cloudflare

Limits / Caveats

Cache keys are content-aware: they include a digest of the resolved image bytes plus every output-affecting option, so a changed file, Blob, or remote URL produces a new entry instead of stale output.
Memory cache is process-local and reset on restart.
Filesystem cache is local-disk scoped.
Cloudflare cache depends on Worker Cache API availability and scope.

Next read