Kotlin: Upload a file

• RLS policy permissions required:
  • buckets table permissions: none
  • objects table permissions: insert
• Refer to the Storage guide on how access control works
• Resumable uploads use a Disk cache by default to store the upload urls. You can customize that in the GoTrue config by changing the resumable.cache property.

Examples

Upload file

val bucket = supabase.storage["avatars"]
bucket.upload("myIcon.png", byteArray, upsert = false)
//on JVM you can use java.io.File
bucket.upload("myIcon.png", file, upsert = false)

Upload file with progress

val bucket = supabase.storage["avatars"]
bucket.uploadAsFlow("test.png", byteArrayOf()).collect \{
    when(it) \{
        is UploadStatus.Progress -> println("Progress: $\{it.totalBytesSend.toFloat() / it.contentLength * 100\}%")
        is UploadStatus.Success -> println("Success")
    \}
\}

Create resumable upload

val bucket = supabase.storage["avatars"]
//JVM/Android:
val upload = bucket.resumable.createOrContinueUpload("icon.png", File("icon.png"))
//Other platforms:
val upload = bucket.resumable.createOrContinueUpload(data = byteArray, source = "this is for continuing previous uploads later", path = "icon.png")
val upload = bucket.resumable.createOrContinueUpload( //Probably better to write an extension function
    channel = \{ offset -> /* create ByteReadChannel and seek to offset */ \},
    source = "this is for continuing previous uploads later",
    size = dataSize,
    path = "icon.png"
)

Start and resumable upload

upload.startOrResumeUploading()

Pause resumable upload

upload.pause()

Cancel resumable upload

upload.cancel()

Listen to the resumable upload state

upload.stateFlow.collect \{
    println("Progress: $\{it.progress * 100\}%")
    println("Paused: $\{it.paused\}")
    println("Is done: $\{it.isDone\}")
\}

Continue previous uploads

val bucket = supabase.storage["avatars"]

//only on JVM/Android:
bucket.resumable.continuePreviousFileUploads()
    .map \{ it.await() \} //await all uploads. This just makes sure the uploads have an update-to-date url. You can also do this in parallel
    .forEach \{ upload ->
        upload.startOrResumeUploading()
    \}

//on other platforms you may have to continue uploads from the source (Probably better to write an extension function):
bucket.resumable.continuePreviousUploads \{ source, offset ->
    //create ByteReadChannel from source and seek to offset
\}
    .map \{ it.await() \} //await all uploads. This just makes sure the uploads have an update-to-date url. You can also do this in parallel
    .forEach \{ upload ->
        upload.startOrResumeUploading()
    \}