Backend API. Server-side code: this runs in a Codename One backend, not in the app on the device.
public final class AzureBlobStorage
- Object
- AzureBlobStorage
ImplementsObjectStorage
Azure Blob Storage, through its REST API, with no Azure library. A bucket here is a container and an object a block blob.
Two ways to authenticate:
- An account key (Shared Key), from the key itself or a connection string. Presigned URLs are service SAS tokens computed here.
- A managed identity (Microsoft Entra ID), on an Azure VM, App Service, Container Apps or Functions, with no secret on the machine. The identity needs a data-plane role such as “Storage Blob Data Contributor”. Presigned URLs are then user-delegation SAS tokens, which costs one call to fetch a delegation key a day.
forConnectionString with UseDevelopmentStorage=true talks to a local
Azurite.
Nested types
class AzureBlobStorage.Factory | Opens azblob:// URLs for ObjectStorages: azblob://<account>. |
Methods
Inherited methods
Method details
forAccountKey
public static AzureBlobStorage forAccountKey(String account, String key)account, authenticated with its base64 access key.forManagedIdentity
public static AzureBlobStorage forManagedIdentity(String account, String clientId)account, authenticated as the machine’s managed identity.
clientId picks a user-assigned identity; null takes the system one.forConnectionString
public static AzureBlobStorage forConnectionString(String connectionString)
throws IOExceptionUseDevelopmentStorage=true. It must carry an AccountKey; a
connection string holding only a SAS is not supported.Throws
IOException- when the string names no account or key
withEndpoint
public AzureBlobStorage withEndpoint(String endpoint)https://<account>.blob.core.chinacloudapi.cn), a private endpoint, or an
emulator, whose URL then includes the account path.getName
public String getName()s3 (Amazon S3 and every S3-compatible
endpoint, R2 included), gcs, azure or memory.createBucket
public void createBucket(String bucket)
throws IOExceptionCreates a bucket, and says nothing when this account already owns it.
Usually infrastructure’s job, but a first run against an emulator or a test fixture needs it.
Throws
putObject
public String putObject(String bucket, String key, byte[] content, String contentType)
throws IOExceptionUploads an object, replacing one of the same key. Returns its ETag, the provider’s receipt for exactly these bytes.
contentType may be null, which stores application/octet-stream.
Throws
getObject
public byte[] getObject(String bucket, String key)
throws IOExceptionThrows
headObject
public ObjectInfo headObject(String bucket, String key)
throws IOExceptionThrows
deleteObject
public void deleteObject(String bucket, String key)
throws IOExceptionThrows
listObjects
public List<String> listObjects(String bucket, String prefix, int max)
throws IOExceptionmax keys under prefix (null or empty for all), following the
provider’s pagination until the listing ends or max is reached. Zero
answers an empty list; a negative max is refused.Throws
presignGet
public String presignGet(String bucket, String key, int seconds)
throws IOExceptionA URL that downloads the object with no credentials, for seconds.
Computed, not requested, wherever the provider allows it: the URL can be handed straight to a client. A lifetime the provider refuses – zero or less, or past seven days on S3, R2 and GCS – throws an IOException on every provider, so moving between them does not change how it fails.
Throws
presignPut
public String presignPut(String bucket, String key, int seconds)
throws IOExceptionThe upload counterpart: a URL a client can PUT the bytes to, for
seconds.
Azure needs one header on that PUT that the others do not:
x-ms-blob-type: BlockBlob. A client that serves more than one provider
can send it everywhere; S3 and GCS ignore it.