API Reference

ChunkLayoutWrapper(parent::Py)

A wrapper around a Python tensorstore.ChunkLayout object. Describes how data is chunked on disk.

ContextSpecWrapper(parent::Py)

A wrapper around a Python tensorstore.Context.Spec object. Represents the configuration for a Context.

ContextWrapper(parent::Py)

A wrapper around a Python tensorstore.Context object. Used for sharing resources (like cache pools) between multiple TensorStore handles.

FutureWrapper{Result}(parent::Py)

A wrapper around a Python tensorstore.Future object. Result is the Julia type that the future will resolve to when .result() is called.

IndexDomainWrapper(parent::Py)

A wrapper around a Python tensorstore.IndexDomain object. Represents the coordinate system and labels of a TensorStore.

SchemaWrapper(parent::Py)

A wrapper around a Python tensorstore.Schema object. Describes the data type, domain, and layout of a TensorStore.

SpecWrapper(parent::Py)

A wrapper around a Python tensorstore.Spec object. Represents a serializable specification of a TensorStore.

TensorStoreWrapper(parent::Py)

A wrapper around a Python tensorstore.TensorStore object. Implements much of the Julia AbstractArray interface.

TransactionWrapper(parent::Py)

A wrapper around a Python tensorstore.Transaction object. Used for atomic multi-write operations.

abort(txn::TransactionWrapper)

Abort the transaction, discarding any changes.

commit_async(txn::TransactionWrapper) -> FutureWrapper{ReturnsNothing}

Start committing the transaction asynchronously.

commit_sync(txn::TransactionWrapper)

Commit the transaction synchronously and wait for completion.

context(spec::Dict) -> ContextWrapper
context() -> ContextWrapper

Create a TensorStore context for shared resources like cache pools.

label(w::TensorStoreWrapper, labels...) -> TensorStoreWrapper

Assign new labels to the dimensions of the TensorStore.

labels(w::IndexDomainWrapper) -> Vector{String}

Return the labels of the dimensions in the index domain.

open(spec; kwargs...) -> FutureWrapper{TensorStoreWrapper}

Open a TensorStore from a configuration dictionary or Spec. Returns a FutureWrapper that resolves to a TensorStoreWrapper.

Common keyword arguments:

• transaction: A TransactionWrapper to use for this open operation.
• context: A ContextWrapper or dictionary for shared resources.
• create: Boolean, whether to create the store if it doesn't exist.
• open: Boolean, whether to open an existing store.
• delete_existing: Boolean, whether to delete an existing store.

result(w::FutureWrapper{Result}) -> Result

Wait for the future to complete and return the result, converted to the Julia type Result.

transaction(f::Function)

Run a function f(txn) within a transaction. The transaction is automatically committed if f returns normally, and aborted if an exception is thrown.

transaction() -> TransactionWrapper

Create a new TensorStore transaction.

translate_by(w::TensorStoreWrapper, offsets...) -> TensorStoreWrapper

Shift the coordinate system of the TensorStore by the given relative offsets. Offsets are relative to the current coordinate system.

translate_to(w::TensorStoreWrapper, coords...) -> TensorStoreWrapper

Shift the coordinate system of the TensorStore so that its origin starts at the specified coordinates. Coordinates are 1-based Julia indices.