UnifiedRead#

class gwpy.io.registry.UnifiedRead(instance: object, cls: type[~gwpy.io.registry.T], registry: ~gwpy.io.registry.UnifiedIORegistry = <gwpy.io.registry.UnifiedIORegistry object>)[source]#

Bases: UnifiedReadWrite, ABC, Generic[T]

Base Class.read() implementation that handles parallel reads.

Each parallel read must return an instance of the same type as the input class, and the results are merged by the provided merge_function which should have the following signature:

def merge_function(self: Type, instances: list[Type], **kwargs) -> Type

i.e take in the type object and a list of instances, and return a single instance of the same type.

Methods Summary

__call__(source, *args[, format, cache, ...])

Execute cls.read().

merge(items, **kwargs)

Merge a sequence of items into a single item.

Methods Documentation

__call__(
source: NamedReadable | list[NamedReadable],
*args,
format: str | None = None,
cache: bool | None = None,
parallel: int = 1,
verbose: str | bool = False,
merge_function: MergeProtocol | None = None,
**kwargs,
) → T[source]#

Execute cls.read().

This method generalises parallel reading of lists of files for any input class.

This method also generalises downloading files from remote URLs over HTTP or via fsspec.

abstractmethod merge(items: Sequence[T], **kwargs) → T[source]#

Merge a sequence of items into a single item.

Parameters:
itemsSequence[T]

The sequence of items to merge.

**kwargs

Additional keyword arguments specific to the merge operation. Subclasses may override this method with specific named parameters; use # type: ignore[override] to suppress mypy warnings about signature compatibility.

Returns:
T

The merged item.

Notes

The base class automatically extracts merge-specific kwargs from the __call__ method by inspecting the merge method signature, so subclasses with specific named parameters will receive them correctly.