xdas.io.Engine

class xdas.io.Engine(vtype=None, ctype=None)

Base class for file format handlers in xdas.

The Engine class provides a plugin architecture for reading and writing various file formats. Each Engine subclass corresponds to a specific file format (e.g., “xdas”, “asn”, “miniseed”) and implements methods to open and save DataArray or DataCollection objects.

Engines are registered in a class-level registry using the __init_subclass__ hook, allowing them to be accessed by name using the Engine[name] syntax. Aliases can also be defined for backwards compatibility or convenience.

Parameters:

• vtype (str, optional) – The virtualization type to use. If vtype is None, the first supported type is used.

• ctype (str or dict, optional) – The coordinate type(s) to use. Can be: - None: uses the first supported ctype for each component - str: uses the same ctype for all components - dict: maps component names to their specific ctypes If None or incomplete, missing ctypes default to the first supported option.

vtype

The version type for this engine instance.

Type:

str

ctype

The component type(s) for this engine instance.

Type:

str or dict

Notes

Subclasses should define class attributes: - _supported_vtypes (list): List of supported virtualization types - _supported_ctypes (dict): Maps component names to lists of supported coordinate types

Examples

Subclass registration (automatic via __init_subclass__):

>>> class NetCDFEngine(Engine, name="netcdf", aliases=["nc"]):
...     _supported_vtypes = ["hdf5"]
...     _supported_ctypes = {
...         "time": ["sampled", "dense"], "distance": ["sampled", "dense"]
...     }
...     def open_dataarray(self, fname, **kwargs):
...         ...

Access registered engines:

>>> engine = Engine["netcdf"](vtype="hdf5")
>>> engine = Engine["nc"](ctype="dense")  # Using alias

__init__(vtype=None, ctype=None)

Methods

__init__([vtype, ctype])
open_dataarray(fname, **kwargs)	Open fname and return a DataArray (abstract).
open_datacollection(fname, **kwargs)	Open fname and return a DataCollection (abstract).
save_dataarray(da, fname, **kwargs)	Write da to fname (abstract).
save_datacollection(dc, fname, **kwargs)	Write dc to fname (abstract).