Skip to content

Asdf

DEFAULT_ASDF_DATA_OBJECT_TYPES module-attribute 

DEFAULT_ASDF_DATA_OBJECT_TYPES = (astropy.table.Table, asdf.tags.core.NDArrayType, np.ndarray, pd.DataFrame, pyarrow.lib.Table)

Types we by default recognize as 'data' in an ASDF file.

asdfopen_generic 

asdfopen_generic(key: str, bucket: Bucket | str | None = None, blocksize: int = DEFAULT_CHUNK_SIZE) -> asdf._asdf.AsdfFile

Open an S3 object or local file as an AsdfFile.

Parameters:

  • key (str) –

    path or S3 key

  • bucket (Bucket | str | None, default: None ) –

    Bucket object or bucket name if this is an S3 object; None if a local file

  • blocksize (int, default: DEFAULT_CHUNK_SIZE ) –

    block size for read operation, relevant only for S3

Returns: an AsdfFile object created from key

extract_objects 

extract_objects(asdf_file: AsdfFile, datatypes: tuple[type] = DEFAULT_ASDF_DATA_OBJECT_TYPES, *, autoload: bool = False, use_full_paths: bool = True) -> tuple[dict[str | bool | int, Any] | dict[tuple[str | bool | int, ...], Any], dict[str | bool | int, Any] | dict[tuple[str | bool | int, ...], Any]]

Search for and return the "data" objects in an AsdfFile, as defined by DEFAULT_ASDF_DATA_OBJECT_TYPE.

Returns a tuple like: ({path: object}, {path: schema description or None if no description}) for each identified object.

If use_full_paths is True, uses full object paths as keys; otherwise, uses last element of path (and logs a warning if any duplicated keys are discarded).

sanitize_data_tag 

sanitize_data_tag(node: Any, *, autoload: bool = False) -> Any

Get the array from a lazy NDArrayType node; otherwise return the node.

sanitize_schema_descr 

sanitize_schema_descr(descr: dict) -> str | dict | None

Pull the "real" description from an ASDF schema description.

unnest_to_pathkeys 

unnest_to_pathkeys(nested: Mapping) -> dict[tuple[Hashable, ...], Any]

Flatten a mapping to a list like

[ {(key level 1, key level 2, ...): value}, ... ]