Add support for generic structured data files (json, yaml, toml)

[mferrera]

Among the many files and formats consumed and produced by different FMU components are yaml files. Within FMU yaml is the go-to medium for configuration. fmu-dataio is not currently able to upload these files as data.

One reason that yaml may have become the go-to configuration for FMU is its readability. These configuration files double as reference material when users are QCing inputs and results. Configuration files may be produced with varied values determined at runtime that differ between ensembles or realizations, and then are used as input to other steps within the experiment. Hence they are valid and important data.

Data type

Currently we do not have a clear class to categorize yaml files.

fmu-dataio/src/fmu/dataio/_model/enums.py

Lines 59 to 73 in 5ecfef9

	class FMUClass(str, Enum):
	"""The class of a data object by FMU convention or standards."""
	case = "case"
	realization = "realization"
	iteration = "iteration"
	surface = "surface"
	table = "table"
	cpgrid = "cpgrid"
	cpgrid_property = "cpgrid_property"
	polygons = "polygons"
	cube = "cube"
	well = "well"
	points = "points"
	dictionary = "dictionary"

Dictionary at first glance seems a viable candidate, but valid yaml will fail against it:

>>> yaml.load("""
... - Hesperiidae
... - Papilionidae
... - Apatelodidae
... - Epiplemidae
... """)

['Hesperiidae', 'Papilionidae', 'Apatelodidae', 'Epiplemidae']

The dictionary data class is made specifically to handle input that comes from a Python dict. Hence this will require another data class.

TODO

[mferrera]

A generic solution could be "serializable", that is, a Python object that is serializable into json. This is inclusive of dictionaries.