split_file references output format that doesn't exist

[ptgolden]

In this method:

koza/src/koza/cli_utils.py

Lines 130 to 143 in fceafe5

	def split_file(file: str,
	fields: str,
	format: OutputFormat = OutputFormat.tsv,
	remove_prefixes: bool = False,
	output_dir: str = "./output"):
	db = duckdb.connect(":memory:")
	#todo: validate that each of the fields is actually a column in the file
	if format == OutputFormat.tsv:
	read_file = f"read_csv('{file}')"
	elif format == OutputFormat.json:
	read_file = f"read_json('{file}')"
	else:
	raise ValueError(f"Format {format} not supported")

split_lines checks if the output format is OutputFormat.json, but no such output format exists:

koza/src/koza/model/config/source_config.py

Lines 77 to 84 in fceafe5

	class OutputFormat(str, Enum):
	"""
	Output formats
	"""
	tsv = "tsv"
	jsonl = "jsonl"
	kgx = "kgx"

Does duckdb have a jsonlines reader?

[ptgolden]

It has since 0.7.0. I'll change to OutputFormat.jsonl and make sure tests pass.

[kevinschaper]

thanks for catching this and cleaning it up!