Refactor query execution into reusable core module

[dbolser]

Summary

• introduce execute_query utility in new core module
• refactor CLI and interactive shell to use the shared executor
• add unit tests for the new helper and update existing tests

Testing

• uv run pytest -q

---

https://chatgpt.com/codex/tasks/task_e_6864ed54821c83279d0d0854639887d0

[cursor]

Bug: JSON Output Format Not Supported

The execute_query function does not support "json" output format. Although the CLI still accepts "json" as a valid option, passing it causes the function to fall through to the else block, resulting in incorrect tab-separated output instead of proper JSON. This is a regression, as the original code included explicit json.dumps() handling that was removed during refactoring.

pksql/core.py#L37-L51

pksql/pksql/core.py

Lines 37 to 51 in 1b90c4c

	output_lines = []
	if output_format == "table":
	if is_query:
	output_lines.append(str(result))
	else:
	output_lines.append("Query executed successfully.")
	else:
	delimiter = "," if output_format == "csv" else "\t"
	if is_query:
	header = delimiter.join(result.columns)
	output_lines.append(header)
	for row in result.fetchall():
	output_lines.append(delimiter.join(map(str, row)))
	else:
	output_lines.append("Query executed successfully.")

Fix in Cursor • Fix in Web

---

Was this report helpful? Give feedback by reacting with 👍 or 👎