sqlwhat
enables you to write Submission Correctness Tests (SCTs) for interactive SQL exercises on DataCamp.
- If you are new to teaching on DataCamp, check out https://authoring.datacamp.com.
- If you want to learn what SCTs are and how they work, visit this article specifically.
- For more information about writing SCTs for SQL exercises, consult https://sqlwhat.readthedocs.io.
pip install sqlwhat # install from pypi
make install # install from source
- API Docs: https://sqlwhat.readthedocs.io
- AST viewer: https://ast-viewer.datacamp.com
- Extensions: https://github.com/datacamp/sqlwhat-ext
Please raise an issue on the respsective parser repo:
from sqlwhat.State import State # State holds info needed for tests
from sqlwhat.Reporter import Reporter
from sqlwhat.checks import * # imports all SCTs
from sqlalchemy import create_engine
code = "SELECT * FROM artists WHERE id < 100"
state = State(
student_code = code,
solution_code = code,
pre_exercise_code = "",
student_conn = create_engine('sqlite:///'),
solution_conn = create_engine('sqlite:///'),
student_result = {'id': [1,2,3], 'name': ['greg', 'jon', 'martha']},
solution_result = {'id': [1,2,3], 'name': ['toby', 'keith', 'deb']},
reporter = Reporter()
)
# test below passes, since code is equal for student and solution
has_equal_ast(state)
# test below raises a TestFail error, since 'name' col of results
# doesn't match between student and solution results
check_result(state)
# shows error data
state.reporter.build_payload()
# can also be done using a chain
from sqlwhat.sct_syntax import Ex
Ex(state).check_result()
pytest -m "not backend"
If you also want to run the backend tests, you need to set a GITHUB_TOKEN
environment variable with access to the (private) sqlbackend
repository.
After this, you can:
make install
pytest
- Running queries is the backend's job
- If a test doesn't run queries, it doesn't need the backend
- Very few tests should run queries
Install sqlwhat and run ..
cd docs
make html