section-underline-after-name (D408)
Derived from the pydocstyle linter.
Fix is always available.
What it does
Checks for section underlines in docstrings that are not on the line immediately following the section name.
Why is this bad?
This rule enforces a consistent style for multiline numpy-style docstrings, and helps prevent incorrect syntax in docstrings using reStructuredText.
Multiline numpy-style docstrings are typically composed of a summary line, followed by a blank line, followed by a series of sections. Each section has a header and a body. There should be a series of underline characters in the line immediately below the header.
This rule enforces a consistent style for multiline numpy-style docstrings with sections. If your docstring uses reStructuredText, the rule also helps protect against incorrect reStructuredText syntax, which would cause errors if you tried to use a tool such as Sphinx to generate documentation from the docstring.
This rule is enabled when using the numpy convention, and disabled
when using the google or pep257 conventions.
Example
def calculate_speed(distance: float, time: float) -> float:
"""Calculate speed as distance divided by time.
Parameters
----------
distance : float
Distance traveled.
time : float
Time spent traveling.
Returns
-------
float
Speed as distance divided by time.
Raises
------
FasterThanLightError
If speed is greater than the speed of light.
"""
try:
return distance / time
except ZeroDivisionError as exc:
raise FasterThanLightError from exc
Use instead:
def calculate_speed(distance: float, time: float) -> float:
"""Calculate speed as distance divided by time.
Parameters
----------
distance : float
Distance traveled.
time : float
Time spent traveling.
Returns
-------
float
Speed as distance divided by time.
Raises
------
FasterThanLightError
If speed is greater than the speed of light.
"""
try:
return distance / time
except ZeroDivisionError as exc:
raise FasterThanLightError from exc