Allows run flake8 on code snippets in docstrings or RST files.
idea proposed by Mike Bayer on zzzeek/sqlalchemy#362
That said, if there was some alternate form of "doctest" that could simply test a code example both for Python syntax, pep8 compliance (which would be AWESOME) as well as symbol consistency, that would be helpful. The tool could be configured with common imports and symbols significant to SQLAlchemy examples and be helpful as a basic sanity check for code examples. As it is, when writing new documentation I have to organize and run the code in a separate .py file to make sure it does the right thing. So this is a problem, just my experience with doctest in writing the tutorials has shown me what it's good at and where it's likely getting in the way.
Realization inspired by https://github.com/asottile/blacken-docs
You can install tool from pip pip install flake8-rst
.
Tool search sourcecode
, code-block
and ipython
blocks, crop and run flake8 on it:
.. sourcecode:: python
class Example(Base):
pass
or
.. code-block:: python
class Example(Base):
pass
Supporting all flake8 arguments and flags except jobs (temporary), with additional one:
flake8-rst --bootstrap "import test"
flake8-rst bootstraps code snippets with this code, useful for fix import errors.
Load configuration from [flake8-rst]
ini sections, like flake8.
In order to use custom roles of flake8-rst
in documentation with Sphinx
, extend sphinx with flake8_rst.sphinxext.custom_roles
in conf.py
.
The roles have no effect on the generated documentation.
extensions = [...,
'flake8_rst.sphinxext.custom_roles'
]
role | example | |
---|---|---|
:flake8-group: |
Blocks with same group are combined to one. | :flake8-group: Group1 |
Blocks with group None are checked individual. |
:flake8-group: None |
|
Blocks with group Ignore are not checked. |
:flake8-group: Ignore |
|
:flake8-set-ignore: |
Overwrites ignore list for current block. | :flake8-set-ignore: F821, E999 |
:flake8-add-ignore: |
Adds arguments to ignore list for current block. | :flake8-add-ignore: E999 |
:flake8-set-select: |
Overwrites select list for current block. | :flake8-set-select: E, F |
:flake8-add-select: |
Adds arguments to select list for current block. | :flake8-add-select: C404 |
:flake8-bootstrap: |
Overwrites --bootstrap for current block |
:flake8-bootstrap: import os; import sys |
Keep in mind:
- Roles added to blocks within the same group (except group
None
) have no effect unless they appear in the first block. - provided bootstrap-code will get split by
;
into individual lines. E999 SyntaxError: invalid syntax
causesflake8
to skipAST
tests. Keep mandatoryE999
issues in blocks with:flake8-group: Ignore
to preserve full testing for the rest of the blocks.
You can specify default groupnames for all directives individually:
flake8-rst --default-groupnames '<file-pattern>-><directive>: <groupname>'
file-pattern
and directive
are matched by Unix filename pattern matching
in the order of appearance.
The default is *.rst->*: default
, so all blocks in *.rst
files are merged, in
other files they stay individual.
But it's also possible to merge only ipython
directives in *.rst
files and leave other directives
treated individually: "*.rst->ipython: default"
Examples:
flake8-rst --default-groupnames "*.rst->*: default"
[flake8-rst]
default-groupnames =
*.rst-*: default
*.py-code-block: default
Disconnected blocks don't know previous defined names:
.. code-block:: python
class Example(Base):
pass
.. code-block:: python
import datetime
obj = Example(datetime.datetime.now()) # F821 undefined name 'Example'
Once blocks are connected, different issues are found:
.. code-block:: python
:flake8-group: ExampleGroup
class Example(Base):
pass
.. code-block:: python
:flake8-group: ExampleGroup
import datetime # E402 module level import not at top of file
obj = Example(datetime.datetime.now())
If appropriate, issues can be ignored for a specific group:
.. code-block:: python
:flake8-group: ExampleGroup1
:flake8-set-ignore: E402
class Example(Base):
pass
.. code-block:: python
:flake8-group: ExampleGroup1
import datetime
obj = Example(datetime.datetime.now())
.. code-block:: python
:flake8-group: ExampleGroup2
class Example(Base):
pass
.. code-block:: python
:flake8-group: ExampleGroup2
:flake8-set-ignore: E402 # no effect, because it's not defined in first
# block of ExampleGroup2
import datetime # E402 module level import not at top of file
obj = Example(datetime.datetime.now())
d.kataev:flake8-rst§ flake8-rst --filename="*.py *.rst" tests/data/* --bootstrap="from sqlalchemy import Table, Column, Sequence, Integer, ForeignKey, String, DateTime"
tests/data/test.py:14:42: F821 undefined name 'metadata'
tests/data/test.py:15:13: E128 continuation line under-indented for visual indent
tests/data/test.py:16:28: F821 undefined name 'JSONB'
tests/data/test.py:19:14: F821 undefined name 'engine'
tests/data/test.py:22:21: E251 unexpected spaces around keyword / parameter equals
tests/data/test.py:22:23: E251 unexpected spaces around keyword / parameter equals
tests/data/test.rst:27:48: F821 undefined name 'metadata'
tests/data/test.rst:41:22: F821 undefined name 'meta'
tests/data/test.rst:56:52: F821 undefined name 'meta'
tests/data/test.rst:57:32: F821 undefined name 'meta'
tests/data/test.rst:69:20: F821 undefined name 'Base'
tests/data/test.rst:72:56: F821 undefined name 'Base'