scipy-doctest 1.6__tar.gz → 1.7.1__tar.gz

{scipy_doctest-1.6 → scipy_doctest-1.7.1}/PKG-INFO

@@ -1,6 +1,6 @@
-Metadata-Version: 2.3
+Metadata-Version: 2.4
 Name: scipy_doctest
-Version: 1.6
+Version: 1.7.1
 Summary: Configurable, whitespace-insensitive, floating-point-aware doctest helpers.
 Maintainer-email: SciPy developers <scipy-dev@python.org>
 Requires-Python: >=3.8
@@ -10,20 +10,36 @@ Classifier: License :: OSI Approved :: BSD License
 Classifier: Intended Audience :: Developers
 Classifier: Programming Language :: Python :: 3
 Classifier: Framework :: Pytest
+License-File: LICENSE
 Requires-Dist: numpy>=1.19.5
 Requires-Dist: pytest
-Requires-Dist: scipy ; extra == "test"
+Requires-Dist: furo==2024.8.6 ; extra == "doc"
+Requires-Dist: myst-parser==4.0.0 ; extra == "doc"
+Requires-Dist: sphinx==8.1.3 ; extra == "doc"
+Requires-Dist: sphinx-copybutton==0.5.2 ; extra == "doc"
+Requires-Dist: scipy <= 1.14.1 ; extra == "test"
 Requires-Dist: matplotlib ; extra == "test"
 Project-URL: Home, https://github.com/scipy/scipy_doctest
+Provides-Extra: doc
 Provides-Extra: test
 
 # Floating-point aware, human readable, numpy-compatible doctesting.
 
+[![PyPI version][pypi-version]][pypi-link]
+[![Conda-Forge][conda-badge]][conda-link]
+
+<!-- prettier-ignore-start -->
+[conda-badge]:              https://img.shields.io/conda/vn/conda-forge/scipy-doctest
+[conda-link]:               https://anaconda.org/conda-forge/scipy-doctest
+[pypi-link]:                https://pypi.org/project/scipy-doctest/
+[pypi-version]:             https://img.shields.io/pypi/v/scipy-doctest
+<!-- prettier-ignore-end -->
+
 ## TL;DR
 
 This project extends the standard library `doctest` module to allow flexibility
 and easy customization of finding, parsing and checking code examples in
-documentation. 
+documentation.
 
 Can be used either as drop-in `doctest` replacement or through the `pytest`
 integration. Uses a floating-point aware doctest checker by default.
@@ -31,7 +47,7 @@ integration. Uses a floating-point aware doctest checker by default.
 ## Motivation and scope
 
 Having examples in the documentation is great. Having wrong examples in the
-documentation is not that great however. 
+documentation is not that great however.
 
 The standard library `doctest` module is great for making sure that docstring
 examples are correct. However, the `doctest` module is limited in several
@@ -46,7 +62,7 @@ This looks reasonably clear but does not work, in three different ways.
 _First_, `1/3` is not equal to 0.333 because floating-point arithmetic.
 _Second_, `numpy` adds whitespace to its output, this whitespace confuses the
 `doctest`, which is whitespace-sensitive. Therefore, we added a magic directive,
- `+SKIP` to avoid a doctest error. _Third_, the example is actually
+`+SKIP` to avoid a doctest error. _Third_, the example is actually
 wrong---notice `0.669` which is not equal to `2/3` to three sig figs. The error
 went unnoticed by the doctester also because of the `+SKIP` directive.
 
@@ -56,44 +72,44 @@ a human reader, and should not be present in the documentation.
 This package defines modified doctesting routines which fix these deficiencies.
 Its main features are
 
-- *Doctesting is floating-point aware.* In a nutshell, the core check is
+- _Doctesting is floating-point aware._ In a nutshell, the core check is
   `np.allclose(want, got, atol=..., rtol=...)`, with user-controllable abs
   and relative tolerances. In the example above (_sans_ `# doctest: +SKIP`),
   `want` is the desired output, `array([0.333, 0.669, 1])` and `got` is the
   actual output from numpy: `array([0.33333333, 0.66666667, 1.        ])`.
 
-- *Human-readable skip markers.* Consider
+- _Human-readable skip markers._ Consider
   ```
   >>> np.random.randint(100)
   42     # may vary
   ```
-Note that the markers (by default, `"# may vary"` and `"# random"`) can be applied
-to either an example's output, or its source.
+  Note that the markers (by default, `"# may vary"` and `"# random"`) can be applied
+  to either an example's output, or its source.
 
 Also note a difference with respect to the standard `# doctest: +SKIP`: the latter
 skips the example entirely, while these additional markers only skip checking
 the output. Thus the example source needs to be valid python code still.
 
-- A user-configurable list of *stopwords*. If an example contains a stopword,
+- A user-configurable list of _stopwords_. If an example contains a stopword,
   it is checked to be valid python, but the output is not checked. This can
   be useful e.g. for not littering the documentation with the output of
   `import matplotlib.pyplot as plt; plt.xlim([2.3, 4.5])`.
 
-- A user-configurable list of *pseudocode* markers. If an example contains one
+- A user-configurable list of _pseudocode_ markers. If an example contains one
   of these markers, it is considered pseudocode and is not checked.
   This is useful for `from example import some_functions` and similar stanzas.
 
 - A `# doctest: +SKIPBLOCK` option flag to skip whole blocks of pseudocode. Here
   a 'block' is a sequence of doctest examples without any intervening text.
 
-- *Doctest discovery* is somewhat more flexible then the standard library
+- _Doctest discovery_ is somewhat more flexible then the standard library
   `doctest` module. Specifically, one can use `testmod(module, strategy='api')`
   to only examine public objects of a module. This is helpful for complex
   packages, with non-trivial internal file structure. Alternatively, the default
   value of `strategy=None` is equivalent to the standard `doctest` module
   behavior.
 
-- *User configuration*. Essentially all aspects of the behavior are user
+- _User configuration_. Essentially all aspects of the behavior are user
   configurable via a `DTConfig` instance attributes. See the `DTConfig`
   docstring for details.
 
@@ -124,7 +140,7 @@ pip install scipy-doctest
 
 2. **Register or load the plugin**
 
-Next, you need to register or load the pytest plugin within your test module or `conftest.py` file. 
+Next, you need to register or load the pytest plugin within your test module or `conftest.py` file.
 
 To do this, add the following line of code:
 
@@ -136,14 +152,16 @@ pytest_plugins = "scipy_doctest"
 
 Check out the [pytest documentation](https://docs.pytest.org/en/stable/how-to/writing_plugins.html#requiring-loading-plugins-in-a-test-module-or-conftest-file) for more information on requiring/loading plugins in a test module or `conftest.py` file.
 
-3. **Run doctests** 
+3. **Run doctests**
 
 Once the plugin is registered, run the doctests by executing the following command:
 
 ```bash
 $ python -m pytest --doctest-modules
 ```
+
 or
+
 ```bash
 $ pytest --pyargs <your-package> --doctest-modules
 ```
@@ -155,10 +173,9 @@ use the command flag
 $ pytest --pyargs <your-package> --doctest-modules --doctest-collect=api
 ```
 
-See [More fine-grained control](https://github.com/scipy/scipy_doctest#More-fine-grained-control) section
+See [More fine-grained control](#more-fine-grained-control) section
 for details on how to customize the behavior.
 
-
 ### Basic usage
 
 The use of `pytest` is optional, and you can use the `doctest` layer API.
@@ -171,6 +188,7 @@ For example,
 >>> res
 TestResults(failed=0, attempted=764)
 ```
+
 The second return value, `hist` is a dict which maps the names of the objects
 to the numbers of failures and attempts for individual examples.
 
@@ -178,10 +196,10 @@ For more details, see the `testmod` docstring. Other useful functions are
 `find_doctests`, `run_docstring_examples` and `testfile` (the latter two mimic
 the behavior of the eponymous functions of the `doctest` module).
 
-
 ### Command-line interface
 
 There is a basic CLI, which also mimics that of the `doctest` module:
+
 ```
 $ python -m scipy_doctest foo.py
 ```
@@ -190,28 +208,30 @@ Note that, just like `$ python -m doctest foo.py`, this may
 fail if `foo.py` is a part of a package due to package imports.
 
 Text files can also be CLI-checked:
+
 ```
 $ python -m scipy_doctest bar.rst
 ```
 
 Notice that the command-line usage only uses the default `DTConfig` settings.
 
+(more-fine-grained-control)=
 
 ## More fine-grained control
 
 More fine-grained control of the functionality is available via the following
 classes
 
-|   Class     |  `doctest` analog  |
-|-------------|--------------------|
-| `DTChecker` | `DocTestChecker`   |
-| `DTParser`  | `DocTestParser`    |
-| `DTRunner`  | `DocTestRunner`    |
-| `DTFinder`  | `DocTestFinder`    |
-| `DTContext` |       --           |
+| Class       | `doctest` analog |
+| ----------- | ---------------- |
+| `DTChecker` | `DocTestChecker` |
+| `DTParser`  | `DocTestParser`  |
+| `DTRunner`  | `DocTestRunner`  |
+| `DTFinder`  | `DocTestFinder`  |
+| `DTContext` | --               |
 
 The `DTContext` class is just a bag class which holds various configuration
-settings as attributes.  An instance of this class is passed around, so user
+settings as attributes. An instance of this class is passed around, so user
 configuration is simply creating an instance, overriding an attribute and
 passing the instance to `testmod` or constructors of `DT*` objects. Defaults
 are provided, based on a long-term usage in SciPy.
@@ -220,7 +240,7 @@ See the [DTConfig docstring](https://github.com/scipy/scipy_doctest/blob/main/sc
 for the full set of attributes that allow you to fine-tune your doctesting experience.
 
 To set any of these attributes, create an instance of `DTConfig` and assign the attributes
-in a usual way. 
+in a usual way.
 
 If using the pytest plugin, it is convenient to use the default instance, which
 is predefined in `scipy_doctest/conftest.py`. This instance will be automatically
@@ -260,14 +280,12 @@ dt_config.skiplist = {
 
 If you don't set these attributes, the [default settings](https://github.com/scipy/scipy_doctest/blob/58ff06a837b7bff1dbac6560013fc6fd07952ae2/scipy_doctest/impl.py#L94) of the attributes are used.
 
-
 #### Alternative Checkers
 
 By default, we use the floating-point aware `DTChecker`. If you want to use an
 alternative checker, all you need to do is to define the corresponding class,
 and add an attribute to the `DTConfig` instance. For example,
 
-
 ```
 class VanillaOutputChecker(doctest.OutputChecker):
     """doctest.OutputChecker to drop in for DTChecker.
@@ -290,10 +308,8 @@ See [a pytest example](https://github.com/scipy/scipy_doctest/blob/main/scipy_do
 and [a doctest example](https://github.com/scipy/scipy_doctest/blob/main/scipy_doctest/tests/test_runner.py#L94)
 for more details.
 
-
 ### NumPy and SciPy wrappers
 
-
 NumPy wraps `scipy-doctest` with the `spin` command
 
 ```
@@ -307,13 +323,11 @@ $ python dev.py smoke-docs    # check docstrings
 $ python dev.py smoke-tutorials   # ReST user guide tutorials
 ```
 
-
-
 ## Rough edges and sharp bits
 
 Here is a (non-exhaustive) list of possible gotchas:
 
-- *In-place development builds*.
+- _In-place development builds_.
 
 Some tools (looking at you `meson-python`) simulate in-place builds with a
 `build-install` directory. If this directory is located under the project root,
@@ -329,12 +343,15 @@ $ pytest build-install/lib/python3.10/site-packages/scipy/ --doctest-modules
 
 instead of `$ pytest --pyargs scipy`.
 
-If push comes to shove, you may try using the magic env variable:
+If you use actual editable installs, of the `pip install --no-build-isolation -e .` variety, you may
+need to add `--import-mode=importlib` to the `pytest` invocation.
+
+If push really comes to shove, you may try using the magic env variable:
 ` PY_IGNORE_IMPORTMISMATCH=1 pytest ...`,
 however the need usually indicates an issue with the package itself.
 (see [gh-107](https://github.com/scipy/scipy_doctest/pull/107) for an example).
 
-- *Optional dependencies are not that optional*
+- _Optional dependencies are not that optional_
 
 If your package contains optional dependencies, doctests do not know about them
 being optional. So you either guard the imports in doctests (yikes!), or
@@ -353,11 +370,11 @@ Note that installed packages are no different:
 $ pytest --pyargs scipy --doctest-modules --ignore=/path/to/installed/scipy/_lib
 ```
 
-- *Doctest collection strategies*
+- _Doctest collection strategies_
 
 The default collection strategy follows `doctest` module and `pytest`. This leads
 to duplicates if your package has the split between public and \_private modules,
-where  public modules re-export things from private ones. The solution is to
+where public modules re-export things from private ones. The solution is to
 use `$ pytest --doctest-collect=api` CLI switch: with this, only public
 objects will be collected.
 
@@ -375,8 +392,7 @@ leads to
 - `scipy.linalg._basic.det`, collected from `scipy/linalg/_basic.py`, is private.
 - `scipy.linalg.det`, collected from `scipy/linalg/__init__.py`, is public.
 
-
-- *`pytest`'s assertion rewriting*
+- _`pytest`'s assertion rewriting_
 
 In some rare cases you may need to either explicitly register the `scipy_doctest`
 package with the `pytest` assertion rewriting machinery, or ask it to avoid rewriting
@@ -388,7 +404,6 @@ In general, rewriting assertions is not very useful for doctests, as the
 output on error is fixed by the doctest machinery anyway. Therefore, we believe
 adding `--assert=plain` is reasonable.
 
-
 ## Prior art and related work
 
 - `pytest` provides some limited floating-point aware `NumericLiteralChecker`.
@@ -409,15 +424,14 @@ adding `--assert=plain` is reasonable.
   to be not easy to reason about, work with, and extend to other projects.
 
   This project is mainly the core functionality of the modified
-  `refguide-check` doctesting, extracted to a separate package. 
+  `refguide-check` doctesting, extracted to a separate package.
   We believe having it separate simplifies both addressing the needs of these
   two packages, and potential adoption by other projects.
 
-
 ### Bug reports, feature requests and contributions
 
 This package is work in progress. Contributions are most welcome!
 Please don't hesitate to open an issue in the tracker or send a pull request.
 
-The current location of the issue tracker is https://github.com/scipy/scipy_doctest.
+The current location of the issue tracker is <https://github.com/scipy/scipy_doctest>.
 

{scipy_doctest-1.6 → scipy_doctest-1.7.1}/README.md

@@ -1,10 +1,20 @@
 # Floating-point aware, human readable, numpy-compatible doctesting.
 
+[![PyPI version][pypi-version]][pypi-link]
+[![Conda-Forge][conda-badge]][conda-link]
+
+<!-- prettier-ignore-start -->
+[conda-badge]:              https://img.shields.io/conda/vn/conda-forge/scipy-doctest
+[conda-link]:               https://anaconda.org/conda-forge/scipy-doctest
+[pypi-link]:                https://pypi.org/project/scipy-doctest/
+[pypi-version]:             https://img.shields.io/pypi/v/scipy-doctest
+<!-- prettier-ignore-end -->
+
 ## TL;DR
 
 This project extends the standard library `doctest` module to allow flexibility
 and easy customization of finding, parsing and checking code examples in
-documentation. 
+documentation.
 
 Can be used either as drop-in `doctest` replacement or through the `pytest`
 integration. Uses a floating-point aware doctest checker by default.
@@ -12,7 +22,7 @@ integration. Uses a floating-point aware doctest checker by default.
 ## Motivation and scope
 
 Having examples in the documentation is great. Having wrong examples in the
-documentation is not that great however. 
+documentation is not that great however.
 
 The standard library `doctest` module is great for making sure that docstring
 examples are correct. However, the `doctest` module is limited in several
@@ -27,7 +37,7 @@ This looks reasonably clear but does not work, in three different ways.
 _First_, `1/3` is not equal to 0.333 because floating-point arithmetic.
 _Second_, `numpy` adds whitespace to its output, this whitespace confuses the
 `doctest`, which is whitespace-sensitive. Therefore, we added a magic directive,
- `+SKIP` to avoid a doctest error. _Third_, the example is actually
+`+SKIP` to avoid a doctest error. _Third_, the example is actually
 wrong---notice `0.669` which is not equal to `2/3` to three sig figs. The error
 went unnoticed by the doctester also because of the `+SKIP` directive.
 
@@ -37,44 +47,44 @@ a human reader, and should not be present in the documentation.
 This package defines modified doctesting routines which fix these deficiencies.
 Its main features are
 
-- *Doctesting is floating-point aware.* In a nutshell, the core check is
+- _Doctesting is floating-point aware._ In a nutshell, the core check is
   `np.allclose(want, got, atol=..., rtol=...)`, with user-controllable abs
   and relative tolerances. In the example above (_sans_ `# doctest: +SKIP`),
   `want` is the desired output, `array([0.333, 0.669, 1])` and `got` is the
   actual output from numpy: `array([0.33333333, 0.66666667, 1.        ])`.
 
-- *Human-readable skip markers.* Consider
+- _Human-readable skip markers._ Consider
   ```
   >>> np.random.randint(100)
   42     # may vary
   ```
-Note that the markers (by default, `"# may vary"` and `"# random"`) can be applied
-to either an example's output, or its source.
+  Note that the markers (by default, `"# may vary"` and `"# random"`) can be applied
+  to either an example's output, or its source.
 
 Also note a difference with respect to the standard `# doctest: +SKIP`: the latter
 skips the example entirely, while these additional markers only skip checking
 the output. Thus the example source needs to be valid python code still.
 
-- A user-configurable list of *stopwords*. If an example contains a stopword,
+- A user-configurable list of _stopwords_. If an example contains a stopword,
   it is checked to be valid python, but the output is not checked. This can
   be useful e.g. for not littering the documentation with the output of
   `import matplotlib.pyplot as plt; plt.xlim([2.3, 4.5])`.
 
-- A user-configurable list of *pseudocode* markers. If an example contains one
+- A user-configurable list of _pseudocode_ markers. If an example contains one
   of these markers, it is considered pseudocode and is not checked.
   This is useful for `from example import some_functions` and similar stanzas.
 
 - A `# doctest: +SKIPBLOCK` option flag to skip whole blocks of pseudocode. Here
   a 'block' is a sequence of doctest examples without any intervening text.
 
-- *Doctest discovery* is somewhat more flexible then the standard library
+- _Doctest discovery_ is somewhat more flexible then the standard library
   `doctest` module. Specifically, one can use `testmod(module, strategy='api')`
   to only examine public objects of a module. This is helpful for complex
   packages, with non-trivial internal file structure. Alternatively, the default
   value of `strategy=None` is equivalent to the standard `doctest` module
   behavior.
 
-- *User configuration*. Essentially all aspects of the behavior are user
+- _User configuration_. Essentially all aspects of the behavior are user
   configurable via a `DTConfig` instance attributes. See the `DTConfig`
   docstring for details.
 
@@ -105,7 +115,7 @@ pip install scipy-doctest
 
 2. **Register or load the plugin**
 
-Next, you need to register or load the pytest plugin within your test module or `conftest.py` file. 
+Next, you need to register or load the pytest plugin within your test module or `conftest.py` file.
 
 To do this, add the following line of code:
 
@@ -117,14 +127,16 @@ pytest_plugins = "scipy_doctest"
 
 Check out the [pytest documentation](https://docs.pytest.org/en/stable/how-to/writing_plugins.html#requiring-loading-plugins-in-a-test-module-or-conftest-file) for more information on requiring/loading plugins in a test module or `conftest.py` file.
 
-3. **Run doctests** 
+3. **Run doctests**
 
 Once the plugin is registered, run the doctests by executing the following command:
 
 ```bash
 $ python -m pytest --doctest-modules
 ```
+
 or
+
 ```bash
 $ pytest --pyargs <your-package> --doctest-modules
 ```
@@ -136,10 +148,9 @@ use the command flag
 $ pytest --pyargs <your-package> --doctest-modules --doctest-collect=api
 ```
 
-See [More fine-grained control](https://github.com/scipy/scipy_doctest#More-fine-grained-control) section
+See [More fine-grained control](#more-fine-grained-control) section
 for details on how to customize the behavior.
 
-
 ### Basic usage
 
 The use of `pytest` is optional, and you can use the `doctest` layer API.
@@ -152,6 +163,7 @@ For example,
 >>> res
 TestResults(failed=0, attempted=764)
 ```
+
 The second return value, `hist` is a dict which maps the names of the objects
 to the numbers of failures and attempts for individual examples.
 
@@ -159,10 +171,10 @@ For more details, see the `testmod` docstring. Other useful functions are
 `find_doctests`, `run_docstring_examples` and `testfile` (the latter two mimic
 the behavior of the eponymous functions of the `doctest` module).
 
-
 ### Command-line interface
 
 There is a basic CLI, which also mimics that of the `doctest` module:
+
 ```
 $ python -m scipy_doctest foo.py
 ```
@@ -171,28 +183,30 @@ Note that, just like `$ python -m doctest foo.py`, this may
 fail if `foo.py` is a part of a package due to package imports.
 
 Text files can also be CLI-checked:
+
 ```
 $ python -m scipy_doctest bar.rst
 ```
 
 Notice that the command-line usage only uses the default `DTConfig` settings.
 
+(more-fine-grained-control)=
 
 ## More fine-grained control
 
 More fine-grained control of the functionality is available via the following
 classes
 
-|   Class     |  `doctest` analog  |
-|-------------|--------------------|
-| `DTChecker` | `DocTestChecker`   |
-| `DTParser`  | `DocTestParser`    |
-| `DTRunner`  | `DocTestRunner`    |
-| `DTFinder`  | `DocTestFinder`    |
-| `DTContext` |       --           |
+| Class       | `doctest` analog |
+| ----------- | ---------------- |
+| `DTChecker` | `DocTestChecker` |
+| `DTParser`  | `DocTestParser`  |
+| `DTRunner`  | `DocTestRunner`  |
+| `DTFinder`  | `DocTestFinder`  |
+| `DTContext` | --               |
 
 The `DTContext` class is just a bag class which holds various configuration
-settings as attributes.  An instance of this class is passed around, so user
+settings as attributes. An instance of this class is passed around, so user
 configuration is simply creating an instance, overriding an attribute and
 passing the instance to `testmod` or constructors of `DT*` objects. Defaults
 are provided, based on a long-term usage in SciPy.
@@ -201,7 +215,7 @@ See the [DTConfig docstring](https://github.com/scipy/scipy_doctest/blob/main/sc
 for the full set of attributes that allow you to fine-tune your doctesting experience.
 
 To set any of these attributes, create an instance of `DTConfig` and assign the attributes
-in a usual way. 
+in a usual way.
 
 If using the pytest plugin, it is convenient to use the default instance, which
 is predefined in `scipy_doctest/conftest.py`. This instance will be automatically
@@ -241,14 +255,12 @@ dt_config.skiplist = {
 
 If you don't set these attributes, the [default settings](https://github.com/scipy/scipy_doctest/blob/58ff06a837b7bff1dbac6560013fc6fd07952ae2/scipy_doctest/impl.py#L94) of the attributes are used.
 
-
 #### Alternative Checkers
 
 By default, we use the floating-point aware `DTChecker`. If you want to use an
 alternative checker, all you need to do is to define the corresponding class,
 and add an attribute to the `DTConfig` instance. For example,
 
-
 ```
 class VanillaOutputChecker(doctest.OutputChecker):
     """doctest.OutputChecker to drop in for DTChecker.
@@ -271,10 +283,8 @@ See [a pytest example](https://github.com/scipy/scipy_doctest/blob/main/scipy_do
 and [a doctest example](https://github.com/scipy/scipy_doctest/blob/main/scipy_doctest/tests/test_runner.py#L94)
 for more details.
 
-
 ### NumPy and SciPy wrappers
 
-
 NumPy wraps `scipy-doctest` with the `spin` command
 
 ```
@@ -288,13 +298,11 @@ $ python dev.py smoke-docs    # check docstrings
 $ python dev.py smoke-tutorials   # ReST user guide tutorials
 ```
 
-
-
 ## Rough edges and sharp bits
 
 Here is a (non-exhaustive) list of possible gotchas:
 
-- *In-place development builds*.
+- _In-place development builds_.
 
 Some tools (looking at you `meson-python`) simulate in-place builds with a
 `build-install` directory. If this directory is located under the project root,
@@ -310,12 +318,15 @@ $ pytest build-install/lib/python3.10/site-packages/scipy/ --doctest-modules
 
 instead of `$ pytest --pyargs scipy`.
 
-If push comes to shove, you may try using the magic env variable:
+If you use actual editable installs, of the `pip install --no-build-isolation -e .` variety, you may
+need to add `--import-mode=importlib` to the `pytest` invocation.
+
+If push really comes to shove, you may try using the magic env variable:
 ` PY_IGNORE_IMPORTMISMATCH=1 pytest ...`,
 however the need usually indicates an issue with the package itself.
 (see [gh-107](https://github.com/scipy/scipy_doctest/pull/107) for an example).
 
-- *Optional dependencies are not that optional*
+- _Optional dependencies are not that optional_
 
 If your package contains optional dependencies, doctests do not know about them
 being optional. So you either guard the imports in doctests (yikes!), or
@@ -334,11 +345,11 @@ Note that installed packages are no different:
 $ pytest --pyargs scipy --doctest-modules --ignore=/path/to/installed/scipy/_lib
 ```
 
-- *Doctest collection strategies*
+- _Doctest collection strategies_
 
 The default collection strategy follows `doctest` module and `pytest`. This leads
 to duplicates if your package has the split between public and \_private modules,
-where  public modules re-export things from private ones. The solution is to
+where public modules re-export things from private ones. The solution is to
 use `$ pytest --doctest-collect=api` CLI switch: with this, only public
 objects will be collected.
 
@@ -356,8 +367,7 @@ leads to
 - `scipy.linalg._basic.det`, collected from `scipy/linalg/_basic.py`, is private.
 - `scipy.linalg.det`, collected from `scipy/linalg/__init__.py`, is public.
 
-
-- *`pytest`'s assertion rewriting*
+- _`pytest`'s assertion rewriting_
 
 In some rare cases you may need to either explicitly register the `scipy_doctest`
 package with the `pytest` assertion rewriting machinery, or ask it to avoid rewriting
@@ -369,7 +379,6 @@ In general, rewriting assertions is not very useful for doctests, as the
 output on error is fixed by the doctest machinery anyway. Therefore, we believe
 adding `--assert=plain` is reasonable.
 
-
 ## Prior art and related work
 
 - `pytest` provides some limited floating-point aware `NumericLiteralChecker`.
@@ -390,14 +399,13 @@ adding `--assert=plain` is reasonable.
   to be not easy to reason about, work with, and extend to other projects.
 
   This project is mainly the core functionality of the modified
-  `refguide-check` doctesting, extracted to a separate package. 
+  `refguide-check` doctesting, extracted to a separate package.
   We believe having it separate simplifies both addressing the needs of these
   two packages, and potential adoption by other projects.
 
-
 ### Bug reports, feature requests and contributions
 
 This package is work in progress. Contributions are most welcome!
 Please don't hesitate to open an issue in the tracker or send a pull request.
 
-The current location of the issue tracker is https://github.com/scipy/scipy_doctest.
+The current location of the issue tracker is <https://github.com/scipy/scipy_doctest>.

{scipy_doctest-1.6 → scipy_doctest-1.7.1}/pyproject.toml

@@ -25,13 +25,44 @@ dependencies  = [
 
 [project.optional-dependencies]
 test = [
-    "scipy",
+    "scipy <= 1.14.1",   # black-list 1.5.1 ?
     "matplotlib"
 ]
+doc = [
+    "furo==2024.8.6",
+    "myst-parser==4.0.0",
+    "sphinx==8.1.3",
+    "sphinx-copybutton==0.5.2"
+]
 
 [project.urls]
 Home = "https://github.com/scipy/scipy_doctest"
 
+[tool.ruff.lint]
+select = [
+    # pycodestyle
+    "E",
+    # Pyflakes
+    "F",
+    # pyupgrade
+    "UP",
+    # flake8-bugbear
+    "B",
+    # isort
+    "I",
+]
+
+[tool.ruff.lint.per-file-ignores]
+"scipy_doctest/tests/module_cases.py" = ["E501"]
+"scipy_doctest/tests/local_file_cases.py" = ["E501"]
+
+[tool.ruff.format]
+# Enable reformatting of code snippets in docstrings.
+docstring-code-format = true
+
+[tool.ruff.lint.pydocstyle]
+convention = "numpy"
+
 [tool.pytest.ini_options]
 addopts = "--verbose --color=yes"
 

{scipy_doctest-1.6 → scipy_doctest-1.7.1}/scipy_doctest/__init__.py

@@ -3,7 +3,7 @@ Configurable, whitespace-insensitive, floating-point-aware doctest helpers.
 """
 
 
-__version__ = "1.6"
+__version__ = "1.7.1"
 
 try:
     # register internal modules with pytest; obscure errors galore otherwise
@@ -19,4 +19,3 @@ except ModuleNotFoundError:
 
 from .impl import DTChecker, DTFinder, DTParser, DTRunner, DebugDTRunner, DTConfig  # noqa
 from .frontend import testmod, testfile, find_doctests, run_docstring_examples      # noqa
-

{scipy_doctest-1.6 → scipy_doctest-1.7.1}/scipy_doctest/conftest.py

@@ -2,4 +2,3 @@ from .impl import DTConfig
 
 
 dt_config = DTConfig()
-

{scipy_doctest-1.6 → scipy_doctest-1.7.1}/scipy_doctest/frontend.py

@@ -46,7 +46,7 @@ def find_doctests(module, strategy=None,
     Returns
     -------
     tests : list
-        A list of `doctest.DocTest`s that are defined by the module docstring,
+        A list of `doctest.DocTest` s that are defined by the module docstring,
         and by its contained objects’ docstrings. The selection is controlled
         by the `strategy` argument.
 
@@ -141,7 +141,7 @@ def testmod(m=None, name=None, globs=None, verbose=None,
         In verbose mode, the summary is detailed, else very brief (in fact,
         empty if all tests passed)
         Default is True.
-   verbose : int
+    verbose : int
         Control the run verbosity:
         0 means only report failures,
         1 means emit object names,
@@ -175,7 +175,7 @@ def testmod(m=None, name=None, globs=None, verbose=None,
     (result, history)
         `result` is a namedtuple ``TestResult(failed, attempted)``
         `history` is a dict with details of which objects were examined (the
-        keys are object names and values are individual objects' ``TestResult``s)
+        keys are object names and values are individual objects' ``TestResult`` s)
 
     Examples
     --------
@@ -204,7 +204,7 @@ def testmod(m=None, name=None, globs=None, verbose=None,
     For complex packages, prefer `strategy='api'`, which works as follows:
     - take the names of public objects from the `__all__` attribute of the package.
     - if `__all__` is not defined, take `dir(module)` and filter out names
-      which start with a leading underscore and dunders.
+    which start with a leading underscore and dunders.
     - filter out deprecated items, i.e. those which raise `DeprecationWarning`.
 
     """
@@ -306,7 +306,7 @@ def testfile(filename, module_relative=True, name=None, package=None,
         In verbose mode, the summary is detailed, else very brief (in fact,
         empty if all tests passed)
         Default is True.
-   verbose : int
+    verbose : int
         Control the run verbosity:
         0 means only report failures,
         1 means emit object names,
@@ -335,7 +335,7 @@ def testfile(filename, module_relative=True, name=None, package=None,
     (result, history)
         `result` is a namedtuple ``TestResult(failed, attempted)``
         `history` is a dict with details of which objects were examined (the
-        keys are object names and values are individual objects' ``TestResult``s)
+        keys are object names and values are individual objects' ``TestResult`` s)
     """
     # initial configuration
     if config is None:
@@ -478,4 +478,3 @@ def _main():
         if result.failed:
             return 1
     return 0
-

{scipy_doctest-1.6 → scipy_doctest-1.7.1}/scipy_doctest/impl.py

@@ -4,19 +4,12 @@ import inspect
 import doctest
 from doctest import NORMALIZE_WHITESPACE, ELLIPSIS, IGNORE_EXCEPTION_DETAIL
 from itertools import zip_longest
+from sys import version_info
 
 import numpy as np
 
 from . import util
 
-
-## shim numpy 1.x vs 2.0
-if np.__version__ < "2":
-    VisibleDeprecationWarning = np.VisibleDeprecationWarning
-else:
-    VisibleDeprecationWarning = np.exceptions.VisibleDeprecationWarning
-
-
 # Register the optionflag to skip whole blocks, i.e.
 # sequences of Examples without an intervening text.
 SKIPBLOCK = doctest.register_optionflag('SKIPBLOCK')
@@ -68,6 +61,7 @@ class DTConfig:
         >>> for test in tests:
         ...     with user_context(test):
         ...         runner.run(test)
+
         Default is a noop.
     local_resources: dict
         If a test needs some local files, list them here. The format is
@@ -190,7 +184,8 @@ class DTConfig:
                  '.bar(', '.title', '.ylabel', '.xlabel', 'set_ylim', 'set_xlim',
                  '# reformatted', '.set_xlabel(', '.set_ylabel(', '.set_zlabel(',
                  '.set(xlim=', '.set(ylim=', '.set(xlabel=', '.set(ylabel=', '.xlim(',
-                 'ax.set('}
+                 'ax.set(', '.text(',
+            }
         self.stopwords = stopwords
 
         if pseudocode is None:
@@ -224,8 +219,13 @@ class DTConfig:
 
 
 def try_convert_namedtuple(got):
-    # suppose that "got" is smth like MoodResult(statistic=10, pvalue=0.1).
-    # Then convert it to the tuple (10, 0.1), so that can later compare tuples.
+    """
+    Converts a string representation of a named tuple into a plain tuple.
+
+    Suppose that "got" is smth like MoodResult(statistic=10, pvalue=0.1).
+    Then convert it to the tuple (10, 0.1), so that can later compare tuples.
+    """
+
     num = got.count('=')
     if num == 0:
         # not a nameduple, bail out
@@ -264,6 +264,8 @@ def try_convert_printed_array(got):
 
 
 def has_masked(got):
+    """Check if a given string represents a NumPy masked array.
+    """
     return 'masked_array' in got and '--' in got
 
 
@@ -291,6 +293,20 @@ def try_split_shape_from_abbrv(s_got):
 
 
 class DTChecker(doctest.OutputChecker):
+    """
+    A drop-in replacement for `doctest.OutputChecker`.
+
+    Allows robust output comparison for numerical values and special
+    cases involving NumPy arrays, masked arrays, namedtuples, and object
+    memory addresses. It is configurable via a `DTConfig` object.
+
+    Parameters:
+    -----------
+    config : DTConfig, optional
+        Configuration object that controls various aspects of output checking.
+        If not provided, a default `DTConfig` instance is used.
+
+    """
     obj_pattern = re.compile(r'at 0x[0-9a-fA-F]+>')
     vanilla = doctest.OutputChecker()
 
@@ -331,13 +347,8 @@ class DTChecker(doctest.OutputChecker):
         # OK then, convert strings to objects
         ns = dict(self.config.check_namespace)
         try:
-            with warnings.catch_warnings():
-                # NumPy's ragged array deprecation of np.array([1, (2, 3)]);
-                # also array abbreviations: try `np.diag(np.arange(1000))`
-                warnings.simplefilter('ignore', VisibleDeprecationWarning)
-
-                a_want = eval(want, dict(ns))
-                a_got = eval(got, dict(ns))
+            a_want = eval(want, dict(ns))
+            a_got = eval(got, dict(ns))
         except Exception:
             # Maybe we're printing a numpy array? This produces invalid python
             # code: `print(np.arange(3))` produces "[0 1 2]" w/o commas between
@@ -351,9 +362,9 @@ class DTChecker(doctest.OutputChecker):
                 s_got = try_convert_printed_array(s_got)
 
                 return self.check_output(s_want, s_got, optionflags)
-            
+
             #handle array abbreviation for n-dimensional arrays, n >= 1
-            ndim_array = (s_want.startswith("array([") and "..." in s_want and 
+            ndim_array = (s_want.startswith("array([") and "..." in s_want and
                           s_got.startswith("array([") and "..." in s_got)
             if ndim_array:
                 s_want, want_shape = try_split_shape_from_abbrv(s_want)
@@ -432,16 +443,31 @@ class DTChecker(doctest.OutputChecker):
         except Exception:
             pass
 
-        with warnings.catch_warnings():
-            # NumPy's ragged array deprecation of np.array([1, (2, 3)])
-            warnings.simplefilter('ignore', VisibleDeprecationWarning)
-
-            # This line is the crux of the whole thing. The rest is mostly scaffolding.
-            result = np.allclose(want, got, atol=self.atol, rtol=self.rtol, equal_nan=True)
+        # This line is the crux of the whole thing. The rest is mostly scaffolding.
+        result = np.allclose(want, got, atol=self.atol, rtol=self.rtol, equal_nan=True)
         return result
 
 
 class DTRunner(doctest.DocTestRunner):
+    """
+    A drop-in replacement for `doctest.DocTestRunner`.
+
+    Improves how test names are reported and allows better control over exception handling.
+    It integrates with `DTConfig` to apply customized settings for doctest execution.
+
+    Parameters:
+    -----------
+    checker : doctest.OutputChecker, optional
+        A custom output checker, defaults to `DTConfig.CheckerKlass(config)`.
+    verbose : bool, optional
+        If `True`, enables verbose output.
+    optionflags : int, optional
+        Bitwise OR of `doctest` option flags; defaults to `DTConfig.optionflags`.
+    config : DTConfig, optional
+        A configuration object controlling doctest behavior; a default instance is used if not provided.
+
+    """
+
     DIVIDER = "\n"
 
     def __init__(self, checker=None, verbose=None, optionflags=None, config=None):
@@ -489,15 +515,20 @@ class DTRunner(doctest.DocTestRunner):
     def get_history(self):
         """Return a dict with names of items which were run.
 
-        Actually the dict is `{name : (f, t)}`, where `name` is the name of
-        an object, and the value is a tuple of the numbers of examples which
-        failed and which were tried.
+        Actually the dict is `{name : (failures, tries, skips)}` (before Python
+        3.13, just `{name : (failures, tries, skips)}`) where `name` is the
+        name of an object, and the value is a tuple of the numbers of examples
+        which failed, were tried, and were skipped, respectively.
         """
-        return self._name2ft
+        if version_info >= (3, 13):
+            return self._stats
+        else:
+            return self._name2ft
 
 
 class DebugDTRunner(DTRunner):
-    """Doctest runner which raises on a first error.
+    """
+    Doctest runner which raises an exception on the first error.
 
     Almost verbatim copy of `doctest.DebugRunner`.
     """
@@ -521,8 +552,24 @@ class DebugDTRunner(DTRunner):
 
 
 class DTFinder(doctest.DocTestFinder):
-    """A Finder with helpful defaults.
     """
+    A drop-in replacement for `doctest.DocTestFinder` with helpful defaults.
+
+    Parameters:
+    -----------
+    verbose : bool, optional
+        If `True`, enables verbose output during doctest discovery.
+    parser : doctest.DocTestParser, optional
+        A custom parser for extracting doctests; defaults to `DTParser(config)`.
+    recurse : bool, default=True
+        Whether to recursively search for doctests in nested objects.
+    exclude_empty : bool, default=True
+        Whether to exclude objects that have no doctests.
+    config : DTConfig, optional
+        A configuration object controlling doctest behavior; a default instance is used if not provided.
+
+    """
+
     def __init__(self, verbose=None, parser=None, recurse=True,
                  exclude_empty=True, config=None):
         if config is None:
@@ -537,7 +584,7 @@ class DTFinder(doctest.DocTestFinder):
         if globs is None:
             globs = dict(self.config.default_namespace)
         # XXX: does this make similar checks in testmod/testfile duplicate?
-        if module not in self.config.skiplist:   
+        if module not in self.config.skiplist:
             tests = super().find(obj, name, module, globs, extraglobs)
 
             if inspect.isclass(obj):
@@ -550,8 +597,21 @@ class DTFinder(doctest.DocTestFinder):
 
 
 class DTParser(doctest.DocTestParser):
-    """A Parser with a stopword list.
     """
+    A drop-in replacement for `doctest.DocTestParser` with a stopword list.
+
+    It filters out stopwords, pseudocode, and random markers from doctests.
+
+    Parameters:
+    -----------
+    config : DTConfig, optional
+        A configuration object containing:
+        - `stopwords`: A list of words that signal a doctest should be ignored.
+        - `pseudocode`: A list of markers indicating non-executable code.
+        - `rndm_markers`: A list of markers indicating results may vary.
+
+    """
+
     def __init__(self, config=None):
         if config is None:
             config = DTConfig()
@@ -608,4 +668,3 @@ class DTParser(doctest.DocTestParser):
                 example.want += "  # _ignore\n"
             examples.append(example)
         return examples
-

{scipy_doctest-1.6 → scipy_doctest-1.7.1}/scipy_doctest/plugin.py

@@ -47,7 +47,7 @@ def pytest_ignore_collect(collection_path, config):
     """
     Determine whether to ignore the specified collection path.
     This function is used to exclude the 'tests' directory and test modules when
-    the '--doctest-modules' option is used.
+    the `--doctest-modules` option is used.
     """
     if config.getoption("--doctest-modules"):
         path_str = str(collection_path)
@@ -62,7 +62,7 @@ def pytest_ignore_collect(collection_path, config):
 def is_private(item):
     """Decide if an DocTestItem `item` is private.
 
-    Private items are ignored in pytest_collect_modifyitem`.
+    Private items are ignored in `pytest_collect_modifyitem`.
     """
     # Here we look at the name of a test module/object. A seemingly less
     # hacky alternative is to populate a set of seen `item.dtest` attributes
@@ -174,7 +174,7 @@ def _is_deprecated(module):
 class DTModule(DoctestModule):
     """
     This class extends the DoctestModule class provided by pytest.
-    
+
     DTModule is responsible for overriding the behavior of the collect method.
     The collect method is called by pytest to collect and generate test items for doctests
     in the specified module or file.
@@ -222,7 +222,7 @@ class DTModule(DoctestModule):
 
         optionflags = dt_config.optionflags
 
-        # Plug in the custom runner: `PytestDTRunner` 
+        # Plug in the custom runner: `PytestDTRunner`
         runner = _get_runner(self.config,
             verbose=False,
             optionflags=optionflags,
@@ -245,7 +245,7 @@ class DTModule(DoctestModule):
 class DTTextfile(DoctestTextfile):
     """
     This class extends the DoctestTextfile class provided by pytest.
-    
+
     DTTextfile is responsible for overriding the behavior of the collect method.
     The collect method is called by pytest to collect and generate test items for doctests
     in the specified text files.
@@ -282,7 +282,7 @@ class DTTextfile(DoctestTextfile):
 def _get_runner(config, verbose, optionflags):
     """
     Override function to return an instance of PytestDTRunner.
-    
+
     This function creates and returns an instance of PytestDTRunner, a custom runner class
     that extends the behavior of DebugDTRunner for running doctests in pytest.
     """
@@ -290,13 +290,13 @@ def _get_runner(config, verbose, optionflags):
         def run(self, test, compileflags=None, out=None, clear_globs=False):
             """
             Run tests in context managers.
-            
+
             Restore the errstate/print state after each docstring.
             Also, make MPL backend non-GUI and close the figures.
-            
+
             The order of context managers is actually relevant. Consider
             user_context_mgr that turns warnings into errors.
-            
+
             Additionally, suppose that MPL deprecates something and plt.something
             starts issuing warnings. Now all of those become errors
             *unless* the `mpl()` context mgr has a chance to filter them out

{scipy_doctest-1.6 → scipy_doctest-1.7.1}/scipy_doctest/tests/failure_cases_2.py

@@ -24,4 +24,3 @@ def func_name_error():
     Further name errors are also suppressed (which is a bug, too):
     >>> raise NameError('Also legit')
     """
-

{scipy_doctest-1.6 → scipy_doctest-1.7.1}/scipy_doctest/tests/finder_cases.py

@@ -61,5 +61,3 @@ def _underscored_private_func():
     >>> 9 + 12
     21
     """
-
-

{scipy_doctest-1.6 → scipy_doctest-1.7.1}/scipy_doctest/tests/local_file_cases.py

@@ -33,5 +33,3 @@ def sio():
                   [ 2.,  5.,  8., 11.],
                   [ 3.,  6.,  9., 12.]]])}
     """
-
-

{scipy_doctest-1.6 → scipy_doctest-1.7.1}/scipy_doctest/tests/module_cases.py

@@ -3,8 +3,6 @@ __all__ = [
     'func7', 'manip_printoptions', 'array_abbreviation'
 ]
 
-import numpy as np
-import pytest
 
 def func():
     """

{scipy_doctest-1.6 → scipy_doctest-1.7.1}/scipy_doctest/tests/scipy_ndimage_tutorial_clone.rst

@@ -215,7 +215,7 @@ Smoothing filters
   corresponds to convolution with a Gaussian kernel. An order of 1, 2,
   or 3 corresponds to convolution with the first, second, or third
   derivatives of a Gaussian. Higher-order derivatives are not
-  implemented. 
+  implemented.
 
 
 
@@ -232,7 +232,7 @@ Smoothing filters
   number, to specify the same order for all axes, or a sequence of
   numbers to specify a different order for each axis. The example below
   shows the filter applied on test data with different values of *sigma*.
-  The *order* parameter is kept at 0. 
+  The *order* parameter is kept at 0.
 
   .. plot:: tutorial/examples/gaussian_filter_plot1.py
       :align: center

{scipy_doctest-1.6 → scipy_doctest-1.7.1}/scipy_doctest/tests/test_finder.py

@@ -86,8 +86,8 @@ class TestSkiplist:
         assert sorted(names) == sorted(wanted_names)
 
     def test_get_doctests_strategy_None(self):
-        # Add a skiplist: strategy=None skips listed items 
-        base = finder_cases.__name__  
+        # Add a skiplist: strategy=None skips listed items
+        base = finder_cases.__name__
         skips = [base + '.func', base + '.Klass.meth_2']
         config = DTConfig(skiplist=skips)
 
@@ -102,8 +102,8 @@ class TestSkiplist:
         assert sorted(names) == sorted(wanted_names)
 
     def test_get_doctests_strategy_api(self):
-        # Add a skiplist: strategy='api' skips listed items 
-        base = finder_cases.__name__  
+        # Add a skiplist: strategy='api' skips listed items
+        base = finder_cases.__name__
         skips = [base + '.func', base + '.Klass.meth_2']
         config = DTConfig(skiplist=skips)
 
@@ -120,8 +120,8 @@ class TestSkiplist:
         assert sorted(names) == sorted(wanted_names)
 
     def test_get_doctests_strategy_list(self):
-        # Add a skiplist: strategy=<list> skips listed items 
-        base = finder_cases.__name__  
+        # Add a skiplist: strategy=<list> skips listed items
+        base = finder_cases.__name__
         skips = [base + '.func', base + '.Klass.meth_2']
         config = DTConfig(skiplist=skips)
 
@@ -217,6 +217,3 @@ def test_private_superclasses_2():
 
     assert len(tests) == len(expected_names)
     assert names == set(expected_names)
-
-
-

{scipy_doctest-1.6 → scipy_doctest-1.7.1}/scipy_doctest/tests/test_parser.py

@@ -31,6 +31,3 @@ def test_config_nocopy():
     config = DTConfig()
     parser = DTParser(config)
     assert parser.config is config
-
-
-

{scipy_doctest-1.6 → scipy_doctest-1.7.1}/scipy_doctest/tests/test_pytest_configuration.py

@@ -39,7 +39,7 @@ def test_failure_cases(pytester):
         result = pytester.inline_run(python_file, "--doctest-modules")
     assert result.ret == pytest.ExitCode.TESTS_FAILED
 
-    
+
 @pytest.mark.skipif(not HAVE_MATPLOTLIB, reason='need matplotlib')
 def test_stopword_cases(pytester):
     """Test that pytest uses the DTParser for doctests."""
@@ -90,4 +90,3 @@ def test_alt_checker(pytester):
     # run all tests with pytest
     result = pytester.inline_run(f, '--doctest-modules')
     assert result.ret == pytest.ExitCode.TESTS_FAILED
-

{scipy_doctest-1.6 → scipy_doctest-1.7.1}/scipy_doctest/tests/test_runner.py

@@ -102,4 +102,3 @@ class TestCheckerDropIn:
         with pytest.raises(doctest.DocTestFailure):
             for t in tests:
                 runner.run(t)
-

{scipy_doctest-1.6 → scipy_doctest-1.7.1}/scipy_doctest/tests/test_skipmarkers.py

@@ -1,4 +1,5 @@
 import doctest
+from sys import version_info
 
 try:
     import matplotlib.pyplot as plt    # noqa
@@ -57,7 +58,8 @@ class TestSyntaxErrors:
                                    lineno=0)
         runner = DebugDTRunner()
         runner.run(test)
-        assert runner.get_history() == {'none : +SKIP': (0, 0)}
+        stats = (0, 1, 1) if version_info >= (3, 13) else (0, 0)
+        assert runner.get_history() == {'none : +SKIP': stats}
 
     def test_invalid_python_pseudocode(self):
         # Marking a test as pseudocode is equivalent to a +SKIP:
@@ -74,7 +76,8 @@ class TestSyntaxErrors:
                                    lineno=0)
         runner = DebugDTRunner()
         runner.run(test)
-        assert runner.get_history() == {'none : pseudocode': (0, 0)}
+        stats = (0, 1, 1) if version_info >= (3, 13) else (0, 0)
+        assert runner.get_history() == {'none : pseudocode': stats}
 
 
 class TestPseudocodeMarkers:
@@ -125,7 +128,8 @@ class TestStopwords:
         runner.run(test)
 
         # one example tried, of which zero failed
-        assert runner.get_history() == {'stopwords_bogus_output': (0, 1)}
+        stats = (0, 1, 0) if version_info >= (3, 13) else (0, 1)
+        assert runner.get_history() == {'stopwords_bogus_output': stats}
 
 
 class TestMayVary:
@@ -148,7 +152,8 @@ class TestMayVary:
         runner.run(test)
 
         # one example tried, of which zero failed
-        assert runner.get_history() == {'may_vary_markers': (0, 1)}
+        stats = (0, 1, 0) if version_info >= (3, 13) else (0, 1)
+        assert runner.get_history() == {'may_vary_markers': stats}
 
     def test_may_vary_source(self):
         # The marker needs to be added to the example output, not source.
@@ -164,7 +169,8 @@ class TestMayVary:
         runner.run(test)
 
         # one example tried, of which zero failed
-        assert runner.get_history() == {'may_vary_source': (0, 1)}
+        stats = (0, 1, 0) if version_info >= (3, 13) else (0, 1)
+        assert runner.get_history() == {'may_vary_source': stats}
 
     def test_may_vary_syntax_error(self):
         # `# may vary` markers do not mask syntax errors, unlike `# doctest: +SKIP`
@@ -193,7 +199,7 @@ below
 
 Note how the block above will fail doctesting unless the second line is
 skipped. A standard solution is to add a +SKIP marker to every line, but this
-is ugly and we skip the whole block instead. 
+is ugly and we skip the whole block instead.
 
 Once the block is over, we get back to usual doctests, which are not skipped
 
@@ -216,4 +222,3 @@ def test_SKIPBLOCK():
     assert test.examples[0].options[SKIP] is True
     assert test.examples[1].options[SKIP] is True
     assert test.examples[2].options == {}    # not skipped
-

{scipy_doctest-1.6 → scipy_doctest-1.7.1}/scipy_doctest/tests/test_testmod.py

@@ -53,7 +53,7 @@ def test_stopwords():
     res, _ = _testmod(stopwords, verbose=_VERBOSE)
     assert res.failed == 0
     assert res.attempted != 0
-    
+
 
 @pytest.mark.skipif(not HAVE_SCIPY, reason='need scipy')
 def test_public_obj_discovery():

{scipy_doctest-1.6 → scipy_doctest-1.7.1}/scipy_doctest/util.py

@@ -260,9 +260,9 @@ modules = []
 def generate_log(module, test):
     """
     Generate a log of the doctested items.
-    
+
     This function logs the items being doctested to a file named 'doctest.log'.
-    
+
     Args:
         module (module): The module being doctested.
         test (str): The name of the doctest item.
@@ -276,4 +276,3 @@ def generate_log(module, test):
             LOGFILE.write(f"{test}\n")
         except AttributeError:
             LOGFILE.write(f"{test}\n")
-