doc: Updated sysutils/dua-cli to 2.20.1

(pin)

sysutils/dua-cli: update to 2.20.1

2.20.1 (2023-05-05)
Bug Fixes
- release definition to use a more recent ubuntu, to get linux binaries.
  Otherwise there are no linux runners who can satisfy the request, and they
  hang forever, instead of saying something useful.

2.20.0 (2023-05-05)
New Features
- TUI now shows performance metrics while scanning and after. This is in
  preparation for the moonwalk upgrade.
- improve CLI help provided with the --format flag. It's now possible to see
  what possible values are without reading a swath of text. Now the default is
  shown as well which is more important now that it changes depending on the
  platform.
- use metric byte format only on MacOS. That way, on linux the binary format
  is used by default which is more common on that platform.

Bug Fixes
- Improve documentation for Marked Items pane to make clearer how to delete
  items.

(pin)

doc: Updated devel/nss to 3.89.1

(wiz)

nss: update to 3.89.1.

Changes:

  - Bug 1804505 - Update the technical constraints for KamuSM.
  - Bug 1822921 - Add BJCA Global Root CA1 and CA2 root certificates.

(wiz)

Updated archivers/xz, devel/py-dogpile-cache

(adam)

py-dogpile-cache: updated to 1.2.0

1.2.0

[feature] [region]
Added new construct api.CantDeserializeException which can be raised by user-defined deserializer functions which would be passed to CacheRegion.deserializer, to indicate a cache value that can窶冲 be deserialized and therefore should be regenerated. This can allow an application that窶冱 been updated to gracefully re-cache old items that were persisted from a previous version of the application. Pull request courtesy Simon Hewitt.

(adam)

xz: updated to 5.4.3

5.4.3 (2023-05-04)

* All fixes from 5.2.12

* Features in the CMake build can now be disabled as CMake cache
  variables, similar to the Autotools build.

* Minor update to the Croatian translation.

(adam)

Revbump all Go packages after go120 security update

(bsiegert)

Added devel/py-scikit-build; Updated textproc/py-Levenshtein

(adam)

py-Levenshtein: updated to 0.21.0

v0.21.0

Changed

relax dependency requirement on rapidfuzz
v0.20.9

Fixed

fix function signature of get_requires_for_build_wheel
v0.20.8

Fixed

type hints for editops/opcoded/matching_blocks did not allow any hashable sequence
v0.20.7

Fixed

type hints did not get installed
v0.20.6

Fixed

fix incorrect result normalization in setratio and seqratio
v0.20.5

Fixed

fix support for cmake versions below 3.17
fix version requirement for rapidfuzz-cpp when building against a previously installed version
v0.20.4

Changed

modernize cmake build to fix most conda-forge builds
v0.20.3

Changed

Added support for Python3.11
v0.20.2

Fixed

fix matching_blocks conversion for empty editops
Changed

added in-tree build backend to install cmake and ninja only when it is not installed yet and only when wheels are available
v0.20.1

Fixed

fix broken matching_blocks conversion
v0.20.0

Changed

use matching_blocks/apply/remove_subsequence/inverse implementation from RapidFuzz
Fixed

stop adding data to wheels
fix segmentation fault on some invalid editop sequences in subtract_edit
detect duplicated entries in editops validation

(adam)

py-scikit-build: added version 0.17.3

Improved build system generator for CPython C/C++/Fortran/Cython extensions.

Better support is available for additional compilers, build systems, cross
compilation, and locating dependencies and determining their build
requirements.

The scikit-build package is fundamentally just glue between the setuptools
Python module and CMake.

(adam)

doc: Updated lang/go120 to 1.20.4

(bsiegert)

go120: update to 1.20.4 (security)

This minor release includes 3 security fixes following the security policy:

* html/template: improper sanitization of CSS values

  Angle brackets (<>) were not considered dangerous characters when inserted
  into CSS contexts. Templates containing multiple actions separated by a '/'
  character could result in unexpectedly closing the CSS context and allowing
  for injection of unexpected HMTL, if executed with untrusted input.

  Thanks to Juho Nurminen of Mattermost for reporting this issue.

  This is CVE-2023-24539 and Go issue https://go.dev/issue/59720.

* html/template: improper handling of JavaScript whitespace

  Not all valid JavaScript whitespace characters were considered to be
  whitespace. Templates containing whitespace characters outside of the
  character set "\t\n\f\r\u0020\u2028\u2029" in JavaScript contexts that also
  contain actions may not be properly sanitized during execution.

  Thanks to Juho Nurminen of Mattermost for reporting this issue.

  This is CVE-2023-24540 and Go issue https://go.dev/issue/59721.

* html/template: improper handling of empty HTML attributes

  Templates containing actions in unquoted HTML attributes (e.g. "attr={{.}}")
  executed with empty input could result in output that would have unexpected
  results when parsed due to HTML normalization rules. This may allow injection
  of arbitrary attributes into tags.

  Thanks to Juho Nurminen of Mattermost for reporting this issue.

  This is CVE-2023-29400 and Go issue https://go.dev/issue/59722.

(bsiegert)

Updated devel/py-requests, net/py-pika

(adam)

py-pika: updated to 1.3.2

1.3.2

Closed issues:

GeventConnection may now work on Windows
ExchangeType should be from str type

Merged pull requests:

Attempt to re-enable GeventConnection on Windows
fix typo
Fix exchange type enum type

(adam)

py-requests: updated to 2.30.0

2.30.0 (2023-05-03)
-------------------

**Dependencies**
- ⚠️ Added support for urllib3 2.0. ⚠️

  This may contain minor breaking changes so we advise careful testing and
  reviewing https://urllib3.readthedocs.io/en/latest/v2-migration-guide.html
  prior to upgrading.

  Users who wish to stay on urllib3 1.x can pin to `urllib3<2`.

(adam)

Updated lang/oracle-jdk17, textproc/py-sphinx-automodapi

(adam)

py-sphinx-automodapi: updated to 0.15.0

0.15.0 (2023-03-13)
-------------------
- Silenced spurious warnings on configuring ``:nosignatures:``.
- Fixed issue with non-ascii characters in object members when the encoding is
  not ``utf8``.
- Allow use of ``:noindex:``, propagating this flag to autodoc.

(adam)

oracle-jdk17: updated to 17.0.7

17.0.7
Who knows?

(adam)

doc: Updated multimedia/streamlink to 5.5.0

(yhardy)

multimedia/streamlink: update to streamlink-5.5.0

5.5.0

Release highlights:

  • Added: --no-config (#5314)
  • Added: --player-external-http-interface (#5295)
  • Fixed: M3U8 attribute parsing issue (#5307)
  • Fixed: various minor plugin issues (#5291, #5299, #5306)
  • Build: bumped urllib3 to >=1.26.0,<3 and fixed compatibility
    issues with urllib3 >=2.0.0 (#5326, #5325)
  • Docs: bumped furo theme to 2023.03.27 (#5301)
  • Docs: bumped build dependencies sphinx >=5.0.0,<7,
    myst-parser >=1.0.0,<2 and sphinx-design >=0.4.1,<1 (#5301)

Full changelog

Ian Cameron:
      plugins.mitele: fix/add Origin header

bastimeyer:
      cli.output: split up module into subpackage
      chore: update GitHub release template
      cli.output: refactor HTTPOutput
      cli: add --player-external-http-interface argument
      plugins.dogan: find direct HLS URL
      script: implement plugin URL tester
      docs: bump furo to 2023.03.27
      revert: docs: remove solus from install docs
      stream.hls: fix M3U8 attribute parser
      docs: fix Linux AppImages link in install docs
      cli: add --no-config argument
      docs: add pull request feedback to CONTRIBUTING.md
      tests: clean up root-logger state
      ci.github: bump deps of useragents workflow
      build: update version range of urllib3
      session: add support for urllib3 2.0

streamlinkbot:
      plugin.api: update useragents

vinyl-umbrella:
      plugins.openrectv: fix 401 error (#5291)

(yhardy)

graphics/freetype2: Work around clang complaining

clang 10 (apple) objects to a fallthrough macro without a statement.
It's really not clear if the fallthrough macro (attribute) is valid,
or just a gcc extension, or if the compiler is right to object to a
lack of statement.  Just add "0;" to not change behavior and avoid the
rabbit hole.

(gdt)

py-bleach: fix filename in REPLACE_SH

(wiz)

py-flask: now py-werkzeug 2.3.3 does exist

(wiz)

doc: Updated www/py-werkzeug to 2.3.3

(wiz)

py-werkzeug: update to 2.3.3.

Version 2.3.3
-------------

Released 2023-05-01

-  Fix parsing of large multipart bodies. Remove invalid leading newline, and restore
    parsing speed. :issue:`2658, 2675`
-  The cookie ``Path`` attribute is set to ``/`` by default again, to prevent clients
    from falling back to RFC 6265's ``default-path`` behavior. :issue:`2672, 2679`

(wiz)

py-flask: py-werkzeug>=2.3.3 does not exist.

(jperkin)

Updated math/py-scipy, www/py-terminado

(adam)

py-terminado: updated to 0.17.1

0.17.1

Bugs fixed

- Don't split manually on linebreaks

Maintenance and upkeep improvements

- Docs and workflow cleanup
- CI Cleanup
- Add ability to release from repo
- Bump actions/checkout from 2 to 3
- Add dependabot

Documentation improvements

- Docs and workflow cleanup

0.17.0

Maintenance and upkeep improvements

- Clean up docs install
- Maintenance cleanup
- Maintenance cleanup

(adam)

py-scipy: updated to 1.10.1

SciPy 1.10.1 is a bug-fix release with no new features
compared to 1.10.0.

SciPy 1.10.0 Release Notes
==========================

SciPy 1.10.0 is the culmination of 6 months of hard work. It contains
many new features, numerous bug-fixes, improved test coverage and better
documentation. There have been a number of deprecations and API changes
in this release, which are documented below. All users are encouraged to
upgrade to this release, as there are a large number of bug-fixes and
optimizations. Before upgrading, we recommend that users check that
their own code does not use deprecated SciPy functionality (to do so,
run your code with ``python -Wd`` and check for ``DeprecationWarning`` s).
Our development attention will now shift to bug-fix releases on the
1.10.x branch, and on adding new features on the main branch.

This release requires Python 3.8+ and NumPy 1.19.5 or greater.

For running on PyPy, PyPy3 6.0+ is required.

**************************
Highlights of this release
**************************

- A new dedicated datasets submodule (`scipy.datasets`) has been added, and is
  now preferred over usage of `scipy.misc` for dataset retrieval.
- A new `scipy.interpolate.make_smoothing_spline` function was added. This
  function constructs a smoothing cubic spline from noisy data, using the
  generalized cross-validation (GCV) criterion to find the tradeoff between
  smoothness and proximity to data points.
- `scipy.stats` has three new distributions, two new hypothesis tests, three
  new sample statistics, a class for greater control over calculations
  involving covariance matrices, and many other enhancements.

************
New features
************

`scipy.datasets` introduction
=============================
- A new dedicated ``datasets`` submodule has been added. The submodules
  is meant for datasets that are relevant to other SciPy submodules ands
  content (tutorials, examples, tests), as well as contain a curated
  set of datasets that are of wider interest. As of this release, all
  the datasets from `scipy.misc` have been added to `scipy.datasets`
  (and deprecated in `scipy.misc`).
- The submodule is based on [Pooch](https://www.fatiando.org/pooch/latest/)
  (a new optional dependency for SciPy), a Python package to simplify fetching
  data files. This move will, in a subsequent release, facilitate SciPy
  to trim down the sdist/wheel sizes, by decoupling the data files and
  moving them out of the SciPy repository, hosting them externally and
  downloading them when requested. After downloading the datasets once,
  the files are cached to avoid network dependence and repeated usage.
- Added datasets from ``scipy.misc``: `scipy.datasets.face`,
  `scipy.datasets.ascent`, `scipy.datasets.electrocardiogram`
- Added download and caching functionality:

  - `scipy.datasets.download_all`: a function to download all the `scipy.datasets`
    associated files at once.
  - `scipy.datasets.clear_cache`: a simple utility function to clear cached dataset
    files from the file system.
  - ``scipy/datasets/_download_all.py`` can be run as a standalone script for
    packaging purposes to avoid any external dependency at build or test time.
    This can be used by SciPy packagers (e.g., for Linux distros) which may
    have to adhere to rules that forbid downloading sources from external
    repositories at package build time.

`scipy.integrate` improvements
==============================
- Added parameter ``complex_func`` to `scipy.integrate.quad`, which can be set
  ``True`` to integrate a complex integrand.

`scipy.interpolate` improvements
================================
- `scipy.interpolate.interpn` now supports tensor-product interpolation methods
  (``slinear``, ``cubic``, ``quintic`` and ``pchip``)
- Tensor-product interpolation methods (``slinear``, ``cubic``, ``quintic`` and
  ``pchip``) in `scipy.interpolate.interpn` and
  `scipy.interpolate.RegularGridInterpolator` now allow values with trailing
  dimensions.
- `scipy.interpolate.RegularGridInterpolator` has a new fast path for
  ``method="linear"`` with 2D data, and ``RegularGridInterpolator`` is now
  easier to subclass
- `scipy.interpolate.interp1d` now can take a single value for non-spline
  methods.
- A new ``extrapolate`` argument is available to `scipy.interpolate.BSpline.design_matrix`,
  allowing extrapolation based on the first and last intervals.
- A new function `scipy.interpolate.make_smoothing_spline` has been added. It is an
  implementation of the generalized cross-validation spline smoothing
  algorithm. The ``lam=None`` (default) mode of this function is a clean-room
  reimplementation of the classic ``gcvspl.f`` Fortran algorithm for
  constructing GCV splines.
- A new ``method="pchip"`` mode was aded to
  `scipy.interpolate.RegularGridInterpolator`. This mode constructs an
  interpolator using tensor products of C1-continuous monotone splines
  (essentially, a `scipy.interpolate.PchipInterpolator` instance per
  dimension).

`scipy.sparse.linalg` improvements
==================================
- The spectral 2-norm is now available in `scipy.sparse.linalg.norm`.
- The performance of `scipy.sparse.linalg.norm` for the default case (Frobenius
  norm) has been improved.
- LAPACK wrappers were added for ``trexc`` and ``trsen``.
- The `scipy.sparse.linalg.lobpcg` algorithm was rewritten, yielding
  the following improvements:

  - a simple tunable restart potentially increases the attainable
    accuracy for edge cases,
  - internal postprocessing runs one final exact Rayleigh-Ritz method
    giving more accurate and orthonormal eigenvectors,
  - output the computed iterate with the smallest max norm of the residual
    and drop the history of subsequent iterations,
  - remove the check for ``LinearOperator`` format input and thus allow
    a simple function handle of a callable object as an input,
  - better handling of common user errors with input data, rather
    than letting the algorithm fail.

`scipy.linalg` improvements
===========================
- `scipy.linalg.lu_factor` now accepts rectangular arrays instead of being restricted
  to square arrays.

`scipy.ndimage` improvements
============================
- The new `scipy.ndimage.value_indices` function provides a time-efficient method to
  search for the locations of individual values with an array of image data.
- A new ``radius`` argument is supported by `scipy.ndimage.gaussian_filter1d` and
  `scipy.ndimage.gaussian_filter` for adjusting the kernel size of the filter.

`scipy.optimize` improvements
=============================
- `scipy.optimize.brute` now coerces non-iterable/single-value ``args`` into a
  tuple.
- `scipy.optimize.least_squares` and `scipy.optimize.curve_fit` now accept
  `scipy.optimize.Bounds` for bounds constraints.
- Added a tutorial for `scipy.optimize.milp`.
- Improved the pretty-printing of `scipy.optimize.OptimizeResult` objects.
- Additional options (``parallel``, ``threads``, ``mip_rel_gap``) can now
  be passed to `scipy.optimize.linprog` with ``method='highs'``.

`scipy.signal` improvements
===========================
- The new window function `scipy.signal.windows.lanczos` was added to compute a
  Lanczos window, also known as a sinc window.

`scipy.sparse.csgraph` improvements
===================================
- the performance of `scipy.sparse.csgraph.dijkstra` has been improved, and
  star graphs in particular see a marked performance improvement

`scipy.special` improvements
============================
- The new function `scipy.special.powm1`, a ufunc with signature
  ``powm1(x, y)``, computes ``x**y - 1``. The function avoids the loss of
  precision that can result when ``y`` is close to 0 or when ``x`` is close to
  1.
- `scipy.special.erfinv` is now more accurate as it leverages the Boost equivalent under
  the hood.

`scipy.stats` improvements
==========================
- Added `scipy.stats.goodness_of_fit`, a generalized goodness-of-fit test for
  use with any univariate distribution, any combination of known and unknown
  parameters, and several choices of test statistic (Kolmogorov-Smirnov,
  Cramer-von Mises, and Anderson-Darling).
- Improved `scipy.stats.bootstrap`: Default method ``'BCa'`` now supports
  multi-sample statistics. Also, the bootstrap distribution is returned in the
  result object, and the result object can be passed into the function as
  parameter ``bootstrap_result`` to add additional resamples or change the
  confidence interval level and type.
- Added maximum spacing estimation to `scipy.stats.fit`.
- Added the Poisson means test ("E-test") as `scipy.stats.poisson_means_test`.
- Added new sample statistics.

  - Added `scipy.stats.contingency.odds_ratio` to compute both the conditional
    and unconditional odds ratios and corresponding confidence intervals for
    2x2 contingency tables.
  - Added `scipy.stats.directional_stats` to compute sample statistics of
    n-dimensional directional data.
  - Added `scipy.stats.expectile`, which generalizes the expected value in the
    same way as quantiles are a generalization of the median.

- Added new statistical distributions.

  - Added `scipy.stats.uniform_direction`, a multivariate distribution to
    sample uniformly from the surface of a hypersphere.
  - Added `scipy.stats.random_table`, a multivariate distribution to sample
    uniformly from m x n contingency tables with provided marginals.
  - Added `scipy.stats.truncpareto`, the truncated Pareto distribution.

- Improved the ``fit`` method of several distributions.

  - `scipy.stats.skewnorm` and `scipy.stats.weibull_min` now use an analytical
    solution when ``method='mm'``, which also serves a starting guess to
    improve the performance of ``method='mle'``.
  - `scipy.stats.gumbel_r` and `scipy.stats.gumbel_l`: analytical maximum
    likelihood estimates have been extended to the cases in which location or
    scale are fixed by the user.
  - Analytical maximum likelihood estimates have been added for
    `scipy.stats.powerlaw`.

- Improved random variate sampling of several distributions.

  - Drawing multiple samples from `scipy.stats.matrix_normal`,
    `scipy.stats.ortho_group`, `scipy.stats.special_ortho_group`, and
    `scipy.stats.unitary_group` is faster.
  - The ``rvs`` method of `scipy.stats.vonmises` now wraps to the interval
    ``[-np.pi, np.pi]``.
  - Improved the reliability of `scipy.stats.loggamma` ``rvs`` method for small
    values of the shape parameter.

- Improved the speed and/or accuracy of functions of several statistical
  distributions.

  - Added `scipy.stats.Covariance` for better speed, accuracy, and user control
    in multivariate normal calculations.
  - `scipy.stats.skewnorm` methods ``cdf``, ``sf``, ``ppf``, and ``isf``
    methods now use the implementations from Boost, improving speed while
    maintaining accuracy. The calculation of higher-order moments is also
    faster and more accurate.
  - `scipy.stats.invgauss` methods ``ppf`` and ``isf`` methods now use the
    implementations from Boost, improving speed and accuracy.
  - `scipy.stats.invweibull` methods ``sf`` and ``isf`` are more accurate for
    small probability masses.
  - `scipy.stats.nct` and `scipy.stats.ncx2` now rely on the implementations
    from Boost, improving speed and accuracy.
  - Implemented the ``logpdf`` method of `scipy.stats.vonmises` for reliability
    in extreme tails.
  - Implemented the ``isf`` method of `scipy.stats.levy` for speed and
    accuracy.
  - Improved the robustness of `scipy.stats.studentized_range` for large ``df``
    by adding an infinite degree-of-freedom approximation.
  - Added a parameter ``lower_limit`` to `scipy.stats.multivariate_normal`,
    allowing the user to change the integration limit from -inf to a desired
    value.
  - Improved the robustness of ``entropy`` of `scipy.stats.vonmises` for large
    concentration values.

- Enhanced `scipy.stats.gaussian_kde`.

  - Added `scipy.stats.gaussian_kde.marginal`, which returns the desired
    marginal distribution of the original kernel density estimate distribution.
  - The ``cdf`` method of `scipy.stats.gaussian_kde` now accepts a
    ``lower_limit`` parameter for integrating the PDF over a rectangular region.
  - Moved calculations for `scipy.stats.gaussian_kde.logpdf` to Cython,
    improving speed.
  - The global interpreter lock is released by the ``pdf`` method of
    `scipy.stats.gaussian_kde` for improved multithreading performance.
  - Replaced explicit matrix inversion with Cholesky decomposition for speed
    and accuracy.

- Enhanced the result objects returned by many `scipy.stats` functions

  - Added a ``confidence_interval`` method to the result object returned by
    `scipy.stats.ttest_1samp` and `scipy.stats.ttest_rel`.
  - The `scipy.stats` functions ``combine_pvalues``, ``fisher_exact``,
    ``chi2_contingency``, ``median_test`` and ``mood`` now return
    bunch objects rather than plain tuples, allowing attributes to be
    accessed by name.
  - Attributes of the result objects returned by ``multiscale_graphcorr``,
    ``anderson_ksamp``, ``binomtest``, ``crosstab``, ``pointbiserialr``,
    ``spearmanr``, ``kendalltau``, and ``weightedtau`` have been renamed to
    ``statistic`` and ``pvalue`` for consistency throughout `scipy.stats`.
    Old attribute names are still allowed for backward compatibility.
  - `scipy.stats.anderson` now returns the parameters of the fitted
    distribution in a `scipy.stats._result_classes.FitResult` object.
  - The ``plot`` method of `scipy.stats._result_classes.FitResult` now accepts
    a ``plot_type`` parameter; the options are ``'hist'`` (histogram, default),
    ``'qq'`` (Q-Q plot), ``'pp'`` (P-P plot), and ``'cdf'`` (empirical CDF
    plot).
  - Kolmogorov-Smirnov tests (e.g. `scipy.stats.kstest`) now return the
    location (argmax) at which the statistic is calculated and the variant
    of the statistic used.

- Improved the performance of several `scipy.stats` functions.

  - Improved the performance of `scipy.stats.cramervonmises_2samp` and
    `scipy.stats.ks_2samp` with ``method='exact'``.
  - Improved the performance of `scipy.stats.siegelslopes`.
  - Improved the performance of `scipy.stats.mstats.hdquantile_sd`.
  - Improved the performance of `scipy.stats.binned_statistic_dd` for several
    NumPy statistics, and binned statistics methods now support complex data.

- Added the ``scramble`` optional argument to `scipy.stats.qmc.LatinHypercube`.
  It replaces ``centered``, which is now deprecated.
- Added a parameter ``optimization`` to all `scipy.stats.qmc.QMCEngine`
  subclasses to improve characteristics of the quasi-random variates.
- Added tie correction to `scipy.stats.mood`.
- Added tutorials for resampling methods in `scipy.stats`.
- `scipy.stats.bootstrap`, `scipy.stats.permutation_test`, and
  `scipy.stats.monte_carlo_test` now automatically detect whether the provided
  ``statistic`` is vectorized, so passing the ``vectorized`` argument
  explicitly is no longer required to take advantage of vectorized statistics.
- Improved the speed of `scipy.stats.permutation_test` for permutation types
  ``'samples'`` and ``'pairings'``.
- Added ``axis``, ``nan_policy``, and masked array support to
  `scipy.stats.jarque_bera`.
- Added the ``nan_policy`` optional argument to `scipy.stats.rankdata`.

*******************
Deprecated features
*******************
- `scipy.misc` module and all the methods in ``misc`` are deprecated in v1.10
  and will be completely removed in SciPy v2.0.0. Users are suggested to
  utilize the `scipy.datasets` module instead for the dataset methods.
- `scipy.stats.qmc.LatinHypercube` parameter ``centered`` has been deprecated.
  It is replaced by the ``scramble`` argument for more consistency with other
  QMC engines.
- `scipy.interpolate.interp2d` class has been deprecated.  The docstring of the
  deprecated routine lists recommended replacements.

********************
Expired Deprecations
********************
- There is an ongoing effort to follow through on long-standing deprecations.
- The following previously deprecated features are affected:

  - Removed ``cond`` & ``rcond`` kwargs in ``linalg.pinv``
  - Removed wrappers ``scipy.linalg.blas.{clapack, flapack}``
  - Removed ``scipy.stats.NumericalInverseHermite`` and removed ``tol`` & ``max_intervals`` kwargs from ``scipy.stats.sampling.NumericalInverseHermite``
  - Removed ``local_search_options`` kwarg frrom ``scipy.optimize.dual_annealing``.

*************
Other changes
*************
- `scipy.stats.bootstrap`, `scipy.stats.permutation_test`, and
  `scipy.stats.monte_carlo_test` now automatically detect whether the provided
  ``statistic`` is vectorized by looking for an ``axis`` parameter in the
  signature of ``statistic``. If an ``axis`` parameter is present in
  ``statistic`` but should not be relied on for vectorized calls, users must
  pass option ``vectorized==False`` explicitly.
- `scipy.stats.multivariate_normal` will now raise a ``ValueError`` when the
  covariance matrix is not positive semidefinite, regardless of which method
  is called.

(adam)

0ad: fix part of the rpath issues + use pkgconfig for miniupnpc

(nikita)

doc: Updated misc/chatgpt-shell-cli to 0.0.20230505

(schmonz)

Update to 20230505. From the changelog:

- Remove broken view count
- Remove sed due to warning, replace with escape function, rename
  variables, autofor>
- Rename to multiline prompt
- formatting
- Keep old OPENAI key name to maintain compatibility with the install
  script, add identation in elif
- Safe escaping using jq
- Refactoring to reduce the use global variables from functions
- Safer quoting
- rename(OPENAI_KEY): OPENAI_API_KEY is standardised across projects
- feature(big-prompt): allow multi-line input during chat mode
- optimize(exit): streamline exit conditions
- add(list_models): Models can be queried via cli argument
- refactor(usage()) Align help output

(schmonz)

doc: Updated net/openconnect to 9.10

(schmonz)

Update to 9.10. From the changelog:

- Fix external browser authentication with KDE plasma-nm < 5.26.
- Always redirect stdout to stderr when spawning external browser.
- Increase default queue length to 32 packets (#582).
- Make the Wintun Layer 3 TUN driver the default on Windows (!427).
- Add support for and bundle Wintun 0.14.1 (!294).
- Fix receiving multiple packets in one TLS frame, and single packets
  split across multiple TLS frames, for Array (#435).
- Fix ESP failures under Windows (#427).
- Add list-system-keys tool to assist Windows/MacOS users in setup.
- Handle idiosyncratic variation in search domain separators for all
  protocols (#433, #443, !388).
- Support region selection field for Pulse authentication (!399).
- Support modified configuration packet from Pulse 9.1R16 servers
  (#472, !401)
- Allow hidden form fields to be populated or converted to text fields
  on the command line (#493, #489, !409)
- Support yet another strange way of encoding challenge-based 2FA for
  GlobalProtect (#495, !411)
- Add --sni option (and corresponding C and Java API functions) to allow
  domain-fronting connections in censored/filtered network environments
  (!297, !451).
- Parrot a GlobalProtect server's software version, if present, as the
  client version (!333)
- Fix NULL pointer dereference that has left Android builds broken since
  v8.20 (!389).
- Fix Fortinet authentication bug where repeated SVPNCOOKIE causes
  segfaults (#514, !418).
- Support F5 VPNs which encode authentication forms only in JSON, not in
  HTML (#512, !431).
- Persist Windows installers for tagged builds (#463, !391).
- Support simultaneous IPv6 and Legacy IP ("dual-stack") for Fortinet
  (#568, !456).
- Support "FTM-push" token mode for Fortinet VPNs (#555, !450).
- Send IPv6-compatible version string in Pulse IF/T session
  establishment, and avoid its ESP/IP version layering idiocy on newer
  servers (#506, !414)
- Add --no-external-auth option to not advertise external-browser
  authentication, as a workaround for servers which behave differently
  when it is advertised (#470, !398)
- Emulate MacOS-specific contents in the HIP report for GlobalProtect (!471).
- Many small improvements in server response parsing, and better logging
  messages and documentation.

(schmonz)

doc: Updated net/vpnc-script to 20230103

(schmonz)

Update to 20230103. From the changelog:

- No need to specify FAMILY in IPv4 only functions
- Avoid "ip route flush cache" for IPv4

(schmonz)

Updated devel/py-test-rerunfailures, www/py-urllib3, devel/py-pdm, www/py-flask

(adam)

py-flask: updated to 2.3.2

Version 2.3.2
-------------
- Set ``Vary: Cookie`` header when the session is accessed, modified, or refreshed.
- Update Werkzeug requirement to >=2.3.3 to apply recent bug fixes.

(adam)

py-pdm: updated to 2.5.5

Release v2.5.5 (2023-05-05)
---------------------------
No significant changes.

(adam)

py-urllib3: updated to 2.0.2

2.0.2 (2023-05-03)
==================
- Fixed ``HTTPResponse.stream()`` to continue yielding bytes if buffered decompressed data
  was still available to be read even if the underlying socket is closed. This prevents
  a compressed response from being truncated.

2.0.1 (2023-04-30)
==================
- Fixed a socket leak when fingerprint or hostname verifications fail.
- Fixed an error when ``HTTPResponse.read(0)`` was the first ``read`` call or when the internal response body buffer was otherwise empty.

(adam)

py-test-rerunfailures: updated to 11.1.2

11.1.2 (2023-03-09)
-------------------

- Execute teardown when test was skipped in setup phase of a fixture.

11.1.1 (2023-02-17)
-------------------

- Fix crash during teardown when runtest protocol hook is overwritten by
  another plugin.
- Fix crash during teardown when TestCase class is used as base class.

11.1 (2023-02-09)
-----------------

Bug fixes

- Run teardown of session, class, ... scoped fixtures only once after rerunning tests

Features

- Expose `reruns` and `reruns_delay` through `pytest.ini` file.

11.0 (2023-01-12)
-----------------

Breaking changes

- Drop support for Python 3.6.
- Drop support for pytest < 6.

Bug fixes

- Fix crash when pytest-xdist is installed but disabled.
- Fix crash when xfail(strict=True) mark is used with --rerun-only flag.

Features

- Added option `--rerun-except` to rerun failed tests those are other than the mentioned Error.
- Add support for Python 3.11.
- Add support for pytest 7.0, 7.1, 7.2.

(adam)

py-cattrs: More stupid python nonsense.

(jperkin)

doc: Updated www/py-autobahn to 23.1.2

(wiz)

py-autobahn: update to 23.1.2.

3 years of development, no summary found.
New binaries (xbrnetwork, xbrnetwork-ui).

(wiz)

doc: Updated www/py-flask to 2.3.1

(wiz)

py-flask: update to 2.3.1.

Version 2.3.1
-------------

Released 2023-04-25

-  Restore deprecated ``from flask import Markup``. :issue:`5084`

Version 2.3.0
-------------

Released 2023-04-25

-  Drop support for Python 3.7. :pr:`5072`
-  Update minimum requirements to the latest versions: Werkzeug>=2.3.0, Jinja2>3.1.2,
    itsdangerous>=2.1.2, click>=8.1.3.
-  Remove previously deprecated code. :pr:`4995`

    -  The ``push`` and ``pop`` methods of the deprecated ``_app_ctx_stack`` and
        ``_request_ctx_stack`` objects are removed. ``top`` still exists to give
        extensions more time to update, but it will be removed.
    -  The ``FLASK_ENV`` environment variable, ``ENV`` config key, and ``app.env``
        property are removed.
    -  The ``session_cookie_name``, ``send_file_max_age_default``, ``use_x_sendfile``,
        ``propagate_exceptions``, and ``templates_auto_reload`` properties on ``app``
        are removed.
    -  The ``JSON_AS_ASCII``, ``JSON_SORT_KEYS``, ``JSONIFY_MIMETYPE``, and
        ``JSONIFY_PRETTYPRINT_REGULAR`` config keys are removed.
    -  The ``app.before_first_request`` and ``bp.before_app_first_request`` decorators
        are removed.
    -  ``json_encoder`` and ``json_decoder`` attributes on app and blueprint, and the
        corresponding ``json.JSONEncoder`` and ``JSONDecoder`` classes, are removed.
    -  The ``json.htmlsafe_dumps`` and ``htmlsafe_dump`` functions are removed.
    -  Calling setup methods on blueprints after registration is an error instead of a
        warning. :pr:`4997`

-  Importing ``escape`` and ``Markup`` from ``flask`` is deprecated. Import them
    directly from ``markupsafe`` instead. :pr:`4996`
-  The ``app.got_first_request`` property is deprecated. :pr:`4997`
-  The ``locked_cached_property`` decorator is deprecated. Use a lock inside the
    decorated function if locking is needed. :issue:`4993`
-  Signals are always available. ``blinker>=1.6.2`` is a required dependency. The
    ``signals_available`` attribute is deprecated. :issue:`5056`
-  Signals support ``async`` subscriber functions. :pr:`5049`
-  Remove uses of locks that could cause requests to block each other very briefly.
    :issue:`4993`
-  Use modern packaging metadata with ``pyproject.toml`` instead of ``setup.cfg``.
    :pr:`4947`
-  Ensure subdomains are applied with nested blueprints. :issue:`4834`
-  ``config.from_file`` can use ``text=False`` to indicate that the parser wants a
    binary file instead. :issue:`4989`
-  If a blueprint is created with an empty name it raises a ``ValueError``.
    :issue:`5010`
-  ``SESSION_COOKIE_DOMAIN`` does not fall back to ``SERVER_NAME``. The default is not
    to set the domain, which modern browsers interpret as an exact match rather than
    a subdomain match. Warnings about ``localhost`` and IP addresses are also removed.
    :issue:`5051`
-  The ``routes`` command shows each rule's ``subdomain`` or ``host`` when domain
    matching is in use. :issue:`5004`
-  Use postponed evaluation of annotations. :pr:`5071`

Version 2.2.4
-------------

Released 2023-04-25

-  Update for compatibility with Werkzeug 2.3.

(wiz)

Updated devel/py-pdm, devel/py-dulwich

(adam)

doc: Updated www/py-werkzeug to 2.3.2

(wiz)

py-dulwich: updated to 0.21.5

0.21.5

* Be more tolerant to non-3-length tuple versions.

0.21.4.1

* Support ``core.symlinks=false``.
* Deprecate ``dulwich.objects.parse_commit``.
* Fix fetching into MemoryRepo.
* Support ``init.defaultBranch`` config.
* Fix ``ObjectStore.iterobjects_subset()`` when
  hex shas are passed for objects that live in packs.
* client: Handle absolute path as redirect location in HTTP client.

(adam)

py-werkzeug: update to 2.3.2.

Version 2.3.2
-------------

Released 2023-04-28

-  Parse the cookie ``Expires`` attribute correctly in the test client. :issue:`2669`
-  ``max_content_length`` can only be enforced on streaming requests if the server
    sets ``wsgi.input_terminated``. :issue:`2668`
-  The cookie ``Path`` attribute is set to ``/`` by default again, to prevent clients
    from falling back to RFC 6265's ``default-path`` behavior. :issue:`2672`

Version 2.3.1
-------------

Released 2023-04-27

-  Percent-encode plus (+) when building URLs and in test requests. :issue:`2657`
-  Cookie values don't quote characters defined in RFC 6265. :issue:`2659`
-  Include ``pyi`` files for ``datastructures`` type annotations. :issue:`2660`
-  ``Authorization`` and ``WWWAuthenticate`` objects can be compared for equality.
    :issue:`2665`

Version 2.3.0
-------------

Released 2023-04-25

-  Drop support for Python 3.7. :pr:`2648`
-  Remove previously deprecated code. :pr:`2592`
-  Passing bytes where strings are expected is deprecated, as well as the ``charset``
    and ``errors`` parameters in many places. Anywhere that was annotated, documented,
    or tested to accept bytes shows a warning. Removing this artifact of the transition
    from Python 2 to 3 removes a significant amount of overhead in instance checks and
    encoding cycles. In general, always work with UTF-8, the modern HTML, URL, and HTTP
    standards all strongly recommend this. :issue:`2602`
-  Deprecate the ``werkzeug.urls`` module, except for the ``uri_to_iri`` and
    ``iri_to_uri`` functions. Use the ``urllib.parse`` library instead. :issue:`2600`
-  Update which characters are considered safe when using percent encoding in URLs,
    based on the WhatWG URL Standard. :issue:`2601`
-  Update which characters are considered safe when using percent encoding for Unicode
    filenames in downloads. :issue:`2598`
-  Deprecate the ``safe_conversion`` parameter of ``iri_to_uri``. The ``Location``
    header is converted to IRI using the same process as everywhere else. :issue:`2609`
-  Deprecate ``werkzeug.wsgi.make_line_iter`` and ``make_chunk_iter``. :pr:`2613`
-  Use modern packaging metadata with ``pyproject.toml`` instead of ``setup.cfg``.
    :pr:`2574`
-  ``Request.get_json()`` will raise a ``415 Unsupported Media Type`` error if the
    ``Content-Type`` header is not ``application/json``, instead of a generic 400.
    :issue:`2550`
-  A URL converter's ``part_isolating`` defaults to ``False`` if its ``regex`` contains
    a ``/``. :issue:`2582`
-  A custom converter's regex can have capturing groups without breaking the router.
    :pr:`2596`
-  The reloader can pick up arguments to ``python`` like ``-X dev``, and does not
    require heuristics to determine how to reload the command. Only available
    on Python >= 3.10. :issue:`2589`
-  The Watchdog reloader ignores file opened events. Bump the minimum version of
    Watchdog to 2.3.0. :issue:`2603`
-  When using a Unix socket for the development server, the path can start with a dot.
    :issue:`2595`
-  Increase default work factor for PBKDF2 to 600,000 iterations. :issue:`2611`
-  ``parse_options_header`` is 2-3 times faster. It conforms to :rfc:`9110`, some
    invalid parts that were previously accepted are now ignored. :issue:`1628`
-  The ``is_filename`` parameter to ``unquote_header_value`` is deprecated. :pr:`2614`
-  Deprecate the ``extra_chars`` parameter and passing bytes to ``quote_header_value``,
    the ``allow_token`` parameter to ``dump_header``, and the ``cls`` parameter and
    passing bytes to ``parse_dict_header``. :pr:`2618`
-  Improve ``parse_accept_header`` implementation. Parse according to :rfc:`9110`.
    Discard items with invalid ``q`` values. :issue:`1623`
-  ``quote_header_value`` quotes the empty string. :pr:`2618`
-  ``dump_options_header`` skips ``None`` values rather than using a bare key.
    :pr:`2618`
-  ``dump_header`` and ``dump_options_header`` will not quote a value if the key ends
    with an asterisk ``*``.
-  ``parse_dict_header`` will decode values with charsets. :pr:`2618`
-  Refactor the ``Authorization`` and ``WWWAuthenticate`` header data structures.
    :issue:`1769`, :pr:`2619`

    -  Both classes have ``type``, ``parameters``, and ``token`` attributes. The
        ``token`` attribute supports auth schemes that use a single opaque token rather
        than ``key=value`` parameters, such as ``Bearer``.
    -  Neither class is a ``dict`` anymore, although they still implement getting,
        setting, and deleting ``auth[key]`` and ``auth.key`` syntax, as well as
        ``auth.get(key)`` and ``key in auth``.
    -  Both classes have a ``from_header`` class method. ``parse_authorization_header``
        and ``parse_www_authenticate_header`` are deprecated.
    -  The methods ``WWWAuthenticate.set_basic`` and ``set_digest`` are deprecated.
        Instead, an instance should be created and assigned to
        ``response.www_authenticate``.
    -  A list of instances can be assigned to ``response.www_authenticate`` to set
        multiple header values. However, accessing the property only returns the first
        instance.

-  Refactor ``parse_cookie`` and ``dump_cookie``. :pr:`2637`

    -  ``parse_cookie`` is up to 40% faster, ``dump_cookie`` is up to 60% faster.
    -  Passing bytes to ``parse_cookie`` and ``dump_cookie`` is deprecated. The
        ``dump_cookie`` ``charset`` parameter is deprecated.
    -  ``dump_cookie`` allows ``domain`` values that do not include a dot ``.``, and
        strips off a leading dot.
    -  ``dump_cookie`` does not set ``path="/"`` unnecessarily by default.

-  Refactor the test client cookie implementation. :issue:`1060, 1680`

    -  The ``cookie_jar`` attribute is deprecated. ``http.cookiejar`` is no longer used
        for storage.
    -  Domain and path matching is used when sending cookies in requests. The
        ``domain`` and ``path`` parameters default to ``localhost`` and ``/``.
    -  Added a ``get_cookie`` method to inspect cookies.
    -  Cookies have ``decoded_key`` and ``decoded_value`` attributes to match what the
        app sees rather than the encoded values a client would see.
    -  The first positional ``server_name`` parameter to ``set_cookie`` and
        ``delete_cookie`` is deprecated. Use the ``domain`` parameter instead.
    -  Other parameters to ``delete_cookie`` besides ``domain``, ``path``, and
        ``value`` are deprecated.

-  If ``request.max_content_length`` is set, it is checked immediately when accessing
    the stream, and while reading from the stream in general, rather than only during
    form parsing. :issue:`1513`
-  The development server, which must not be used in production, will exhaust the
    request stream up to 10GB or 1000 reads. This allows clients to see a 413 error if
    ``max_content_length`` is exceeded, instead of a "connection reset" failure.
    :pr:`2620`
-  The development server discards header keys that contain underscores ``_``, as they
    are ambiguous with dashes ``-`` in WSGI. :pr:`2622`
-  ``secure_filename`` looks for more Windows reserved file names. :pr:`2623`
-  Update type annotation for ``best_match`` to make ``default`` parameter clearer.
    :issue:`2625`
-  Multipart parser handles empty fields correctly. :issue:`2632`
-  The ``Map`` ``charset`` parameter and ``Request.url_charset`` property are
    deprecated. Percent encoding in URLs must always represent UTF-8 bytes. Invalid
    bytes are left percent encoded rather than replaced. :issue:`2602`
-  The ``Request.charset``, ``Request.encoding_errors``, ``Response.charset``, and
    ``Client.charset`` attributes are deprecated. Request and response data must always
    use UTF-8. :issue:`2602`
-  Header values that have charset information only allow ASCII, UTF-8, and ISO-8859-1.
    :pr:`2614, 2641`
-  Update type annotation for ``ProfilerMiddleware`` ``stream`` parameter.
    :issue:`2642`
-  Use postponed evaluation of annotations. :pr:`2644`
-  The development server escapes ASCII control characters in decoded URLs before
    logging the request to the terminal. :pr:`2652`
-  The ``FormDataParser`` ``parse_functions`` attribute and ``get_parse_func`` method,
    and the invalid ``application/x-url-encoded`` content type, are deprecated.
    :pr:`2653`
-  ``generate_password_hash`` supports scrypt. Plain hash methods are deprecated, only
    scrypt and pbkdf2 are supported. :issue:`2654`

(wiz)

py-pdm: updated to 2.5.4

Release v2.5.4 (2023-05-05)

Bug Fixes
- Pin the urllib3 to `<2.0` to avoid incompatibility with `cachecontrol`.

(adam)

doc: Updated sysutils/py-watchdog to 3.0.0

(wiz)

py-watchdog: update to 3.0.0.

3.0.0
~~~~~

- Drop support for Python 3.6.
- ``watchdog`` is now PEP 561 compatible, and tested with ``mypy``
- Fix missing ``>`` in ``FileSystemEvent.__repr__()``  (`#980 <https://github.com/gorakhargosh/watchdog/pull/980>`__)
- [ci] Lots of improvements
- [inotify] Return from ``InotifyEmitter.queue_events()`` if not launched when thread is inactive (`#963 <https://github.com/gorakhargosh/watchdog/pull/963>`__)
- [tests] Stability improvements
- [utils] Remove handling of ``threading.Event.isSet`` spelling (`#962 <https://github.com/gorakhargosh/watchdog/pull/962>`__)
- [watchmedo] Fixed tricks YAML generation (`#965 <https://github.com/gorakhargosh/watchdog/pull/965>`__)
- Thanks to our beloved contributors: @kurtmckee, @altendky, @agroszer, @BoboTiG

2.3.1
~~~~~

- Run ``black`` on the entire source code
- Bundle the ``requirements-tests.txt`` file in the source distribution (`#939 <https://github.com/gorakhargosh/watchdog/pull/939>`__)
- [watchmedo] Exclude ``FileOpenedEvent`` events from ``AutoRestartTrick``, and ``ShellCommandTrick``, to restore watchdog < 2.3.0 behavior. A better solution should be found in the future. (`#949 <https://github.com/gorakhargosh/watchdog/pull/949>`__)
- [watchmedo] Log ``FileOpenedEvent``, and ``FileClosedEvent``, events in ``LoggerTrick``
- Thanks to our beloved contributors: @BoboTiG

2.3.0
~~~~~

- [inotify] Add support for ``IN_OPEN`` events: a ``FileOpenedEvent`` event will be fired (`#941 <https://github.com/gorakhargosh/watchdog/pull/941>`__)
- [watchmedo] Add optional event debouncing for ``auto-restart``, only restarting once if many events happen in quick succession (``--debounce-interval``) (`#940 <https://github.com/gorakhargosh/watchdog/pull/940>`__)
- [watchmedo] Exit gracefully on ``KeyboardInterrupt`` exception (Ctrl+C) (`#945 <https://github.com/gorakhargosh/watchdog/pull/945>`__)
- [watchmedo] Add option to not auto-restart the command after it exits (``--no-restart-on-command-exit``) (`#946 <https://github.com/gorakhargosh/watchdog/pull/946>`__)
- Thanks to our beloved contributors: @BoboTiG, @dstaple, @taleinat, @cernekj

2.2.1
~~~~~

- Enable ``mypy`` to discover type hints as specified in PEP 561 (`#933 <https://github.com/gorakhargosh/watchdog/pull/933>`__)
- [ci] Set the expected Python version when building release files
- [ci] Update actions versions in use
- [watchmedo] [regression] Fix usage of missing ``signal.SIGHUP`` attribute on non-Unix OSes (`#935 <https://github.com/gorakhargosh/watchdog/pull/935>`__)
- Thanks to our beloved contributors: @BoboTiG, @simon04, @piotrpdev

2.2.0
~~~~~

- [build] Wheels are now available for Python 3.11 (`#932 <https://github.com/gorakhargosh/watchdog/pull/932>`__)
- [documentation] HTML documentation builds are now tested for errors (`#902 <https://github.com/gorakhargosh/watchdog/pull/902>`__)
- [documentation] Fix typos here, and there (`#910 <https://github.com/gorakhargosh/watchdog/pull/910>`__)
- [fsevents2] The ``fsevents2`` observer is now deprecated (`#909 <https://github.com/gorakhargosh/watchdog/pull/909>`__)
- [tests] The error message returned by musl libc for error code ``-1`` is now allowed (`#923 <https://github.com/gorakhargosh/watchdog/pull/923>`__)
- [utils] Remove unnecessary code in ``dirsnapshot.py`` (`#930 <https://github.com/gorakhargosh/watchdog/pull/930>`__)
- [watchmedo] Handle shutdown events from ``SIGHUP`` (`#912 <https://github.com/gorakhargosh/watchdog/pull/912>`__)
- Thanks to our beloved contributors: @kurtmckee, @babymastodon, @QuantumEnergyE, @timgates42, @BoboTiG

(wiz)

Updated www/py-notebook, www/py-widgetsnbextension

(adam)

py-widgetsnbextension: updated to 4.0.7

4.0.7
Unknown changes

(adam)

py-notebook: updated to 6.5.4

6.5.4

Enhancements made

- Add show_banner trait to control the banner display

6.5.3

Enhancements made

- Add a banner and log to information about the migration to Notebook 7 plan
- Add sys_info to page template for 6.5.x

Bugs fixed

- Add .mo and .json files for translations
- Fix rename_file and delete_file to handle hidden files properly

Other merged PRs

- Fix ru_RU translation
- Update kernel translation

6.5.2

Bugs fixed

- Ensure custom preload is correctly handled
- Fix: jQuery-UI 404 Error by updating dependency path in static template

Maintenance and upkeep improvements

- Depend on nbclassic 0.4.7

6.5.1

Merged PRs

- fix: pin temporary to nbclassic 0.4.5

6.5.0

Bugs fixed

- Fix a typo in exception text
- Normalise `os_path`

(adam)

doc: Updated www/py-urllib3 to 2.0.0

(wiz)

py-urllib3: update to 2.0.0.

2.0.0 (2023-04-26)
==================

Read the `v2.0 migration guide <https://urllib3.readthedocs.io/en/latest/v2-migration-guide.html>`__ for help upgrading to the latest version of urllib3.

Removed
-------

* Removed support for Python 2.7, 3.5, and 3.6 (`#883 <https://github.com/urllib3/urllib3/issues/883>`__, `#2336 <https://github.com/urllib3/urllib3/issues/2336>`__).
* Removed fallback on certificate ``commonName`` in ``match_hostname()`` function.
  This behavior was deprecated in May 2000 in RFC 2818. Instead only ``subjectAltName``
  is used to verify the hostname by default. To enable verifying the hostname against
  ``commonName`` use ``SSLContext.hostname_checks_common_name = True`` (`#2113 <https://github.com/urllib3/urllib3/issues/2113>`__).
* Removed support for Python with an ``ssl`` module compiled with LibreSSL, CiscoSSL,
  wolfSSL, and all other OpenSSL alternatives. Python is moving to require OpenSSL with PEP 644 (`#2168 <https://github.com/urllib3/urllib3/issues/2168>`__).
* Removed support for OpenSSL versions earlier than 1.1.1 or that don't have SNI support.
  When an incompatible OpenSSL version is detected an ``ImportError`` is raised (`#2168 <https://github.com/urllib3/urllib3/issues/2168>`__).
* Removed the list of default ciphers for OpenSSL 1.1.1+ and SecureTransport as their own defaults are already secure (`#2082 <https://github.com/urllib3/urllib3/issues/2082>`__).
* Removed ``urllib3.contrib.appengine.AppEngineManager`` and support for Google App Engine Standard Environment (`#2044 <https://github.com/urllib3/urllib3/issues/2044>`__).
* Removed deprecated ``Retry`` options ``method_whitelist``, ``DEFAULT_REDIRECT_HEADERS_BLACKLIST`` (`#2086 <https://github.com/urllib3/urllib3/issues/2086>`__).
* Removed ``urllib3.HTTPResponse.from_httplib`` (`#2648 <https://github.com/urllib3/urllib3/issues/2648>`__).
* Removed default value of ``None`` for the ``request_context`` parameter of ``urllib3.PoolManager.connection_from_pool_key``. This change should have no effect on users as the default value of ``None`` was an invalid option and was never used (`#1897 <https://github.com/urllib3/urllib3/issues/1897>`__).
* Removed the ``urllib3.request`` module. ``urllib3.request.RequestMethods`` has been made a private API.
  This change was made to ensure that ``from urllib3 import request`` imported the top-level ``request()``
  function instead of the ``urllib3.request`` module (`#2269 <https://github.com/urllib3/urllib3/issues/2269>`__).
* Removed support for SSLv3.0 from the ``urllib3.contrib.pyopenssl`` even when support is available from the compiled OpenSSL library (`#2233 <https://github.com/urllib3/urllib3/issues/2233>`__).
* Removed the deprecated ``urllib3.contrib.ntlmpool`` module (`#2339 <https://github.com/urllib3/urllib3/issues/2339>`__).
* Removed ``DEFAULT_CIPHERS``, ``HAS_SNI``, ``USE_DEFAULT_SSLCONTEXT_CIPHERS``, from the private module ``urllib3.util.ssl_`` (`#2168 <https://github.com/urllib3/urllib3/issues/2168>`__).
* Removed ``urllib3.exceptions.SNIMissingWarning`` (`#2168 <https://github.com/urllib3/urllib3/issues/2168>`__).
* Removed the ``_prepare_conn`` method from ``HTTPConnectionPool``. Previously this was only used to call ``HTTPSConnection.set_cert()`` by ``HTTPSConnectionPool`` (`#1985 <https://github.com/urllib3/urllib3/issues/1985>`__).
* Removed ``tls_in_tls_required`` property from ``HTTPSConnection``. This is now determined from the ``scheme`` parameter in ``HTTPConnection.set_tunnel()`` (`#1985 <https://github.com/urllib3/urllib3/issues/1985>`__).

Deprecated
----------

* Deprecated ``HTTPResponse.getheaders()`` and ``HTTPResponse.getheader()`` which will be removed in urllib3 v2.1.0. Instead use ``HTTPResponse.headers`` and ``HTTPResponse.headers.get(name, default)``. (`#1543 <https://github.com/urllib3/urllib3/issues/1543>`__, `#2814 <https://github.com/urllib3/urllib3/issues/2814>`__).
* Deprecated ``urllib3.contrib.pyopenssl`` module which will be removed in urllib3 v2.1.0 (`#2691 <https://github.com/urllib3/urllib3/issues/2691>`__).
* Deprecated ``urllib3.contrib.securetransport`` module which will be removed in urllib3 v2.1.0 (`#2692 <https://github.com/urllib3/urllib3/issues/2692>`__).
* Deprecated ``ssl_version`` option in favor of ``ssl_minimum_version``. ``ssl_version`` will be removed in urllib3 v2.1.0 (`#2110 <https://github.com/urllib3/urllib3/issues/2110>`__).
* Deprecated the ``strict`` parameter as it's not longer needed in Python 3.x. It will be removed in urllib3 v2.1.0 (`#2267 <https://github.com/urllib3/urllib3/issues/2267>`__)
* Deprecated the ``NewConnectionError.pool`` attribute which will be removed in urllib3 v2.1.0 (`#2271 <https://github.com/urllib3/urllib3/issues/2271>`__).
* Deprecated ``format_header_param_html5`` and ``format_header_param`` in favor of ``format_multipart_header_param`` (`#2257 <https://github.com/urllib3/urllib3/issues/2257>`__).
* Deprecated ``RequestField.header_formatter`` parameter which will be removed in urllib3 v2.1.0 (`#2257 <https://github.com/urllib3/urllib3/issues/2257>`__).
* Deprecated ``HTTPSConnection.set_cert()`` method. Instead pass parameters to the ``HTTPSConnection`` constructor (`#1985 <https://github.com/urllib3/urllib3/issues/1985>`__).
* Deprecated ``HTTPConnection.request_chunked()`` method which will be removed in urllib3 v2.1.0. Instead pass ``chunked=True`` to ``HTTPConnection.request()`` (`#1985 <https://github.com/urllib3/urllib3/issues/1985>`__).

Added
-----

* Added top-level ``urllib3.request`` function which uses a preconfigured module-global ``PoolManager`` instance (`#2150 <https://github.com/urllib3/urllib3/issues/2150>`__).
* Added the ``json`` parameter to ``urllib3.request()``, ``PoolManager.request()``, and ``ConnectionPool.request()`` methods to send JSON bodies in requests. Using this parameter will set the header ``Content-Type: application/json`` if ``Content-Type`` isn't already defined.
  Added support for parsing JSON response bodies with ``HTTPResponse.json()`` method (`#2243 <https://github.com/urllib3/urllib3/issues/2243>`__).
* Added type hints to the ``urllib3`` module (`#1897 <https://github.com/urllib3/urllib3/issues/1897>`__).
* Added ``ssl_minimum_version`` and ``ssl_maximum_version`` options which set
  ``SSLContext.minimum_version`` and ``SSLContext.maximum_version`` (`#2110 <https://github.com/urllib3/urllib3/issues/2110>`__).
* Added support for Zstandard (RFC 8878) when ``zstandard`` 1.18.0 or later is installed.
  Added the ``zstd`` extra which installs the ``zstandard`` package (`#1992 <https://github.com/urllib3/urllib3/issues/1992>`__).
* Added ``urllib3.response.BaseHTTPResponse`` class. All future response classes will be subclasses of ``BaseHTTPResponse`` (`#2083 <https://github.com/urllib3/urllib3/issues/2083>`__).
* Added ``FullPoolError`` which is raised when ``PoolManager(block=True)`` and a connection is returned to a full pool (`#2197 <https://github.com/urllib3/urllib3/issues/2197>`__).
* Added ``HTTPHeaderDict`` to the top-level ``urllib3`` namespace (`#2216 <https://github.com/urllib3/urllib3/issues/2216>`__).
* Added support for configuring header merging behavior with HTTPHeaderDict
  When using a ``HTTPHeaderDict`` to provide headers for a request, by default duplicate
  header values will be repeated. But if ``combine=True`` is passed into a call to
  ``HTTPHeaderDict.add``, then the added header value will be merged in with an existing
  value into a comma-separated list (``X-My-Header: foo, bar``) (`#2242 <https://github.com/urllib3/urllib3/issues/2242>`__).
* Added ``NameResolutionError`` exception when a DNS error occurs (`#2305 <https://github.com/urllib3/urllib3/issues/2305>`__).
* Added ``proxy_assert_hostname`` and ``proxy_assert_fingerprint`` kwargs to ``ProxyManager`` (`#2409 <https://github.com/urllib3/urllib3/issues/2409>`__).
* Added a configurable ``backoff_max`` parameter to the ``Retry`` class.
  If a custom ``backoff_max`` is provided to the ``Retry`` class, it
  will replace the ``Retry.DEFAULT_BACKOFF_MAX`` (`#2494 <https://github.com/urllib3/urllib3/issues/2494>`__).
* Added the ``authority`` property to the Url class as per RFC 3986 3.2. This property should be used in place of ``netloc`` for users who want to include the userinfo (auth) component of the URI (`#2520 <https://github.com/urllib3/urllib3/issues/2520>`__).
* Added the ``scheme`` parameter to ``HTTPConnection.set_tunnel`` to configure the scheme of the origin being tunnelled to (`#1985 <https://github.com/urllib3/urllib3/issues/1985>`__).
* Added the ``is_closed``, ``is_connected`` and ``has_connected_to_proxy`` properties to ``HTTPConnection`` (`#1985 <https://github.com/urllib3/urllib3/issues/1985>`__).
* Added optional ``backoff_jitter`` parameter to ``Retry``. (`#2952 <https://github.com/urllib3/urllib3/issues/2952>`__)

Changed
-------

* Changed ``urllib3.response.HTTPResponse.read`` to respect the semantics of ``io.BufferedIOBase`` regardless of compression. Specifically, this method:

  * Only returns an empty bytes object to indicate EOF (that is, the response has been fully consumed).
  * Never returns more bytes than requested.
  * Can issue any number of system calls: zero, one or multiple.

  If you want each ``urllib3.response.HTTPResponse.read`` call to issue a single system call, you need to disable decompression by setting ``decode_content=False`` (`#2128 <https://github.com/urllib3/urllib3/issues/2128>`__).
* Changed ``urllib3.HTTPConnection.getresponse`` to return an instance of ``urllib3.HTTPResponse`` instead of ``http.client.HTTPResponse`` (`#2648 <https://github.com/urllib3/urllib3/issues/2648>`__).
* Changed ``ssl_version`` to instead set the corresponding ``SSLContext.minimum_version``
  and ``SSLContext.maximum_version`` values.  Regardless of ``ssl_version`` passed
  ``SSLContext`` objects are now constructed using ``ssl.PROTOCOL_TLS_CLIENT`` (`#2110 <https://github.com/urllib3/urllib3/issues/2110>`__).
* Changed default ``SSLContext.minimum_version`` to be ``TLSVersion.TLSv1_2`` in line with Python 3.10 (`#2373 <https://github.com/urllib3/urllib3/issues/2373>`__).
* Changed ``ProxyError`` to wrap any connection error (timeout, TLS, DNS) that occurs when connecting to the proxy (`#2482 <https://github.com/urllib3/urllib3/pull/2482>`__).
* Changed ``urllib3.util.create_urllib3_context`` to not override the system cipher suites with
  a default value. The new default will be cipher suites configured by the operating system (`#2168 <https://github.com/urllib3/urllib3/issues/2168>`__).
* Changed ``multipart/form-data`` header parameter formatting matches the WHATWG HTML Standard as of 2021-06-10. Control characters in filenames are no longer percent encoded (`#2257 <https://github.com/urllib3/urllib3/issues/2257>`__).
* Changed the error raised when connecting via HTTPS when the ``ssl`` module isn't available from ``SSLError`` to ``ImportError`` (`#2589 <https://github.com/urllib3/urllib3/issues/2589>`__).
* Changed ``HTTPConnection.request()`` to always use lowercase chunk boundaries when sending requests with ``Transfer-Encoding: chunked`` (`#2515 <https://github.com/urllib3/urllib3/issues/2515>`__).
* Changed ``enforce_content_length`` default to True, preventing silent data loss when reading streamed responses (`#2514 <https://github.com/urllib3/urllib3/issues/2514>`__).
* Changed internal implementation of ``HTTPHeaderDict`` to use ``dict`` instead of ``collections.OrderedDict`` for better performance (`#2080 <https://github.com/urllib3/urllib3/issues/2080>`__).
* Changed the ``urllib3.contrib.pyopenssl`` module to wrap ``OpenSSL.SSL.Error`` with ``ssl.SSLError`` in ``PyOpenSSLContext.load_cert_chain`` (`#2628 <https://github.com/urllib3/urllib3/issues/2628>`__).
* Changed usage of the deprecated ``socket.error`` to ``OSError`` (`#2120 <https://github.com/urllib3/urllib3/issues/2120>`__).
* Changed all parameters in the ``HTTPConnection`` and ``HTTPSConnection`` constructors to be keyword-only except ``host`` and ``port`` (`#1985 <https://github.com/urllib3/urllib3/issues/1985>`__).
* Changed ``HTTPConnection.getresponse()`` to set the socket timeout from ``HTTPConnection.timeout`` value before reading
  data from the socket. This previously was done manually by the ``HTTPConnectionPool`` calling ``HTTPConnection.sock.settimeout(...)`` (`#1985 <https://github.com/urllib3/urllib3/issues/1985>`__).
* Changed the ``_proxy_host`` property to ``_tunnel_host`` in ``HTTPConnectionPool`` to more closely match how the property is used (value in ``HTTPConnection.set_tunnel()``) (`#1985 <https://github.com/urllib3/urllib3/issues/1985>`__).
* Changed name of ``Retry.BACK0FF_MAX`` to be ``Retry.DEFAULT_BACKOFF_MAX``.
* Changed TLS handshakes to use ``SSLContext.check_hostname`` when possible (`#2452 <https://github.com/urllib3/urllib3/pull/2452>`__).
* Changed ``server_hostname`` to behave like other parameters only used by ``HTTPSConnectionPool`` (`#2537 <https://github.com/urllib3/urllib3/pull/2537>`__).
* Changed the default ``blocksize`` to 16KB to match OpenSSL's default read amounts (`#2348 <https://github.com/urllib3/urllib3/pull/2348>`__).
* Changed ``HTTPResponse.read()`` to raise an error when calling with ``decode_content=False`` after using ``decode_content=True`` to prevent data loss (`#2800 <https://github.com/urllib3/urllib3/issues/2800>`__).

Fixed
-----

* Fixed thread-safety issue where accessing a ``PoolManager`` with many distinct origins would cause connection pools to be closed while requests are in progress (`#1252 <https://github.com/urllib3/urllib3/issues/1252>`__).
* Fixed an issue where an ``HTTPConnection`` instance would erroneously reuse the socket read timeout value from reading the previous response instead of a newly configured connect timeout.
  Instead now if ``HTTPConnection.timeout`` is updated before sending the next request the new timeout value will be used (`#2645 <https://github.com/urllib3/urllib3/issues/2645>`__).
* Fixed ``socket.error.errno`` when raised from pyOpenSSL's ``OpenSSL.SSL.SysCallError`` (`#2118 <https://github.com/urllib3/urllib3/issues/2118>`__).
* Fixed the default value of ``HTTPSConnection.socket_options`` to match ``HTTPConnection`` (`#2213 <https://github.com/urllib3/urllib3/issues/2213>`__).
* Fixed a bug where ``headers`` would be modified by the ``remove_headers_on_redirect`` feature (`#2272 <https://github.com/urllib3/urllib3/issues/2272>`__).
* Fixed a reference cycle bug in ``urllib3.util.connection.create_connection()`` (`#2277 <https://github.com/urllib3/urllib3/issues/2277>`__).
* Fixed a socket leak if ``HTTPConnection.connect()`` fails (`#2571 <https://github.com/urllib3/urllib3/pull/2571>`__).
* Fixed ``urllib3.contrib.pyopenssl.WrappedSocket`` and ``urllib3.contrib.securetransport.WrappedSocket`` close methods (`#2970 <https://github.com/urllib3/urllib3/issues/2970>`__)

(wiz)

Added www/py-nbclassic, www/py-notebook_shim

(adam)

py-notebook_shim: added version 0.2.3

This project provides a way for JupyterLab and other frontends to switch to
Jupyter Server for their Python Web application backend.

(adam)

py-nbclassic: added version 1.0.0

Jupyter NBClassic is a web application for making and sharing live codes
with documents, equations, and plots. It supports multiple programming
languages.

(adam)

doc: Added devel/py-click-default-group version 1.2.2

(wiz)

devel/Makefile: + 1

(wiz)

Added devel/py-hatch-jupyter-builder, www/py-jupyter_server

(adam)

devel/py-click-default-group: import py-click-default-group-1.2.2

`DefaultGroup` is a sub class of click.Group, but it invokes a
default subcommand instead of showing a help message when a subcommand
is not passed.

(wiz)

py-jupyter_server: added version 2.5.0

The Jupyter Server provides the backend (i.e. the core services, APIs, and REST
endpoints) for Jupyter web applications like Jupyter notebook, JupyterLab, and
Voila.

(adam)

py-hatch-jupyter-builder: added version 0.8.3

This provides a build hook plugin for Hatch that adds a build step for use with
Jupyter packages.

(adam)

doc: Updated devel/py-coverage to 7.2.5

(wiz)

py-coverage: update to 7.2.5.

Version 7.2.5 — 2023-04-30
--------------------------

- Fix: ``html_report()`` could fail with an AttributeError on ``isatty`` if run
  in an unusual environment where sys.stdout had been replaced.  This is now
  fixed.

Version 7.2.4 — 2023-04-28
--------------------------

PyCon 2023 sprint fixes!

- Fix: with ``relative_files = true``, specifying a specific file to include or
  omit wouldn't work correctly (`issue 1604`_).  This is now fixed, with
  testing help by `Marc Gibbons <pull 1608_>`_.

- Fix: the XML report would have an incorrect ``<source>`` element when using
  relative files and the source option ended with a slash (`issue 1541`_).
  This is now fixed, thanks to `Kevin Brown-Silva <pull 1608_>`_.

- When the HTML report location is printed to the terminal, it's now a
  terminal-compatible URL, so that you can click the location to open the HTML
  file in your browser.  Finishes `issue 1523`_ thanks to `Ricardo Newbery
  <pull 1613_>`_.

- Docs: a new :ref:`Migrating page <migrating>` with details about how to
  migrate between major versions of coverage.py.  It currently covers the
  wildcard changes in 7.x.  Thanks, `Brian Grohe <pull 1610_>`_.

.. _issue 1523: https://github.com/nedbat/coveragepy/issues/1523
.. _issue 1541: https://github.com/nedbat/coveragepy/issues/1541
.. _issue 1604: https://github.com/nedbat/coveragepy/issues/1604
.. _pull 1608: https://github.com/nedbat/coveragepy/pull/1608
.. _pull 1609: https://github.com/nedbat/coveragepy/pull/1609
.. _pull 1610: https://github.com/nedbat/coveragepy/pull/1610
.. _pull 1613: https://github.com/nedbat/coveragepy/pull/1613

Version 7.2.3 — 2023-04-06
--------------------------

- Fix: the :ref:`config_run_sigterm` setting was meant to capture data if a
  process was terminated with a SIGTERM signal, but it didn't always.  This was
  fixed thanks to `Lewis Gaul <pull 1600_>`_, closing `issue 1599`_.

- Performance: HTML reports with context information are now much more compact.
  File sizes are typically as small as one-third the previous size, but can be
  dramatically smaller. This closes `issue 1584`_ thanks to `Oleh Krehel
  <pull 1587_>`_.

- Development dependencies no longer use hashed pins, closing `issue 1592`_.

.. _issue 1584: https://github.com/nedbat/coveragepy/issues/1584
.. _pull 1587: https://github.com/nedbat/coveragepy/pull/1587
.. _issue 1592: https://github.com/nedbat/coveragepy/issues/1592
.. _issue 1599: https://github.com/nedbat/coveragepy/issues/1599
.. _pull 1600: https://github.com/nedbat/coveragepy/pull/1600

Version 7.2.2 — 2023-03-16
--------------------------

- Fix: if a virtualenv was created inside a source directory, and a sourced
  package was installed inside the virtualenv, then all of the third-party
  packages inside the virtualenv would be measured.  This was incorrect, but
  has now been fixed: only the specified packages will be measured, thanks to
  `Manuel Jacob <pull 1560_>`_.

- Fix: the ``coverage lcov`` command could create a .lcov file with incorrect
  LF (lines found) and LH (lines hit) totals.  This is now fixed, thanks to
  `Ian Moore <pull 1583_>`_.

- Fix: the ``coverage xml`` command on Windows could create a .xml file with
  duplicate ``<package>`` elements. This is now fixed, thanks to `Benjamin
  Parzella <pull 1574_>`_, closing `issue 1573`_.

.. _pull 1560: https://github.com/nedbat/coveragepy/pull/1560
.. _issue 1573: https://github.com/nedbat/coveragepy/issues/1573
.. _pull 1574: https://github.com/nedbat/coveragepy/pull/1574
.. _pull 1583: https://github.com/nedbat/coveragepy/pull/1583

Version 7.2.1 — 2023-02-26
--------------------------

- Fix: the PyPI page had broken links to documentation pages, but no longer
  does, closing `issue 1566`_.

- Fix: public members of the coverage module are now properly indicated so that
  mypy will find them, fixing `issue 1564`_.

.. _issue 1564: https://github.com/nedbat/coveragepy/issues/1564
.. _issue 1566: https://github.com/nedbat/coveragepy/issues/1566

Version 7.2.0 — 2023-02-22
--------------------------

- Added a new setting ``[report] exclude_also`` to let you add more exclusions
  without overwriting the defaults.  Thanks, `Alpha Chen <pull 1557_>`_,
  closing `issue 1391`_.

- Added a :meth:`.CoverageData.purge_files` method to remove recorded data for
  a particular file.  Contributed by `Stephan Deibel <pull 1547_>`_.

- Fix: when reporting commands fail, they will no longer congratulate
  themselves with messages like "Wrote XML report to file.xml" before spewing a
  traceback about their failure.

- Fix: arguments in the public API that name file paths now accept pathlib.Path
  objects.  This includes the ``data_file`` and ``config_file`` arguments to
  the Coverage constructor and the ``basename`` argument to CoverageData.
  Closes `issue 1552`_.

- Fix: In some embedded environments, an IndexError could occur on stop() when
  the originating thread exits before completion.  This is now fixed, thanks to
  `Russell Keith-Magee <pull 1543_>`_, closing `issue 1542`_.

- Added a ``py.typed`` file to announce our type-hintedness.  Thanks,
  `KotlinIsland <pull 1550_>`_.

.. _issue 1391: https://github.com/nedbat/coveragepy/issues/1391
.. _issue 1542: https://github.com/nedbat/coveragepy/issues/1542
.. _pull 1543: https://github.com/nedbat/coveragepy/pull/1543
.. _pull 1547: https://github.com/nedbat/coveragepy/pull/1547
.. _pull 1550: https://github.com/nedbat/coveragepy/pull/1550
.. _issue 1552: https://github.com/nedbat/coveragepy/issues/1552
.. _pull 1557: https://github.com/nedbat/coveragepy/pull/1557

Version 7.1.0 — 2023-01-24
--------------------------

- Added: the debug output file can now be specified with ``[run] debug_file``
  in the configuration file.  Closes `issue 1319`_.

- Performance: fixed a slowdown with dynamic contexts that's been around since
  6.4.3.  The fix closes `issue 1538`_.  Thankfully this doesn't break the
  `Cython change`_ that fixed `issue 972`_.  Thanks to Mathieu Kniewallner for
  the deep investigative work and comprehensive issue report.

- Typing: all product and test code has type annotations.

.. _Cython change: https://github.com/nedbat/coveragepy/pull/1347
.. _issue 972: https://github.com/nedbat/coveragepy/issues/972
.. _issue 1319: https://github.com/nedbat/coveragepy/issues/1319
.. _issue 1538: https://github.com/nedbat/coveragepy/issues/1538

Version 7.0.5 — 2023-01-10
--------------------------

- Fix: On Python 3.7, a file with type annotations but no ``from __future__
  import annotations`` would be missing statements in the coverage report. This
  is now fixed, closing `issue 1524`_.

.. _issue 1524: https://github.com/nedbat/coveragepy/issues/1524

Version 7.0.4 — 2023-01-07
--------------------------

- Performance: an internal cache of file names was accidentally disabled,
  resulting in sometimes drastic reductions in performance.  This is now fixed,
  closing `issue 1527`_.  Thanks to Ivan Ciuvalschii for the reproducible test
  case.

.. _issue 1527: https://github.com/nedbat/coveragepy/issues/1527

Version 7.0.3 — 2023-01-03
--------------------------

- Fix: when using pytest-cov or pytest-xdist, or perhaps both, the combining
  step could fail with ``assert row is not None`` using 7.0.2.  This was due to
  a race condition that has always been possible and is still possible. In
  7.0.1 and before, the error was silently swallowed by the combining code.
  Now it will produce a message "Couldn't combine data file" and ignore the
  data file as it used to do before 7.0.2.  Closes `issue 1522`_.

.. _issue 1522: https://github.com/nedbat/coveragepy/issues/1522

Version 7.0.2 — 2023-01-02
--------------------------

- Fix: when using the ``[run] relative_files = True`` setting, a relative
  ``[paths]`` pattern was still being made absolute.  This is now fixed,
  closing `issue 1519`_.

- Fix: if Python doesn't provide tomllib, then TOML configuration files can
  only be read if coverage.py is installed with the ``[toml]`` extra.
  Coverage.py will raise an error if TOML support is not installed when it sees
  your settings are in a .toml file. But it didn't understand that
  ``[tools.coverage]`` was a valid section header, so the error wasn't reported
  if you used that header, and settings were silently ignored.  This is now
  fixed, closing `issue 1516`_.

- Fix: adjusted how decorators are traced on PyPy 7.3.10, fixing `issue 1515`_.

- Fix: the ``coverage lcov`` report did not properly implement the
  ``--fail-under=MIN`` option.  This has been fixed.

- Refactor: added many type annotations, including a number of refactorings.
  This should not affect outward behavior, but they were a bit invasive in some
  places, so keep your eyes peeled for oddities.

- Refactor: removed the vestigial and long untested support for Jython and
  IronPython.

.. _issue 1515: https://github.com/nedbat/coveragepy/issues/1515
.. _issue 1516: https://github.com/nedbat/coveragepy/issues/1516
.. _issue 1519: https://github.com/nedbat/coveragepy/issues/1519

Version 7.0.1 — 2022-12-23
--------------------------

- When checking if a file mapping resolved to a file that exists, we weren't
  considering files in .whl files.  This is now fixed, closing `issue 1511`_.

- File pattern rules were too strict, forbidding plus signs and curly braces in
  directory and file names.  This is now fixed, closing `issue 1513`_.

- Unusual Unicode or control characters in source files could prevent
  reporting.  This is now fixed, closing `issue 1512`_.

- The PyPy wheel now installs on PyPy 3.7, 3.8, and 3.9, closing `issue 1510`_.

.. _issue 1510: https://github.com/nedbat/coveragepy/issues/1510
.. _issue 1511: https://github.com/nedbat/coveragepy/issues/1511
.. _issue 1512: https://github.com/nedbat/coveragepy/issues/1512
.. _issue 1513: https://github.com/nedbat/coveragepy/issues/1513

Version 7.0.0 — 2022-12-18
--------------------------

Nothing new beyond 7.0.0b1.

Version 7.0.0b1 — 2022-12-03
----------------------------

A number of changes have been made to file path handling, including pattern
matching and path remapping with the ``[paths]`` setting (see
:ref:`config_paths`).  These changes might affect you, and require you to
update your settings.

(This release includes the changes from `6.6.0b1 <changes_6-6-0b1_>`_, since
6.6.0 was never released.)

- Changes to file pattern matching, which might require updating your
  configuration:

  - Previously, ``*`` would incorrectly match directory separators, making
    precise matching difficult.  This is now fixed, closing `issue 1407`_.

  - Now ``**`` matches any number of nested directories, including none.

- Improvements to combining data files when using the
  :ref:`config_run_relative_files` setting, which might require updating your
  configuration:

  - During ``coverage combine``, relative file paths are implicitly combined
    without needing a ``[paths]`` configuration setting.  This also fixed
    `issue 991`_.

  - A ``[paths]`` setting like ``*/foo`` will now match ``foo/bar.py`` so that
    relative file paths can be combined more easily.

  - The :ref:`config_run_relative_files` setting is properly interpreted in
    more places, fixing `issue 1280`_.

- When remapping file paths with ``[paths]``, a path will be remapped only if
  the resulting path exists.  The documentation has long said the prefix had to
  exist, but it was never enforced.  This fixes `issue 608`_, improves `issue
  649`_, and closes `issue 757`_.

- Reporting operations now implicitly use the ``[paths]`` setting to remap file
  paths within a single data file.  Combining multiple files still requires the
  ``coverage combine`` step, but this simplifies some single-file situations.
  Closes `issue 1212`_ and `issue 713`_.

- The ``coverage report`` command now has a ``--format=`` option.  The original
  style is now ``--format=text``, and is the default.

  - Using ``--format=markdown`` will write the table in Markdown format, thanks
    to `Steve Oswald <pull 1479_>`_, closing `issue 1418`_.

  - Using ``--format=total`` will write a single total number to the
    output.  This can be useful for making badges or writing status updates.

- Combining data files with ``coverage combine`` now hashes the data files to
  skip files that add no new information.  This can reduce the time needed.
  Many details affect the speed-up, but for coverage.py's own test suite,
  combining is about 40% faster. Closes `issue 1483`_.

- When searching for completely un-executed files, coverage.py uses the
  presence of ``__init__.py`` files to determine which directories have source
  that could have been imported.  However, `implicit namespace packages`_ don't
  require ``__init__.py``.  A new setting ``[report]
  include_namespace_packages`` tells coverage.py to consider these directories
  during reporting.  Thanks to `Felix Horvat <pull 1387_>`_ for the
  contribution.  Closes `issue 1383`_ and `issue 1024`_.

- Fixed environment variable expansion in pyproject.toml files.  It was overly
  broad, causing errors outside of coverage.py settings, as described in `issue
  1481`_ and `issue 1345`_.  This is now fixed, but in rare cases will require
  changing your pyproject.toml to quote non-string values that use environment
  substitution.

- An empty file has a coverage total of 100%, but used to fail with
  ``--fail-under``.  This has been fixed, closing `issue 1470`_.

- The text report table no longer writes out two separator lines if there are
  no files listed in the table.  One is plenty.

- Fixed a mis-measurement of a strange use of wildcard alternatives in
  match/case statements, closing `issue 1421`_.

- Fixed internal logic that prevented coverage.py from running on
  implementations other than CPython or PyPy (`issue 1474`_).

- The deprecated ``[run] note`` setting has been completely removed.

.. _implicit namespace packages: https://peps.python.org/pep-0420/
.. _issue 608: https://github.com/nedbat/coveragepy/issues/608
.. _issue 649: https://github.com/nedbat/coveragepy/issues/649
.. _issue 713: https://github.com/nedbat/coveragepy/issues/713
.. _issue 757: https://github.com/nedbat/coveragepy/issues/757
.. _issue 991: https://github.com/nedbat/coveragepy/issues/991
.. _issue 1024: https://github.com/nedbat/coveragepy/issues/1024
.. _issue 1212: https://github.com/nedbat/coveragepy/issues/1212
.. _issue 1280: https://github.com/nedbat/coveragepy/issues/1280
.. _issue 1345: https://github.com/nedbat/coveragepy/issues/1345
.. _issue 1383: https://github.com/nedbat/coveragepy/issues/1383
.. _issue 1407: https://github.com/nedbat/coveragepy/issues/1407
.. _issue 1418: https://github.com/nedbat/coveragepy/issues/1418
.. _issue 1421: https://github.com/nedbat/coveragepy/issues/1421
.. _issue 1470: https://github.com/nedbat/coveragepy/issues/1470
.. _issue 1474: https://github.com/nedbat/coveragepy/issues/1474
.. _issue 1481: https://github.com/nedbat/coveragepy/issues/1481
.. _issue 1483: https://github.com/nedbat/coveragepy/issues/1483
.. _pull 1387: https://github.com/nedbat/coveragepy/pull/1387
.. _pull 1479: https://github.com/nedbat/coveragepy/pull/1479

Version 6.6.0b1 — 2022-10-31
----------------------------

(Note: 6.6.0 final was never released. These changes are part of `7.0.0b1
<changes_7-0-0b1_>`_.)

- Changes to file pattern matching, which might require updating your
  configuration:

  - Previously, ``*`` would incorrectly match directory separators, making
    precise matching difficult.  This is now fixed, closing `issue 1407`_.

  - Now ``**`` matches any number of nested directories, including none.

- Improvements to combining data files when using the
  :ref:`config_run_relative_files` setting:

  - During ``coverage combine``, relative file paths are implicitly combined
    without needing a ``[paths]`` configuration setting.  This also fixed
    `issue 991`_.

  - A ``[paths]`` setting like ``*/foo`` will now match ``foo/bar.py`` so that
    relative file paths can be combined more easily.

  - The setting is properly interpreted in more places, fixing `issue 1280`_.

- Fixed environment variable expansion in pyproject.toml files.  It was overly
  broad, causing errors outside of coverage.py settings, as described in `issue
  1481`_ and `issue 1345`_.  This is now fixed, but in rare cases will require
  changing your pyproject.toml to quote non-string values that use environment
  substitution.

- Fixed internal logic that prevented coverage.py from running on
  implementations other than CPython or PyPy (`issue 1474`_).

.. _issue 991: https://github.com/nedbat/coveragepy/issues/991
.. _issue 1280: https://github.com/nedbat/coveragepy/issues/1280
.. _issue 1345: https://github.com/nedbat/coveragepy/issues/1345
.. _issue 1407: https://github.com/nedbat/coveragepy/issues/1407
.. _issue 1474: https://github.com/nedbat/coveragepy/issues/1474
.. _issue 1481: https://github.com/nedbat/coveragepy/issues/1481

(wiz)