This commit is contained in:
2025-04-06 03:14:47 +02:00
parent aaf9ab523b
commit b9c99befab
2263 changed files with 401112 additions and 20 deletions

View File

@ -0,0 +1 @@
pip

View File

@ -0,0 +1,174 @@
Apache License
Version 2.0, January 2004
http://www.apache.org/licenses/
TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
1. Definitions.
"License" shall mean the terms and conditions for use, reproduction,
and distribution as defined by Sections 1 through 9 of this document.
"Licensor" shall mean the copyright owner or entity authorized by
the copyright owner that is granting the License.
"Legal Entity" shall mean the union of the acting entity and all
other entities that control, are controlled by, or are under common
control with that entity. For the purposes of this definition,
"control" means (i) the power, direct or indirect, to cause the
direction or management of such entity, whether by contract or
otherwise, or (ii) ownership of fifty percent (50%) or more of the
outstanding shares, or (iii) beneficial ownership of such entity.
"You" (or "Your") shall mean an individual or Legal Entity
exercising permissions granted by this License.
"Source" form shall mean the preferred form for making modifications,
including but not limited to software source code, documentation
source, and configuration files.
"Object" form shall mean any form resulting from mechanical
transformation or translation of a Source form, including but
not limited to compiled object code, generated documentation,
and conversions to other media types.
"Work" shall mean the work of authorship, whether in Source or
Object form, made available under the License, as indicated by a
copyright notice that is included in or attached to the work
(an example is provided in the Appendix below).
"Derivative Works" shall mean any work, whether in Source or Object
form, that is based on (or derived from) the Work and for which the
editorial revisions, annotations, elaborations, or other modifications
represent, as a whole, an original work of authorship. For the purposes
of this License, Derivative Works shall not include works that remain
separable from, or merely link (or bind by name) to the interfaces of,
the Work and Derivative Works thereof.
"Contribution" shall mean any work of authorship, including
the original version of the Work and any modifications or additions
to that Work or Derivative Works thereof, that is intentionally
submitted to Licensor for inclusion in the Work by the copyright owner
or by an individual or Legal Entity authorized to submit on behalf of
the copyright owner. For the purposes of this definition, "submitted"
means any form of electronic, verbal, or written communication sent
to the Licensor or its representatives, including but not limited to
communication on electronic mailing lists, source code control systems,
and issue tracking systems that are managed by, or on behalf of, the
Licensor for the purpose of discussing and improving the Work, but
excluding communication that is conspicuously marked or otherwise
designated in writing by the copyright owner as "Not a Contribution."
"Contributor" shall mean Licensor and any individual or Legal Entity
on behalf of whom a Contribution has been received by Licensor and
subsequently incorporated within the Work.
2. Grant of Copyright License. Subject to the terms and conditions of
this License, each Contributor hereby grants to You a perpetual,
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
copyright license to reproduce, prepare Derivative Works of,
publicly display, publicly perform, sublicense, and distribute the
Work and such Derivative Works in Source or Object form.
3. Grant of Patent License. Subject to the terms and conditions of
this License, each Contributor hereby grants to You a perpetual,
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
(except as stated in this section) patent license to make, have made,
use, offer to sell, sell, import, and otherwise transfer the Work,
where such license applies only to those patent claims licensable
by such Contributor that are necessarily infringed by their
Contribution(s) alone or by combination of their Contribution(s)
with the Work to which such Contribution(s) was submitted. If You
institute patent litigation against any entity (including a
cross-claim or counterclaim in a lawsuit) alleging that the Work
or a Contribution incorporated within the Work constitutes direct
or contributory patent infringement, then any patent licenses
granted to You under this License for that Work shall terminate
as of the date such litigation is filed.
4. Redistribution. You may reproduce and distribute copies of the
Work or Derivative Works thereof in any medium, with or without
modifications, and in Source or Object form, provided that You
meet the following conditions:
(a) You must give any other recipients of the Work or
Derivative Works a copy of this License; and
(b) You must cause any modified files to carry prominent notices
stating that You changed the files; and
(c) You must retain, in the Source form of any Derivative Works
that You distribute, all copyright, patent, trademark, and
attribution notices from the Source form of the Work,
excluding those notices that do not pertain to any part of
the Derivative Works; and
(d) If the Work includes a "NOTICE" text file as part of its
distribution, then any Derivative Works that You distribute must
include a readable copy of the attribution notices contained
within such NOTICE file, excluding those notices that do not
pertain to any part of the Derivative Works, in at least one
of the following places: within a NOTICE text file distributed
as part of the Derivative Works; within the Source form or
documentation, if provided along with the Derivative Works; or,
within a display generated by the Derivative Works, if and
wherever such third-party notices normally appear. The contents
of the NOTICE file are for informational purposes only and
do not modify the License. You may add Your own attribution
notices within Derivative Works that You distribute, alongside
or as an addendum to the NOTICE text from the Work, provided
that such additional attribution notices cannot be construed
as modifying the License.
You may add Your own copyright statement to Your modifications and
may provide additional or different license terms and conditions
for use, reproduction, or distribution of Your modifications, or
for any such Derivative Works as a whole, provided Your use,
reproduction, and distribution of the Work otherwise complies with
the conditions stated in this License.
5. Submission of Contributions. Unless You explicitly state otherwise,
any Contribution intentionally submitted for inclusion in the Work
by You to the Licensor shall be under the terms and conditions of
this License, without any additional terms or conditions.
Notwithstanding the above, nothing herein shall supersede or modify
the terms of any separate license agreement you may have executed
with Licensor regarding such Contributions.
6. Trademarks. This License does not grant permission to use the trade
names, trademarks, service marks, or product names of the Licensor,
except as required for reasonable and customary use in describing the
origin of the Work and reproducing the content of the NOTICE file.
7. Disclaimer of Warranty. Unless required by applicable law or
agreed to in writing, Licensor provides the Work (and each
Contributor provides its Contributions) on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
implied, including, without limitation, any warranties or conditions
of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
PARTICULAR PURPOSE. You are solely responsible for determining the
appropriateness of using or redistributing the Work and assume any
risks associated with Your exercise of permissions under this License.
8. Limitation of Liability. In no event and under no legal theory,
whether in tort (including negligence), contract, or otherwise,
unless required by applicable law (such as deliberate and grossly
negligent acts) or agreed to in writing, shall any Contributor be
liable to You for damages, including any direct, indirect, special,
incidental, or consequential damages of any character arising as a
result of this License or out of the use or inability to use the
Work (including but not limited to damages for loss of goodwill,
work stoppage, computer failure or malfunction, or any and all
other commercial damages or losses), even if such Contributor
has been advised of the possibility of such damages.
9. Accepting Warranty or Additional Liability. While redistributing
the Work or Derivative Works thereof, You may choose to offer,
and charge a fee for, acceptance of support, warranty, indemnity,
or other liability obligations and/or rights consistent with this
License. However, in accepting such obligations, You may act only
on Your own behalf and on Your sole responsibility, not on behalf
of any other Contributor, and only if You agree to indemnify,
defend, and hold each Contributor harmless for any liability
incurred by, or claims asserted against, such Contributor by reason
of your accepting any such warranty or additional liability.

View File

@ -0,0 +1,245 @@
Metadata-Version: 2.1
Name: PyNaCl
Version: 1.5.0
Summary: Python binding to the Networking and Cryptography (NaCl) library
Home-page: https://github.com/pyca/pynacl/
Author: The PyNaCl developers
Author-email: cryptography-dev@python.org
License: Apache License 2.0
Platform: UNKNOWN
Classifier: Programming Language :: Python :: Implementation :: CPython
Classifier: Programming Language :: Python :: Implementation :: PyPy
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.6
Classifier: Programming Language :: Python :: 3.7
Classifier: Programming Language :: Python :: 3.8
Classifier: Programming Language :: Python :: 3.9
Classifier: Programming Language :: Python :: 3.10
Requires-Python: >=3.6
Requires-Dist: cffi (>=1.4.1)
Provides-Extra: docs
Requires-Dist: sphinx (>=1.6.5) ; extra == 'docs'
Requires-Dist: sphinx-rtd-theme ; extra == 'docs'
Provides-Extra: tests
Requires-Dist: pytest (!=3.3.0,>=3.2.1) ; extra == 'tests'
Requires-Dist: hypothesis (>=3.27.0) ; extra == 'tests'
===============================================
PyNaCl: Python binding to the libsodium library
===============================================
.. image:: https://img.shields.io/pypi/v/pynacl.svg
:target: https://pypi.org/project/PyNaCl/
:alt: Latest Version
.. image:: https://codecov.io/github/pyca/pynacl/coverage.svg?branch=main
:target: https://codecov.io/github/pyca/pynacl?branch=main
.. image:: https://img.shields.io/pypi/pyversions/pynacl.svg
:target: https://pypi.org/project/PyNaCl/
:alt: Compatible Python Versions
PyNaCl is a Python binding to `libsodium`_, which is a fork of the
`Networking and Cryptography library`_. These libraries have a stated goal of
improving usability, security and speed. It supports Python 3.6+ as well as
PyPy 3.
.. _libsodium: https://github.com/jedisct1/libsodium
.. _Networking and Cryptography library: https://nacl.cr.yp.to/
Features
--------
* Digital signatures
* Secret-key encryption
* Public-key encryption
* Hashing and message authentication
* Password based key derivation and password hashing
`Changelog`_
------------
.. _Changelog: https://pynacl.readthedocs.io/en/stable/changelog/
Installation
============
Binary wheel install
--------------------
PyNaCl ships as a binary wheel on macOS, Windows and Linux ``manylinux1`` [#many]_ ,
so all dependencies are included. Make sure you have an up-to-date pip
and run:
.. code-block:: console
$ pip install pynacl
Faster wheel build
------------------
You can define the environment variable ``LIBSODIUM_MAKE_ARGS`` to pass arguments to ``make``
and enable `parallelization`_:
.. code-block:: console
$ LIBSODIUM_MAKE_ARGS=-j4 pip install pynacl
Linux source build
------------------
PyNaCl relies on `libsodium`_, a portable C library. A copy is bundled
with PyNaCl so to install you can run:
.. code-block:: console
$ pip install pynacl
If you'd prefer to use the version of ``libsodium`` provided by your
distribution, you can disable the bundled copy during install by running:
.. code-block:: console
$ SODIUM_INSTALL=system pip install pynacl
.. warning:: Usage of the legacy ``easy_install`` command provided by setuptools
is generally discouraged, and is completely unsupported in PyNaCl's case.
.. _parallelization: https://www.gnu.org/software/make/manual/html_node/Parallel.html
.. _libsodium: https://github.com/jedisct1/libsodium
.. [#many] `manylinux1 wheels <https://www.python.org/dev/peps/pep-0513/>`_
are built on a baseline linux environment based on Centos 5.11
and should work on most x86 and x86_64 glibc based linux environments.
Changelog
=========
1.5.0 (2022-01-07)
------------------
* **BACKWARDS INCOMPATIBLE:** Removed support for Python 2.7 and Python 3.5.
* **BACKWARDS INCOMPATIBLE:** We no longer distribute ``manylinux1``
wheels.
* Added ``manylinux2014``, ``manylinux_2_24``, ``musllinux``, and macOS
``universal2`` wheels (the latter supports macOS ``arm64``).
* Update ``libsodium`` to 1.0.18-stable (July 25, 2021 release).
* Add inline type hints.
1.4.0 (2020-05-25)
------------------
* Update ``libsodium`` to 1.0.18.
* **BACKWARDS INCOMPATIBLE:** We no longer distribute 32-bit ``manylinux1``
wheels. Continuing to produce them was a maintenance burden.
* Added support for Python 3.8, and removed support for Python 3.4.
* Add low level bindings for extracting the seed and the public key
from crypto_sign_ed25519 secret key
* Add low level bindings for deterministic random generation.
* Add ``wheel`` and ``setuptools`` setup_requirements in ``setup.py`` (#485)
* Fix checks on very slow builders (#481, #495)
* Add low-level bindings to ed25519 arithmetic functions
* Update low-level blake2b state implementation
* Fix wrong short-input behavior of SealedBox.decrypt() (#517)
* Raise CryptPrefixError exception instead of InvalidkeyError when trying
to check a password against a verifier stored in a unknown format (#519)
* Add support for minimal builds of libsodium. Trying to call functions
not available in a minimal build will raise an UnavailableError
exception. To compile a minimal build of the bundled libsodium, set
the SODIUM_INSTALL_MINIMAL environment variable to any non-empty
string (e.g. ``SODIUM_INSTALL_MINIMAL=1``) for setup.
1.3.0 2018-09-26
----------------
* Added support for Python 3.7.
* Update ``libsodium`` to 1.0.16.
* Run and test all code examples in PyNaCl docs through sphinx's
doctest builder.
* Add low-level bindings for chacha20-poly1305 AEAD constructions.
* Add low-level bindings for the chacha20-poly1305 secretstream constructions.
* Add low-level bindings for ed25519ph pre-hashed signing construction.
* Add low-level bindings for constant-time increment and addition
on fixed-precision big integers represented as little-endian
byte sequences.
* Add low-level bindings for the ISO/IEC 7816-4 compatible padding API.
* Add low-level bindings for libsodium's crypto_kx... key exchange
construction.
* Set hypothesis deadline to None in tests/test_pwhash.py to avoid
incorrect test failures on slower processor architectures. GitHub
issue #370
1.2.1 - 2017-12-04
------------------
* Update hypothesis minimum allowed version.
* Infrastructure: add proper configuration for readthedocs builder
runtime environment.
1.2.0 - 2017-11-01
------------------
* Update ``libsodium`` to 1.0.15.
* Infrastructure: add jenkins support for automatic build of
``manylinux1`` binary wheels
* Added support for ``SealedBox`` construction.
* Added support for ``argon2i`` and ``argon2id`` password hashing constructs
and restructured high-level password hashing implementation to expose
the same interface for all hashers.
* Added support for 128 bit ``siphashx24`` variant of ``siphash24``.
* Added support for ``from_seed`` APIs for X25519 keypair generation.
* Dropped support for Python 3.3.
1.1.2 - 2017-03-31
------------------
* reorder link time library search path when using bundled
libsodium
1.1.1 - 2017-03-15
------------------
* Fixed a circular import bug in ``nacl.utils``.
1.1.0 - 2017-03-14
------------------
* Dropped support for Python 2.6.
* Added ``shared_key()`` method on ``Box``.
* You can now pass ``None`` to ``nonce`` when encrypting with ``Box`` or
``SecretBox`` and it will automatically generate a random nonce.
* Added support for ``siphash24``.
* Added support for ``blake2b``.
* Added support for ``scrypt``.
* Update ``libsodium`` to 1.0.11.
* Default to the bundled ``libsodium`` when compiling.
* All raised exceptions are defined mixing-in
``nacl.exceptions.CryptoError``
1.0.1 - 2016-01-24
------------------
* Fix an issue with absolute paths that prevented the creation of wheels.
1.0 - 2016-01-23
----------------
* PyNaCl has been ported to use the new APIs available in cffi 1.0+.
Due to this change we no longer support PyPy releases older than 2.6.
* Python 3.2 support has been dropped.
* Functions to convert between Ed25519 and Curve25519 keys have been added.
0.3.0 - 2015-03-04
------------------
* The low-level API (`nacl.c.*`) has been changed to match the
upstream NaCl C/C++ conventions (as well as those of other NaCl bindings).
The order of arguments and return values has changed significantly. To
avoid silent failures, `nacl.c` has been removed, and replaced with
`nacl.bindings` (with the new argument ordering). If you have code which
calls these functions (e.g. `nacl.c.crypto_box_keypair()`), you must review
the new docstrings and update your code/imports to match the new
conventions.

View File

@ -0,0 +1,68 @@
PyNaCl-1.5.0.dist-info/INSTALLER,sha256=zuuue4knoyJ-UwPPXg8fezS7VCrXJQrAP7zeNuwvFQg,4
PyNaCl-1.5.0.dist-info/LICENSE,sha256=0xdK1j5yHUydzLitQyCEiZLTFDabxGMZcgtYAskVP-k,9694
PyNaCl-1.5.0.dist-info/METADATA,sha256=OJaXCiHgNRywLY9cj3X2euddUPZ4dnyyqAQMU01X4j0,8634
PyNaCl-1.5.0.dist-info/RECORD,,
PyNaCl-1.5.0.dist-info/WHEEL,sha256=TIQeZFe3DwXBO5UGlCH1aKpf5Cx6FJLbIUqd-Sq2juI,185
PyNaCl-1.5.0.dist-info/top_level.txt,sha256=wfdEOI_G2RIzmzsMyhpqP17HUh6Jcqi99to9aHLEslo,13
nacl/__init__.py,sha256=0IUunzBT8_Jn0DUdHacBExOYeAEMggo8slkfjo7O0XM,1116
nacl/__pycache__/__init__.cpython-311.pyc,,
nacl/__pycache__/encoding.cpython-311.pyc,,
nacl/__pycache__/exceptions.cpython-311.pyc,,
nacl/__pycache__/hash.cpython-311.pyc,,
nacl/__pycache__/hashlib.cpython-311.pyc,,
nacl/__pycache__/public.cpython-311.pyc,,
nacl/__pycache__/secret.cpython-311.pyc,,
nacl/__pycache__/signing.cpython-311.pyc,,
nacl/__pycache__/utils.cpython-311.pyc,,
nacl/_sodium.abi3.so,sha256=uJ6RwSnbb9wO4esR0bVUqrfFHtBOGm34IQIdmaE1fGY,2740136
nacl/bindings/__init__.py,sha256=BDlStrds2EuUS4swOL4pnf92PWVS_CHRCptX3KhEX-s,16997
nacl/bindings/__pycache__/__init__.cpython-311.pyc,,
nacl/bindings/__pycache__/crypto_aead.cpython-311.pyc,,
nacl/bindings/__pycache__/crypto_box.cpython-311.pyc,,
nacl/bindings/__pycache__/crypto_core.cpython-311.pyc,,
nacl/bindings/__pycache__/crypto_generichash.cpython-311.pyc,,
nacl/bindings/__pycache__/crypto_hash.cpython-311.pyc,,
nacl/bindings/__pycache__/crypto_kx.cpython-311.pyc,,
nacl/bindings/__pycache__/crypto_pwhash.cpython-311.pyc,,
nacl/bindings/__pycache__/crypto_scalarmult.cpython-311.pyc,,
nacl/bindings/__pycache__/crypto_secretbox.cpython-311.pyc,,
nacl/bindings/__pycache__/crypto_secretstream.cpython-311.pyc,,
nacl/bindings/__pycache__/crypto_shorthash.cpython-311.pyc,,
nacl/bindings/__pycache__/crypto_sign.cpython-311.pyc,,
nacl/bindings/__pycache__/randombytes.cpython-311.pyc,,
nacl/bindings/__pycache__/sodium_core.cpython-311.pyc,,
nacl/bindings/__pycache__/utils.cpython-311.pyc,,
nacl/bindings/crypto_aead.py,sha256=BIw1k_JCfr5ylZk0RF5rCFIM1fhfLkEa-aiWkrfffNE,15597
nacl/bindings/crypto_box.py,sha256=Ox0NG2t4MsGhBAa7Kgah4o0gc99ULMsqkdX56ofOouY,10139
nacl/bindings/crypto_core.py,sha256=6u9G3y7H-QrawO785UkFFFtwDoCkeHE63GOUl9p5-eA,13736
nacl/bindings/crypto_generichash.py,sha256=9mX0DGIIzicr-uXrqFM1nU4tirasbixDwbcdfV7W1fc,8852
nacl/bindings/crypto_hash.py,sha256=Rg1rsEwE3azhsQT-dNVPA4NB9VogJAKn1EfxYt0pPe0,2175
nacl/bindings/crypto_kx.py,sha256=oZNVlNgROpHOa1XQ_uZe0tqIkdfuApeJlRnwR23_74k,6723
nacl/bindings/crypto_pwhash.py,sha256=laVDo4xFUuGyEjtZAU510AklBF6ablBy7Z3HN1WDYjY,18848
nacl/bindings/crypto_scalarmult.py,sha256=_DX-mst2uCnzjo6fP5HRTnhv1BC95B9gmJc3L_or16g,8244
nacl/bindings/crypto_secretbox.py,sha256=KgZ1VvkCJDlQ85jtfe9c02VofPvuEgZEhWni-aX3MsM,2914
nacl/bindings/crypto_secretstream.py,sha256=G0FgZS01qA5RzWzm5Bdms8Yy_lvgdZDoUYYBActPmvQ,11165
nacl/bindings/crypto_shorthash.py,sha256=PQU7djHTLDGdVs-w_TsivjFHHp5EK5k2Yh6p-6z0T60,2603
nacl/bindings/crypto_sign.py,sha256=53j2im9E4F79qT_2U8IfCAc3lzg0VMwEjvAPEUccVDg,10342
nacl/bindings/randombytes.py,sha256=uBK3W4WcjgnjZdWanrX0fjYZpr9KHbBgNMl9rui-Ojc,1563
nacl/bindings/sodium_core.py,sha256=9Y9CX--sq-TaPaQRPRpx8SWDSS9PJOja_Cqb-yqyJNQ,1039
nacl/bindings/utils.py,sha256=KDwQnadXeNMbqEA1SmpNyCVo5k8MiUQa07QM66VzfXM,4298
nacl/encoding.py,sha256=qTAPc2MXSkdh4cqDVY0ra6kHyViHMCmEo_re7cgGk5w,2915
nacl/exceptions.py,sha256=GZH32aJtZgqCO4uz0LRsev8z0WyvAYuV3YVqT9AAQq4,2451
nacl/hash.py,sha256=EYBOe6UVc9SUQINEmyuRSa1QGRSvdwdrBzTL1tdFLU8,6392
nacl/hashlib.py,sha256=L5Fv75St8AMPvb-GhA4YqX5p1mC_Sb4HhC1NxNQMpJA,4400
nacl/public.py,sha256=RVGCWQRjIJOmW-8sNrVLtsDjMMGx30i6UyfViGCnQNA,14792
nacl/pwhash/__init__.py,sha256=XSDXd7wQHNLEHl0mkHfVb5lFQsp6ygHkhen718h0BSM,2675
nacl/pwhash/__pycache__/__init__.cpython-311.pyc,,
nacl/pwhash/__pycache__/_argon2.cpython-311.pyc,,
nacl/pwhash/__pycache__/argon2i.cpython-311.pyc,,
nacl/pwhash/__pycache__/argon2id.cpython-311.pyc,,
nacl/pwhash/__pycache__/scrypt.cpython-311.pyc,,
nacl/pwhash/_argon2.py,sha256=jL1ChR9biwYh3RSuc-LJ2-W4DlVLHpir-XHGX8cpeJQ,1779
nacl/pwhash/argon2i.py,sha256=IIvIuO9siKUu5-Wpz0SGiltLQv7Du_mi9BUE8INRK_4,4405
nacl/pwhash/argon2id.py,sha256=H22i8O4j9Ws4L3JsXl9TRcJzDcyaVumhQRPzINAgJWM,4433
nacl/pwhash/scrypt.py,sha256=fMr3Qht1a1EY8aebNNntfLRjinIPXtKYKKrrBhY5LDc,6986
nacl/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
nacl/secret.py,sha256=kauBNuP-0rb3TjU2EMBMu5Vnmzjnscp1bRqMspy5LzU,12108
nacl/signing.py,sha256=kbTEUyHLUMaNLv1nCjxzGxCs82Qs5w8gxE_CnEwPuIU,8337
nacl/utils.py,sha256=gmlTD1x9ZNwzHd8LpALH1CHud-Htv8ejRb3y7TyS9f0,2341

View File

@ -0,0 +1,7 @@
Wheel-Version: 1.0
Generator: bdist_wheel (0.37.1)
Root-Is-Purelib: false
Tag: cp36-abi3-manylinux_2_17_x86_64
Tag: cp36-abi3-manylinux2014_x86_64
Tag: cp36-abi3-manylinux_2_24_x86_64

View File

@ -0,0 +1,2 @@
_sodium
nacl

View File

@ -0,0 +1 @@
pip

View File

@ -0,0 +1,20 @@
Copyright (c) 2017-2021 Ingy döt Net
Copyright (c) 2006-2016 Kirill Simonov
Permission is hereby granted, free of charge, to any person obtaining a copy of
this software and associated documentation files (the "Software"), to deal in
the Software without restriction, including without limitation the rights to
use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies
of the Software, and to permit persons to whom the Software is furnished to do
so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.

View File

@ -0,0 +1,46 @@
Metadata-Version: 2.1
Name: PyYAML
Version: 6.0.2
Summary: YAML parser and emitter for Python
Home-page: https://pyyaml.org/
Download-URL: https://pypi.org/project/PyYAML/
Author: Kirill Simonov
Author-email: xi@resolvent.net
License: MIT
Project-URL: Bug Tracker, https://github.com/yaml/pyyaml/issues
Project-URL: CI, https://github.com/yaml/pyyaml/actions
Project-URL: Documentation, https://pyyaml.org/wiki/PyYAMLDocumentation
Project-URL: Mailing lists, http://lists.sourceforge.net/lists/listinfo/yaml-core
Project-URL: Source Code, https://github.com/yaml/pyyaml
Platform: Any
Classifier: Development Status :: 5 - Production/Stable
Classifier: Intended Audience :: Developers
Classifier: License :: OSI Approved :: MIT License
Classifier: Operating System :: OS Independent
Classifier: Programming Language :: Cython
Classifier: Programming Language :: Python
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.8
Classifier: Programming Language :: Python :: 3.9
Classifier: Programming Language :: Python :: 3.10
Classifier: Programming Language :: Python :: 3.11
Classifier: Programming Language :: Python :: 3.12
Classifier: Programming Language :: Python :: 3.13
Classifier: Programming Language :: Python :: Implementation :: CPython
Classifier: Programming Language :: Python :: Implementation :: PyPy
Classifier: Topic :: Software Development :: Libraries :: Python Modules
Classifier: Topic :: Text Processing :: Markup
Requires-Python: >=3.8
License-File: LICENSE
YAML is a data serialization format designed for human readability
and interaction with scripting languages. PyYAML is a YAML parser
and emitter for Python.
PyYAML features a complete YAML 1.1 parser, Unicode support, pickle
support, capable extension API, and sensible error messages. PyYAML
supports standard YAML tags and provides Python-specific tags that
allow to represent an arbitrary Python object.
PyYAML is applicable for a broad range of tasks from complex
configuration files to object serialization and persistence.

View File

@ -0,0 +1,44 @@
PyYAML-6.0.2.dist-info/INSTALLER,sha256=zuuue4knoyJ-UwPPXg8fezS7VCrXJQrAP7zeNuwvFQg,4
PyYAML-6.0.2.dist-info/LICENSE,sha256=jTko-dxEkP1jVwfLiOsmvXZBAqcoKVQwfT5RZ6V36KQ,1101
PyYAML-6.0.2.dist-info/METADATA,sha256=9-odFB5seu4pGPcEv7E8iyxNF51_uKnaNGjLAhz2lto,2060
PyYAML-6.0.2.dist-info/RECORD,,
PyYAML-6.0.2.dist-info/REQUESTED,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
PyYAML-6.0.2.dist-info/WHEEL,sha256=YWWHkv6sHhBDPNqgSfLklIm4KZnZJH4x2lIHOwCoU7Q,152
PyYAML-6.0.2.dist-info/top_level.txt,sha256=rpj0IVMTisAjh_1vG3Ccf9v5jpCQwAz6cD1IVU5ZdhQ,11
_yaml/__init__.py,sha256=04Ae_5osxahpJHa3XBZUAf4wi6XX32gR8D6X6p64GEA,1402
_yaml/__pycache__/__init__.cpython-311.pyc,,
yaml/__init__.py,sha256=N35S01HMesFTe0aRRMWkPj0Pa8IEbHpE9FK7cr5Bdtw,12311
yaml/__pycache__/__init__.cpython-311.pyc,,
yaml/__pycache__/composer.cpython-311.pyc,,
yaml/__pycache__/constructor.cpython-311.pyc,,
yaml/__pycache__/cyaml.cpython-311.pyc,,
yaml/__pycache__/dumper.cpython-311.pyc,,
yaml/__pycache__/emitter.cpython-311.pyc,,
yaml/__pycache__/error.cpython-311.pyc,,
yaml/__pycache__/events.cpython-311.pyc,,
yaml/__pycache__/loader.cpython-311.pyc,,
yaml/__pycache__/nodes.cpython-311.pyc,,
yaml/__pycache__/parser.cpython-311.pyc,,
yaml/__pycache__/reader.cpython-311.pyc,,
yaml/__pycache__/representer.cpython-311.pyc,,
yaml/__pycache__/resolver.cpython-311.pyc,,
yaml/__pycache__/scanner.cpython-311.pyc,,
yaml/__pycache__/serializer.cpython-311.pyc,,
yaml/__pycache__/tokens.cpython-311.pyc,,
yaml/_yaml.cpython-311-x86_64-linux-gnu.so,sha256=sZBsAqPs6VM8YzOkHpNL0qKIfR0zNM9gttjzjoVPaiI,2466120
yaml/composer.py,sha256=_Ko30Wr6eDWUeUpauUGT3Lcg9QPBnOPVlTnIMRGJ9FM,4883
yaml/constructor.py,sha256=kNgkfaeLUkwQYY_Q6Ff1Tz2XVw_pG1xVE9Ak7z-viLA,28639
yaml/cyaml.py,sha256=6ZrAG9fAYvdVe2FK_w0hmXoG7ZYsoYUwapG8CiC72H0,3851
yaml/dumper.py,sha256=PLctZlYwZLp7XmeUdwRuv4nYOZ2UBnDIUy8-lKfLF-o,2837
yaml/emitter.py,sha256=jghtaU7eFwg31bG0B7RZea_29Adi9CKmXq_QjgQpCkQ,43006
yaml/error.py,sha256=Ah9z-toHJUbE9j-M8YpxgSRM5CgLCcwVzJgLLRF2Fxo,2533
yaml/events.py,sha256=50_TksgQiE4up-lKo_V-nBy-tAIxkIPQxY5qDhKCeHw,2445
yaml/loader.py,sha256=UVa-zIqmkFSCIYq_PgSGm4NSJttHY2Rf_zQ4_b1fHN0,2061
yaml/nodes.py,sha256=gPKNj8pKCdh2d4gr3gIYINnPOaOxGhJAUiYhGRnPE84,1440
yaml/parser.py,sha256=ilWp5vvgoHFGzvOZDItFoGjD6D42nhlZrZyjAwa0oJo,25495
yaml/reader.py,sha256=0dmzirOiDG4Xo41RnuQS7K9rkY3xjHiVasfDMNTqCNw,6794
yaml/representer.py,sha256=IuWP-cAW9sHKEnS0gCqSa894k1Bg4cgTxaDwIcbRQ-Y,14190
yaml/resolver.py,sha256=9L-VYfm4mWHxUD1Vg4X7rjDRK_7VZd6b92wzq7Y2IKY,9004
yaml/scanner.py,sha256=YEM3iLZSaQwXcQRg2l2R4MdT0zGP2F9eHkKGKnHyWQY,51279
yaml/serializer.py,sha256=ChuFgmhU01hj4xgI8GaKv6vfM2Bujwa9i7d2FAHj7cA,4165
yaml/tokens.py,sha256=lTQIzSVw8Mg9wv459-TjiOQe6wVziqaRlqX2_89rp54,2573

View File

@ -0,0 +1,6 @@
Wheel-Version: 1.0
Generator: bdist_wheel (0.44.0)
Root-Is-Purelib: false
Tag: cp311-cp311-manylinux_2_17_x86_64
Tag: cp311-cp311-manylinux2014_x86_64

View File

@ -0,0 +1,2 @@
_yaml
yaml

View File

@ -0,0 +1,222 @@
# don't import any costly modules
import sys
import os
is_pypy = '__pypy__' in sys.builtin_module_names
def warn_distutils_present():
if 'distutils' not in sys.modules:
return
if is_pypy and sys.version_info < (3, 7):
# PyPy for 3.6 unconditionally imports distutils, so bypass the warning
# https://foss.heptapod.net/pypy/pypy/-/blob/be829135bc0d758997b3566062999ee8b23872b4/lib-python/3/site.py#L250
return
import warnings
warnings.warn(
"Distutils was imported before Setuptools, but importing Setuptools "
"also replaces the `distutils` module in `sys.modules`. This may lead "
"to undesirable behaviors or errors. To avoid these issues, avoid "
"using distutils directly, ensure that setuptools is installed in the "
"traditional way (e.g. not an editable install), and/or make sure "
"that setuptools is always imported before distutils."
)
def clear_distutils():
if 'distutils' not in sys.modules:
return
import warnings
warnings.warn("Setuptools is replacing distutils.")
mods = [
name
for name in sys.modules
if name == "distutils" or name.startswith("distutils.")
]
for name in mods:
del sys.modules[name]
def enabled():
"""
Allow selection of distutils by environment variable.
"""
which = os.environ.get('SETUPTOOLS_USE_DISTUTILS', 'local')
return which == 'local'
def ensure_local_distutils():
import importlib
clear_distutils()
# With the DistutilsMetaFinder in place,
# perform an import to cause distutils to be
# loaded from setuptools._distutils. Ref #2906.
with shim():
importlib.import_module('distutils')
# check that submodules load as expected
core = importlib.import_module('distutils.core')
assert '_distutils' in core.__file__, core.__file__
assert 'setuptools._distutils.log' not in sys.modules
def do_override():
"""
Ensure that the local copy of distutils is preferred over stdlib.
See https://github.com/pypa/setuptools/issues/417#issuecomment-392298401
for more motivation.
"""
if enabled():
warn_distutils_present()
ensure_local_distutils()
class _TrivialRe:
def __init__(self, *patterns):
self._patterns = patterns
def match(self, string):
return all(pat in string for pat in self._patterns)
class DistutilsMetaFinder:
def find_spec(self, fullname, path, target=None):
# optimization: only consider top level modules and those
# found in the CPython test suite.
if path is not None and not fullname.startswith('test.'):
return
method_name = 'spec_for_{fullname}'.format(**locals())
method = getattr(self, method_name, lambda: None)
return method()
def spec_for_distutils(self):
if self.is_cpython():
return
import importlib
import importlib.abc
import importlib.util
try:
mod = importlib.import_module('setuptools._distutils')
except Exception:
# There are a couple of cases where setuptools._distutils
# may not be present:
# - An older Setuptools without a local distutils is
# taking precedence. Ref #2957.
# - Path manipulation during sitecustomize removes
# setuptools from the path but only after the hook
# has been loaded. Ref #2980.
# In either case, fall back to stdlib behavior.
return
class DistutilsLoader(importlib.abc.Loader):
def create_module(self, spec):
mod.__name__ = 'distutils'
return mod
def exec_module(self, module):
pass
return importlib.util.spec_from_loader(
'distutils', DistutilsLoader(), origin=mod.__file__
)
@staticmethod
def is_cpython():
"""
Suppress supplying distutils for CPython (build and tests).
Ref #2965 and #3007.
"""
return os.path.isfile('pybuilddir.txt')
def spec_for_pip(self):
"""
Ensure stdlib distutils when running under pip.
See pypa/pip#8761 for rationale.
"""
if self.pip_imported_during_build():
return
clear_distutils()
self.spec_for_distutils = lambda: None
@classmethod
def pip_imported_during_build(cls):
"""
Detect if pip is being imported in a build script. Ref #2355.
"""
import traceback
return any(
cls.frame_file_is_setup(frame) for frame, line in traceback.walk_stack(None)
)
@staticmethod
def frame_file_is_setup(frame):
"""
Return True if the indicated frame suggests a setup.py file.
"""
# some frames may not have __file__ (#2940)
return frame.f_globals.get('__file__', '').endswith('setup.py')
def spec_for_sensitive_tests(self):
"""
Ensure stdlib distutils when running select tests under CPython.
python/cpython#91169
"""
clear_distutils()
self.spec_for_distutils = lambda: None
sensitive_tests = (
[
'test.test_distutils',
'test.test_peg_generator',
'test.test_importlib',
]
if sys.version_info < (3, 10)
else [
'test.test_distutils',
]
)
for name in DistutilsMetaFinder.sensitive_tests:
setattr(
DistutilsMetaFinder,
f'spec_for_{name}',
DistutilsMetaFinder.spec_for_sensitive_tests,
)
DISTUTILS_FINDER = DistutilsMetaFinder()
def add_shim():
DISTUTILS_FINDER in sys.meta_path or insert_shim()
class shim:
def __enter__(self):
insert_shim()
def __exit__(self, exc, value, tb):
remove_shim()
def insert_shim():
sys.meta_path.insert(0, DISTUTILS_FINDER)
def remove_shim():
try:
sys.meta_path.remove(DISTUTILS_FINDER)
except ValueError:
pass

View File

@ -0,0 +1 @@
__import__('_distutils_hack').do_override()

View File

@ -0,0 +1,33 @@
# This is a stub package designed to roughly emulate the _yaml
# extension module, which previously existed as a standalone module
# and has been moved into the `yaml` package namespace.
# It does not perfectly mimic its old counterpart, but should get
# close enough for anyone who's relying on it even when they shouldn't.
import yaml
# in some circumstances, the yaml module we imoprted may be from a different version, so we need
# to tread carefully when poking at it here (it may not have the attributes we expect)
if not getattr(yaml, '__with_libyaml__', False):
from sys import version_info
exc = ModuleNotFoundError if version_info >= (3, 6) else ImportError
raise exc("No module named '_yaml'")
else:
from yaml._yaml import *
import warnings
warnings.warn(
'The _yaml extension module is now located at yaml._yaml'
' and its location is subject to change. To use the'
' LibYAML-based parser and emitter, import from `yaml`:'
' `from yaml import CLoader as Loader, CDumper as Dumper`.',
DeprecationWarning
)
del warnings
# Don't `del yaml` here because yaml is actually an existing
# namespace member of _yaml.
__name__ = '_yaml'
# If the module is top-level (i.e. not a part of any specific package)
# then the attribute should be set to ''.
# https://docs.python.org/3.8/library/types.html
__package__ = ''

View File

@ -0,0 +1,422 @@
Metadata-Version: 2.1
Name: autorandr
Version: 1.15.post1
Summary: Automatically select a display configuration based on connected devices
Home-page: https://github.com/phillipberndt/autorandr
Author: Phillip Berndt
Author-email: phillip.berndt@googlemail.com
License: GPLv3
Keywords: xrandr
Platform: UNKNOWN
Classifier: Environment :: Console
Classifier: Intended Audience :: End Users/Desktop
Classifier: License :: OSI Approved :: GNU General Public License v3 (GPLv3)
Classifier: Programming Language :: Python :: 2
Classifier: Programming Language :: Python :: 2.7
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.3
Classifier: Programming Language :: Python :: 3.4
Classifier: Programming Language :: Python :: 3.6
Classifier: Programming Language :: Python :: 3.7
Classifier: Programming Language :: Python :: 3.8
Classifier: Programming Language :: Python :: 3.9
Classifier: Programming Language :: Python :: 3.10
Description-Content-Type: text/markdown
# autorandr
Automatically select a display configuration based on connected devices
## Branch information
This is a compatible Python rewrite of
[wertarbyte/autorandr](https://github.com/wertarbyte/autorandr). Contributions
for bash-completion, fd.o/XDG autostart, Nitrogen, pm-utils, and systemd can be
found under [contrib](contrib/).
The original [wertarbyte/autorandr](https://github.com/wertarbyte/autorandr)
tree is unmaintained, with lots of open pull requests and issues. I forked it
and merged what I thought were the most important changes. If you are searching
for that version, see the [`legacy` branch](https://github.com/phillipberndt/autorandr/tree/legacy).
Note that the Python version is better suited for non-standard configurations,
like if you use `--transform` or `--reflect`. If you use `auto-disper`, you
have to use the bash version, as there is no disper support in the Python
version (yet). Both versions use a compatible configuration file format, so
you can, to some extent, switch between them. I will maintain the `legacy`
branch until @wertarbyte finds the time to maintain his branch again.
If you are interested in why there are two versions around, see
[#7](https://github.com/phillipberndt/autorandr/issues/7),
[#8](https://github.com/phillipberndt/autorandr/issues/8) and
especially
[#12](https://github.com/phillipberndt/autorandr/issues/12)
if you are unhappy with this version and would like to contribute to the bash
version.
## License information and authors
autorandr is available under the terms of the GNU General Public License
(version 3).
Contributors to this version of autorandr are:
* Adrián López
* andersonjacob
* Alexander Lochmann
* Alexander Wirt
* Brice Waegeneire
* Chris Dunder
* Christoph Gysin
* Christophe-Marie Duquesne
* Daniel Hahler
* Maciej Sitarz
* Mathias Svensson
* Matthew R Johnson
* Nazar Mokrynskyi
* Phillip Berndt
* Rasmus Wriedt Larsen
* Sam Coulter
* Simon Wydooghe
* Stefan Tomanek
* stormc
* tachylatus
* Timo Bingmann
* Timo Kaufmann
* Tomasz Bogdal
* Victor Häggqvist
* Jan-Oliver Kaiser
* Alexandre Viau
## Installation/removal
You can use the `autorandr.py` script as a stand-alone binary. If you'd like to
install it as a system-wide application, there is a Makefile included that also
places some configuration files in appropriate directories such that autorandr
is invoked automatically when a monitor is connected or removed, the system
wakes up from suspend, or a user logs into an X11 session. Run `make install`
as root to install it.
If you prefer to have a system wide install managed by your package manager,
you can
* Use the [official Arch package](https://archlinux.org/packages/extra/any/autorandr/).
* Use the [official Debian package](https://packages.debian.org/sid/x11/autorandr) on sid
* Use the [FreeBSD Ports Collection](https://www.freshports.org/x11/autorandr/) on FreeBSD.
* Use the [official Gentoo package](https://packages.gentoo.org/packages/x11-misc/autorandr).
* Use the
[nix package](https://github.com/NixOS/nixpkgs/blob/master/nixos/modules/services/misc/autorandr.nix)
on NixOS.
* Use the
[guix package](https://git.savannah.gnu.org/cgit/guix.git/log/gnu/packages/xdisorg.scm?qt=grep&q=autorandr)
on Guix.
* Use the [SlackBuild](https://slackbuilds.org/repository/14.2/desktop/autorandr/) on Slackware.
* Use the automated nightlies generated by the
[openSUSE build service](https://build.opensuse.org/package/show/home:phillipberndt/autorandr)
for various distributions (RPM and DEB based).
* Use the [X binary package system](https://wiki.voidlinux.eu/XBPS)' on Void Linux
* Build a .deb-file from the source tree using `make deb`.
* Build a .rpm-file from the source tree using `make rpm`.
We appreciate packaging scripts for other distributions, please file a pull
request if you write one.
If you prefer `pip` over your package manager, you can install autorandr with:
sudo pip install "git+http://github.com/phillipberndt/autorandr#egg=autorandr"
or simply
sudo pip install autorandr
if you prefer to use a stable version.
## How to use
Save your current display configuration and setup with:
autorandr --save mobile
Connect an additional display, configure your setup and save it:
autorandr --save docked
Now autorandr can detect which hardware setup is active:
$ autorandr
mobile
docked (detected)
To automatically reload your setup:
$ autorandr --change
To manually load a profile:
$ autorandr --load <profile>
or simply:
$ autorandr <profile>
autorandr tries to avoid reloading an identical configuration. To force the
(re)configuration:
$ autorandr --load <profile> --force
To prevent a profile from being loaded, place a script call _block_ in its
directory. The script is evaluated before the screen setup is inspected, and
in case of it returning a value of 0 the profile is skipped. This can be used
to query the status of a docking station you are about to leave.
If no suitable profile can be identified, the current configuration is kept.
To change this behaviour and switch to a fallback configuration, specify
`--default <profile>`. The system-wide installation of autorandr by default
calls autorandr with a parameter `--default default`. There are three special,
virtual configurations called `horizontal`, `vertical` and `common`. They
automatically generate a configuration that incorporates all screens
connected to the computer. You can symlink `default` to one of these
names in your configuration directory to have autorandr use any of them
as the default configuration without you having to change the system-wide
configuration.
You can store default values for any option in an INI-file located at
`~/.config/autorandr/settings.ini`. In a `config` section, you may place any
default values in the form `option-name=option-argument`.
A common and effective use of this is to specify default `skip-options`, for
instance skipping the `gamma` setting if using
[`redshift`](https://github.com/jonls/redshift) as a daemon. To implement
the equivalent of `--skip-options gamma`, your `settings.ini` file should look
like this:
```
[config]
skip-options=gamma
```
## Advanced usage
### Hook scripts
Three more scripts can be placed in the configuration directory
(as defined by the [XDG spec](https://specifications.freedesktop.org/basedir-spec/basedir-spec-latest.html),
usually `~/.config/autorandr` or `~/.autorandr` if you have an old installation
for user configuration and `/etc/xdg/autorandr` for system wide configuration):
- `postswitch` is executed *after* a mode switch has taken place. This can be
used to notify window managers or other applications about the switch.
- `preswitch` is executed *before* a mode switch takes place.
- `postsave` is executed after a profile was stored or altered.
- `predetect` is executed before autorandr attempts to run xrandr.
These scripts must be executable and can be placed directly in the configuration
directory, where they will always be executed, or in the profile subdirectories,
where they will only be executed on changes regarding that specific profile.
Instead (or in addition) to these scripts, you can also place as many executable
files as you like in subdirectories called `script_name.d` (e.g. `postswitch.d`).
The order of execution of scripts in these directories is by file name, you can
force a certain ordering by naming them `10-wallpaper`, `20-restart-wm`, etc.
If a script with the same name occurs multiple times, user configuration
takes precedence over system configuration (as specified by the
[XDG spec](https://specifications.freedesktop.org/basedir-spec/basedir-spec-latest.html))
and profile configuration over general configuration.
As a concrete example, suppose you have the files
- `/etc/xdg/autorandr/postswitch`
- `~/.config/autorandr/postswitch`
- `~/.config/autorandr/postswitch.d/notify-herbstluftwm`
- `~/.config/autorandr/docked/postswitch`
and switch from `mobile` to `docked`. Then
`~/.config/autorandr/docked/postswitch` is executed, since the profile specific
configuration takes precedence, and
`~/.config/autorandr/postswitch.d/notify-herbstluftwm` is executed, since
it has a unique name.
If you switch back from `docked` to `mobile`, `~/.config/autorandr/postswitch`
is executed instead of the `docked` specific `postswitch`.
If you experience issues with xrandr being executed too early after connecting
a new monitor, then you can use a `predetect` script to delay the execution.
Write e.g. `sleep 1` into that file to make autorandr wait a second before
running `xrandr`.
#### Variables
Some of autorandr's state is exposed as environment variables
prefixed with `AUTORANDR_`, such as:
- `AUTORANDR_CURRENT_PROFILE`
- `AUTORANDR_CURRENT_PROFILES`
- `AUTORANDR_PROFILE_FOLDER`
- `AUTORANDR_MONITORS`
with the intention that they can be used within the hook scripts.
For instance, you might display which profile has just been activated by
including the following in a `postswitch` script:
```sh
notify-send -i display "Display profile" "$AUTORANDR_CURRENT_PROFILE"
```
The one kink is that during `preswitch`, `AUTORANDR_CURRENT_PROFILE` is
reporting the *upcoming* profile rather than the *current* one.
### Wildcard EDID matching
The EDID strings in the `~/.config/autorandr/*/setup` files may contain an
asterisk to enable wildcard matching: Such EDIDs are matched against connected
monitors using the usual file name globbing rules. This can be used to create
profiles matching multiple (or any) monitors.
### udev triggers with NVidia cards
In order for `udev` to detect `drm` events from the native NVidia driver, the
kernel parameter `nvidia-drm.modeset` must be set to 1. For example, add a file
`/etc/modprobe.d/nvidia-drm-modeset.conf`:
```
options nvidia_drm modeset=1
```
### Wayland
Before running autorandr will check the environment for the `WAYLAND_DISPLAY`
variable to check if the program is running in a Wayland session. This is to
avoid issues between usage of xrandr in Wayland environments.
If you need to run autorandr in a Wayland environment, one workaround is to
unset the `WAYLAND_DISPLAY` variable before running the program, such as:
```
WAYLAND_DISPLAY= autorandr
```
## Changelog
**autorandr 1.15**
* *2023-11-27* Several regex literal bug fixes
* *2023-12-27* Fix #375: Listen to correct events in launcher
* *2024-03-03* Fix #367: Skip profiles without outputs
**autorandr 1.14**
* *2023-06-22* Direct --match-edid renaming of output messages to stderr
* *2023-06-22* Add Wayland awareness
* *2023-06-22* Various minor auxiliary tooling bug fixes, see git-log
**autorandr 1.13.3**
* *2023-01-24* Revert udev rule to rely on "change" event (see #324)
**autorandr 1.13.2**
* *2023-01-23* Fix autostart in KDE (see #320)
* *2023-01-23* Match add/remove rather than change in udev rule (see #321)
* *2023-01-23* Fix wildcard use in EDIDs (see #322)
* *2023-01-23* Do a final xrandr call to set the frame buffer size (see #319)
**autorandr 1.13.1**
* *2023-01-16* Fix bug with Version comparison
**autorandr 1.13**
* *2023-01-15* Add reversed horizontal/vertical profiles
* *2023-01-15* Fix distutils deprecation warning
* *2023-01-15* Print error when user script fails
* *2022-12-01* Support `--skip-options set` to skip setting properties
**autorandr 1.12.1**
* *2021-12-22* Fix `--match-edid` (see #273)
**autorandr 1.12**
* *2021-12-16* Switch default interpreter to Python 3
* *2021-12-16* Add `--list` to list all profiles
* *2021-12-16* Add `--cycle` to cycle all detected profiles
* *2021-12-16* Store display properties (see #204)
**autorandr 1.11**
* *2020-05-23* Handle empty sys.executable
* *2020-06-08* Fix Python 2 compatibility
* *2020-10-06* Set group membership of users in batch mode
**autorandr 1.10.1**
* *2020-05-04* Revert making the launcher the default (fixes #195)
**autorandr 1.10**
* *2020-04-23* Fix hook script execution order to match description from readme
* *2020-04-11* Handle negative gamma values (fixes #188)
* *2020-04-11* Sort approximate matches in detected profiles by quality of match
* *2020-01-31* Handle non-ASCII environment variables (fixes #180)
* *2019-12-31* Fix output positioning if the top-left output is not the first
* *2019-12-31* Accept negative gamma values (and interpret them as 0)
* *2019-12-31* Prefer the X11 launcher over systemd/udev configuration
**autorandr 1.9**
* *2019-11-10* Count closed lids as disconnected outputs
* *2019-10-05* Do not overwrite existing configurations without `--force`
* *2019-08-16* Accept modes that don't match the WWWxHHH pattern
* *2019-03-22* Improve bash autocompletion
* *2019-03-21* Store CRTC values in configurations
* *2019-03-24* Fix handling of recently disconnected outputs (See #128 and #143)
**autorandr 1.8.1**
* *2019-03-18* Removed mandb call from Makefile
**autorandr 1.8**
* *2019-02-17* Add an X11 daemon that runs autorandr when a display connects (by @rliou92, #127)
* *2019-02-17* Replace width=0 check with disconnected to detect disconnected monitors (by @joseph-jones, #139)
* *2019-02-17* Fix handling of empty padding (by @jschwab, #138)
* *2019-02-17* Add a man page (by @somers-all-the-time, #133)
**autorandr 1.7**
* *2018-09-25* Fix FB size computation with rotated screens (by @Janno, #117)
**autorandr 1.6**
* *2018-04-19* Bugfix: Do not load default profile unless --change is set
* *2018-04-30* Added a `AUTORANDR_MONITORS` variable to hooks (by @bricewge, #106)
* *2018-06-29* Fix detection of current configuration if extra monitors are active
* *2018-07-11* Bugfix in the latest change: Correctly handle "off" minitors when comparing
* *2018-07-19* Do not kill spawned user processes from systemd unit
* *2018-07-20* Correctly handle "off" monitors when comparing -- fixup for another bug.
**autorandr 1.5**
* *2018-01-03* Add --version
* *2018-01-04* Fixed vertical/horizontal/clone-largest virtual profiles
* *2018-03-07* Output all non-error messages to stdout instead of stderr
* *2018-03-25* Add --detected and --current to filter the profile list output
* *2018-03-25* Allow wildcard matching in EDIDs
**autorandr 1.4**
* *2017-12-22* Fixed broken virtual profile support
* *2017-12-14* Added support for a settings file
* *2017-12-14* Added a virtual profile `off`, which disables all screens
**autorandr 1.3**
* *2017-11-13* Add a short form for `--load`
* *2017-11-21* Fix environment stealing in `--batch` mode (See #87)
**autorandr 1.2**
* *2017-07-16* Skip `--panning` unless it is required (See #72)
* *2017-10-13* Add `clone-largest` virtual profile
**autorandr 1.1**
* *2017-06-07* Call systemctl with `--no-block` from udev rule (See #61)
* *2017-01-20* New script hook, `predetect`
* *2017-01-18* Accept comments (lines starting with `#`) in config/setup files
**autorandr 1.0**
* *2016-12-07* Tag the current code as version 1.0.0; see github issue #54
* *2016-10-03* Install a desktop file to `/etc/xdg/autostart` by default

View File

@ -0,0 +1,10 @@
../../../bin/autorandr,sha256=b7-Kfx617lTXecYwDuiiUgSizM-HhcqaIKkFNj2U7_Y,273
__pycache__/autorandr.cpython-311.pyc,,
autorandr-1.15.post1.dist-info/INSTALLER,sha256=zuuue4knoyJ-UwPPXg8fezS7VCrXJQrAP7zeNuwvFQg,4
autorandr-1.15.post1.dist-info/METADATA,sha256=WU3tSPIh4uxAwXtLlUMhAhf-D1CkTaRHmoqWpOoIrtA,16035
autorandr-1.15.post1.dist-info/RECORD,,
autorandr-1.15.post1.dist-info/REQUESTED,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
autorandr-1.15.post1.dist-info/WHEEL,sha256=G16H4A3IeoQmnOrYV4ueZGKSjhipXx8zc8nu9FGlvMA,92
autorandr-1.15.post1.dist-info/entry_points.txt,sha256=nIbh59SOctw_HjgX7a8iEKc-TawiF-tiwbIs7Wh9H5w,64
autorandr-1.15.post1.dist-info/top_level.txt,sha256=EZfqdN1OiuXc20FuQhKSxk4gieAvzWOXie8MgR29Aek,10
autorandr.py,sha256=rRxKO3qME5aTlZ0zdkLsL9kP-pUFV1y_CZZ6-ukoJJo,75378

View File

@ -0,0 +1,5 @@
Wheel-Version: 1.0
Generator: bdist_wheel (0.37.1)
Root-Is-Purelib: true
Tag: py3-none-any

View File

@ -0,0 +1,3 @@
[console_scripts]
autorandr = autorandr:exception_handled_main

File diff suppressed because it is too large Load Diff

View File

@ -0,0 +1 @@
pip

View File

@ -0,0 +1,201 @@
Apache License
Version 2.0, January 2004
http://www.apache.org/licenses/
TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
1. Definitions.
"License" shall mean the terms and conditions for use, reproduction,
and distribution as defined by Sections 1 through 9 of this document.
"Licensor" shall mean the copyright owner or entity authorized by
the copyright owner that is granting the License.
"Legal Entity" shall mean the union of the acting entity and all
other entities that control, are controlled by, or are under common
control with that entity. For the purposes of this definition,
"control" means (i) the power, direct or indirect, to cause the
direction or management of such entity, whether by contract or
otherwise, or (ii) ownership of fifty percent (50%) or more of the
outstanding shares, or (iii) beneficial ownership of such entity.
"You" (or "Your") shall mean an individual or Legal Entity
exercising permissions granted by this License.
"Source" form shall mean the preferred form for making modifications,
including but not limited to software source code, documentation
source, and configuration files.
"Object" form shall mean any form resulting from mechanical
transformation or translation of a Source form, including but
not limited to compiled object code, generated documentation,
and conversions to other media types.
"Work" shall mean the work of authorship, whether in Source or
Object form, made available under the License, as indicated by a
copyright notice that is included in or attached to the work
(an example is provided in the Appendix below).
"Derivative Works" shall mean any work, whether in Source or Object
form, that is based on (or derived from) the Work and for which the
editorial revisions, annotations, elaborations, or other modifications
represent, as a whole, an original work of authorship. For the purposes
of this License, Derivative Works shall not include works that remain
separable from, or merely link (or bind by name) to the interfaces of,
the Work and Derivative Works thereof.
"Contribution" shall mean any work of authorship, including
the original version of the Work and any modifications or additions
to that Work or Derivative Works thereof, that is intentionally
submitted to Licensor for inclusion in the Work by the copyright owner
or by an individual or Legal Entity authorized to submit on behalf of
the copyright owner. For the purposes of this definition, "submitted"
means any form of electronic, verbal, or written communication sent
to the Licensor or its representatives, including but not limited to
communication on electronic mailing lists, source code control systems,
and issue tracking systems that are managed by, or on behalf of, the
Licensor for the purpose of discussing and improving the Work, but
excluding communication that is conspicuously marked or otherwise
designated in writing by the copyright owner as "Not a Contribution."
"Contributor" shall mean Licensor and any individual or Legal Entity
on behalf of whom a Contribution has been received by Licensor and
subsequently incorporated within the Work.
2. Grant of Copyright License. Subject to the terms and conditions of
this License, each Contributor hereby grants to You a perpetual,
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
copyright license to reproduce, prepare Derivative Works of,
publicly display, publicly perform, sublicense, and distribute the
Work and such Derivative Works in Source or Object form.
3. Grant of Patent License. Subject to the terms and conditions of
this License, each Contributor hereby grants to You a perpetual,
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
(except as stated in this section) patent license to make, have made,
use, offer to sell, sell, import, and otherwise transfer the Work,
where such license applies only to those patent claims licensable
by such Contributor that are necessarily infringed by their
Contribution(s) alone or by combination of their Contribution(s)
with the Work to which such Contribution(s) was submitted. If You
institute patent litigation against any entity (including a
cross-claim or counterclaim in a lawsuit) alleging that the Work
or a Contribution incorporated within the Work constitutes direct
or contributory patent infringement, then any patent licenses
granted to You under this License for that Work shall terminate
as of the date such litigation is filed.
4. Redistribution. You may reproduce and distribute copies of the
Work or Derivative Works thereof in any medium, with or without
modifications, and in Source or Object form, provided that You
meet the following conditions:
(a) You must give any other recipients of the Work or
Derivative Works a copy of this License; and
(b) You must cause any modified files to carry prominent notices
stating that You changed the files; and
(c) You must retain, in the Source form of any Derivative Works
that You distribute, all copyright, patent, trademark, and
attribution notices from the Source form of the Work,
excluding those notices that do not pertain to any part of
the Derivative Works; and
(d) If the Work includes a "NOTICE" text file as part of its
distribution, then any Derivative Works that You distribute must
include a readable copy of the attribution notices contained
within such NOTICE file, excluding those notices that do not
pertain to any part of the Derivative Works, in at least one
of the following places: within a NOTICE text file distributed
as part of the Derivative Works; within the Source form or
documentation, if provided along with the Derivative Works; or,
within a display generated by the Derivative Works, if and
wherever such third-party notices normally appear. The contents
of the NOTICE file are for informational purposes only and
do not modify the License. You may add Your own attribution
notices within Derivative Works that You distribute, alongside
or as an addendum to the NOTICE text from the Work, provided
that such additional attribution notices cannot be construed
as modifying the License.
You may add Your own copyright statement to Your modifications and
may provide additional or different license terms and conditions
for use, reproduction, or distribution of Your modifications, or
for any such Derivative Works as a whole, provided Your use,
reproduction, and distribution of the Work otherwise complies with
the conditions stated in this License.
5. Submission of Contributions. Unless You explicitly state otherwise,
any Contribution intentionally submitted for inclusion in the Work
by You to the Licensor shall be under the terms and conditions of
this License, without any additional terms or conditions.
Notwithstanding the above, nothing herein shall supersede or modify
the terms of any separate license agreement you may have executed
with Licensor regarding such Contributions.
6. Trademarks. This License does not grant permission to use the trade
names, trademarks, service marks, or product names of the Licensor,
except as required for reasonable and customary use in describing the
origin of the Work and reproducing the content of the NOTICE file.
7. Disclaimer of Warranty. Unless required by applicable law or
agreed to in writing, Licensor provides the Work (and each
Contributor provides its Contributions) on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
implied, including, without limitation, any warranties or conditions
of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
PARTICULAR PURPOSE. You are solely responsible for determining the
appropriateness of using or redistributing the Work and assume any
risks associated with Your exercise of permissions under this License.
8. Limitation of Liability. In no event and under no legal theory,
whether in tort (including negligence), contract, or otherwise,
unless required by applicable law (such as deliberate and grossly
negligent acts) or agreed to in writing, shall any Contributor be
liable to You for damages, including any direct, indirect, special,
incidental, or consequential damages of any character arising as a
result of this License or out of the use or inability to use the
Work (including but not limited to damages for loss of goodwill,
work stoppage, computer failure or malfunction, or any and all
other commercial damages or losses), even if such Contributor
has been advised of the possibility of such damages.
9. Accepting Warranty or Additional Liability. While redistributing
the Work or Derivative Works thereof, You may choose to offer,
and charge a fee for, acceptance of support, warranty, indemnity,
or other liability obligations and/or rights consistent with this
License. However, in accepting such obligations, You may act only
on Your own behalf and on Your sole responsibility, not on behalf
of any other Contributor, and only if You agree to indemnify,
defend, and hold each Contributor harmless for any liability
incurred by, or claims asserted against, such Contributor by reason
of your accepting any such warranty or additional liability.
END OF TERMS AND CONDITIONS
APPENDIX: How to apply the Apache License to your work.
To apply the Apache License to your work, attach the following
boilerplate notice, with the fields enclosed by brackets "[]"
replaced with your own identifying information. (Don't include
the brackets!) The text should be enclosed in the appropriate
comment syntax for the file format. We also recommend that a
file or class name and description of purpose be included on the
same "printed page" as the copyright notice for easier
identification within third-party archives.
Copyright [yyyy] [name of copyright owner]
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.

View File

@ -0,0 +1,330 @@
Metadata-Version: 2.2
Name: bcrypt
Version: 4.3.0
Summary: Modern password hashing for your software and your servers
Author-email: The Python Cryptographic Authority developers <cryptography-dev@python.org>
License: Apache-2.0
Project-URL: homepage, https://github.com/pyca/bcrypt/
Classifier: Development Status :: 5 - Production/Stable
Classifier: License :: OSI Approved :: Apache Software License
Classifier: Programming Language :: Python :: Implementation :: CPython
Classifier: Programming Language :: Python :: Implementation :: PyPy
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3 :: Only
Classifier: Programming Language :: Python :: 3.8
Classifier: Programming Language :: Python :: 3.9
Classifier: Programming Language :: Python :: 3.10
Classifier: Programming Language :: Python :: 3.11
Classifier: Programming Language :: Python :: 3.12
Classifier: Programming Language :: Python :: 3.13
Requires-Python: >=3.8
Description-Content-Type: text/x-rst
License-File: LICENSE
Provides-Extra: tests
Requires-Dist: pytest!=3.3.0,>=3.2.1; extra == "tests"
Provides-Extra: typecheck
Requires-Dist: mypy; extra == "typecheck"
bcrypt
======
.. image:: https://img.shields.io/pypi/v/bcrypt.svg
:target: https://pypi.org/project/bcrypt/
:alt: Latest Version
.. image:: https://github.com/pyca/bcrypt/workflows/CI/badge.svg?branch=main
:target: https://github.com/pyca/bcrypt/actions?query=workflow%3ACI+branch%3Amain
Acceptable password hashing for your software and your servers (but you should
really use argon2id or scrypt)
Installation
============
To install bcrypt, simply:
.. code:: console
$ pip install bcrypt
Note that bcrypt should build very easily on Linux provided you have a C
compiler and a Rust compiler (the minimum supported Rust version is 1.56.0).
For Debian and Ubuntu, the following command will ensure that the required dependencies are installed:
.. code:: console
$ sudo apt-get install build-essential cargo
For Fedora and RHEL-derivatives, the following command will ensure that the required dependencies are installed:
.. code:: console
$ sudo yum install gcc cargo
For Alpine, the following command will ensure that the required dependencies are installed:
.. code:: console
$ apk add --update musl-dev gcc cargo
Alternatives
============
While bcrypt remains an acceptable choice for password storage, depending on your specific use case you may also want to consider using scrypt (either via `standard library`_ or `cryptography`_) or argon2id via `argon2_cffi`_.
Changelog
=========
Unreleased
----------
* Dropped support for Python 3.7.
* We now support free-threaded Python 3.13.
* We now support PyPy 3.11.
* We now publish wheels for free-threaded Python 3.13, for PyPy 3.11 on
``manylinux``, and for ARMv7l on ``manylinux``.
4.2.1
-----
* Bump Rust dependency versions - this should resolve crashes on Python 3.13
free-threaded builds.
* We no longer build ``manylinux`` wheels for PyPy 3.9.
4.2.0
-----
* Bump Rust dependency versions
* Removed the ``BCRYPT_ALLOW_RUST_163`` environment variable.
4.1.3
-----
* Bump Rust dependency versions
4.1.2
-----
* Publish both ``py37`` and ``py39`` wheels. This should resolve some errors
relating to initializing a module multiple times per process.
4.1.1
-----
* Fixed the type signature on the ``kdf`` method.
* Fixed packaging bug on Windows.
* Fixed incompatibility with passlib package detection assumptions.
4.1.0
-----
* Dropped support for Python 3.6.
* Bumped MSRV to 1.64. (Note: Rust 1.63 can be used by setting the ``BCRYPT_ALLOW_RUST_163`` environment variable)
4.0.1
-----
* We now build PyPy ``manylinux`` wheels.
* Fixed a bug where passing an invalid ``salt`` to ``checkpw`` could result in
a ``pyo3_runtime.PanicException``. It now correctly raises a ``ValueError``.
4.0.0
-----
* ``bcrypt`` is now implemented in Rust. Users building from source will need
to have a Rust compiler available. Nothing will change for users downloading
wheels.
* We no longer ship ``manylinux2010`` wheels. Users should upgrade to the latest
``pip`` to ensure this doesnt cause issues downloading wheels on their
platform. We now ship ``manylinux_2_28`` wheels for users on new enough platforms.
* ``NUL`` bytes are now allowed in inputs.
3.2.2
-----
* Fixed packaging of ``py.typed`` files in wheels so that ``mypy`` works.
3.2.1
-----
* Added support for compilation on z/OS
* The next release of ``bcrypt`` with be 4.0 and it will require Rust at
compile time, for users building from source. There will be no additional
requirement for users who are installing from wheels. Users on most
platforms will be able to obtain a wheel by making sure they have an up to
date ``pip``. The minimum supported Rust version will be 1.56.0.
* This will be the final release for which we ship ``manylinux2010`` wheels.
Going forward the minimum supported manylinux ABI for our wheels will be
``manylinux2014``. The vast majority of users will continue to receive
``manylinux`` wheels provided they have an up to date ``pip``.
3.2.0
-----
* Added typehints for library functions.
* Dropped support for Python versions less than 3.6 (2.7, 3.4, 3.5).
* Shipped ``abi3`` Windows wheels (requires pip >= 20).
3.1.7
-----
* Set a ``setuptools`` lower bound for PEP517 wheel building.
* We no longer distribute 32-bit ``manylinux1`` wheels. Continuing to produce
them was a maintenance burden.
3.1.6
-----
* Added support for compilation on Haiku.
3.1.5
-----
* Added support for compilation on AIX.
* Dropped Python 2.6 and 3.3 support.
* Switched to using ``abi3`` wheels for Python 3. If you are not getting a
wheel on a compatible platform please upgrade your ``pip`` version.
3.1.4
-----
* Fixed compilation with mingw and on illumos.
3.1.3
-----
* Fixed a compilation issue on Solaris.
* Added a warning when using too few rounds with ``kdf``.
3.1.2
-----
* Fixed a compile issue affecting big endian platforms.
* Fixed invalid escape sequence warnings on Python 3.6.
* Fixed building in non-UTF8 environments on Python 2.
3.1.1
-----
* Resolved a ``UserWarning`` when used with ``cffi`` 1.8.3.
3.1.0
-----
* Added support for ``checkpw``, a convenience method for verifying a password.
* Ensure that you get a ``$2y$`` hash when you input a ``$2y$`` salt.
* Fixed a regression where ``$2a`` hashes were vulnerable to a wraparound bug.
* Fixed compilation under Alpine Linux.
3.0.0
-----
* Switched the C backend to code obtained from the OpenBSD project rather than
openwall.
* Added support for ``bcrypt_pbkdf`` via the ``kdf`` function.
2.0.0
-----
* Added support for an adjustible prefix when calling ``gensalt``.
* Switched to CFFI 1.0+
Usage
-----
Password Hashing
~~~~~~~~~~~~~~~~
Hashing and then later checking that a password matches the previous hashed
password is very simple:
.. code:: pycon
>>> import bcrypt
>>> password = b"super secret password"
>>> # Hash a password for the first time, with a randomly-generated salt
>>> hashed = bcrypt.hashpw(password, bcrypt.gensalt())
>>> # Check that an unhashed password matches one that has previously been
>>> # hashed
>>> if bcrypt.checkpw(password, hashed):
... print("It Matches!")
... else:
... print("It Does not Match :(")
KDF
~~~
As of 3.0.0 ``bcrypt`` now offers a ``kdf`` function which does ``bcrypt_pbkdf``.
This KDF is used in OpenSSH's newer encrypted private key format.
.. code:: pycon
>>> import bcrypt
>>> key = bcrypt.kdf(
... password=b'password',
... salt=b'salt',
... desired_key_bytes=32,
... rounds=100)
Adjustable Work Factor
~~~~~~~~~~~~~~~~~~~~~~
One of bcrypt's features is an adjustable logarithmic work factor. To adjust
the work factor merely pass the desired number of rounds to
``bcrypt.gensalt(rounds=12)`` which defaults to 12):
.. code:: pycon
>>> import bcrypt
>>> password = b"super secret password"
>>> # Hash a password for the first time, with a certain number of rounds
>>> hashed = bcrypt.hashpw(password, bcrypt.gensalt(14))
>>> # Check that a unhashed password matches one that has previously been
>>> # hashed
>>> if bcrypt.checkpw(password, hashed):
... print("It Matches!")
... else:
... print("It Does not Match :(")
Adjustable Prefix
~~~~~~~~~~~~~~~~~
Another one of bcrypt's features is an adjustable prefix to let you define what
libraries you'll remain compatible with. To adjust this, pass either ``2a`` or
``2b`` (the default) to ``bcrypt.gensalt(prefix=b"2b")`` as a bytes object.
As of 3.0.0 the ``$2y$`` prefix is still supported in ``hashpw`` but deprecated.
Maximum Password Length
~~~~~~~~~~~~~~~~~~~~~~~
The bcrypt algorithm only handles passwords up to 72 characters, any characters
beyond that are ignored. To work around this, a common approach is to hash a
password with a cryptographic hash (such as ``sha256``) and then base64
encode it to prevent NULL byte problems before hashing the result with
``bcrypt``:
.. code:: pycon
>>> password = b"an incredibly long password" * 10
>>> hashed = bcrypt.hashpw(
... base64.b64encode(hashlib.sha256(password).digest()),
... bcrypt.gensalt()
... )
Compatibility
-------------
This library should be compatible with py-bcrypt and it will run on Python
3.8+ (including free-threaded builds), and PyPy 3.
Security
--------
``bcrypt`` follows the `same security policy as cryptography`_, if you
identify a vulnerability, we ask you to contact us privately.
.. _`same security policy as cryptography`: https://cryptography.io/en/latest/security.html
.. _`standard library`: https://docs.python.org/3/library/hashlib.html#hashlib.scrypt
.. _`argon2_cffi`: https://argon2-cffi.readthedocs.io
.. _`cryptography`: https://cryptography.io/en/latest/hazmat/primitives/key-derivation-functions/#cryptography.hazmat.primitives.kdf.scrypt.Scrypt

View File

@ -0,0 +1,11 @@
bcrypt-4.3.0.dist-info/INSTALLER,sha256=zuuue4knoyJ-UwPPXg8fezS7VCrXJQrAP7zeNuwvFQg,4
bcrypt-4.3.0.dist-info/LICENSE,sha256=gXPVwptPlW1TJ4HSuG5OMPg-a3h43OGMkZRR1rpwfJA,10850
bcrypt-4.3.0.dist-info/METADATA,sha256=95qX7ziIfmOF0kNM95YZuWhLVfFy-6EtssVvf1ZgeWg,10042
bcrypt-4.3.0.dist-info/RECORD,,
bcrypt-4.3.0.dist-info/WHEEL,sha256=XlovOtcAZFqrc4OSNBtc5R3yDeRHyhWP24RdDnylFpY,111
bcrypt-4.3.0.dist-info/top_level.txt,sha256=BkR_qBzDbSuycMzHWE1vzXrfYecAzUVmQs6G2CukqNI,7
bcrypt/__init__.py,sha256=cv-NupIX6P7o6A4PK_F0ur6IZoDr3GnvyzFO9k16wKQ,1000
bcrypt/__init__.pyi,sha256=ITUCB9mPVU8sKUbJQMDUH5YfQXZb1O55F9qvKZR_o8I,333
bcrypt/__pycache__/__init__.cpython-311.pyc,,
bcrypt/_bcrypt.abi3.so,sha256=oMArVCuY_atg2H4SGNfM-zbfEgUOkd4qSiWn2nPqmXc,644928
bcrypt/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0

View File

@ -0,0 +1,5 @@
Wheel-Version: 1.0
Generator: setuptools (75.8.2)
Root-Is-Purelib: false
Tag: cp39-abi3-manylinux_2_34_x86_64

View File

@ -0,0 +1 @@
bcrypt

View File

@ -0,0 +1,43 @@
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
from ._bcrypt import (
__author__,
__copyright__,
__email__,
__license__,
__summary__,
__title__,
__uri__,
checkpw,
gensalt,
hashpw,
kdf,
)
from ._bcrypt import (
__version_ex__ as __version__,
)
__all__ = [
"__author__",
"__copyright__",
"__email__",
"__license__",
"__summary__",
"__title__",
"__uri__",
"__version__",
"checkpw",
"gensalt",
"hashpw",
"kdf",
]

View File

@ -0,0 +1,10 @@
def gensalt(rounds: int = 12, prefix: bytes = b"2b") -> bytes: ...
def hashpw(password: bytes, salt: bytes) -> bytes: ...
def checkpw(password: bytes, hashed_password: bytes) -> bool: ...
def kdf(
password: bytes,
salt: bytes,
desired_key_bytes: int,
rounds: int,
ignore_few_rounds: bool = False,
) -> bytes: ...

Binary file not shown.

View File

@ -0,0 +1,20 @@
This package contains a modified version of ca-bundle.crt:
ca-bundle.crt -- Bundle of CA Root Certificates
This is a bundle of X.509 certificates of public Certificate Authorities
(CA). These were automatically extracted from Mozilla's root certificates
file (certdata.txt). This file can be found in the mozilla source tree:
https://hg.mozilla.org/mozilla-central/file/tip/security/nss/lib/ckfw/builtins/certdata.txt
It contains the certificates in PEM format and therefore
can be directly used with curl / libcurl / php_curl, or with
an Apache+mod_ssl webserver for SSL client authentication.
Just configure this file as the SSLCACertificateFile.#
***** BEGIN LICENSE BLOCK *****
This Source Code Form is subject to the terms of the Mozilla Public License,
v. 2.0. If a copy of the MPL was not distributed with this file, You can obtain
one at http://mozilla.org/MPL/2.0/.
***** END LICENSE BLOCK *****
@(#) $RCSfile: certdata.txt,v $ $Revision: 1.80 $ $Date: 2011/11/03 15:11:58 $

View File

@ -0,0 +1,77 @@
Metadata-Version: 2.2
Name: certifi
Version: 2025.1.31
Summary: Python package for providing Mozilla's CA Bundle.
Home-page: https://github.com/certifi/python-certifi
Author: Kenneth Reitz
Author-email: me@kennethreitz.com
License: MPL-2.0
Project-URL: Source, https://github.com/certifi/python-certifi
Classifier: Development Status :: 5 - Production/Stable
Classifier: Intended Audience :: Developers
Classifier: License :: OSI Approved :: Mozilla Public License 2.0 (MPL 2.0)
Classifier: Natural Language :: English
Classifier: Programming Language :: Python
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3 :: Only
Classifier: Programming Language :: Python :: 3.6
Classifier: Programming Language :: Python :: 3.7
Classifier: Programming Language :: Python :: 3.8
Classifier: Programming Language :: Python :: 3.9
Classifier: Programming Language :: Python :: 3.10
Classifier: Programming Language :: Python :: 3.11
Classifier: Programming Language :: Python :: 3.12
Classifier: Programming Language :: Python :: 3.13
Requires-Python: >=3.6
License-File: LICENSE
Dynamic: author
Dynamic: author-email
Dynamic: classifier
Dynamic: description
Dynamic: home-page
Dynamic: license
Dynamic: project-url
Dynamic: requires-python
Dynamic: summary
Certifi: Python SSL Certificates
================================
Certifi provides Mozilla's carefully curated collection of Root Certificates for
validating the trustworthiness of SSL certificates while verifying the identity
of TLS hosts. It has been extracted from the `Requests`_ project.
Installation
------------
``certifi`` is available on PyPI. Simply install it with ``pip``::
$ pip install certifi
Usage
-----
To reference the installed certificate authority (CA) bundle, you can use the
built-in function::
>>> import certifi
>>> certifi.where()
'/usr/local/lib/python3.7/site-packages/certifi/cacert.pem'
Or from the command line::
$ python -m certifi
/usr/local/lib/python3.7/site-packages/certifi/cacert.pem
Enjoy!
.. _`Requests`: https://requests.readthedocs.io/en/master/
Addition/Removal of Certificates
--------------------------------
Certifi does not support any addition/removal or other modification of the
CA trust store content. This project is intended to provide a reliable and
highly portable root of trust to python deployments. Look to upstream projects
for methods to use alternate trust.

View File

@ -0,0 +1,14 @@
certifi-2025.1.31.dist-info/INSTALLER,sha256=zuuue4knoyJ-UwPPXg8fezS7VCrXJQrAP7zeNuwvFQg,4
certifi-2025.1.31.dist-info/LICENSE,sha256=6TcW2mucDVpKHfYP5pWzcPBpVgPSH2-D8FPkLPwQyvc,989
certifi-2025.1.31.dist-info/METADATA,sha256=t5kcT5aGu0dQ6_psUNZYTqnC0uCRnponewm3uYjeHbg,2451
certifi-2025.1.31.dist-info/RECORD,,
certifi-2025.1.31.dist-info/WHEEL,sha256=In9FTNxeP60KnTkGw7wk6mJPYd_dQSjEZmXdBdMCI-8,91
certifi-2025.1.31.dist-info/top_level.txt,sha256=KMu4vUCfsjLrkPbSNdgdekS-pVJzBAJFO__nI8NF6-U,8
certifi/__init__.py,sha256=neIaAf7BM36ygmQCmy-ZsSyjnvjWghFeu13wwEAnjj0,94
certifi/__main__.py,sha256=xBBoj905TUWBLRGANOcf7oi6e-3dMP4cEoG9OyMs11g,243
certifi/__pycache__/__init__.cpython-311.pyc,,
certifi/__pycache__/__main__.cpython-311.pyc,,
certifi/__pycache__/core.cpython-311.pyc,,
certifi/cacert.pem,sha256=xVsh-Qf3-G1IrdCTVS-1ZRdJ_1-GBQjMu0I9bB-9gMc,297255
certifi/core.py,sha256=qRDDFyXVJwTB_EmoGppaXU_R9qCZvhl-EzxPMuV3nTA,4426
certifi/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0

View File

@ -0,0 +1,5 @@
Wheel-Version: 1.0
Generator: setuptools (75.8.0)
Root-Is-Purelib: true
Tag: py3-none-any

View File

@ -0,0 +1,4 @@
from .core import contents, where
__all__ = ["contents", "where"]
__version__ = "2025.01.31"

View File

@ -0,0 +1,12 @@
import argparse
from certifi import contents, where
parser = argparse.ArgumentParser()
parser.add_argument("-c", "--contents", action="store_true")
args = parser.parse_args()
if args.contents:
print(contents())
else:
print(where())

File diff suppressed because it is too large Load Diff

View File

@ -0,0 +1,114 @@
"""
certifi.py
~~~~~~~~~~
This module returns the installation location of cacert.pem or its contents.
"""
import sys
import atexit
def exit_cacert_ctx() -> None:
_CACERT_CTX.__exit__(None, None, None) # type: ignore[union-attr]
if sys.version_info >= (3, 11):
from importlib.resources import as_file, files
_CACERT_CTX = None
_CACERT_PATH = None
def where() -> str:
# This is slightly terrible, but we want to delay extracting the file
# in cases where we're inside of a zipimport situation until someone
# actually calls where(), but we don't want to re-extract the file
# on every call of where(), so we'll do it once then store it in a
# global variable.
global _CACERT_CTX
global _CACERT_PATH
if _CACERT_PATH is None:
# This is slightly janky, the importlib.resources API wants you to
# manage the cleanup of this file, so it doesn't actually return a
# path, it returns a context manager that will give you the path
# when you enter it and will do any cleanup when you leave it. In
# the common case of not needing a temporary file, it will just
# return the file system location and the __exit__() is a no-op.
#
# We also have to hold onto the actual context manager, because
# it will do the cleanup whenever it gets garbage collected, so
# we will also store that at the global level as well.
_CACERT_CTX = as_file(files("certifi").joinpath("cacert.pem"))
_CACERT_PATH = str(_CACERT_CTX.__enter__())
atexit.register(exit_cacert_ctx)
return _CACERT_PATH
def contents() -> str:
return files("certifi").joinpath("cacert.pem").read_text(encoding="ascii")
elif sys.version_info >= (3, 7):
from importlib.resources import path as get_path, read_text
_CACERT_CTX = None
_CACERT_PATH = None
def where() -> str:
# This is slightly terrible, but we want to delay extracting the
# file in cases where we're inside of a zipimport situation until
# someone actually calls where(), but we don't want to re-extract
# the file on every call of where(), so we'll do it once then store
# it in a global variable.
global _CACERT_CTX
global _CACERT_PATH
if _CACERT_PATH is None:
# This is slightly janky, the importlib.resources API wants you
# to manage the cleanup of this file, so it doesn't actually
# return a path, it returns a context manager that will give
# you the path when you enter it and will do any cleanup when
# you leave it. In the common case of not needing a temporary
# file, it will just return the file system location and the
# __exit__() is a no-op.
#
# We also have to hold onto the actual context manager, because
# it will do the cleanup whenever it gets garbage collected, so
# we will also store that at the global level as well.
_CACERT_CTX = get_path("certifi", "cacert.pem")
_CACERT_PATH = str(_CACERT_CTX.__enter__())
atexit.register(exit_cacert_ctx)
return _CACERT_PATH
def contents() -> str:
return read_text("certifi", "cacert.pem", encoding="ascii")
else:
import os
import types
from typing import Union
Package = Union[types.ModuleType, str]
Resource = Union[str, "os.PathLike"]
# This fallback will work for Python versions prior to 3.7 that lack the
# importlib.resources module but relies on the existing `where` function
# so won't address issues with environments like PyOxidizer that don't set
# __file__ on modules.
def read_text(
package: Package,
resource: Resource,
encoding: str = 'utf-8',
errors: str = 'strict'
) -> str:
with open(where(), encoding=encoding) as data:
return data.read()
# If we don't have importlib.resources, then we will just do the old logic
# of assuming we're on the filesystem and munge the path directly.
def where() -> str:
f = os.path.dirname(__file__)
return os.path.join(f, "cacert.pem")
def contents() -> str:
return read_text("certifi", "cacert.pem", encoding="ascii")

View File

@ -0,0 +1 @@
pip

View File

@ -0,0 +1,26 @@
Except when otherwise stated (look for LICENSE files in directories or
information at the beginning of each file) all software and
documentation is licensed as follows:
The MIT License
Permission is hereby granted, free of charge, to any person
obtaining a copy of this software and associated documentation
files (the "Software"), to deal in the Software without
restriction, including without limitation the rights to use,
copy, modify, merge, publish, distribute, sublicense, and/or
sell copies of the Software, and to permit persons to whom the
Software is furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included
in all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
DEALINGS IN THE SOFTWARE.

View File

@ -0,0 +1,40 @@
Metadata-Version: 2.1
Name: cffi
Version: 1.17.1
Summary: Foreign Function Interface for Python calling C code.
Home-page: http://cffi.readthedocs.org
Author: Armin Rigo, Maciej Fijalkowski
Author-email: python-cffi@googlegroups.com
License: MIT
Project-URL: Documentation, http://cffi.readthedocs.org/
Project-URL: Source Code, https://github.com/python-cffi/cffi
Project-URL: Issue Tracker, https://github.com/python-cffi/cffi/issues
Project-URL: Changelog, https://cffi.readthedocs.io/en/latest/whatsnew.html
Project-URL: Downloads, https://github.com/python-cffi/cffi/releases
Project-URL: Contact, https://groups.google.com/forum/#!forum/python-cffi
Classifier: Programming Language :: Python
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.8
Classifier: Programming Language :: Python :: 3.9
Classifier: Programming Language :: Python :: 3.10
Classifier: Programming Language :: Python :: 3.11
Classifier: Programming Language :: Python :: 3.12
Classifier: Programming Language :: Python :: 3.13
Classifier: Programming Language :: Python :: Implementation :: CPython
Classifier: Programming Language :: Python :: Implementation :: PyPy
Classifier: License :: OSI Approved :: MIT License
Requires-Python: >=3.8
License-File: LICENSE
Requires-Dist: pycparser
CFFI
====
Foreign Function Interface for Python calling C code.
Please see the `Documentation <http://cffi.readthedocs.org/>`_.
Contact
-------
`Mailing list <https://groups.google.com/forum/#!forum/python-cffi>`_

View File

@ -0,0 +1,48 @@
_cffi_backend.cpython-311-x86_64-linux-gnu.so,sha256=K3Ig76G2fNGS7ef9yadiP-gNjpCHXd-J1ZNzvv6jfQs,1068624
cffi-1.17.1.dist-info/INSTALLER,sha256=zuuue4knoyJ-UwPPXg8fezS7VCrXJQrAP7zeNuwvFQg,4
cffi-1.17.1.dist-info/LICENSE,sha256=BLgPWwd7vtaICM_rreteNSPyqMmpZJXFh72W3x6sKjM,1294
cffi-1.17.1.dist-info/METADATA,sha256=u6nuvP_qPJKu2zvIbi2zkGzVu7KjnnRIYUFyIrOY3j4,1531
cffi-1.17.1.dist-info/RECORD,,
cffi-1.17.1.dist-info/WHEEL,sha256=JyEZ6Cxo51rQOYRMkM7cW9w1CnY9FuHxSeIJCGoxjS4,151
cffi-1.17.1.dist-info/entry_points.txt,sha256=y6jTxnyeuLnL-XJcDv8uML3n6wyYiGRg8MTp_QGJ9Ho,75
cffi-1.17.1.dist-info/top_level.txt,sha256=rE7WR3rZfNKxWI9-jn6hsHCAl7MDkB-FmuQbxWjFehQ,19
cffi/__init__.py,sha256=H6t_ebva6EeHpUuItFLW1gbRp94eZRNJODLaWKdbx1I,513
cffi/__pycache__/__init__.cpython-311.pyc,,
cffi/__pycache__/_imp_emulation.cpython-311.pyc,,
cffi/__pycache__/_shimmed_dist_utils.cpython-311.pyc,,
cffi/__pycache__/api.cpython-311.pyc,,
cffi/__pycache__/backend_ctypes.cpython-311.pyc,,
cffi/__pycache__/cffi_opcode.cpython-311.pyc,,
cffi/__pycache__/commontypes.cpython-311.pyc,,
cffi/__pycache__/cparser.cpython-311.pyc,,
cffi/__pycache__/error.cpython-311.pyc,,
cffi/__pycache__/ffiplatform.cpython-311.pyc,,
cffi/__pycache__/lock.cpython-311.pyc,,
cffi/__pycache__/model.cpython-311.pyc,,
cffi/__pycache__/pkgconfig.cpython-311.pyc,,
cffi/__pycache__/recompiler.cpython-311.pyc,,
cffi/__pycache__/setuptools_ext.cpython-311.pyc,,
cffi/__pycache__/vengine_cpy.cpython-311.pyc,,
cffi/__pycache__/vengine_gen.cpython-311.pyc,,
cffi/__pycache__/verifier.cpython-311.pyc,,
cffi/_cffi_errors.h,sha256=zQXt7uR_m8gUW-fI2hJg0KoSkJFwXv8RGUkEDZ177dQ,3908
cffi/_cffi_include.h,sha256=Exhmgm9qzHWzWivjfTe0D7Xp4rPUkVxdNuwGhMTMzbw,15055
cffi/_embedding.h,sha256=EDKw5QrLvQoe3uosXB3H1xPVTYxsn33eV3A43zsA_Fw,18787
cffi/_imp_emulation.py,sha256=RxREG8zAbI2RPGBww90u_5fi8sWdahpdipOoPzkp7C0,2960
cffi/_shimmed_dist_utils.py,sha256=Bjj2wm8yZbvFvWEx5AEfmqaqZyZFhYfoyLLQHkXZuao,2230
cffi/api.py,sha256=alBv6hZQkjpmZplBphdaRn2lPO9-CORs_M7ixabvZWI,42169
cffi/backend_ctypes.py,sha256=h5ZIzLc6BFVXnGyc9xPqZWUS7qGy7yFSDqXe68Sa8z4,42454
cffi/cffi_opcode.py,sha256=JDV5l0R0_OadBX_uE7xPPTYtMdmpp8I9UYd6av7aiDU,5731
cffi/commontypes.py,sha256=7N6zPtCFlvxXMWhHV08psUjdYIK2XgsN3yo5dgua_v4,2805
cffi/cparser.py,sha256=0qI3mEzZSNVcCangoyXOoAcL-RhpQL08eG8798T024s,44789
cffi/error.py,sha256=v6xTiS4U0kvDcy4h_BDRo5v39ZQuj-IMRYLv5ETddZs,877
cffi/ffiplatform.py,sha256=avxFjdikYGJoEtmJO7ewVmwG_VEVl6EZ_WaNhZYCqv4,3584
cffi/lock.py,sha256=l9TTdwMIMpi6jDkJGnQgE9cvTIR7CAntIJr8EGHt3pY,747
cffi/model.py,sha256=W30UFQZE73jL5Mx5N81YT77us2W2iJjTm0XYfnwz1cg,21797
cffi/parse_c_type.h,sha256=OdwQfwM9ktq6vlCB43exFQmxDBtj2MBNdK8LYl15tjw,5976
cffi/pkgconfig.py,sha256=LP1w7vmWvmKwyqLaU1Z243FOWGNQMrgMUZrvgFuOlco,4374
cffi/recompiler.py,sha256=sim4Tm7lamt2Jn8uzKN0wMYp6ODByk3g7of47-h9LD4,65367
cffi/setuptools_ext.py,sha256=-ebj79lO2_AUH-kRcaja2pKY1Z_5tloGwsJgzK8P3Cc,8871
cffi/vengine_cpy.py,sha256=8UagT6ZEOZf6Dju7_CfNulue8CnsHLEzJYhnqUhoF04,43752
cffi/vengine_gen.py,sha256=DUlEIrDiVin1Pnhn1sfoamnS5NLqfJcOdhRoeSNeJRg,26939
cffi/verifier.py,sha256=oX8jpaohg2Qm3aHcznidAdvrVm5N4sQYG0a3Eo5mIl4,11182

View File

@ -0,0 +1,6 @@
Wheel-Version: 1.0
Generator: setuptools (74.1.1)
Root-Is-Purelib: false
Tag: cp311-cp311-manylinux_2_17_x86_64
Tag: cp311-cp311-manylinux2014_x86_64

View File

@ -0,0 +1,2 @@
[distutils.setup_keywords]
cffi_modules = cffi.setuptools_ext:cffi_modules

View File

@ -0,0 +1,2 @@
_cffi_backend
cffi

View File

@ -0,0 +1,14 @@
__all__ = ['FFI', 'VerificationError', 'VerificationMissing', 'CDefError',
'FFIError']
from .api import FFI
from .error import CDefError, FFIError, VerificationError, VerificationMissing
from .error import PkgConfigError
__version__ = "1.17.1"
__version_info__ = (1, 17, 1)
# The verifier module file names are based on the CRC32 of a string that
# contains the following version number. It may be older than __version__
# if nothing is clearly incompatible.
__version_verifier_modules__ = "0.8.6"

View File

@ -0,0 +1,149 @@
#ifndef CFFI_MESSAGEBOX
# ifdef _MSC_VER
# define CFFI_MESSAGEBOX 1
# else
# define CFFI_MESSAGEBOX 0
# endif
#endif
#if CFFI_MESSAGEBOX
/* Windows only: logic to take the Python-CFFI embedding logic
initialization errors and display them in a background thread
with MessageBox. The idea is that if the whole program closes
as a result of this problem, then likely it is already a console
program and you can read the stderr output in the console too.
If it is not a console program, then it will likely show its own
dialog to complain, or generally not abruptly close, and for this
case the background thread should stay alive.
*/
static void *volatile _cffi_bootstrap_text;
static PyObject *_cffi_start_error_capture(void)
{
PyObject *result = NULL;
PyObject *x, *m, *bi;
if (InterlockedCompareExchangePointer(&_cffi_bootstrap_text,
(void *)1, NULL) != NULL)
return (PyObject *)1;
m = PyImport_AddModule("_cffi_error_capture");
if (m == NULL)
goto error;
result = PyModule_GetDict(m);
if (result == NULL)
goto error;
#if PY_MAJOR_VERSION >= 3
bi = PyImport_ImportModule("builtins");
#else
bi = PyImport_ImportModule("__builtin__");
#endif
if (bi == NULL)
goto error;
PyDict_SetItemString(result, "__builtins__", bi);
Py_DECREF(bi);
x = PyRun_String(
"import sys\n"
"class FileLike:\n"
" def write(self, x):\n"
" try:\n"
" of.write(x)\n"
" except: pass\n"
" self.buf += x\n"
" def flush(self):\n"
" pass\n"
"fl = FileLike()\n"
"fl.buf = ''\n"
"of = sys.stderr\n"
"sys.stderr = fl\n"
"def done():\n"
" sys.stderr = of\n"
" return fl.buf\n", /* make sure the returned value stays alive */
Py_file_input,
result, result);
Py_XDECREF(x);
error:
if (PyErr_Occurred())
{
PyErr_WriteUnraisable(Py_None);
PyErr_Clear();
}
return result;
}
#pragma comment(lib, "user32.lib")
static DWORD WINAPI _cffi_bootstrap_dialog(LPVOID ignored)
{
Sleep(666); /* may be interrupted if the whole process is closing */
#if PY_MAJOR_VERSION >= 3
MessageBoxW(NULL, (wchar_t *)_cffi_bootstrap_text,
L"Python-CFFI error",
MB_OK | MB_ICONERROR);
#else
MessageBoxA(NULL, (char *)_cffi_bootstrap_text,
"Python-CFFI error",
MB_OK | MB_ICONERROR);
#endif
_cffi_bootstrap_text = NULL;
return 0;
}
static void _cffi_stop_error_capture(PyObject *ecap)
{
PyObject *s;
void *text;
if (ecap == (PyObject *)1)
return;
if (ecap == NULL)
goto error;
s = PyRun_String("done()", Py_eval_input, ecap, ecap);
if (s == NULL)
goto error;
/* Show a dialog box, but in a background thread, and
never show multiple dialog boxes at once. */
#if PY_MAJOR_VERSION >= 3
text = PyUnicode_AsWideCharString(s, NULL);
#else
text = PyString_AsString(s);
#endif
_cffi_bootstrap_text = text;
if (text != NULL)
{
HANDLE h;
h = CreateThread(NULL, 0, _cffi_bootstrap_dialog,
NULL, 0, NULL);
if (h != NULL)
CloseHandle(h);
}
/* decref the string, but it should stay alive as 'fl.buf'
in the small module above. It will really be freed only if
we later get another similar error. So it's a leak of at
most one copy of the small module. That's fine for this
situation which is usually a "fatal error" anyway. */
Py_DECREF(s);
PyErr_Clear();
return;
error:
_cffi_bootstrap_text = NULL;
PyErr_Clear();
}
#else
static PyObject *_cffi_start_error_capture(void) { return NULL; }
static void _cffi_stop_error_capture(PyObject *ecap) { }
#endif

View File

@ -0,0 +1,389 @@
#define _CFFI_
/* We try to define Py_LIMITED_API before including Python.h.
Mess: we can only define it if Py_DEBUG, Py_TRACE_REFS and
Py_REF_DEBUG are not defined. This is a best-effort approximation:
we can learn about Py_DEBUG from pyconfig.h, but it is unclear if
the same works for the other two macros. Py_DEBUG implies them,
but not the other way around.
The implementation is messy (issue #350): on Windows, with _MSC_VER,
we have to define Py_LIMITED_API even before including pyconfig.h.
In that case, we guess what pyconfig.h will do to the macros above,
and check our guess after the #include.
Note that on Windows, with CPython 3.x, you need >= 3.5 and virtualenv
version >= 16.0.0. With older versions of either, you don't get a
copy of PYTHON3.DLL in the virtualenv. We can't check the version of
CPython *before* we even include pyconfig.h. ffi.set_source() puts
a ``#define _CFFI_NO_LIMITED_API'' at the start of this file if it is
running on Windows < 3.5, as an attempt at fixing it, but that's
arguably wrong because it may not be the target version of Python.
Still better than nothing I guess. As another workaround, you can
remove the definition of Py_LIMITED_API here.
See also 'py_limited_api' in cffi/setuptools_ext.py.
*/
#if !defined(_CFFI_USE_EMBEDDING) && !defined(Py_LIMITED_API)
# ifdef _MSC_VER
# if !defined(_DEBUG) && !defined(Py_DEBUG) && !defined(Py_TRACE_REFS) && !defined(Py_REF_DEBUG) && !defined(_CFFI_NO_LIMITED_API)
# define Py_LIMITED_API
# endif
# include <pyconfig.h>
/* sanity-check: Py_LIMITED_API will cause crashes if any of these
are also defined. Normally, the Python file PC/pyconfig.h does not
cause any of these to be defined, with the exception that _DEBUG
causes Py_DEBUG. Double-check that. */
# ifdef Py_LIMITED_API
# if defined(Py_DEBUG)
# error "pyconfig.h unexpectedly defines Py_DEBUG, but Py_LIMITED_API is set"
# endif
# if defined(Py_TRACE_REFS)
# error "pyconfig.h unexpectedly defines Py_TRACE_REFS, but Py_LIMITED_API is set"
# endif
# if defined(Py_REF_DEBUG)
# error "pyconfig.h unexpectedly defines Py_REF_DEBUG, but Py_LIMITED_API is set"
# endif
# endif
# else
# include <pyconfig.h>
# if !defined(Py_DEBUG) && !defined(Py_TRACE_REFS) && !defined(Py_REF_DEBUG) && !defined(_CFFI_NO_LIMITED_API)
# define Py_LIMITED_API
# endif
# endif
#endif
#include <Python.h>
#ifdef __cplusplus
extern "C" {
#endif
#include <stddef.h>
#include "parse_c_type.h"
/* this block of #ifs should be kept exactly identical between
c/_cffi_backend.c, cffi/vengine_cpy.py, cffi/vengine_gen.py
and cffi/_cffi_include.h */
#if defined(_MSC_VER)
# include <malloc.h> /* for alloca() */
# if _MSC_VER < 1600 /* MSVC < 2010 */
typedef __int8 int8_t;
typedef __int16 int16_t;
typedef __int32 int32_t;
typedef __int64 int64_t;
typedef unsigned __int8 uint8_t;
typedef unsigned __int16 uint16_t;
typedef unsigned __int32 uint32_t;
typedef unsigned __int64 uint64_t;
typedef __int8 int_least8_t;
typedef __int16 int_least16_t;
typedef __int32 int_least32_t;
typedef __int64 int_least64_t;
typedef unsigned __int8 uint_least8_t;
typedef unsigned __int16 uint_least16_t;
typedef unsigned __int32 uint_least32_t;
typedef unsigned __int64 uint_least64_t;
typedef __int8 int_fast8_t;
typedef __int16 int_fast16_t;
typedef __int32 int_fast32_t;
typedef __int64 int_fast64_t;
typedef unsigned __int8 uint_fast8_t;
typedef unsigned __int16 uint_fast16_t;
typedef unsigned __int32 uint_fast32_t;
typedef unsigned __int64 uint_fast64_t;
typedef __int64 intmax_t;
typedef unsigned __int64 uintmax_t;
# else
# include <stdint.h>
# endif
# if _MSC_VER < 1800 /* MSVC < 2013 */
# ifndef __cplusplus
typedef unsigned char _Bool;
# endif
# endif
# define _cffi_float_complex_t _Fcomplex /* include <complex.h> for it */
# define _cffi_double_complex_t _Dcomplex /* include <complex.h> for it */
#else
# include <stdint.h>
# if (defined (__SVR4) && defined (__sun)) || defined(_AIX) || defined(__hpux)
# include <alloca.h>
# endif
# define _cffi_float_complex_t float _Complex
# define _cffi_double_complex_t double _Complex
#endif
#ifdef __GNUC__
# define _CFFI_UNUSED_FN __attribute__((unused))
#else
# define _CFFI_UNUSED_FN /* nothing */
#endif
#ifdef __cplusplus
# ifndef _Bool
typedef bool _Bool; /* semi-hackish: C++ has no _Bool; bool is builtin */
# endif
#endif
/********** CPython-specific section **********/
#ifndef PYPY_VERSION
#if PY_MAJOR_VERSION >= 3
# define PyInt_FromLong PyLong_FromLong
#endif
#define _cffi_from_c_double PyFloat_FromDouble
#define _cffi_from_c_float PyFloat_FromDouble
#define _cffi_from_c_long PyInt_FromLong
#define _cffi_from_c_ulong PyLong_FromUnsignedLong
#define _cffi_from_c_longlong PyLong_FromLongLong
#define _cffi_from_c_ulonglong PyLong_FromUnsignedLongLong
#define _cffi_from_c__Bool PyBool_FromLong
#define _cffi_to_c_double PyFloat_AsDouble
#define _cffi_to_c_float PyFloat_AsDouble
#define _cffi_from_c_int(x, type) \
(((type)-1) > 0 ? /* unsigned */ \
(sizeof(type) < sizeof(long) ? \
PyInt_FromLong((long)x) : \
sizeof(type) == sizeof(long) ? \
PyLong_FromUnsignedLong((unsigned long)x) : \
PyLong_FromUnsignedLongLong((unsigned long long)x)) : \
(sizeof(type) <= sizeof(long) ? \
PyInt_FromLong((long)x) : \
PyLong_FromLongLong((long long)x)))
#define _cffi_to_c_int(o, type) \
((type)( \
sizeof(type) == 1 ? (((type)-1) > 0 ? (type)_cffi_to_c_u8(o) \
: (type)_cffi_to_c_i8(o)) : \
sizeof(type) == 2 ? (((type)-1) > 0 ? (type)_cffi_to_c_u16(o) \
: (type)_cffi_to_c_i16(o)) : \
sizeof(type) == 4 ? (((type)-1) > 0 ? (type)_cffi_to_c_u32(o) \
: (type)_cffi_to_c_i32(o)) : \
sizeof(type) == 8 ? (((type)-1) > 0 ? (type)_cffi_to_c_u64(o) \
: (type)_cffi_to_c_i64(o)) : \
(Py_FatalError("unsupported size for type " #type), (type)0)))
#define _cffi_to_c_i8 \
((int(*)(PyObject *))_cffi_exports[1])
#define _cffi_to_c_u8 \
((int(*)(PyObject *))_cffi_exports[2])
#define _cffi_to_c_i16 \
((int(*)(PyObject *))_cffi_exports[3])
#define _cffi_to_c_u16 \
((int(*)(PyObject *))_cffi_exports[4])
#define _cffi_to_c_i32 \
((int(*)(PyObject *))_cffi_exports[5])
#define _cffi_to_c_u32 \
((unsigned int(*)(PyObject *))_cffi_exports[6])
#define _cffi_to_c_i64 \
((long long(*)(PyObject *))_cffi_exports[7])
#define _cffi_to_c_u64 \
((unsigned long long(*)(PyObject *))_cffi_exports[8])
#define _cffi_to_c_char \
((int(*)(PyObject *))_cffi_exports[9])
#define _cffi_from_c_pointer \
((PyObject *(*)(char *, struct _cffi_ctypedescr *))_cffi_exports[10])
#define _cffi_to_c_pointer \
((char *(*)(PyObject *, struct _cffi_ctypedescr *))_cffi_exports[11])
#define _cffi_get_struct_layout \
not used any more
#define _cffi_restore_errno \
((void(*)(void))_cffi_exports[13])
#define _cffi_save_errno \
((void(*)(void))_cffi_exports[14])
#define _cffi_from_c_char \
((PyObject *(*)(char))_cffi_exports[15])
#define _cffi_from_c_deref \
((PyObject *(*)(char *, struct _cffi_ctypedescr *))_cffi_exports[16])
#define _cffi_to_c \
((int(*)(char *, struct _cffi_ctypedescr *, PyObject *))_cffi_exports[17])
#define _cffi_from_c_struct \
((PyObject *(*)(char *, struct _cffi_ctypedescr *))_cffi_exports[18])
#define _cffi_to_c_wchar_t \
((_cffi_wchar_t(*)(PyObject *))_cffi_exports[19])
#define _cffi_from_c_wchar_t \
((PyObject *(*)(_cffi_wchar_t))_cffi_exports[20])
#define _cffi_to_c_long_double \
((long double(*)(PyObject *))_cffi_exports[21])
#define _cffi_to_c__Bool \
((_Bool(*)(PyObject *))_cffi_exports[22])
#define _cffi_prepare_pointer_call_argument \
((Py_ssize_t(*)(struct _cffi_ctypedescr *, \
PyObject *, char **))_cffi_exports[23])
#define _cffi_convert_array_from_object \
((int(*)(char *, struct _cffi_ctypedescr *, PyObject *))_cffi_exports[24])
#define _CFFI_CPIDX 25
#define _cffi_call_python \
((void(*)(struct _cffi_externpy_s *, char *))_cffi_exports[_CFFI_CPIDX])
#define _cffi_to_c_wchar3216_t \
((int(*)(PyObject *))_cffi_exports[26])
#define _cffi_from_c_wchar3216_t \
((PyObject *(*)(int))_cffi_exports[27])
#define _CFFI_NUM_EXPORTS 28
struct _cffi_ctypedescr;
static void *_cffi_exports[_CFFI_NUM_EXPORTS];
#define _cffi_type(index) ( \
assert((((uintptr_t)_cffi_types[index]) & 1) == 0), \
(struct _cffi_ctypedescr *)_cffi_types[index])
static PyObject *_cffi_init(const char *module_name, Py_ssize_t version,
const struct _cffi_type_context_s *ctx)
{
PyObject *module, *o_arg, *new_module;
void *raw[] = {
(void *)module_name,
(void *)version,
(void *)_cffi_exports,
(void *)ctx,
};
module = PyImport_ImportModule("_cffi_backend");
if (module == NULL)
goto failure;
o_arg = PyLong_FromVoidPtr((void *)raw);
if (o_arg == NULL)
goto failure;
new_module = PyObject_CallMethod(
module, (char *)"_init_cffi_1_0_external_module", (char *)"O", o_arg);
Py_DECREF(o_arg);
Py_DECREF(module);
return new_module;
failure:
Py_XDECREF(module);
return NULL;
}
#ifdef HAVE_WCHAR_H
typedef wchar_t _cffi_wchar_t;
#else
typedef uint16_t _cffi_wchar_t; /* same random pick as _cffi_backend.c */
#endif
_CFFI_UNUSED_FN static uint16_t _cffi_to_c_char16_t(PyObject *o)
{
if (sizeof(_cffi_wchar_t) == 2)
return (uint16_t)_cffi_to_c_wchar_t(o);
else
return (uint16_t)_cffi_to_c_wchar3216_t(o);
}
_CFFI_UNUSED_FN static PyObject *_cffi_from_c_char16_t(uint16_t x)
{
if (sizeof(_cffi_wchar_t) == 2)
return _cffi_from_c_wchar_t((_cffi_wchar_t)x);
else
return _cffi_from_c_wchar3216_t((int)x);
}
_CFFI_UNUSED_FN static int _cffi_to_c_char32_t(PyObject *o)
{
if (sizeof(_cffi_wchar_t) == 4)
return (int)_cffi_to_c_wchar_t(o);
else
return (int)_cffi_to_c_wchar3216_t(o);
}
_CFFI_UNUSED_FN static PyObject *_cffi_from_c_char32_t(unsigned int x)
{
if (sizeof(_cffi_wchar_t) == 4)
return _cffi_from_c_wchar_t((_cffi_wchar_t)x);
else
return _cffi_from_c_wchar3216_t((int)x);
}
union _cffi_union_alignment_u {
unsigned char m_char;
unsigned short m_short;
unsigned int m_int;
unsigned long m_long;
unsigned long long m_longlong;
float m_float;
double m_double;
long double m_longdouble;
};
struct _cffi_freeme_s {
struct _cffi_freeme_s *next;
union _cffi_union_alignment_u alignment;
};
_CFFI_UNUSED_FN static int
_cffi_convert_array_argument(struct _cffi_ctypedescr *ctptr, PyObject *arg,
char **output_data, Py_ssize_t datasize,
struct _cffi_freeme_s **freeme)
{
char *p;
if (datasize < 0)
return -1;
p = *output_data;
if (p == NULL) {
struct _cffi_freeme_s *fp = (struct _cffi_freeme_s *)PyObject_Malloc(
offsetof(struct _cffi_freeme_s, alignment) + (size_t)datasize);
if (fp == NULL)
return -1;
fp->next = *freeme;
*freeme = fp;
p = *output_data = (char *)&fp->alignment;
}
memset((void *)p, 0, (size_t)datasize);
return _cffi_convert_array_from_object(p, ctptr, arg);
}
_CFFI_UNUSED_FN static void
_cffi_free_array_arguments(struct _cffi_freeme_s *freeme)
{
do {
void *p = (void *)freeme;
freeme = freeme->next;
PyObject_Free(p);
} while (freeme != NULL);
}
/********** end CPython-specific section **********/
#else
_CFFI_UNUSED_FN
static void (*_cffi_call_python_org)(struct _cffi_externpy_s *, char *);
# define _cffi_call_python _cffi_call_python_org
#endif
#define _cffi_array_len(array) (sizeof(array) / sizeof((array)[0]))
#define _cffi_prim_int(size, sign) \
((size) == 1 ? ((sign) ? _CFFI_PRIM_INT8 : _CFFI_PRIM_UINT8) : \
(size) == 2 ? ((sign) ? _CFFI_PRIM_INT16 : _CFFI_PRIM_UINT16) : \
(size) == 4 ? ((sign) ? _CFFI_PRIM_INT32 : _CFFI_PRIM_UINT32) : \
(size) == 8 ? ((sign) ? _CFFI_PRIM_INT64 : _CFFI_PRIM_UINT64) : \
_CFFI__UNKNOWN_PRIM)
#define _cffi_prim_float(size) \
((size) == sizeof(float) ? _CFFI_PRIM_FLOAT : \
(size) == sizeof(double) ? _CFFI_PRIM_DOUBLE : \
(size) == sizeof(long double) ? _CFFI__UNKNOWN_LONG_DOUBLE : \
_CFFI__UNKNOWN_FLOAT_PRIM)
#define _cffi_check_int(got, got_nonpos, expected) \
((got_nonpos) == (expected <= 0) && \
(got) == (unsigned long long)expected)
#ifdef MS_WIN32
# define _cffi_stdcall __stdcall
#else
# define _cffi_stdcall /* nothing */
#endif
#ifdef __cplusplus
}
#endif

View File

@ -0,0 +1,550 @@
/***** Support code for embedding *****/
#ifdef __cplusplus
extern "C" {
#endif
#if defined(_WIN32)
# define CFFI_DLLEXPORT __declspec(dllexport)
#elif defined(__GNUC__)
# define CFFI_DLLEXPORT __attribute__((visibility("default")))
#else
# define CFFI_DLLEXPORT /* nothing */
#endif
/* There are two global variables of type _cffi_call_python_fnptr:
* _cffi_call_python, which we declare just below, is the one called
by ``extern "Python"`` implementations.
* _cffi_call_python_org, which on CPython is actually part of the
_cffi_exports[] array, is the function pointer copied from
_cffi_backend. If _cffi_start_python() fails, then this is set
to NULL; otherwise, it should never be NULL.
After initialization is complete, both are equal. However, the
first one remains equal to &_cffi_start_and_call_python until the
very end of initialization, when we are (or should be) sure that
concurrent threads also see a completely initialized world, and
only then is it changed.
*/
#undef _cffi_call_python
typedef void (*_cffi_call_python_fnptr)(struct _cffi_externpy_s *, char *);
static void _cffi_start_and_call_python(struct _cffi_externpy_s *, char *);
static _cffi_call_python_fnptr _cffi_call_python = &_cffi_start_and_call_python;
#ifndef _MSC_VER
/* --- Assuming a GCC not infinitely old --- */
# define cffi_compare_and_swap(l,o,n) __sync_bool_compare_and_swap(l,o,n)
# define cffi_write_barrier() __sync_synchronize()
# if !defined(__amd64__) && !defined(__x86_64__) && \
!defined(__i386__) && !defined(__i386)
# define cffi_read_barrier() __sync_synchronize()
# else
# define cffi_read_barrier() (void)0
# endif
#else
/* --- Windows threads version --- */
# include <Windows.h>
# define cffi_compare_and_swap(l,o,n) \
(InterlockedCompareExchangePointer(l,n,o) == (o))
# define cffi_write_barrier() InterlockedCompareExchange(&_cffi_dummy,0,0)
# define cffi_read_barrier() (void)0
static volatile LONG _cffi_dummy;
#endif
#ifdef WITH_THREAD
# ifndef _MSC_VER
# include <pthread.h>
static pthread_mutex_t _cffi_embed_startup_lock;
# else
static CRITICAL_SECTION _cffi_embed_startup_lock;
# endif
static char _cffi_embed_startup_lock_ready = 0;
#endif
static void _cffi_acquire_reentrant_mutex(void)
{
static void *volatile lock = NULL;
while (!cffi_compare_and_swap(&lock, NULL, (void *)1)) {
/* should ideally do a spin loop instruction here, but
hard to do it portably and doesn't really matter I
think: pthread_mutex_init() should be very fast, and
this is only run at start-up anyway. */
}
#ifdef WITH_THREAD
if (!_cffi_embed_startup_lock_ready) {
# ifndef _MSC_VER
pthread_mutexattr_t attr;
pthread_mutexattr_init(&attr);
pthread_mutexattr_settype(&attr, PTHREAD_MUTEX_RECURSIVE);
pthread_mutex_init(&_cffi_embed_startup_lock, &attr);
# else
InitializeCriticalSection(&_cffi_embed_startup_lock);
# endif
_cffi_embed_startup_lock_ready = 1;
}
#endif
while (!cffi_compare_and_swap(&lock, (void *)1, NULL))
;
#ifndef _MSC_VER
pthread_mutex_lock(&_cffi_embed_startup_lock);
#else
EnterCriticalSection(&_cffi_embed_startup_lock);
#endif
}
static void _cffi_release_reentrant_mutex(void)
{
#ifndef _MSC_VER
pthread_mutex_unlock(&_cffi_embed_startup_lock);
#else
LeaveCriticalSection(&_cffi_embed_startup_lock);
#endif
}
/********** CPython-specific section **********/
#ifndef PYPY_VERSION
#include "_cffi_errors.h"
#define _cffi_call_python_org _cffi_exports[_CFFI_CPIDX]
PyMODINIT_FUNC _CFFI_PYTHON_STARTUP_FUNC(void); /* forward */
static void _cffi_py_initialize(void)
{
/* XXX use initsigs=0, which "skips initialization registration of
signal handlers, which might be useful when Python is
embedded" according to the Python docs. But review and think
if it should be a user-controllable setting.
XXX we should also give a way to write errors to a buffer
instead of to stderr.
XXX if importing 'site' fails, CPython (any version) calls
exit(). Should we try to work around this behavior here?
*/
Py_InitializeEx(0);
}
static int _cffi_initialize_python(void)
{
/* This initializes Python, imports _cffi_backend, and then the
present .dll/.so is set up as a CPython C extension module.
*/
int result;
PyGILState_STATE state;
PyObject *pycode=NULL, *global_dict=NULL, *x;
PyObject *builtins;
state = PyGILState_Ensure();
/* Call the initxxx() function from the present module. It will
create and initialize us as a CPython extension module, instead
of letting the startup Python code do it---it might reimport
the same .dll/.so and get maybe confused on some platforms.
It might also have troubles locating the .dll/.so again for all
I know.
*/
(void)_CFFI_PYTHON_STARTUP_FUNC();
if (PyErr_Occurred())
goto error;
/* Now run the Python code provided to ffi.embedding_init_code().
*/
pycode = Py_CompileString(_CFFI_PYTHON_STARTUP_CODE,
"<init code for '" _CFFI_MODULE_NAME "'>",
Py_file_input);
if (pycode == NULL)
goto error;
global_dict = PyDict_New();
if (global_dict == NULL)
goto error;
builtins = PyEval_GetBuiltins();
if (builtins == NULL)
goto error;
if (PyDict_SetItemString(global_dict, "__builtins__", builtins) < 0)
goto error;
x = PyEval_EvalCode(
#if PY_MAJOR_VERSION < 3
(PyCodeObject *)
#endif
pycode, global_dict, global_dict);
if (x == NULL)
goto error;
Py_DECREF(x);
/* Done! Now if we've been called from
_cffi_start_and_call_python() in an ``extern "Python"``, we can
only hope that the Python code did correctly set up the
corresponding @ffi.def_extern() function. Otherwise, the
general logic of ``extern "Python"`` functions (inside the
_cffi_backend module) will find that the reference is still
missing and print an error.
*/
result = 0;
done:
Py_XDECREF(pycode);
Py_XDECREF(global_dict);
PyGILState_Release(state);
return result;
error:;
{
/* Print as much information as potentially useful.
Debugging load-time failures with embedding is not fun
*/
PyObject *ecap;
PyObject *exception, *v, *tb, *f, *modules, *mod;
PyErr_Fetch(&exception, &v, &tb);
ecap = _cffi_start_error_capture();
f = PySys_GetObject((char *)"stderr");
if (f != NULL && f != Py_None) {
PyFile_WriteString(
"Failed to initialize the Python-CFFI embedding logic:\n\n", f);
}
if (exception != NULL) {
PyErr_NormalizeException(&exception, &v, &tb);
PyErr_Display(exception, v, tb);
}
Py_XDECREF(exception);
Py_XDECREF(v);
Py_XDECREF(tb);
if (f != NULL && f != Py_None) {
PyFile_WriteString("\nFrom: " _CFFI_MODULE_NAME
"\ncompiled with cffi version: 1.17.1"
"\n_cffi_backend module: ", f);
modules = PyImport_GetModuleDict();
mod = PyDict_GetItemString(modules, "_cffi_backend");
if (mod == NULL) {
PyFile_WriteString("not loaded", f);
}
else {
v = PyObject_GetAttrString(mod, "__file__");
PyFile_WriteObject(v, f, 0);
Py_XDECREF(v);
}
PyFile_WriteString("\nsys.path: ", f);
PyFile_WriteObject(PySys_GetObject((char *)"path"), f, 0);
PyFile_WriteString("\n\n", f);
}
_cffi_stop_error_capture(ecap);
}
result = -1;
goto done;
}
#if PY_VERSION_HEX < 0x03080000
PyAPI_DATA(char *) _PyParser_TokenNames[]; /* from CPython */
#endif
static int _cffi_carefully_make_gil(void)
{
/* This does the basic initialization of Python. It can be called
completely concurrently from unrelated threads. It assumes
that we don't hold the GIL before (if it exists), and we don't
hold it afterwards.
(What it really does used to be completely different in Python 2
and Python 3, with the Python 2 solution avoiding the spin-lock
around the Py_InitializeEx() call. However, after recent changes
to CPython 2.7 (issue #358) it no longer works. So we use the
Python 3 solution everywhere.)
This initializes Python by calling Py_InitializeEx().
Important: this must not be called concurrently at all.
So we use a global variable as a simple spin lock. This global
variable must be from 'libpythonX.Y.so', not from this
cffi-based extension module, because it must be shared from
different cffi-based extension modules.
In Python < 3.8, we choose
_PyParser_TokenNames[0] as a completely arbitrary pointer value
that is never written to. The default is to point to the
string "ENDMARKER". We change it temporarily to point to the
next character in that string. (Yes, I know it's REALLY
obscure.)
In Python >= 3.8, this string array is no longer writable, so
instead we pick PyCapsuleType.tp_version_tag. We can't change
Python < 3.8 because someone might use a mixture of cffi
embedded modules, some of which were compiled before this file
changed.
In Python >= 3.12, this stopped working because that particular
tp_version_tag gets modified during interpreter startup. It's
arguably a bad idea before 3.12 too, but again we can't change
that because someone might use a mixture of cffi embedded
modules, and no-one reported a bug so far. In Python >= 3.12
we go instead for PyCapsuleType.tp_as_buffer, which is supposed
to always be NULL. We write to it temporarily a pointer to
a struct full of NULLs, which is semantically the same.
*/
#ifdef WITH_THREAD
# if PY_VERSION_HEX < 0x03080000
char *volatile *lock = (char *volatile *)_PyParser_TokenNames;
char *old_value, *locked_value;
while (1) { /* spin loop */
old_value = *lock;
locked_value = old_value + 1;
if (old_value[0] == 'E') {
assert(old_value[1] == 'N');
if (cffi_compare_and_swap(lock, old_value, locked_value))
break;
}
else {
assert(old_value[0] == 'N');
/* should ideally do a spin loop instruction here, but
hard to do it portably and doesn't really matter I
think: PyEval_InitThreads() should be very fast, and
this is only run at start-up anyway. */
}
}
# else
# if PY_VERSION_HEX < 0x030C0000
int volatile *lock = (int volatile *)&PyCapsule_Type.tp_version_tag;
int old_value, locked_value = -42;
assert(!(PyCapsule_Type.tp_flags & Py_TPFLAGS_HAVE_VERSION_TAG));
# else
static struct ebp_s { PyBufferProcs buf; int mark; } empty_buffer_procs;
empty_buffer_procs.mark = -42;
PyBufferProcs *volatile *lock = (PyBufferProcs *volatile *)
&PyCapsule_Type.tp_as_buffer;
PyBufferProcs *old_value, *locked_value = &empty_buffer_procs.buf;
# endif
while (1) { /* spin loop */
old_value = *lock;
if (old_value == 0) {
if (cffi_compare_and_swap(lock, old_value, locked_value))
break;
}
else {
# if PY_VERSION_HEX < 0x030C0000
assert(old_value == locked_value);
# else
/* The pointer should point to a possibly different
empty_buffer_procs from another C extension module */
assert(((struct ebp_s *)old_value)->mark == -42);
# endif
/* should ideally do a spin loop instruction here, but
hard to do it portably and doesn't really matter I
think: PyEval_InitThreads() should be very fast, and
this is only run at start-up anyway. */
}
}
# endif
#endif
/* call Py_InitializeEx() */
if (!Py_IsInitialized()) {
_cffi_py_initialize();
#if PY_VERSION_HEX < 0x03070000
PyEval_InitThreads();
#endif
PyEval_SaveThread(); /* release the GIL */
/* the returned tstate must be the one that has been stored into the
autoTLSkey by _PyGILState_Init() called from Py_Initialize(). */
}
else {
#if PY_VERSION_HEX < 0x03070000
/* PyEval_InitThreads() is always a no-op from CPython 3.7 */
PyGILState_STATE state = PyGILState_Ensure();
PyEval_InitThreads();
PyGILState_Release(state);
#endif
}
#ifdef WITH_THREAD
/* release the lock */
while (!cffi_compare_and_swap(lock, locked_value, old_value))
;
#endif
return 0;
}
/********** end CPython-specific section **********/
#else
/********** PyPy-specific section **********/
PyMODINIT_FUNC _CFFI_PYTHON_STARTUP_FUNC(const void *[]); /* forward */
static struct _cffi_pypy_init_s {
const char *name;
void *func; /* function pointer */
const char *code;
} _cffi_pypy_init = {
_CFFI_MODULE_NAME,
_CFFI_PYTHON_STARTUP_FUNC,
_CFFI_PYTHON_STARTUP_CODE,
};
extern int pypy_carefully_make_gil(const char *);
extern int pypy_init_embedded_cffi_module(int, struct _cffi_pypy_init_s *);
static int _cffi_carefully_make_gil(void)
{
return pypy_carefully_make_gil(_CFFI_MODULE_NAME);
}
static int _cffi_initialize_python(void)
{
return pypy_init_embedded_cffi_module(0xB011, &_cffi_pypy_init);
}
/********** end PyPy-specific section **********/
#endif
#ifdef __GNUC__
__attribute__((noinline))
#endif
static _cffi_call_python_fnptr _cffi_start_python(void)
{
/* Delicate logic to initialize Python. This function can be
called multiple times concurrently, e.g. when the process calls
its first ``extern "Python"`` functions in multiple threads at
once. It can also be called recursively, in which case we must
ignore it. We also have to consider what occurs if several
different cffi-based extensions reach this code in parallel
threads---it is a different copy of the code, then, and we
can't have any shared global variable unless it comes from
'libpythonX.Y.so'.
Idea:
* _cffi_carefully_make_gil(): "carefully" call
PyEval_InitThreads() (possibly with Py_InitializeEx() first).
* then we use a (local) custom lock to make sure that a call to this
cffi-based extension will wait if another call to the *same*
extension is running the initialization in another thread.
It is reentrant, so that a recursive call will not block, but
only one from a different thread.
* then we grab the GIL and (Python 2) we call Py_InitializeEx().
At this point, concurrent calls to Py_InitializeEx() are not
possible: we have the GIL.
* do the rest of the specific initialization, which may
temporarily release the GIL but not the custom lock.
Only release the custom lock when we are done.
*/
static char called = 0;
if (_cffi_carefully_make_gil() != 0)
return NULL;
_cffi_acquire_reentrant_mutex();
/* Here the GIL exists, but we don't have it. We're only protected
from concurrency by the reentrant mutex. */
/* This file only initializes the embedded module once, the first
time this is called, even if there are subinterpreters. */
if (!called) {
called = 1; /* invoke _cffi_initialize_python() only once,
but don't set '_cffi_call_python' right now,
otherwise concurrent threads won't call
this function at all (we need them to wait) */
if (_cffi_initialize_python() == 0) {
/* now initialization is finished. Switch to the fast-path. */
/* We would like nobody to see the new value of
'_cffi_call_python' without also seeing the rest of the
data initialized. However, this is not possible. But
the new value of '_cffi_call_python' is the function
'cffi_call_python()' from _cffi_backend. So: */
cffi_write_barrier();
/* ^^^ we put a write barrier here, and a corresponding
read barrier at the start of cffi_call_python(). This
ensures that after that read barrier, we see everything
done here before the write barrier.
*/
assert(_cffi_call_python_org != NULL);
_cffi_call_python = (_cffi_call_python_fnptr)_cffi_call_python_org;
}
else {
/* initialization failed. Reset this to NULL, even if it was
already set to some other value. Future calls to
_cffi_start_python() are still forced to occur, and will
always return NULL from now on. */
_cffi_call_python_org = NULL;
}
}
_cffi_release_reentrant_mutex();
return (_cffi_call_python_fnptr)_cffi_call_python_org;
}
static
void _cffi_start_and_call_python(struct _cffi_externpy_s *externpy, char *args)
{
_cffi_call_python_fnptr fnptr;
int current_err = errno;
#ifdef _MSC_VER
int current_lasterr = GetLastError();
#endif
fnptr = _cffi_start_python();
if (fnptr == NULL) {
fprintf(stderr, "function %s() called, but initialization code "
"failed. Returning 0.\n", externpy->name);
memset(args, 0, externpy->size_of_result);
}
#ifdef _MSC_VER
SetLastError(current_lasterr);
#endif
errno = current_err;
if (fnptr != NULL)
fnptr(externpy, args);
}
/* The cffi_start_python() function makes sure Python is initialized
and our cffi module is set up. It can be called manually from the
user C code. The same effect is obtained automatically from any
dll-exported ``extern "Python"`` function. This function returns
-1 if initialization failed, 0 if all is OK. */
_CFFI_UNUSED_FN
static int cffi_start_python(void)
{
if (_cffi_call_python == &_cffi_start_and_call_python) {
if (_cffi_start_python() == NULL)
return -1;
}
cffi_read_barrier();
return 0;
}
#undef cffi_compare_and_swap
#undef cffi_write_barrier
#undef cffi_read_barrier
#ifdef __cplusplus
}
#endif

View File

@ -0,0 +1,83 @@
try:
# this works on Python < 3.12
from imp import *
except ImportError:
# this is a limited emulation for Python >= 3.12.
# Note that this is used only for tests or for the old ffi.verify().
# This is copied from the source code of Python 3.11.
from _imp import (acquire_lock, release_lock,
is_builtin, is_frozen)
from importlib._bootstrap import _load
from importlib import machinery
import os
import sys
import tokenize
SEARCH_ERROR = 0
PY_SOURCE = 1
PY_COMPILED = 2
C_EXTENSION = 3
PY_RESOURCE = 4
PKG_DIRECTORY = 5
C_BUILTIN = 6
PY_FROZEN = 7
PY_CODERESOURCE = 8
IMP_HOOK = 9
def get_suffixes():
extensions = [(s, 'rb', C_EXTENSION)
for s in machinery.EXTENSION_SUFFIXES]
source = [(s, 'r', PY_SOURCE) for s in machinery.SOURCE_SUFFIXES]
bytecode = [(s, 'rb', PY_COMPILED) for s in machinery.BYTECODE_SUFFIXES]
return extensions + source + bytecode
def find_module(name, path=None):
if not isinstance(name, str):
raise TypeError("'name' must be a str, not {}".format(type(name)))
elif not isinstance(path, (type(None), list)):
# Backwards-compatibility
raise RuntimeError("'path' must be None or a list, "
"not {}".format(type(path)))
if path is None:
if is_builtin(name):
return None, None, ('', '', C_BUILTIN)
elif is_frozen(name):
return None, None, ('', '', PY_FROZEN)
else:
path = sys.path
for entry in path:
package_directory = os.path.join(entry, name)
for suffix in ['.py', machinery.BYTECODE_SUFFIXES[0]]:
package_file_name = '__init__' + suffix
file_path = os.path.join(package_directory, package_file_name)
if os.path.isfile(file_path):
return None, package_directory, ('', '', PKG_DIRECTORY)
for suffix, mode, type_ in get_suffixes():
file_name = name + suffix
file_path = os.path.join(entry, file_name)
if os.path.isfile(file_path):
break
else:
continue
break # Break out of outer loop when breaking out of inner loop.
else:
raise ImportError(name, name=name)
encoding = None
if 'b' not in mode:
with open(file_path, 'rb') as file:
encoding = tokenize.detect_encoding(file.readline)[0]
file = open(file_path, mode, encoding=encoding)
return file, file_path, (suffix, mode, type_)
def load_dynamic(name, path, file=None):
loader = machinery.ExtensionFileLoader(name, path)
spec = machinery.ModuleSpec(name=name, loader=loader, origin=path)
return _load(spec)

View File

@ -0,0 +1,45 @@
"""
Temporary shim module to indirect the bits of distutils we need from setuptools/distutils while providing useful
error messages beyond `No module named 'distutils' on Python >= 3.12, or when setuptools' vendored distutils is broken.
This is a compromise to avoid a hard-dep on setuptools for Python >= 3.12, since many users don't need runtime compilation support from CFFI.
"""
import sys
try:
# import setuptools first; this is the most robust way to ensure its embedded distutils is available
# (the .pth shim should usually work, but this is even more robust)
import setuptools
except Exception as ex:
if sys.version_info >= (3, 12):
# Python 3.12 has no built-in distutils to fall back on, so any import problem is fatal
raise Exception("This CFFI feature requires setuptools on Python >= 3.12. The setuptools module is missing or non-functional.") from ex
# silently ignore on older Pythons (support fallback to stdlib distutils where available)
else:
del setuptools
try:
# bring in just the bits of distutils we need, whether they really came from setuptools or stdlib-embedded distutils
from distutils import log, sysconfig
from distutils.ccompiler import CCompiler
from distutils.command.build_ext import build_ext
from distutils.core import Distribution, Extension
from distutils.dir_util import mkpath
from distutils.errors import DistutilsSetupError, CompileError, LinkError
from distutils.log import set_threshold, set_verbosity
if sys.platform == 'win32':
try:
# FUTURE: msvc9compiler module was removed in setuptools 74; consider removing, as it's only used by an ancient patch in `recompiler`
from distutils.msvc9compiler import MSVCCompiler
except ImportError:
MSVCCompiler = None
except Exception as ex:
if sys.version_info >= (3, 12):
raise Exception("This CFFI feature requires setuptools on Python >= 3.12. Please install the setuptools package.") from ex
# anything older, just let the underlying distutils import error fly
raise Exception("This CFFI feature requires distutils. Please install the distutils or setuptools package.") from ex
del sys

View File

@ -0,0 +1,967 @@
import sys, types
from .lock import allocate_lock
from .error import CDefError
from . import model
try:
callable
except NameError:
# Python 3.1
from collections import Callable
callable = lambda x: isinstance(x, Callable)
try:
basestring
except NameError:
# Python 3.x
basestring = str
_unspecified = object()
class FFI(object):
r'''
The main top-level class that you instantiate once, or once per module.
Example usage:
ffi = FFI()
ffi.cdef("""
int printf(const char *, ...);
""")
C = ffi.dlopen(None) # standard library
-or-
C = ffi.verify() # use a C compiler: verify the decl above is right
C.printf("hello, %s!\n", ffi.new("char[]", "world"))
'''
def __init__(self, backend=None):
"""Create an FFI instance. The 'backend' argument is used to
select a non-default backend, mostly for tests.
"""
if backend is None:
# You need PyPy (>= 2.0 beta), or a CPython (>= 2.6) with
# _cffi_backend.so compiled.
import _cffi_backend as backend
from . import __version__
if backend.__version__ != __version__:
# bad version! Try to be as explicit as possible.
if hasattr(backend, '__file__'):
# CPython
raise Exception("Version mismatch: this is the 'cffi' package version %s, located in %r. When we import the top-level '_cffi_backend' extension module, we get version %s, located in %r. The two versions should be equal; check your installation." % (
__version__, __file__,
backend.__version__, backend.__file__))
else:
# PyPy
raise Exception("Version mismatch: this is the 'cffi' package version %s, located in %r. This interpreter comes with a built-in '_cffi_backend' module, which is version %s. The two versions should be equal; check your installation." % (
__version__, __file__, backend.__version__))
# (If you insist you can also try to pass the option
# 'backend=backend_ctypes.CTypesBackend()', but don't
# rely on it! It's probably not going to work well.)
from . import cparser
self._backend = backend
self._lock = allocate_lock()
self._parser = cparser.Parser()
self._cached_btypes = {}
self._parsed_types = types.ModuleType('parsed_types').__dict__
self._new_types = types.ModuleType('new_types').__dict__
self._function_caches = []
self._libraries = []
self._cdefsources = []
self._included_ffis = []
self._windows_unicode = None
self._init_once_cache = {}
self._cdef_version = None
self._embedding = None
self._typecache = model.get_typecache(backend)
if hasattr(backend, 'set_ffi'):
backend.set_ffi(self)
for name in list(backend.__dict__):
if name.startswith('RTLD_'):
setattr(self, name, getattr(backend, name))
#
with self._lock:
self.BVoidP = self._get_cached_btype(model.voidp_type)
self.BCharA = self._get_cached_btype(model.char_array_type)
if isinstance(backend, types.ModuleType):
# _cffi_backend: attach these constants to the class
if not hasattr(FFI, 'NULL'):
FFI.NULL = self.cast(self.BVoidP, 0)
FFI.CData, FFI.CType = backend._get_types()
else:
# ctypes backend: attach these constants to the instance
self.NULL = self.cast(self.BVoidP, 0)
self.CData, self.CType = backend._get_types()
self.buffer = backend.buffer
def cdef(self, csource, override=False, packed=False, pack=None):
"""Parse the given C source. This registers all declared functions,
types, and global variables. The functions and global variables can
then be accessed via either 'ffi.dlopen()' or 'ffi.verify()'.
The types can be used in 'ffi.new()' and other functions.
If 'packed' is specified as True, all structs declared inside this
cdef are packed, i.e. laid out without any field alignment at all.
Alternatively, 'pack' can be a small integer, and requests for
alignment greater than that are ignored (pack=1 is equivalent to
packed=True).
"""
self._cdef(csource, override=override, packed=packed, pack=pack)
def embedding_api(self, csource, packed=False, pack=None):
self._cdef(csource, packed=packed, pack=pack, dllexport=True)
if self._embedding is None:
self._embedding = ''
def _cdef(self, csource, override=False, **options):
if not isinstance(csource, str): # unicode, on Python 2
if not isinstance(csource, basestring):
raise TypeError("cdef() argument must be a string")
csource = csource.encode('ascii')
with self._lock:
self._cdef_version = object()
self._parser.parse(csource, override=override, **options)
self._cdefsources.append(csource)
if override:
for cache in self._function_caches:
cache.clear()
finishlist = self._parser._recomplete
if finishlist:
self._parser._recomplete = []
for tp in finishlist:
tp.finish_backend_type(self, finishlist)
def dlopen(self, name, flags=0):
"""Load and return a dynamic library identified by 'name'.
The standard C library can be loaded by passing None.
Note that functions and types declared by 'ffi.cdef()' are not
linked to a particular library, just like C headers; in the
library we only look for the actual (untyped) symbols.
"""
if not (isinstance(name, basestring) or
name is None or
isinstance(name, self.CData)):
raise TypeError("dlopen(name): name must be a file name, None, "
"or an already-opened 'void *' handle")
with self._lock:
lib, function_cache = _make_ffi_library(self, name, flags)
self._function_caches.append(function_cache)
self._libraries.append(lib)
return lib
def dlclose(self, lib):
"""Close a library obtained with ffi.dlopen(). After this call,
access to functions or variables from the library will fail
(possibly with a segmentation fault).
"""
type(lib).__cffi_close__(lib)
def _typeof_locked(self, cdecl):
# call me with the lock!
key = cdecl
if key in self._parsed_types:
return self._parsed_types[key]
#
if not isinstance(cdecl, str): # unicode, on Python 2
cdecl = cdecl.encode('ascii')
#
type = self._parser.parse_type(cdecl)
really_a_function_type = type.is_raw_function
if really_a_function_type:
type = type.as_function_pointer()
btype = self._get_cached_btype(type)
result = btype, really_a_function_type
self._parsed_types[key] = result
return result
def _typeof(self, cdecl, consider_function_as_funcptr=False):
# string -> ctype object
try:
result = self._parsed_types[cdecl]
except KeyError:
with self._lock:
result = self._typeof_locked(cdecl)
#
btype, really_a_function_type = result
if really_a_function_type and not consider_function_as_funcptr:
raise CDefError("the type %r is a function type, not a "
"pointer-to-function type" % (cdecl,))
return btype
def typeof(self, cdecl):
"""Parse the C type given as a string and return the
corresponding <ctype> object.
It can also be used on 'cdata' instance to get its C type.
"""
if isinstance(cdecl, basestring):
return self._typeof(cdecl)
if isinstance(cdecl, self.CData):
return self._backend.typeof(cdecl)
if isinstance(cdecl, types.BuiltinFunctionType):
res = _builtin_function_type(cdecl)
if res is not None:
return res
if (isinstance(cdecl, types.FunctionType)
and hasattr(cdecl, '_cffi_base_type')):
with self._lock:
return self._get_cached_btype(cdecl._cffi_base_type)
raise TypeError(type(cdecl))
def sizeof(self, cdecl):
"""Return the size in bytes of the argument. It can be a
string naming a C type, or a 'cdata' instance.
"""
if isinstance(cdecl, basestring):
BType = self._typeof(cdecl)
return self._backend.sizeof(BType)
else:
return self._backend.sizeof(cdecl)
def alignof(self, cdecl):
"""Return the natural alignment size in bytes of the C type
given as a string.
"""
if isinstance(cdecl, basestring):
cdecl = self._typeof(cdecl)
return self._backend.alignof(cdecl)
def offsetof(self, cdecl, *fields_or_indexes):
"""Return the offset of the named field inside the given
structure or array, which must be given as a C type name.
You can give several field names in case of nested structures.
You can also give numeric values which correspond to array
items, in case of an array type.
"""
if isinstance(cdecl, basestring):
cdecl = self._typeof(cdecl)
return self._typeoffsetof(cdecl, *fields_or_indexes)[1]
def new(self, cdecl, init=None):
"""Allocate an instance according to the specified C type and
return a pointer to it. The specified C type must be either a
pointer or an array: ``new('X *')`` allocates an X and returns
a pointer to it, whereas ``new('X[n]')`` allocates an array of
n X'es and returns an array referencing it (which works
mostly like a pointer, like in C). You can also use
``new('X[]', n)`` to allocate an array of a non-constant
length n.
The memory is initialized following the rules of declaring a
global variable in C: by default it is zero-initialized, but
an explicit initializer can be given which can be used to
fill all or part of the memory.
When the returned <cdata> object goes out of scope, the memory
is freed. In other words the returned <cdata> object has
ownership of the value of type 'cdecl' that it points to. This
means that the raw data can be used as long as this object is
kept alive, but must not be used for a longer time. Be careful
about that when copying the pointer to the memory somewhere
else, e.g. into another structure.
"""
if isinstance(cdecl, basestring):
cdecl = self._typeof(cdecl)
return self._backend.newp(cdecl, init)
def new_allocator(self, alloc=None, free=None,
should_clear_after_alloc=True):
"""Return a new allocator, i.e. a function that behaves like ffi.new()
but uses the provided low-level 'alloc' and 'free' functions.
'alloc' is called with the size as argument. If it returns NULL, a
MemoryError is raised. 'free' is called with the result of 'alloc'
as argument. Both can be either Python function or directly C
functions. If 'free' is None, then no free function is called.
If both 'alloc' and 'free' are None, the default is used.
If 'should_clear_after_alloc' is set to False, then the memory
returned by 'alloc' is assumed to be already cleared (or you are
fine with garbage); otherwise CFFI will clear it.
"""
compiled_ffi = self._backend.FFI()
allocator = compiled_ffi.new_allocator(alloc, free,
should_clear_after_alloc)
def allocate(cdecl, init=None):
if isinstance(cdecl, basestring):
cdecl = self._typeof(cdecl)
return allocator(cdecl, init)
return allocate
def cast(self, cdecl, source):
"""Similar to a C cast: returns an instance of the named C
type initialized with the given 'source'. The source is
casted between integers or pointers of any type.
"""
if isinstance(cdecl, basestring):
cdecl = self._typeof(cdecl)
return self._backend.cast(cdecl, source)
def string(self, cdata, maxlen=-1):
"""Return a Python string (or unicode string) from the 'cdata'.
If 'cdata' is a pointer or array of characters or bytes, returns
the null-terminated string. The returned string extends until
the first null character, or at most 'maxlen' characters. If
'cdata' is an array then 'maxlen' defaults to its length.
If 'cdata' is a pointer or array of wchar_t, returns a unicode
string following the same rules.
If 'cdata' is a single character or byte or a wchar_t, returns
it as a string or unicode string.
If 'cdata' is an enum, returns the value of the enumerator as a
string, or 'NUMBER' if the value is out of range.
"""
return self._backend.string(cdata, maxlen)
def unpack(self, cdata, length):
"""Unpack an array of C data of the given length,
returning a Python string/unicode/list.
If 'cdata' is a pointer to 'char', returns a byte string.
It does not stop at the first null. This is equivalent to:
ffi.buffer(cdata, length)[:]
If 'cdata' is a pointer to 'wchar_t', returns a unicode string.
'length' is measured in wchar_t's; it is not the size in bytes.
If 'cdata' is a pointer to anything else, returns a list of
'length' items. This is a faster equivalent to:
[cdata[i] for i in range(length)]
"""
return self._backend.unpack(cdata, length)
#def buffer(self, cdata, size=-1):
# """Return a read-write buffer object that references the raw C data
# pointed to by the given 'cdata'. The 'cdata' must be a pointer or
# an array. Can be passed to functions expecting a buffer, or directly
# manipulated with:
#
# buf[:] get a copy of it in a regular string, or
# buf[idx] as a single character
# buf[:] = ...
# buf[idx] = ... change the content
# """
# note that 'buffer' is a type, set on this instance by __init__
def from_buffer(self, cdecl, python_buffer=_unspecified,
require_writable=False):
"""Return a cdata of the given type pointing to the data of the
given Python object, which must support the buffer interface.
Note that this is not meant to be used on the built-in types
str or unicode (you can build 'char[]' arrays explicitly)
but only on objects containing large quantities of raw data
in some other format, like 'array.array' or numpy arrays.
The first argument is optional and default to 'char[]'.
"""
if python_buffer is _unspecified:
cdecl, python_buffer = self.BCharA, cdecl
elif isinstance(cdecl, basestring):
cdecl = self._typeof(cdecl)
return self._backend.from_buffer(cdecl, python_buffer,
require_writable)
def memmove(self, dest, src, n):
"""ffi.memmove(dest, src, n) copies n bytes of memory from src to dest.
Like the C function memmove(), the memory areas may overlap;
apart from that it behaves like the C function memcpy().
'src' can be any cdata ptr or array, or any Python buffer object.
'dest' can be any cdata ptr or array, or a writable Python buffer
object. The size to copy, 'n', is always measured in bytes.
Unlike other methods, this one supports all Python buffer including
byte strings and bytearrays---but it still does not support
non-contiguous buffers.
"""
return self._backend.memmove(dest, src, n)
def callback(self, cdecl, python_callable=None, error=None, onerror=None):
"""Return a callback object or a decorator making such a
callback object. 'cdecl' must name a C function pointer type.
The callback invokes the specified 'python_callable' (which may
be provided either directly or via a decorator). Important: the
callback object must be manually kept alive for as long as the
callback may be invoked from the C level.
"""
def callback_decorator_wrap(python_callable):
if not callable(python_callable):
raise TypeError("the 'python_callable' argument "
"is not callable")
return self._backend.callback(cdecl, python_callable,
error, onerror)
if isinstance(cdecl, basestring):
cdecl = self._typeof(cdecl, consider_function_as_funcptr=True)
if python_callable is None:
return callback_decorator_wrap # decorator mode
else:
return callback_decorator_wrap(python_callable) # direct mode
def getctype(self, cdecl, replace_with=''):
"""Return a string giving the C type 'cdecl', which may be itself
a string or a <ctype> object. If 'replace_with' is given, it gives
extra text to append (or insert for more complicated C types), like
a variable name, or '*' to get actually the C type 'pointer-to-cdecl'.
"""
if isinstance(cdecl, basestring):
cdecl = self._typeof(cdecl)
replace_with = replace_with.strip()
if (replace_with.startswith('*')
and '&[' in self._backend.getcname(cdecl, '&')):
replace_with = '(%s)' % replace_with
elif replace_with and not replace_with[0] in '[(':
replace_with = ' ' + replace_with
return self._backend.getcname(cdecl, replace_with)
def gc(self, cdata, destructor, size=0):
"""Return a new cdata object that points to the same
data. Later, when this new cdata object is garbage-collected,
'destructor(old_cdata_object)' will be called.
The optional 'size' gives an estimate of the size, used to
trigger the garbage collection more eagerly. So far only used
on PyPy. It tells the GC that the returned object keeps alive
roughly 'size' bytes of external memory.
"""
return self._backend.gcp(cdata, destructor, size)
def _get_cached_btype(self, type):
assert self._lock.acquire(False) is False
# call me with the lock!
try:
BType = self._cached_btypes[type]
except KeyError:
finishlist = []
BType = type.get_cached_btype(self, finishlist)
for type in finishlist:
type.finish_backend_type(self, finishlist)
return BType
def verify(self, source='', tmpdir=None, **kwargs):
"""Verify that the current ffi signatures compile on this
machine, and return a dynamic library object. The dynamic
library can be used to call functions and access global
variables declared in this 'ffi'. The library is compiled
by the C compiler: it gives you C-level API compatibility
(including calling macros). This is unlike 'ffi.dlopen()',
which requires binary compatibility in the signatures.
"""
from .verifier import Verifier, _caller_dir_pycache
#
# If set_unicode(True) was called, insert the UNICODE and
# _UNICODE macro declarations
if self._windows_unicode:
self._apply_windows_unicode(kwargs)
#
# Set the tmpdir here, and not in Verifier.__init__: it picks
# up the caller's directory, which we want to be the caller of
# ffi.verify(), as opposed to the caller of Veritier().
tmpdir = tmpdir or _caller_dir_pycache()
#
# Make a Verifier() and use it to load the library.
self.verifier = Verifier(self, source, tmpdir, **kwargs)
lib = self.verifier.load_library()
#
# Save the loaded library for keep-alive purposes, even
# if the caller doesn't keep it alive itself (it should).
self._libraries.append(lib)
return lib
def _get_errno(self):
return self._backend.get_errno()
def _set_errno(self, errno):
self._backend.set_errno(errno)
errno = property(_get_errno, _set_errno, None,
"the value of 'errno' from/to the C calls")
def getwinerror(self, code=-1):
return self._backend.getwinerror(code)
def _pointer_to(self, ctype):
with self._lock:
return model.pointer_cache(self, ctype)
def addressof(self, cdata, *fields_or_indexes):
"""Return the address of a <cdata 'struct-or-union'>.
If 'fields_or_indexes' are given, returns the address of that
field or array item in the structure or array, recursively in
case of nested structures.
"""
try:
ctype = self._backend.typeof(cdata)
except TypeError:
if '__addressof__' in type(cdata).__dict__:
return type(cdata).__addressof__(cdata, *fields_or_indexes)
raise
if fields_or_indexes:
ctype, offset = self._typeoffsetof(ctype, *fields_or_indexes)
else:
if ctype.kind == "pointer":
raise TypeError("addressof(pointer)")
offset = 0
ctypeptr = self._pointer_to(ctype)
return self._backend.rawaddressof(ctypeptr, cdata, offset)
def _typeoffsetof(self, ctype, field_or_index, *fields_or_indexes):
ctype, offset = self._backend.typeoffsetof(ctype, field_or_index)
for field1 in fields_or_indexes:
ctype, offset1 = self._backend.typeoffsetof(ctype, field1, 1)
offset += offset1
return ctype, offset
def include(self, ffi_to_include):
"""Includes the typedefs, structs, unions and enums defined
in another FFI instance. Usage is similar to a #include in C,
where a part of the program might include types defined in
another part for its own usage. Note that the include()
method has no effect on functions, constants and global
variables, which must anyway be accessed directly from the
lib object returned by the original FFI instance.
"""
if not isinstance(ffi_to_include, FFI):
raise TypeError("ffi.include() expects an argument that is also of"
" type cffi.FFI, not %r" % (
type(ffi_to_include).__name__,))
if ffi_to_include is self:
raise ValueError("self.include(self)")
with ffi_to_include._lock:
with self._lock:
self._parser.include(ffi_to_include._parser)
self._cdefsources.append('[')
self._cdefsources.extend(ffi_to_include._cdefsources)
self._cdefsources.append(']')
self._included_ffis.append(ffi_to_include)
def new_handle(self, x):
return self._backend.newp_handle(self.BVoidP, x)
def from_handle(self, x):
return self._backend.from_handle(x)
def release(self, x):
self._backend.release(x)
def set_unicode(self, enabled_flag):
"""Windows: if 'enabled_flag' is True, enable the UNICODE and
_UNICODE defines in C, and declare the types like TCHAR and LPTCSTR
to be (pointers to) wchar_t. If 'enabled_flag' is False,
declare these types to be (pointers to) plain 8-bit characters.
This is mostly for backward compatibility; you usually want True.
"""
if self._windows_unicode is not None:
raise ValueError("set_unicode() can only be called once")
enabled_flag = bool(enabled_flag)
if enabled_flag:
self.cdef("typedef wchar_t TBYTE;"
"typedef wchar_t TCHAR;"
"typedef const wchar_t *LPCTSTR;"
"typedef const wchar_t *PCTSTR;"
"typedef wchar_t *LPTSTR;"
"typedef wchar_t *PTSTR;"
"typedef TBYTE *PTBYTE;"
"typedef TCHAR *PTCHAR;")
else:
self.cdef("typedef char TBYTE;"
"typedef char TCHAR;"
"typedef const char *LPCTSTR;"
"typedef const char *PCTSTR;"
"typedef char *LPTSTR;"
"typedef char *PTSTR;"
"typedef TBYTE *PTBYTE;"
"typedef TCHAR *PTCHAR;")
self._windows_unicode = enabled_flag
def _apply_windows_unicode(self, kwds):
defmacros = kwds.get('define_macros', ())
if not isinstance(defmacros, (list, tuple)):
raise TypeError("'define_macros' must be a list or tuple")
defmacros = list(defmacros) + [('UNICODE', '1'),
('_UNICODE', '1')]
kwds['define_macros'] = defmacros
def _apply_embedding_fix(self, kwds):
# must include an argument like "-lpython2.7" for the compiler
def ensure(key, value):
lst = kwds.setdefault(key, [])
if value not in lst:
lst.append(value)
#
if '__pypy__' in sys.builtin_module_names:
import os
if sys.platform == "win32":
# we need 'libpypy-c.lib'. Current distributions of
# pypy (>= 4.1) contain it as 'libs/python27.lib'.
pythonlib = "python{0[0]}{0[1]}".format(sys.version_info)
if hasattr(sys, 'prefix'):
ensure('library_dirs', os.path.join(sys.prefix, 'libs'))
else:
# we need 'libpypy-c.{so,dylib}', which should be by
# default located in 'sys.prefix/bin' for installed
# systems.
if sys.version_info < (3,):
pythonlib = "pypy-c"
else:
pythonlib = "pypy3-c"
if hasattr(sys, 'prefix'):
ensure('library_dirs', os.path.join(sys.prefix, 'bin'))
# On uninstalled pypy's, the libpypy-c is typically found in
# .../pypy/goal/.
if hasattr(sys, 'prefix'):
ensure('library_dirs', os.path.join(sys.prefix, 'pypy', 'goal'))
else:
if sys.platform == "win32":
template = "python%d%d"
if hasattr(sys, 'gettotalrefcount'):
template += '_d'
else:
try:
import sysconfig
except ImportError: # 2.6
from cffi._shimmed_dist_utils import sysconfig
template = "python%d.%d"
if sysconfig.get_config_var('DEBUG_EXT'):
template += sysconfig.get_config_var('DEBUG_EXT')
pythonlib = (template %
(sys.hexversion >> 24, (sys.hexversion >> 16) & 0xff))
if hasattr(sys, 'abiflags'):
pythonlib += sys.abiflags
ensure('libraries', pythonlib)
if sys.platform == "win32":
ensure('extra_link_args', '/MANIFEST')
def set_source(self, module_name, source, source_extension='.c', **kwds):
import os
if hasattr(self, '_assigned_source'):
raise ValueError("set_source() cannot be called several times "
"per ffi object")
if not isinstance(module_name, basestring):
raise TypeError("'module_name' must be a string")
if os.sep in module_name or (os.altsep and os.altsep in module_name):
raise ValueError("'module_name' must not contain '/': use a dotted "
"name to make a 'package.module' location")
self._assigned_source = (str(module_name), source,
source_extension, kwds)
def set_source_pkgconfig(self, module_name, pkgconfig_libs, source,
source_extension='.c', **kwds):
from . import pkgconfig
if not isinstance(pkgconfig_libs, list):
raise TypeError("the pkgconfig_libs argument must be a list "
"of package names")
kwds2 = pkgconfig.flags_from_pkgconfig(pkgconfig_libs)
pkgconfig.merge_flags(kwds, kwds2)
self.set_source(module_name, source, source_extension, **kwds)
def distutils_extension(self, tmpdir='build', verbose=True):
from cffi._shimmed_dist_utils import mkpath
from .recompiler import recompile
#
if not hasattr(self, '_assigned_source'):
if hasattr(self, 'verifier'): # fallback, 'tmpdir' ignored
return self.verifier.get_extension()
raise ValueError("set_source() must be called before"
" distutils_extension()")
module_name, source, source_extension, kwds = self._assigned_source
if source is None:
raise TypeError("distutils_extension() is only for C extension "
"modules, not for dlopen()-style pure Python "
"modules")
mkpath(tmpdir)
ext, updated = recompile(self, module_name,
source, tmpdir=tmpdir, extradir=tmpdir,
source_extension=source_extension,
call_c_compiler=False, **kwds)
if verbose:
if updated:
sys.stderr.write("regenerated: %r\n" % (ext.sources[0],))
else:
sys.stderr.write("not modified: %r\n" % (ext.sources[0],))
return ext
def emit_c_code(self, filename):
from .recompiler import recompile
#
if not hasattr(self, '_assigned_source'):
raise ValueError("set_source() must be called before emit_c_code()")
module_name, source, source_extension, kwds = self._assigned_source
if source is None:
raise TypeError("emit_c_code() is only for C extension modules, "
"not for dlopen()-style pure Python modules")
recompile(self, module_name, source,
c_file=filename, call_c_compiler=False,
uses_ffiplatform=False, **kwds)
def emit_python_code(self, filename):
from .recompiler import recompile
#
if not hasattr(self, '_assigned_source'):
raise ValueError("set_source() must be called before emit_c_code()")
module_name, source, source_extension, kwds = self._assigned_source
if source is not None:
raise TypeError("emit_python_code() is only for dlopen()-style "
"pure Python modules, not for C extension modules")
recompile(self, module_name, source,
c_file=filename, call_c_compiler=False,
uses_ffiplatform=False, **kwds)
def compile(self, tmpdir='.', verbose=0, target=None, debug=None):
"""The 'target' argument gives the final file name of the
compiled DLL. Use '*' to force distutils' choice, suitable for
regular CPython C API modules. Use a file name ending in '.*'
to ask for the system's default extension for dynamic libraries
(.so/.dll/.dylib).
The default is '*' when building a non-embedded C API extension,
and (module_name + '.*') when building an embedded library.
"""
from .recompiler import recompile
#
if not hasattr(self, '_assigned_source'):
raise ValueError("set_source() must be called before compile()")
module_name, source, source_extension, kwds = self._assigned_source
return recompile(self, module_name, source, tmpdir=tmpdir,
target=target, source_extension=source_extension,
compiler_verbose=verbose, debug=debug, **kwds)
def init_once(self, func, tag):
# Read _init_once_cache[tag], which is either (False, lock) if
# we're calling the function now in some thread, or (True, result).
# Don't call setdefault() in most cases, to avoid allocating and
# immediately freeing a lock; but still use setdefaut() to avoid
# races.
try:
x = self._init_once_cache[tag]
except KeyError:
x = self._init_once_cache.setdefault(tag, (False, allocate_lock()))
# Common case: we got (True, result), so we return the result.
if x[0]:
return x[1]
# Else, it's a lock. Acquire it to serialize the following tests.
with x[1]:
# Read again from _init_once_cache the current status.
x = self._init_once_cache[tag]
if x[0]:
return x[1]
# Call the function and store the result back.
result = func()
self._init_once_cache[tag] = (True, result)
return result
def embedding_init_code(self, pysource):
if self._embedding:
raise ValueError("embedding_init_code() can only be called once")
# fix 'pysource' before it gets dumped into the C file:
# - remove empty lines at the beginning, so it starts at "line 1"
# - dedent, if all non-empty lines are indented
# - check for SyntaxErrors
import re
match = re.match(r'\s*\n', pysource)
if match:
pysource = pysource[match.end():]
lines = pysource.splitlines() or ['']
prefix = re.match(r'\s*', lines[0]).group()
for i in range(1, len(lines)):
line = lines[i]
if line.rstrip():
while not line.startswith(prefix):
prefix = prefix[:-1]
i = len(prefix)
lines = [line[i:]+'\n' for line in lines]
pysource = ''.join(lines)
#
compile(pysource, "cffi_init", "exec")
#
self._embedding = pysource
def def_extern(self, *args, **kwds):
raise ValueError("ffi.def_extern() is only available on API-mode FFI "
"objects")
def list_types(self):
"""Returns the user type names known to this FFI instance.
This returns a tuple containing three lists of names:
(typedef_names, names_of_structs, names_of_unions)
"""
typedefs = []
structs = []
unions = []
for key in self._parser._declarations:
if key.startswith('typedef '):
typedefs.append(key[8:])
elif key.startswith('struct '):
structs.append(key[7:])
elif key.startswith('union '):
unions.append(key[6:])
typedefs.sort()
structs.sort()
unions.sort()
return (typedefs, structs, unions)
def _load_backend_lib(backend, name, flags):
import os
if not isinstance(name, basestring):
if sys.platform != "win32" or name is not None:
return backend.load_library(name, flags)
name = "c" # Windows: load_library(None) fails, but this works
# on Python 2 (backward compatibility hack only)
first_error = None
if '.' in name or '/' in name or os.sep in name:
try:
return backend.load_library(name, flags)
except OSError as e:
first_error = e
import ctypes.util
path = ctypes.util.find_library(name)
if path is None:
if name == "c" and sys.platform == "win32" and sys.version_info >= (3,):
raise OSError("dlopen(None) cannot work on Windows for Python 3 "
"(see http://bugs.python.org/issue23606)")
msg = ("ctypes.util.find_library() did not manage "
"to locate a library called %r" % (name,))
if first_error is not None:
msg = "%s. Additionally, %s" % (first_error, msg)
raise OSError(msg)
return backend.load_library(path, flags)
def _make_ffi_library(ffi, libname, flags):
backend = ffi._backend
backendlib = _load_backend_lib(backend, libname, flags)
#
def accessor_function(name):
key = 'function ' + name
tp, _ = ffi._parser._declarations[key]
BType = ffi._get_cached_btype(tp)
value = backendlib.load_function(BType, name)
library.__dict__[name] = value
#
def accessor_variable(name):
key = 'variable ' + name
tp, _ = ffi._parser._declarations[key]
BType = ffi._get_cached_btype(tp)
read_variable = backendlib.read_variable
write_variable = backendlib.write_variable
setattr(FFILibrary, name, property(
lambda self: read_variable(BType, name),
lambda self, value: write_variable(BType, name, value)))
#
def addressof_var(name):
try:
return addr_variables[name]
except KeyError:
with ffi._lock:
if name not in addr_variables:
key = 'variable ' + name
tp, _ = ffi._parser._declarations[key]
BType = ffi._get_cached_btype(tp)
if BType.kind != 'array':
BType = model.pointer_cache(ffi, BType)
p = backendlib.load_function(BType, name)
addr_variables[name] = p
return addr_variables[name]
#
def accessor_constant(name):
raise NotImplementedError("non-integer constant '%s' cannot be "
"accessed from a dlopen() library" % (name,))
#
def accessor_int_constant(name):
library.__dict__[name] = ffi._parser._int_constants[name]
#
accessors = {}
accessors_version = [False]
addr_variables = {}
#
def update_accessors():
if accessors_version[0] is ffi._cdef_version:
return
#
for key, (tp, _) in ffi._parser._declarations.items():
if not isinstance(tp, model.EnumType):
tag, name = key.split(' ', 1)
if tag == 'function':
accessors[name] = accessor_function
elif tag == 'variable':
accessors[name] = accessor_variable
elif tag == 'constant':
accessors[name] = accessor_constant
else:
for i, enumname in enumerate(tp.enumerators):
def accessor_enum(name, tp=tp, i=i):
tp.check_not_partial()
library.__dict__[name] = tp.enumvalues[i]
accessors[enumname] = accessor_enum
for name in ffi._parser._int_constants:
accessors.setdefault(name, accessor_int_constant)
accessors_version[0] = ffi._cdef_version
#
def make_accessor(name):
with ffi._lock:
if name in library.__dict__ or name in FFILibrary.__dict__:
return # added by another thread while waiting for the lock
if name not in accessors:
update_accessors()
if name not in accessors:
raise AttributeError(name)
accessors[name](name)
#
class FFILibrary(object):
def __getattr__(self, name):
make_accessor(name)
return getattr(self, name)
def __setattr__(self, name, value):
try:
property = getattr(self.__class__, name)
except AttributeError:
make_accessor(name)
setattr(self, name, value)
else:
property.__set__(self, value)
def __dir__(self):
with ffi._lock:
update_accessors()
return accessors.keys()
def __addressof__(self, name):
if name in library.__dict__:
return library.__dict__[name]
if name in FFILibrary.__dict__:
return addressof_var(name)
make_accessor(name)
if name in library.__dict__:
return library.__dict__[name]
if name in FFILibrary.__dict__:
return addressof_var(name)
raise AttributeError("cffi library has no function or "
"global variable named '%s'" % (name,))
def __cffi_close__(self):
backendlib.close_lib()
self.__dict__.clear()
#
if isinstance(libname, basestring):
try:
if not isinstance(libname, str): # unicode, on Python 2
libname = libname.encode('utf-8')
FFILibrary.__name__ = 'FFILibrary_%s' % libname
except UnicodeError:
pass
library = FFILibrary()
return library, library.__dict__
def _builtin_function_type(func):
# a hack to make at least ffi.typeof(builtin_function) work,
# if the builtin function was obtained by 'vengine_cpy'.
import sys
try:
module = sys.modules[func.__module__]
ffi = module._cffi_original_ffi
types_of_builtin_funcs = module._cffi_types_of_builtin_funcs
tp = types_of_builtin_funcs[func]
except (KeyError, AttributeError, TypeError):
return None
else:
with ffi._lock:
return ffi._get_cached_btype(tp)

File diff suppressed because it is too large Load Diff

View File

@ -0,0 +1,187 @@
from .error import VerificationError
class CffiOp(object):
def __init__(self, op, arg):
self.op = op
self.arg = arg
def as_c_expr(self):
if self.op is None:
assert isinstance(self.arg, str)
return '(_cffi_opcode_t)(%s)' % (self.arg,)
classname = CLASS_NAME[self.op]
return '_CFFI_OP(_CFFI_OP_%s, %s)' % (classname, self.arg)
def as_python_bytes(self):
if self.op is None and self.arg.isdigit():
value = int(self.arg) # non-negative: '-' not in self.arg
if value >= 2**31:
raise OverflowError("cannot emit %r: limited to 2**31-1"
% (self.arg,))
return format_four_bytes(value)
if isinstance(self.arg, str):
raise VerificationError("cannot emit to Python: %r" % (self.arg,))
return format_four_bytes((self.arg << 8) | self.op)
def __str__(self):
classname = CLASS_NAME.get(self.op, self.op)
return '(%s %s)' % (classname, self.arg)
def format_four_bytes(num):
return '\\x%02X\\x%02X\\x%02X\\x%02X' % (
(num >> 24) & 0xFF,
(num >> 16) & 0xFF,
(num >> 8) & 0xFF,
(num ) & 0xFF)
OP_PRIMITIVE = 1
OP_POINTER = 3
OP_ARRAY = 5
OP_OPEN_ARRAY = 7
OP_STRUCT_UNION = 9
OP_ENUM = 11
OP_FUNCTION = 13
OP_FUNCTION_END = 15
OP_NOOP = 17
OP_BITFIELD = 19
OP_TYPENAME = 21
OP_CPYTHON_BLTN_V = 23 # varargs
OP_CPYTHON_BLTN_N = 25 # noargs
OP_CPYTHON_BLTN_O = 27 # O (i.e. a single arg)
OP_CONSTANT = 29
OP_CONSTANT_INT = 31
OP_GLOBAL_VAR = 33
OP_DLOPEN_FUNC = 35
OP_DLOPEN_CONST = 37
OP_GLOBAL_VAR_F = 39
OP_EXTERN_PYTHON = 41
PRIM_VOID = 0
PRIM_BOOL = 1
PRIM_CHAR = 2
PRIM_SCHAR = 3
PRIM_UCHAR = 4
PRIM_SHORT = 5
PRIM_USHORT = 6
PRIM_INT = 7
PRIM_UINT = 8
PRIM_LONG = 9
PRIM_ULONG = 10
PRIM_LONGLONG = 11
PRIM_ULONGLONG = 12
PRIM_FLOAT = 13
PRIM_DOUBLE = 14
PRIM_LONGDOUBLE = 15
PRIM_WCHAR = 16
PRIM_INT8 = 17
PRIM_UINT8 = 18
PRIM_INT16 = 19
PRIM_UINT16 = 20
PRIM_INT32 = 21
PRIM_UINT32 = 22
PRIM_INT64 = 23
PRIM_UINT64 = 24
PRIM_INTPTR = 25
PRIM_UINTPTR = 26
PRIM_PTRDIFF = 27
PRIM_SIZE = 28
PRIM_SSIZE = 29
PRIM_INT_LEAST8 = 30
PRIM_UINT_LEAST8 = 31
PRIM_INT_LEAST16 = 32
PRIM_UINT_LEAST16 = 33
PRIM_INT_LEAST32 = 34
PRIM_UINT_LEAST32 = 35
PRIM_INT_LEAST64 = 36
PRIM_UINT_LEAST64 = 37
PRIM_INT_FAST8 = 38
PRIM_UINT_FAST8 = 39
PRIM_INT_FAST16 = 40
PRIM_UINT_FAST16 = 41
PRIM_INT_FAST32 = 42
PRIM_UINT_FAST32 = 43
PRIM_INT_FAST64 = 44
PRIM_UINT_FAST64 = 45
PRIM_INTMAX = 46
PRIM_UINTMAX = 47
PRIM_FLOATCOMPLEX = 48
PRIM_DOUBLECOMPLEX = 49
PRIM_CHAR16 = 50
PRIM_CHAR32 = 51
_NUM_PRIM = 52
_UNKNOWN_PRIM = -1
_UNKNOWN_FLOAT_PRIM = -2
_UNKNOWN_LONG_DOUBLE = -3
_IO_FILE_STRUCT = -1
PRIMITIVE_TO_INDEX = {
'char': PRIM_CHAR,
'short': PRIM_SHORT,
'int': PRIM_INT,
'long': PRIM_LONG,
'long long': PRIM_LONGLONG,
'signed char': PRIM_SCHAR,
'unsigned char': PRIM_UCHAR,
'unsigned short': PRIM_USHORT,
'unsigned int': PRIM_UINT,
'unsigned long': PRIM_ULONG,
'unsigned long long': PRIM_ULONGLONG,
'float': PRIM_FLOAT,
'double': PRIM_DOUBLE,
'long double': PRIM_LONGDOUBLE,
'_cffi_float_complex_t': PRIM_FLOATCOMPLEX,
'_cffi_double_complex_t': PRIM_DOUBLECOMPLEX,
'_Bool': PRIM_BOOL,
'wchar_t': PRIM_WCHAR,
'char16_t': PRIM_CHAR16,
'char32_t': PRIM_CHAR32,
'int8_t': PRIM_INT8,
'uint8_t': PRIM_UINT8,
'int16_t': PRIM_INT16,
'uint16_t': PRIM_UINT16,
'int32_t': PRIM_INT32,
'uint32_t': PRIM_UINT32,
'int64_t': PRIM_INT64,
'uint64_t': PRIM_UINT64,
'intptr_t': PRIM_INTPTR,
'uintptr_t': PRIM_UINTPTR,
'ptrdiff_t': PRIM_PTRDIFF,
'size_t': PRIM_SIZE,
'ssize_t': PRIM_SSIZE,
'int_least8_t': PRIM_INT_LEAST8,
'uint_least8_t': PRIM_UINT_LEAST8,
'int_least16_t': PRIM_INT_LEAST16,
'uint_least16_t': PRIM_UINT_LEAST16,
'int_least32_t': PRIM_INT_LEAST32,
'uint_least32_t': PRIM_UINT_LEAST32,
'int_least64_t': PRIM_INT_LEAST64,
'uint_least64_t': PRIM_UINT_LEAST64,
'int_fast8_t': PRIM_INT_FAST8,
'uint_fast8_t': PRIM_UINT_FAST8,
'int_fast16_t': PRIM_INT_FAST16,
'uint_fast16_t': PRIM_UINT_FAST16,
'int_fast32_t': PRIM_INT_FAST32,
'uint_fast32_t': PRIM_UINT_FAST32,
'int_fast64_t': PRIM_INT_FAST64,
'uint_fast64_t': PRIM_UINT_FAST64,
'intmax_t': PRIM_INTMAX,
'uintmax_t': PRIM_UINTMAX,
}
F_UNION = 0x01
F_CHECK_FIELDS = 0x02
F_PACKED = 0x04
F_EXTERNAL = 0x08
F_OPAQUE = 0x10
G_FLAGS = dict([('_CFFI_' + _key, globals()[_key])
for _key in ['F_UNION', 'F_CHECK_FIELDS', 'F_PACKED',
'F_EXTERNAL', 'F_OPAQUE']])
CLASS_NAME = {}
for _name, _value in list(globals().items()):
if _name.startswith('OP_') and isinstance(_value, int):
CLASS_NAME[_value] = _name[3:]

View File

@ -0,0 +1,82 @@
import sys
from . import model
from .error import FFIError
COMMON_TYPES = {}
try:
# fetch "bool" and all simple Windows types
from _cffi_backend import _get_common_types
_get_common_types(COMMON_TYPES)
except ImportError:
pass
COMMON_TYPES['FILE'] = model.unknown_type('FILE', '_IO_FILE')
COMMON_TYPES['bool'] = '_Bool' # in case we got ImportError above
COMMON_TYPES['float _Complex'] = '_cffi_float_complex_t'
COMMON_TYPES['double _Complex'] = '_cffi_double_complex_t'
for _type in model.PrimitiveType.ALL_PRIMITIVE_TYPES:
if _type.endswith('_t'):
COMMON_TYPES[_type] = _type
del _type
_CACHE = {}
def resolve_common_type(parser, commontype):
try:
return _CACHE[commontype]
except KeyError:
cdecl = COMMON_TYPES.get(commontype, commontype)
if not isinstance(cdecl, str):
result, quals = cdecl, 0 # cdecl is already a BaseType
elif cdecl in model.PrimitiveType.ALL_PRIMITIVE_TYPES:
result, quals = model.PrimitiveType(cdecl), 0
elif cdecl == 'set-unicode-needed':
raise FFIError("The Windows type %r is only available after "
"you call ffi.set_unicode()" % (commontype,))
else:
if commontype == cdecl:
raise FFIError(
"Unsupported type: %r. Please look at "
"http://cffi.readthedocs.io/en/latest/cdef.html#ffi-cdef-limitations "
"and file an issue if you think this type should really "
"be supported." % (commontype,))
result, quals = parser.parse_type_and_quals(cdecl) # recursive
assert isinstance(result, model.BaseTypeByIdentity)
_CACHE[commontype] = result, quals
return result, quals
# ____________________________________________________________
# extra types for Windows (most of them are in commontypes.c)
def win_common_types():
return {
"UNICODE_STRING": model.StructType(
"_UNICODE_STRING",
["Length",
"MaximumLength",
"Buffer"],
[model.PrimitiveType("unsigned short"),
model.PrimitiveType("unsigned short"),
model.PointerType(model.PrimitiveType("wchar_t"))],
[-1, -1, -1]),
"PUNICODE_STRING": "UNICODE_STRING *",
"PCUNICODE_STRING": "const UNICODE_STRING *",
"TBYTE": "set-unicode-needed",
"TCHAR": "set-unicode-needed",
"LPCTSTR": "set-unicode-needed",
"PCTSTR": "set-unicode-needed",
"LPTSTR": "set-unicode-needed",
"PTSTR": "set-unicode-needed",
"PTBYTE": "set-unicode-needed",
"PTCHAR": "set-unicode-needed",
}
if sys.platform == 'win32':
COMMON_TYPES.update(win_common_types())

File diff suppressed because it is too large Load Diff

View File

@ -0,0 +1,31 @@
class FFIError(Exception):
__module__ = 'cffi'
class CDefError(Exception):
__module__ = 'cffi'
def __str__(self):
try:
current_decl = self.args[1]
filename = current_decl.coord.file
linenum = current_decl.coord.line
prefix = '%s:%d: ' % (filename, linenum)
except (AttributeError, TypeError, IndexError):
prefix = ''
return '%s%s' % (prefix, self.args[0])
class VerificationError(Exception):
""" An error raised when verification fails
"""
__module__ = 'cffi'
class VerificationMissing(Exception):
""" An error raised when incomplete structures are passed into
cdef, but no verification has been done
"""
__module__ = 'cffi'
class PkgConfigError(Exception):
""" An error raised for missing modules in pkg-config
"""
__module__ = 'cffi'

View File

@ -0,0 +1,113 @@
import sys, os
from .error import VerificationError
LIST_OF_FILE_NAMES = ['sources', 'include_dirs', 'library_dirs',
'extra_objects', 'depends']
def get_extension(srcfilename, modname, sources=(), **kwds):
from cffi._shimmed_dist_utils import Extension
allsources = [srcfilename]
for src in sources:
allsources.append(os.path.normpath(src))
return Extension(name=modname, sources=allsources, **kwds)
def compile(tmpdir, ext, compiler_verbose=0, debug=None):
"""Compile a C extension module using distutils."""
saved_environ = os.environ.copy()
try:
outputfilename = _build(tmpdir, ext, compiler_verbose, debug)
outputfilename = os.path.abspath(outputfilename)
finally:
# workaround for a distutils bugs where some env vars can
# become longer and longer every time it is used
for key, value in saved_environ.items():
if os.environ.get(key) != value:
os.environ[key] = value
return outputfilename
def _build(tmpdir, ext, compiler_verbose=0, debug=None):
# XXX compact but horrible :-(
from cffi._shimmed_dist_utils import Distribution, CompileError, LinkError, set_threshold, set_verbosity
dist = Distribution({'ext_modules': [ext]})
dist.parse_config_files()
options = dist.get_option_dict('build_ext')
if debug is None:
debug = sys.flags.debug
options['debug'] = ('ffiplatform', debug)
options['force'] = ('ffiplatform', True)
options['build_lib'] = ('ffiplatform', tmpdir)
options['build_temp'] = ('ffiplatform', tmpdir)
#
try:
old_level = set_threshold(0) or 0
try:
set_verbosity(compiler_verbose)
dist.run_command('build_ext')
cmd_obj = dist.get_command_obj('build_ext')
[soname] = cmd_obj.get_outputs()
finally:
set_threshold(old_level)
except (CompileError, LinkError) as e:
raise VerificationError('%s: %s' % (e.__class__.__name__, e))
#
return soname
try:
from os.path import samefile
except ImportError:
def samefile(f1, f2):
return os.path.abspath(f1) == os.path.abspath(f2)
def maybe_relative_path(path):
if not os.path.isabs(path):
return path # already relative
dir = path
names = []
while True:
prevdir = dir
dir, name = os.path.split(prevdir)
if dir == prevdir or not dir:
return path # failed to make it relative
names.append(name)
try:
if samefile(dir, os.curdir):
names.reverse()
return os.path.join(*names)
except OSError:
pass
# ____________________________________________________________
try:
int_or_long = (int, long)
import cStringIO
except NameError:
int_or_long = int # Python 3
import io as cStringIO
def _flatten(x, f):
if isinstance(x, str):
f.write('%ds%s' % (len(x), x))
elif isinstance(x, dict):
keys = sorted(x.keys())
f.write('%dd' % len(keys))
for key in keys:
_flatten(key, f)
_flatten(x[key], f)
elif isinstance(x, (list, tuple)):
f.write('%dl' % len(x))
for value in x:
_flatten(value, f)
elif isinstance(x, int_or_long):
f.write('%di' % (x,))
else:
raise TypeError(
"the keywords to verify() contains unsupported object %r" % (x,))
def flatten(x):
f = cStringIO.StringIO()
_flatten(x, f)
return f.getvalue()

View File

@ -0,0 +1,30 @@
import sys
if sys.version_info < (3,):
try:
from thread import allocate_lock
except ImportError:
from dummy_thread import allocate_lock
else:
try:
from _thread import allocate_lock
except ImportError:
from _dummy_thread import allocate_lock
##import sys
##l1 = allocate_lock
##class allocate_lock(object):
## def __init__(self):
## self._real = l1()
## def __enter__(self):
## for i in range(4, 0, -1):
## print sys._getframe(i).f_code
## print
## return self._real.__enter__()
## def __exit__(self, *args):
## return self._real.__exit__(*args)
## def acquire(self, f):
## assert f is False
## return self._real.acquire(f)

View File

@ -0,0 +1,618 @@
import types
import weakref
from .lock import allocate_lock
from .error import CDefError, VerificationError, VerificationMissing
# type qualifiers
Q_CONST = 0x01
Q_RESTRICT = 0x02
Q_VOLATILE = 0x04
def qualify(quals, replace_with):
if quals & Q_CONST:
replace_with = ' const ' + replace_with.lstrip()
if quals & Q_VOLATILE:
replace_with = ' volatile ' + replace_with.lstrip()
if quals & Q_RESTRICT:
# It seems that __restrict is supported by gcc and msvc.
# If you hit some different compiler, add a #define in
# _cffi_include.h for it (and in its copies, documented there)
replace_with = ' __restrict ' + replace_with.lstrip()
return replace_with
class BaseTypeByIdentity(object):
is_array_type = False
is_raw_function = False
def get_c_name(self, replace_with='', context='a C file', quals=0):
result = self.c_name_with_marker
assert result.count('&') == 1
# some logic duplication with ffi.getctype()... :-(
replace_with = replace_with.strip()
if replace_with:
if replace_with.startswith('*') and '&[' in result:
replace_with = '(%s)' % replace_with
elif not replace_with[0] in '[(':
replace_with = ' ' + replace_with
replace_with = qualify(quals, replace_with)
result = result.replace('&', replace_with)
if '$' in result:
raise VerificationError(
"cannot generate '%s' in %s: unknown type name"
% (self._get_c_name(), context))
return result
def _get_c_name(self):
return self.c_name_with_marker.replace('&', '')
def has_c_name(self):
return '$' not in self._get_c_name()
def is_integer_type(self):
return False
def get_cached_btype(self, ffi, finishlist, can_delay=False):
try:
BType = ffi._cached_btypes[self]
except KeyError:
BType = self.build_backend_type(ffi, finishlist)
BType2 = ffi._cached_btypes.setdefault(self, BType)
assert BType2 is BType
return BType
def __repr__(self):
return '<%s>' % (self._get_c_name(),)
def _get_items(self):
return [(name, getattr(self, name)) for name in self._attrs_]
class BaseType(BaseTypeByIdentity):
def __eq__(self, other):
return (self.__class__ == other.__class__ and
self._get_items() == other._get_items())
def __ne__(self, other):
return not self == other
def __hash__(self):
return hash((self.__class__, tuple(self._get_items())))
class VoidType(BaseType):
_attrs_ = ()
def __init__(self):
self.c_name_with_marker = 'void&'
def build_backend_type(self, ffi, finishlist):
return global_cache(self, ffi, 'new_void_type')
void_type = VoidType()
class BasePrimitiveType(BaseType):
def is_complex_type(self):
return False
class PrimitiveType(BasePrimitiveType):
_attrs_ = ('name',)
ALL_PRIMITIVE_TYPES = {
'char': 'c',
'short': 'i',
'int': 'i',
'long': 'i',
'long long': 'i',
'signed char': 'i',
'unsigned char': 'i',
'unsigned short': 'i',
'unsigned int': 'i',
'unsigned long': 'i',
'unsigned long long': 'i',
'float': 'f',
'double': 'f',
'long double': 'f',
'_cffi_float_complex_t': 'j',
'_cffi_double_complex_t': 'j',
'_Bool': 'i',
# the following types are not primitive in the C sense
'wchar_t': 'c',
'char16_t': 'c',
'char32_t': 'c',
'int8_t': 'i',
'uint8_t': 'i',
'int16_t': 'i',
'uint16_t': 'i',
'int32_t': 'i',
'uint32_t': 'i',
'int64_t': 'i',
'uint64_t': 'i',
'int_least8_t': 'i',
'uint_least8_t': 'i',
'int_least16_t': 'i',
'uint_least16_t': 'i',
'int_least32_t': 'i',
'uint_least32_t': 'i',
'int_least64_t': 'i',
'uint_least64_t': 'i',
'int_fast8_t': 'i',
'uint_fast8_t': 'i',
'int_fast16_t': 'i',
'uint_fast16_t': 'i',
'int_fast32_t': 'i',
'uint_fast32_t': 'i',
'int_fast64_t': 'i',
'uint_fast64_t': 'i',
'intptr_t': 'i',
'uintptr_t': 'i',
'intmax_t': 'i',
'uintmax_t': 'i',
'ptrdiff_t': 'i',
'size_t': 'i',
'ssize_t': 'i',
}
def __init__(self, name):
assert name in self.ALL_PRIMITIVE_TYPES
self.name = name
self.c_name_with_marker = name + '&'
def is_char_type(self):
return self.ALL_PRIMITIVE_TYPES[self.name] == 'c'
def is_integer_type(self):
return self.ALL_PRIMITIVE_TYPES[self.name] == 'i'
def is_float_type(self):
return self.ALL_PRIMITIVE_TYPES[self.name] == 'f'
def is_complex_type(self):
return self.ALL_PRIMITIVE_TYPES[self.name] == 'j'
def build_backend_type(self, ffi, finishlist):
return global_cache(self, ffi, 'new_primitive_type', self.name)
class UnknownIntegerType(BasePrimitiveType):
_attrs_ = ('name',)
def __init__(self, name):
self.name = name
self.c_name_with_marker = name + '&'
def is_integer_type(self):
return True
def build_backend_type(self, ffi, finishlist):
raise NotImplementedError("integer type '%s' can only be used after "
"compilation" % self.name)
class UnknownFloatType(BasePrimitiveType):
_attrs_ = ('name', )
def __init__(self, name):
self.name = name
self.c_name_with_marker = name + '&'
def build_backend_type(self, ffi, finishlist):
raise NotImplementedError("float type '%s' can only be used after "
"compilation" % self.name)
class BaseFunctionType(BaseType):
_attrs_ = ('args', 'result', 'ellipsis', 'abi')
def __init__(self, args, result, ellipsis, abi=None):
self.args = args
self.result = result
self.ellipsis = ellipsis
self.abi = abi
#
reprargs = [arg._get_c_name() for arg in self.args]
if self.ellipsis:
reprargs.append('...')
reprargs = reprargs or ['void']
replace_with = self._base_pattern % (', '.join(reprargs),)
if abi is not None:
replace_with = replace_with[:1] + abi + ' ' + replace_with[1:]
self.c_name_with_marker = (
self.result.c_name_with_marker.replace('&', replace_with))
class RawFunctionType(BaseFunctionType):
# Corresponds to a C type like 'int(int)', which is the C type of
# a function, but not a pointer-to-function. The backend has no
# notion of such a type; it's used temporarily by parsing.
_base_pattern = '(&)(%s)'
is_raw_function = True
def build_backend_type(self, ffi, finishlist):
raise CDefError("cannot render the type %r: it is a function "
"type, not a pointer-to-function type" % (self,))
def as_function_pointer(self):
return FunctionPtrType(self.args, self.result, self.ellipsis, self.abi)
class FunctionPtrType(BaseFunctionType):
_base_pattern = '(*&)(%s)'
def build_backend_type(self, ffi, finishlist):
result = self.result.get_cached_btype(ffi, finishlist)
args = []
for tp in self.args:
args.append(tp.get_cached_btype(ffi, finishlist))
abi_args = ()
if self.abi == "__stdcall":
if not self.ellipsis: # __stdcall ignored for variadic funcs
try:
abi_args = (ffi._backend.FFI_STDCALL,)
except AttributeError:
pass
return global_cache(self, ffi, 'new_function_type',
tuple(args), result, self.ellipsis, *abi_args)
def as_raw_function(self):
return RawFunctionType(self.args, self.result, self.ellipsis, self.abi)
class PointerType(BaseType):
_attrs_ = ('totype', 'quals')
def __init__(self, totype, quals=0):
self.totype = totype
self.quals = quals
extra = " *&"
if totype.is_array_type:
extra = "(%s)" % (extra.lstrip(),)
extra = qualify(quals, extra)
self.c_name_with_marker = totype.c_name_with_marker.replace('&', extra)
def build_backend_type(self, ffi, finishlist):
BItem = self.totype.get_cached_btype(ffi, finishlist, can_delay=True)
return global_cache(self, ffi, 'new_pointer_type', BItem)
voidp_type = PointerType(void_type)
def ConstPointerType(totype):
return PointerType(totype, Q_CONST)
const_voidp_type = ConstPointerType(void_type)
class NamedPointerType(PointerType):
_attrs_ = ('totype', 'name')
def __init__(self, totype, name, quals=0):
PointerType.__init__(self, totype, quals)
self.name = name
self.c_name_with_marker = name + '&'
class ArrayType(BaseType):
_attrs_ = ('item', 'length')
is_array_type = True
def __init__(self, item, length):
self.item = item
self.length = length
#
if length is None:
brackets = '&[]'
elif length == '...':
brackets = '&[/*...*/]'
else:
brackets = '&[%s]' % length
self.c_name_with_marker = (
self.item.c_name_with_marker.replace('&', brackets))
def length_is_unknown(self):
return isinstance(self.length, str)
def resolve_length(self, newlength):
return ArrayType(self.item, newlength)
def build_backend_type(self, ffi, finishlist):
if self.length_is_unknown():
raise CDefError("cannot render the type %r: unknown length" %
(self,))
self.item.get_cached_btype(ffi, finishlist) # force the item BType
BPtrItem = PointerType(self.item).get_cached_btype(ffi, finishlist)
return global_cache(self, ffi, 'new_array_type', BPtrItem, self.length)
char_array_type = ArrayType(PrimitiveType('char'), None)
class StructOrUnionOrEnum(BaseTypeByIdentity):
_attrs_ = ('name',)
forcename = None
def build_c_name_with_marker(self):
name = self.forcename or '%s %s' % (self.kind, self.name)
self.c_name_with_marker = name + '&'
def force_the_name(self, forcename):
self.forcename = forcename
self.build_c_name_with_marker()
def get_official_name(self):
assert self.c_name_with_marker.endswith('&')
return self.c_name_with_marker[:-1]
class StructOrUnion(StructOrUnionOrEnum):
fixedlayout = None
completed = 0
partial = False
packed = 0
def __init__(self, name, fldnames, fldtypes, fldbitsize, fldquals=None):
self.name = name
self.fldnames = fldnames
self.fldtypes = fldtypes
self.fldbitsize = fldbitsize
self.fldquals = fldquals
self.build_c_name_with_marker()
def anonymous_struct_fields(self):
if self.fldtypes is not None:
for name, type in zip(self.fldnames, self.fldtypes):
if name == '' and isinstance(type, StructOrUnion):
yield type
def enumfields(self, expand_anonymous_struct_union=True):
fldquals = self.fldquals
if fldquals is None:
fldquals = (0,) * len(self.fldnames)
for name, type, bitsize, quals in zip(self.fldnames, self.fldtypes,
self.fldbitsize, fldquals):
if (name == '' and isinstance(type, StructOrUnion)
and expand_anonymous_struct_union):
# nested anonymous struct/union
for result in type.enumfields():
yield result
else:
yield (name, type, bitsize, quals)
def force_flatten(self):
# force the struct or union to have a declaration that lists
# directly all fields returned by enumfields(), flattening
# nested anonymous structs/unions.
names = []
types = []
bitsizes = []
fldquals = []
for name, type, bitsize, quals in self.enumfields():
names.append(name)
types.append(type)
bitsizes.append(bitsize)
fldquals.append(quals)
self.fldnames = tuple(names)
self.fldtypes = tuple(types)
self.fldbitsize = tuple(bitsizes)
self.fldquals = tuple(fldquals)
def get_cached_btype(self, ffi, finishlist, can_delay=False):
BType = StructOrUnionOrEnum.get_cached_btype(self, ffi, finishlist,
can_delay)
if not can_delay:
self.finish_backend_type(ffi, finishlist)
return BType
def finish_backend_type(self, ffi, finishlist):
if self.completed:
if self.completed != 2:
raise NotImplementedError("recursive structure declaration "
"for '%s'" % (self.name,))
return
BType = ffi._cached_btypes[self]
#
self.completed = 1
#
if self.fldtypes is None:
pass # not completing it: it's an opaque struct
#
elif self.fixedlayout is None:
fldtypes = [tp.get_cached_btype(ffi, finishlist)
for tp in self.fldtypes]
lst = list(zip(self.fldnames, fldtypes, self.fldbitsize))
extra_flags = ()
if self.packed:
if self.packed == 1:
extra_flags = (8,) # SF_PACKED
else:
extra_flags = (0, self.packed)
ffi._backend.complete_struct_or_union(BType, lst, self,
-1, -1, *extra_flags)
#
else:
fldtypes = []
fieldofs, fieldsize, totalsize, totalalignment = self.fixedlayout
for i in range(len(self.fldnames)):
fsize = fieldsize[i]
ftype = self.fldtypes[i]
#
if isinstance(ftype, ArrayType) and ftype.length_is_unknown():
# fix the length to match the total size
BItemType = ftype.item.get_cached_btype(ffi, finishlist)
nlen, nrest = divmod(fsize, ffi.sizeof(BItemType))
if nrest != 0:
self._verification_error(
"field '%s.%s' has a bogus size?" % (
self.name, self.fldnames[i] or '{}'))
ftype = ftype.resolve_length(nlen)
self.fldtypes = (self.fldtypes[:i] + (ftype,) +
self.fldtypes[i+1:])
#
BFieldType = ftype.get_cached_btype(ffi, finishlist)
if isinstance(ftype, ArrayType) and ftype.length is None:
assert fsize == 0
else:
bitemsize = ffi.sizeof(BFieldType)
if bitemsize != fsize:
self._verification_error(
"field '%s.%s' is declared as %d bytes, but is "
"really %d bytes" % (self.name,
self.fldnames[i] or '{}',
bitemsize, fsize))
fldtypes.append(BFieldType)
#
lst = list(zip(self.fldnames, fldtypes, self.fldbitsize, fieldofs))
ffi._backend.complete_struct_or_union(BType, lst, self,
totalsize, totalalignment)
self.completed = 2
def _verification_error(self, msg):
raise VerificationError(msg)
def check_not_partial(self):
if self.partial and self.fixedlayout is None:
raise VerificationMissing(self._get_c_name())
def build_backend_type(self, ffi, finishlist):
self.check_not_partial()
finishlist.append(self)
#
return global_cache(self, ffi, 'new_%s_type' % self.kind,
self.get_official_name(), key=self)
class StructType(StructOrUnion):
kind = 'struct'
class UnionType(StructOrUnion):
kind = 'union'
class EnumType(StructOrUnionOrEnum):
kind = 'enum'
partial = False
partial_resolved = False
def __init__(self, name, enumerators, enumvalues, baseinttype=None):
self.name = name
self.enumerators = enumerators
self.enumvalues = enumvalues
self.baseinttype = baseinttype
self.build_c_name_with_marker()
def force_the_name(self, forcename):
StructOrUnionOrEnum.force_the_name(self, forcename)
if self.forcename is None:
name = self.get_official_name()
self.forcename = '$' + name.replace(' ', '_')
def check_not_partial(self):
if self.partial and not self.partial_resolved:
raise VerificationMissing(self._get_c_name())
def build_backend_type(self, ffi, finishlist):
self.check_not_partial()
base_btype = self.build_baseinttype(ffi, finishlist)
return global_cache(self, ffi, 'new_enum_type',
self.get_official_name(),
self.enumerators, self.enumvalues,
base_btype, key=self)
def build_baseinttype(self, ffi, finishlist):
if self.baseinttype is not None:
return self.baseinttype.get_cached_btype(ffi, finishlist)
#
if self.enumvalues:
smallest_value = min(self.enumvalues)
largest_value = max(self.enumvalues)
else:
import warnings
try:
# XXX! The goal is to ensure that the warnings.warn()
# will not suppress the warning. We want to get it
# several times if we reach this point several times.
__warningregistry__.clear()
except NameError:
pass
warnings.warn("%r has no values explicitly defined; "
"guessing that it is equivalent to 'unsigned int'"
% self._get_c_name())
smallest_value = largest_value = 0
if smallest_value < 0: # needs a signed type
sign = 1
candidate1 = PrimitiveType("int")
candidate2 = PrimitiveType("long")
else:
sign = 0
candidate1 = PrimitiveType("unsigned int")
candidate2 = PrimitiveType("unsigned long")
btype1 = candidate1.get_cached_btype(ffi, finishlist)
btype2 = candidate2.get_cached_btype(ffi, finishlist)
size1 = ffi.sizeof(btype1)
size2 = ffi.sizeof(btype2)
if (smallest_value >= ((-1) << (8*size1-1)) and
largest_value < (1 << (8*size1-sign))):
return btype1
if (smallest_value >= ((-1) << (8*size2-1)) and
largest_value < (1 << (8*size2-sign))):
return btype2
raise CDefError("%s values don't all fit into either 'long' "
"or 'unsigned long'" % self._get_c_name())
def unknown_type(name, structname=None):
if structname is None:
structname = '$%s' % name
tp = StructType(structname, None, None, None)
tp.force_the_name(name)
tp.origin = "unknown_type"
return tp
def unknown_ptr_type(name, structname=None):
if structname is None:
structname = '$$%s' % name
tp = StructType(structname, None, None, None)
return NamedPointerType(tp, name)
global_lock = allocate_lock()
_typecache_cffi_backend = weakref.WeakValueDictionary()
def get_typecache(backend):
# returns _typecache_cffi_backend if backend is the _cffi_backend
# module, or type(backend).__typecache if backend is an instance of
# CTypesBackend (or some FakeBackend class during tests)
if isinstance(backend, types.ModuleType):
return _typecache_cffi_backend
with global_lock:
if not hasattr(type(backend), '__typecache'):
type(backend).__typecache = weakref.WeakValueDictionary()
return type(backend).__typecache
def global_cache(srctype, ffi, funcname, *args, **kwds):
key = kwds.pop('key', (funcname, args))
assert not kwds
try:
return ffi._typecache[key]
except KeyError:
pass
try:
res = getattr(ffi._backend, funcname)(*args)
except NotImplementedError as e:
raise NotImplementedError("%s: %r: %s" % (funcname, srctype, e))
# note that setdefault() on WeakValueDictionary is not atomic
# and contains a rare bug (http://bugs.python.org/issue19542);
# we have to use a lock and do it ourselves
cache = ffi._typecache
with global_lock:
res1 = cache.get(key)
if res1 is None:
cache[key] = res
return res
else:
return res1
def pointer_cache(ffi, BType):
return global_cache('?', ffi, 'new_pointer_type', BType)
def attach_exception_info(e, name):
if e.args and type(e.args[0]) is str:
e.args = ('%s: %s' % (name, e.args[0]),) + e.args[1:]

View File

@ -0,0 +1,181 @@
/* This part is from file 'cffi/parse_c_type.h'. It is copied at the
beginning of C sources generated by CFFI's ffi.set_source(). */
typedef void *_cffi_opcode_t;
#define _CFFI_OP(opcode, arg) (_cffi_opcode_t)(opcode | (((uintptr_t)(arg)) << 8))
#define _CFFI_GETOP(cffi_opcode) ((unsigned char)(uintptr_t)cffi_opcode)
#define _CFFI_GETARG(cffi_opcode) (((intptr_t)cffi_opcode) >> 8)
#define _CFFI_OP_PRIMITIVE 1
#define _CFFI_OP_POINTER 3
#define _CFFI_OP_ARRAY 5
#define _CFFI_OP_OPEN_ARRAY 7
#define _CFFI_OP_STRUCT_UNION 9
#define _CFFI_OP_ENUM 11
#define _CFFI_OP_FUNCTION 13
#define _CFFI_OP_FUNCTION_END 15
#define _CFFI_OP_NOOP 17
#define _CFFI_OP_BITFIELD 19
#define _CFFI_OP_TYPENAME 21
#define _CFFI_OP_CPYTHON_BLTN_V 23 // varargs
#define _CFFI_OP_CPYTHON_BLTN_N 25 // noargs
#define _CFFI_OP_CPYTHON_BLTN_O 27 // O (i.e. a single arg)
#define _CFFI_OP_CONSTANT 29
#define _CFFI_OP_CONSTANT_INT 31
#define _CFFI_OP_GLOBAL_VAR 33
#define _CFFI_OP_DLOPEN_FUNC 35
#define _CFFI_OP_DLOPEN_CONST 37
#define _CFFI_OP_GLOBAL_VAR_F 39
#define _CFFI_OP_EXTERN_PYTHON 41
#define _CFFI_PRIM_VOID 0
#define _CFFI_PRIM_BOOL 1
#define _CFFI_PRIM_CHAR 2
#define _CFFI_PRIM_SCHAR 3
#define _CFFI_PRIM_UCHAR 4
#define _CFFI_PRIM_SHORT 5
#define _CFFI_PRIM_USHORT 6
#define _CFFI_PRIM_INT 7
#define _CFFI_PRIM_UINT 8
#define _CFFI_PRIM_LONG 9
#define _CFFI_PRIM_ULONG 10
#define _CFFI_PRIM_LONGLONG 11
#define _CFFI_PRIM_ULONGLONG 12
#define _CFFI_PRIM_FLOAT 13
#define _CFFI_PRIM_DOUBLE 14
#define _CFFI_PRIM_LONGDOUBLE 15
#define _CFFI_PRIM_WCHAR 16
#define _CFFI_PRIM_INT8 17
#define _CFFI_PRIM_UINT8 18
#define _CFFI_PRIM_INT16 19
#define _CFFI_PRIM_UINT16 20
#define _CFFI_PRIM_INT32 21
#define _CFFI_PRIM_UINT32 22
#define _CFFI_PRIM_INT64 23
#define _CFFI_PRIM_UINT64 24
#define _CFFI_PRIM_INTPTR 25
#define _CFFI_PRIM_UINTPTR 26
#define _CFFI_PRIM_PTRDIFF 27
#define _CFFI_PRIM_SIZE 28
#define _CFFI_PRIM_SSIZE 29
#define _CFFI_PRIM_INT_LEAST8 30
#define _CFFI_PRIM_UINT_LEAST8 31
#define _CFFI_PRIM_INT_LEAST16 32
#define _CFFI_PRIM_UINT_LEAST16 33
#define _CFFI_PRIM_INT_LEAST32 34
#define _CFFI_PRIM_UINT_LEAST32 35
#define _CFFI_PRIM_INT_LEAST64 36
#define _CFFI_PRIM_UINT_LEAST64 37
#define _CFFI_PRIM_INT_FAST8 38
#define _CFFI_PRIM_UINT_FAST8 39
#define _CFFI_PRIM_INT_FAST16 40
#define _CFFI_PRIM_UINT_FAST16 41
#define _CFFI_PRIM_INT_FAST32 42
#define _CFFI_PRIM_UINT_FAST32 43
#define _CFFI_PRIM_INT_FAST64 44
#define _CFFI_PRIM_UINT_FAST64 45
#define _CFFI_PRIM_INTMAX 46
#define _CFFI_PRIM_UINTMAX 47
#define _CFFI_PRIM_FLOATCOMPLEX 48
#define _CFFI_PRIM_DOUBLECOMPLEX 49
#define _CFFI_PRIM_CHAR16 50
#define _CFFI_PRIM_CHAR32 51
#define _CFFI__NUM_PRIM 52
#define _CFFI__UNKNOWN_PRIM (-1)
#define _CFFI__UNKNOWN_FLOAT_PRIM (-2)
#define _CFFI__UNKNOWN_LONG_DOUBLE (-3)
#define _CFFI__IO_FILE_STRUCT (-1)
struct _cffi_global_s {
const char *name;
void *address;
_cffi_opcode_t type_op;
void *size_or_direct_fn; // OP_GLOBAL_VAR: size, or 0 if unknown
// OP_CPYTHON_BLTN_*: addr of direct function
};
struct _cffi_getconst_s {
unsigned long long value;
const struct _cffi_type_context_s *ctx;
int gindex;
};
struct _cffi_struct_union_s {
const char *name;
int type_index; // -> _cffi_types, on a OP_STRUCT_UNION
int flags; // _CFFI_F_* flags below
size_t size;
int alignment;
int first_field_index; // -> _cffi_fields array
int num_fields;
};
#define _CFFI_F_UNION 0x01 // is a union, not a struct
#define _CFFI_F_CHECK_FIELDS 0x02 // complain if fields are not in the
// "standard layout" or if some are missing
#define _CFFI_F_PACKED 0x04 // for CHECK_FIELDS, assume a packed struct
#define _CFFI_F_EXTERNAL 0x08 // in some other ffi.include()
#define _CFFI_F_OPAQUE 0x10 // opaque
struct _cffi_field_s {
const char *name;
size_t field_offset;
size_t field_size;
_cffi_opcode_t field_type_op;
};
struct _cffi_enum_s {
const char *name;
int type_index; // -> _cffi_types, on a OP_ENUM
int type_prim; // _CFFI_PRIM_xxx
const char *enumerators; // comma-delimited string
};
struct _cffi_typename_s {
const char *name;
int type_index; /* if opaque, points to a possibly artificial
OP_STRUCT which is itself opaque */
};
struct _cffi_type_context_s {
_cffi_opcode_t *types;
const struct _cffi_global_s *globals;
const struct _cffi_field_s *fields;
const struct _cffi_struct_union_s *struct_unions;
const struct _cffi_enum_s *enums;
const struct _cffi_typename_s *typenames;
int num_globals;
int num_struct_unions;
int num_enums;
int num_typenames;
const char *const *includes;
int num_types;
int flags; /* future extension */
};
struct _cffi_parse_info_s {
const struct _cffi_type_context_s *ctx;
_cffi_opcode_t *output;
unsigned int output_size;
size_t error_location;
const char *error_message;
};
struct _cffi_externpy_s {
const char *name;
size_t size_of_result;
void *reserved1, *reserved2;
};
#ifdef _CFFI_INTERNAL
static int parse_c_type(struct _cffi_parse_info_s *info, const char *input);
static int search_in_globals(const struct _cffi_type_context_s *ctx,
const char *search, size_t search_len);
static int search_in_struct_unions(const struct _cffi_type_context_s *ctx,
const char *search, size_t search_len);
#endif

View File

@ -0,0 +1,121 @@
# pkg-config, https://www.freedesktop.org/wiki/Software/pkg-config/ integration for cffi
import sys, os, subprocess
from .error import PkgConfigError
def merge_flags(cfg1, cfg2):
"""Merge values from cffi config flags cfg2 to cf1
Example:
merge_flags({"libraries": ["one"]}, {"libraries": ["two"]})
{"libraries": ["one", "two"]}
"""
for key, value in cfg2.items():
if key not in cfg1:
cfg1[key] = value
else:
if not isinstance(cfg1[key], list):
raise TypeError("cfg1[%r] should be a list of strings" % (key,))
if not isinstance(value, list):
raise TypeError("cfg2[%r] should be a list of strings" % (key,))
cfg1[key].extend(value)
return cfg1
def call(libname, flag, encoding=sys.getfilesystemencoding()):
"""Calls pkg-config and returns the output if found
"""
a = ["pkg-config", "--print-errors"]
a.append(flag)
a.append(libname)
try:
pc = subprocess.Popen(a, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
except EnvironmentError as e:
raise PkgConfigError("cannot run pkg-config: %s" % (str(e).strip(),))
bout, berr = pc.communicate()
if pc.returncode != 0:
try:
berr = berr.decode(encoding)
except Exception:
pass
raise PkgConfigError(berr.strip())
if sys.version_info >= (3,) and not isinstance(bout, str): # Python 3.x
try:
bout = bout.decode(encoding)
except UnicodeDecodeError:
raise PkgConfigError("pkg-config %s %s returned bytes that cannot "
"be decoded with encoding %r:\n%r" %
(flag, libname, encoding, bout))
if os.altsep != '\\' and '\\' in bout:
raise PkgConfigError("pkg-config %s %s returned an unsupported "
"backslash-escaped output:\n%r" %
(flag, libname, bout))
return bout
def flags_from_pkgconfig(libs):
r"""Return compiler line flags for FFI.set_source based on pkg-config output
Usage
...
ffibuilder.set_source("_foo", pkgconfig = ["libfoo", "libbar >= 1.8.3"])
If pkg-config is installed on build machine, then arguments include_dirs,
library_dirs, libraries, define_macros, extra_compile_args and
extra_link_args are extended with an output of pkg-config for libfoo and
libbar.
Raises PkgConfigError in case the pkg-config call fails.
"""
def get_include_dirs(string):
return [x[2:] for x in string.split() if x.startswith("-I")]
def get_library_dirs(string):
return [x[2:] for x in string.split() if x.startswith("-L")]
def get_libraries(string):
return [x[2:] for x in string.split() if x.startswith("-l")]
# convert -Dfoo=bar to list of tuples [("foo", "bar")] expected by distutils
def get_macros(string):
def _macro(x):
x = x[2:] # drop "-D"
if '=' in x:
return tuple(x.split("=", 1)) # "-Dfoo=bar" => ("foo", "bar")
else:
return (x, None) # "-Dfoo" => ("foo", None)
return [_macro(x) for x in string.split() if x.startswith("-D")]
def get_other_cflags(string):
return [x for x in string.split() if not x.startswith("-I") and
not x.startswith("-D")]
def get_other_libs(string):
return [x for x in string.split() if not x.startswith("-L") and
not x.startswith("-l")]
# return kwargs for given libname
def kwargs(libname):
fse = sys.getfilesystemencoding()
all_cflags = call(libname, "--cflags")
all_libs = call(libname, "--libs")
return {
"include_dirs": get_include_dirs(all_cflags),
"library_dirs": get_library_dirs(all_libs),
"libraries": get_libraries(all_libs),
"define_macros": get_macros(all_cflags),
"extra_compile_args": get_other_cflags(all_cflags),
"extra_link_args": get_other_libs(all_libs),
}
# merge all arguments together
ret = {}
for libname in libs:
lib_flags = kwargs(libname)
merge_flags(ret, lib_flags)
return ret

File diff suppressed because it is too large Load Diff

View File

@ -0,0 +1,216 @@
import os
import sys
try:
basestring
except NameError:
# Python 3.x
basestring = str
def error(msg):
from cffi._shimmed_dist_utils import DistutilsSetupError
raise DistutilsSetupError(msg)
def execfile(filename, glob):
# We use execfile() (here rewritten for Python 3) instead of
# __import__() to load the build script. The problem with
# a normal import is that in some packages, the intermediate
# __init__.py files may already try to import the file that
# we are generating.
with open(filename) as f:
src = f.read()
src += '\n' # Python 2.6 compatibility
code = compile(src, filename, 'exec')
exec(code, glob, glob)
def add_cffi_module(dist, mod_spec):
from cffi.api import FFI
if not isinstance(mod_spec, basestring):
error("argument to 'cffi_modules=...' must be a str or a list of str,"
" not %r" % (type(mod_spec).__name__,))
mod_spec = str(mod_spec)
try:
build_file_name, ffi_var_name = mod_spec.split(':')
except ValueError:
error("%r must be of the form 'path/build.py:ffi_variable'" %
(mod_spec,))
if not os.path.exists(build_file_name):
ext = ''
rewritten = build_file_name.replace('.', '/') + '.py'
if os.path.exists(rewritten):
ext = ' (rewrite cffi_modules to [%r])' % (
rewritten + ':' + ffi_var_name,)
error("%r does not name an existing file%s" % (build_file_name, ext))
mod_vars = {'__name__': '__cffi__', '__file__': build_file_name}
execfile(build_file_name, mod_vars)
try:
ffi = mod_vars[ffi_var_name]
except KeyError:
error("%r: object %r not found in module" % (mod_spec,
ffi_var_name))
if not isinstance(ffi, FFI):
ffi = ffi() # maybe it's a function instead of directly an ffi
if not isinstance(ffi, FFI):
error("%r is not an FFI instance (got %r)" % (mod_spec,
type(ffi).__name__))
if not hasattr(ffi, '_assigned_source'):
error("%r: the set_source() method was not called" % (mod_spec,))
module_name, source, source_extension, kwds = ffi._assigned_source
if ffi._windows_unicode:
kwds = kwds.copy()
ffi._apply_windows_unicode(kwds)
if source is None:
_add_py_module(dist, ffi, module_name)
else:
_add_c_module(dist, ffi, module_name, source, source_extension, kwds)
def _set_py_limited_api(Extension, kwds):
"""
Add py_limited_api to kwds if setuptools >= 26 is in use.
Do not alter the setting if it already exists.
Setuptools takes care of ignoring the flag on Python 2 and PyPy.
CPython itself should ignore the flag in a debugging version
(by not listing .abi3.so in the extensions it supports), but
it doesn't so far, creating troubles. That's why we check
for "not hasattr(sys, 'gettotalrefcount')" (the 2.7 compatible equivalent
of 'd' not in sys.abiflags). (http://bugs.python.org/issue28401)
On Windows, with CPython <= 3.4, it's better not to use py_limited_api
because virtualenv *still* doesn't copy PYTHON3.DLL on these versions.
Recently (2020) we started shipping only >= 3.5 wheels, though. So
we'll give it another try and set py_limited_api on Windows >= 3.5.
"""
from cffi import recompiler
if ('py_limited_api' not in kwds and not hasattr(sys, 'gettotalrefcount')
and recompiler.USE_LIMITED_API):
import setuptools
try:
setuptools_major_version = int(setuptools.__version__.partition('.')[0])
if setuptools_major_version >= 26:
kwds['py_limited_api'] = True
except ValueError: # certain development versions of setuptools
# If we don't know the version number of setuptools, we
# try to set 'py_limited_api' anyway. At worst, we get a
# warning.
kwds['py_limited_api'] = True
return kwds
def _add_c_module(dist, ffi, module_name, source, source_extension, kwds):
# We are a setuptools extension. Need this build_ext for py_limited_api.
from setuptools.command.build_ext import build_ext
from cffi._shimmed_dist_utils import Extension, log, mkpath
from cffi import recompiler
allsources = ['$PLACEHOLDER']
allsources.extend(kwds.pop('sources', []))
kwds = _set_py_limited_api(Extension, kwds)
ext = Extension(name=module_name, sources=allsources, **kwds)
def make_mod(tmpdir, pre_run=None):
c_file = os.path.join(tmpdir, module_name + source_extension)
log.info("generating cffi module %r" % c_file)
mkpath(tmpdir)
# a setuptools-only, API-only hook: called with the "ext" and "ffi"
# arguments just before we turn the ffi into C code. To use it,
# subclass the 'distutils.command.build_ext.build_ext' class and
# add a method 'def pre_run(self, ext, ffi)'.
if pre_run is not None:
pre_run(ext, ffi)
updated = recompiler.make_c_source(ffi, module_name, source, c_file)
if not updated:
log.info("already up-to-date")
return c_file
if dist.ext_modules is None:
dist.ext_modules = []
dist.ext_modules.append(ext)
base_class = dist.cmdclass.get('build_ext', build_ext)
class build_ext_make_mod(base_class):
def run(self):
if ext.sources[0] == '$PLACEHOLDER':
pre_run = getattr(self, 'pre_run', None)
ext.sources[0] = make_mod(self.build_temp, pre_run)
base_class.run(self)
dist.cmdclass['build_ext'] = build_ext_make_mod
# NB. multiple runs here will create multiple 'build_ext_make_mod'
# classes. Even in this case the 'build_ext' command should be
# run once; but just in case, the logic above does nothing if
# called again.
def _add_py_module(dist, ffi, module_name):
from setuptools.command.build_py import build_py
from setuptools.command.build_ext import build_ext
from cffi._shimmed_dist_utils import log, mkpath
from cffi import recompiler
def generate_mod(py_file):
log.info("generating cffi module %r" % py_file)
mkpath(os.path.dirname(py_file))
updated = recompiler.make_py_source(ffi, module_name, py_file)
if not updated:
log.info("already up-to-date")
base_class = dist.cmdclass.get('build_py', build_py)
class build_py_make_mod(base_class):
def run(self):
base_class.run(self)
module_path = module_name.split('.')
module_path[-1] += '.py'
generate_mod(os.path.join(self.build_lib, *module_path))
def get_source_files(self):
# This is called from 'setup.py sdist' only. Exclude
# the generate .py module in this case.
saved_py_modules = self.py_modules
try:
if saved_py_modules:
self.py_modules = [m for m in saved_py_modules
if m != module_name]
return base_class.get_source_files(self)
finally:
self.py_modules = saved_py_modules
dist.cmdclass['build_py'] = build_py_make_mod
# distutils and setuptools have no notion I could find of a
# generated python module. If we don't add module_name to
# dist.py_modules, then things mostly work but there are some
# combination of options (--root and --record) that will miss
# the module. So we add it here, which gives a few apparently
# harmless warnings about not finding the file outside the
# build directory.
# Then we need to hack more in get_source_files(); see above.
if dist.py_modules is None:
dist.py_modules = []
dist.py_modules.append(module_name)
# the following is only for "build_ext -i"
base_class_2 = dist.cmdclass.get('build_ext', build_ext)
class build_ext_make_mod(base_class_2):
def run(self):
base_class_2.run(self)
if self.inplace:
# from get_ext_fullpath() in distutils/command/build_ext.py
module_path = module_name.split('.')
package = '.'.join(module_path[:-1])
build_py = self.get_finalized_command('build_py')
package_dir = build_py.get_package_dir(package)
file_name = module_path[-1] + '.py'
generate_mod(os.path.join(package_dir, file_name))
dist.cmdclass['build_ext'] = build_ext_make_mod
def cffi_modules(dist, attr, value):
assert attr == 'cffi_modules'
if isinstance(value, basestring):
value = [value]
for cffi_module in value:
add_cffi_module(dist, cffi_module)

File diff suppressed because it is too large Load Diff

View File

@ -0,0 +1,679 @@
#
# DEPRECATED: implementation for ffi.verify()
#
import sys, os
import types
from . import model
from .error import VerificationError
class VGenericEngine(object):
_class_key = 'g'
_gen_python_module = False
def __init__(self, verifier):
self.verifier = verifier
self.ffi = verifier.ffi
self.export_symbols = []
self._struct_pending_verification = {}
def patch_extension_kwds(self, kwds):
# add 'export_symbols' to the dictionary. Note that we add the
# list before filling it. When we fill it, it will thus also show
# up in kwds['export_symbols'].
kwds.setdefault('export_symbols', self.export_symbols)
def find_module(self, module_name, path, so_suffixes):
for so_suffix in so_suffixes:
basename = module_name + so_suffix
if path is None:
path = sys.path
for dirname in path:
filename = os.path.join(dirname, basename)
if os.path.isfile(filename):
return filename
def collect_types(self):
pass # not needed in the generic engine
def _prnt(self, what=''):
self._f.write(what + '\n')
def write_source_to_f(self):
prnt = self._prnt
# first paste some standard set of lines that are mostly '#include'
prnt(cffimod_header)
# then paste the C source given by the user, verbatim.
prnt(self.verifier.preamble)
#
# call generate_gen_xxx_decl(), for every xxx found from
# ffi._parser._declarations. This generates all the functions.
self._generate('decl')
#
# on Windows, distutils insists on putting init_cffi_xyz in
# 'export_symbols', so instead of fighting it, just give up and
# give it one
if sys.platform == 'win32':
if sys.version_info >= (3,):
prefix = 'PyInit_'
else:
prefix = 'init'
modname = self.verifier.get_module_name()
prnt("void %s%s(void) { }\n" % (prefix, modname))
def load_library(self, flags=0):
# import it with the CFFI backend
backend = self.ffi._backend
# needs to make a path that contains '/', on Posix
filename = os.path.join(os.curdir, self.verifier.modulefilename)
module = backend.load_library(filename, flags)
#
# call loading_gen_struct() to get the struct layout inferred by
# the C compiler
self._load(module, 'loading')
# build the FFILibrary class and instance, this is a module subclass
# because modules are expected to have usually-constant-attributes and
# in PyPy this means the JIT is able to treat attributes as constant,
# which we want.
class FFILibrary(types.ModuleType):
_cffi_generic_module = module
_cffi_ffi = self.ffi
_cffi_dir = []
def __dir__(self):
return FFILibrary._cffi_dir
library = FFILibrary("")
#
# finally, call the loaded_gen_xxx() functions. This will set
# up the 'library' object.
self._load(module, 'loaded', library=library)
return library
def _get_declarations(self):
lst = [(key, tp) for (key, (tp, qual)) in
self.ffi._parser._declarations.items()]
lst.sort()
return lst
def _generate(self, step_name):
for name, tp in self._get_declarations():
kind, realname = name.split(' ', 1)
try:
method = getattr(self, '_generate_gen_%s_%s' % (kind,
step_name))
except AttributeError:
raise VerificationError(
"not implemented in verify(): %r" % name)
try:
method(tp, realname)
except Exception as e:
model.attach_exception_info(e, name)
raise
def _load(self, module, step_name, **kwds):
for name, tp in self._get_declarations():
kind, realname = name.split(' ', 1)
method = getattr(self, '_%s_gen_%s' % (step_name, kind))
try:
method(tp, realname, module, **kwds)
except Exception as e:
model.attach_exception_info(e, name)
raise
def _generate_nothing(self, tp, name):
pass
def _loaded_noop(self, tp, name, module, **kwds):
pass
# ----------
# typedefs: generates no code so far
_generate_gen_typedef_decl = _generate_nothing
_loading_gen_typedef = _loaded_noop
_loaded_gen_typedef = _loaded_noop
# ----------
# function declarations
def _generate_gen_function_decl(self, tp, name):
assert isinstance(tp, model.FunctionPtrType)
if tp.ellipsis:
# cannot support vararg functions better than this: check for its
# exact type (including the fixed arguments), and build it as a
# constant function pointer (no _cffi_f_%s wrapper)
self._generate_gen_const(False, name, tp)
return
prnt = self._prnt
numargs = len(tp.args)
argnames = []
for i, type in enumerate(tp.args):
indirection = ''
if isinstance(type, model.StructOrUnion):
indirection = '*'
argnames.append('%sx%d' % (indirection, i))
context = 'argument of %s' % name
arglist = [type.get_c_name(' %s' % arg, context)
for type, arg in zip(tp.args, argnames)]
tpresult = tp.result
if isinstance(tpresult, model.StructOrUnion):
arglist.insert(0, tpresult.get_c_name(' *r', context))
tpresult = model.void_type
arglist = ', '.join(arglist) or 'void'
wrappername = '_cffi_f_%s' % name
self.export_symbols.append(wrappername)
if tp.abi:
abi = tp.abi + ' '
else:
abi = ''
funcdecl = ' %s%s(%s)' % (abi, wrappername, arglist)
context = 'result of %s' % name
prnt(tpresult.get_c_name(funcdecl, context))
prnt('{')
#
if isinstance(tp.result, model.StructOrUnion):
result_code = '*r = '
elif not isinstance(tp.result, model.VoidType):
result_code = 'return '
else:
result_code = ''
prnt(' %s%s(%s);' % (result_code, name, ', '.join(argnames)))
prnt('}')
prnt()
_loading_gen_function = _loaded_noop
def _loaded_gen_function(self, tp, name, module, library):
assert isinstance(tp, model.FunctionPtrType)
if tp.ellipsis:
newfunction = self._load_constant(False, tp, name, module)
else:
indirections = []
base_tp = tp
if (any(isinstance(typ, model.StructOrUnion) for typ in tp.args)
or isinstance(tp.result, model.StructOrUnion)):
indirect_args = []
for i, typ in enumerate(tp.args):
if isinstance(typ, model.StructOrUnion):
typ = model.PointerType(typ)
indirections.append((i, typ))
indirect_args.append(typ)
indirect_result = tp.result
if isinstance(indirect_result, model.StructOrUnion):
if indirect_result.fldtypes is None:
raise TypeError("'%s' is used as result type, "
"but is opaque" % (
indirect_result._get_c_name(),))
indirect_result = model.PointerType(indirect_result)
indirect_args.insert(0, indirect_result)
indirections.insert(0, ("result", indirect_result))
indirect_result = model.void_type
tp = model.FunctionPtrType(tuple(indirect_args),
indirect_result, tp.ellipsis)
BFunc = self.ffi._get_cached_btype(tp)
wrappername = '_cffi_f_%s' % name
newfunction = module.load_function(BFunc, wrappername)
for i, typ in indirections:
newfunction = self._make_struct_wrapper(newfunction, i, typ,
base_tp)
setattr(library, name, newfunction)
type(library)._cffi_dir.append(name)
def _make_struct_wrapper(self, oldfunc, i, tp, base_tp):
backend = self.ffi._backend
BType = self.ffi._get_cached_btype(tp)
if i == "result":
ffi = self.ffi
def newfunc(*args):
res = ffi.new(BType)
oldfunc(res, *args)
return res[0]
else:
def newfunc(*args):
args = args[:i] + (backend.newp(BType, args[i]),) + args[i+1:]
return oldfunc(*args)
newfunc._cffi_base_type = base_tp
return newfunc
# ----------
# named structs
def _generate_gen_struct_decl(self, tp, name):
assert name == tp.name
self._generate_struct_or_union_decl(tp, 'struct', name)
def _loading_gen_struct(self, tp, name, module):
self._loading_struct_or_union(tp, 'struct', name, module)
def _loaded_gen_struct(self, tp, name, module, **kwds):
self._loaded_struct_or_union(tp)
def _generate_gen_union_decl(self, tp, name):
assert name == tp.name
self._generate_struct_or_union_decl(tp, 'union', name)
def _loading_gen_union(self, tp, name, module):
self._loading_struct_or_union(tp, 'union', name, module)
def _loaded_gen_union(self, tp, name, module, **kwds):
self._loaded_struct_or_union(tp)
def _generate_struct_or_union_decl(self, tp, prefix, name):
if tp.fldnames is None:
return # nothing to do with opaque structs
checkfuncname = '_cffi_check_%s_%s' % (prefix, name)
layoutfuncname = '_cffi_layout_%s_%s' % (prefix, name)
cname = ('%s %s' % (prefix, name)).strip()
#
prnt = self._prnt
prnt('static void %s(%s *p)' % (checkfuncname, cname))
prnt('{')
prnt(' /* only to generate compile-time warnings or errors */')
prnt(' (void)p;')
for fname, ftype, fbitsize, fqual in tp.enumfields():
if (isinstance(ftype, model.PrimitiveType)
and ftype.is_integer_type()) or fbitsize >= 0:
# accept all integers, but complain on float or double
prnt(' (void)((p->%s) << 1);' % fname)
else:
# only accept exactly the type declared.
try:
prnt(' { %s = &p->%s; (void)tmp; }' % (
ftype.get_c_name('*tmp', 'field %r'%fname, quals=fqual),
fname))
except VerificationError as e:
prnt(' /* %s */' % str(e)) # cannot verify it, ignore
prnt('}')
self.export_symbols.append(layoutfuncname)
prnt('intptr_t %s(intptr_t i)' % (layoutfuncname,))
prnt('{')
prnt(' struct _cffi_aligncheck { char x; %s y; };' % cname)
prnt(' static intptr_t nums[] = {')
prnt(' sizeof(%s),' % cname)
prnt(' offsetof(struct _cffi_aligncheck, y),')
for fname, ftype, fbitsize, fqual in tp.enumfields():
if fbitsize >= 0:
continue # xxx ignore fbitsize for now
prnt(' offsetof(%s, %s),' % (cname, fname))
if isinstance(ftype, model.ArrayType) and ftype.length is None:
prnt(' 0, /* %s */' % ftype._get_c_name())
else:
prnt(' sizeof(((%s *)0)->%s),' % (cname, fname))
prnt(' -1')
prnt(' };')
prnt(' return nums[i];')
prnt(' /* the next line is not executed, but compiled */')
prnt(' %s(0);' % (checkfuncname,))
prnt('}')
prnt()
def _loading_struct_or_union(self, tp, prefix, name, module):
if tp.fldnames is None:
return # nothing to do with opaque structs
layoutfuncname = '_cffi_layout_%s_%s' % (prefix, name)
#
BFunc = self.ffi._typeof_locked("intptr_t(*)(intptr_t)")[0]
function = module.load_function(BFunc, layoutfuncname)
layout = []
num = 0
while True:
x = function(num)
if x < 0: break
layout.append(x)
num += 1
if isinstance(tp, model.StructOrUnion) and tp.partial:
# use the function()'s sizes and offsets to guide the
# layout of the struct
totalsize = layout[0]
totalalignment = layout[1]
fieldofs = layout[2::2]
fieldsize = layout[3::2]
tp.force_flatten()
assert len(fieldofs) == len(fieldsize) == len(tp.fldnames)
tp.fixedlayout = fieldofs, fieldsize, totalsize, totalalignment
else:
cname = ('%s %s' % (prefix, name)).strip()
self._struct_pending_verification[tp] = layout, cname
def _loaded_struct_or_union(self, tp):
if tp.fldnames is None:
return # nothing to do with opaque structs
self.ffi._get_cached_btype(tp) # force 'fixedlayout' to be considered
if tp in self._struct_pending_verification:
# check that the layout sizes and offsets match the real ones
def check(realvalue, expectedvalue, msg):
if realvalue != expectedvalue:
raise VerificationError(
"%s (we have %d, but C compiler says %d)"
% (msg, expectedvalue, realvalue))
ffi = self.ffi
BStruct = ffi._get_cached_btype(tp)
layout, cname = self._struct_pending_verification.pop(tp)
check(layout[0], ffi.sizeof(BStruct), "wrong total size")
check(layout[1], ffi.alignof(BStruct), "wrong total alignment")
i = 2
for fname, ftype, fbitsize, fqual in tp.enumfields():
if fbitsize >= 0:
continue # xxx ignore fbitsize for now
check(layout[i], ffi.offsetof(BStruct, fname),
"wrong offset for field %r" % (fname,))
if layout[i+1] != 0:
BField = ffi._get_cached_btype(ftype)
check(layout[i+1], ffi.sizeof(BField),
"wrong size for field %r" % (fname,))
i += 2
assert i == len(layout)
# ----------
# 'anonymous' declarations. These are produced for anonymous structs
# or unions; the 'name' is obtained by a typedef.
def _generate_gen_anonymous_decl(self, tp, name):
if isinstance(tp, model.EnumType):
self._generate_gen_enum_decl(tp, name, '')
else:
self._generate_struct_or_union_decl(tp, '', name)
def _loading_gen_anonymous(self, tp, name, module):
if isinstance(tp, model.EnumType):
self._loading_gen_enum(tp, name, module, '')
else:
self._loading_struct_or_union(tp, '', name, module)
def _loaded_gen_anonymous(self, tp, name, module, **kwds):
if isinstance(tp, model.EnumType):
self._loaded_gen_enum(tp, name, module, **kwds)
else:
self._loaded_struct_or_union(tp)
# ----------
# constants, likely declared with '#define'
def _generate_gen_const(self, is_int, name, tp=None, category='const',
check_value=None):
prnt = self._prnt
funcname = '_cffi_%s_%s' % (category, name)
self.export_symbols.append(funcname)
if check_value is not None:
assert is_int
assert category == 'const'
prnt('int %s(char *out_error)' % funcname)
prnt('{')
self._check_int_constant_value(name, check_value)
prnt(' return 0;')
prnt('}')
elif is_int:
assert category == 'const'
prnt('int %s(long long *out_value)' % funcname)
prnt('{')
prnt(' *out_value = (long long)(%s);' % (name,))
prnt(' return (%s) <= 0;' % (name,))
prnt('}')
else:
assert tp is not None
assert check_value is None
if category == 'var':
ampersand = '&'
else:
ampersand = ''
extra = ''
if category == 'const' and isinstance(tp, model.StructOrUnion):
extra = 'const *'
ampersand = '&'
prnt(tp.get_c_name(' %s%s(void)' % (extra, funcname), name))
prnt('{')
prnt(' return (%s%s);' % (ampersand, name))
prnt('}')
prnt()
def _generate_gen_constant_decl(self, tp, name):
is_int = isinstance(tp, model.PrimitiveType) and tp.is_integer_type()
self._generate_gen_const(is_int, name, tp)
_loading_gen_constant = _loaded_noop
def _load_constant(self, is_int, tp, name, module, check_value=None):
funcname = '_cffi_const_%s' % name
if check_value is not None:
assert is_int
self._load_known_int_constant(module, funcname)
value = check_value
elif is_int:
BType = self.ffi._typeof_locked("long long*")[0]
BFunc = self.ffi._typeof_locked("int(*)(long long*)")[0]
function = module.load_function(BFunc, funcname)
p = self.ffi.new(BType)
negative = function(p)
value = int(p[0])
if value < 0 and not negative:
BLongLong = self.ffi._typeof_locked("long long")[0]
value += (1 << (8*self.ffi.sizeof(BLongLong)))
else:
assert check_value is None
fntypeextra = '(*)(void)'
if isinstance(tp, model.StructOrUnion):
fntypeextra = '*' + fntypeextra
BFunc = self.ffi._typeof_locked(tp.get_c_name(fntypeextra, name))[0]
function = module.load_function(BFunc, funcname)
value = function()
if isinstance(tp, model.StructOrUnion):
value = value[0]
return value
def _loaded_gen_constant(self, tp, name, module, library):
is_int = isinstance(tp, model.PrimitiveType) and tp.is_integer_type()
value = self._load_constant(is_int, tp, name, module)
setattr(library, name, value)
type(library)._cffi_dir.append(name)
# ----------
# enums
def _check_int_constant_value(self, name, value):
prnt = self._prnt
if value <= 0:
prnt(' if ((%s) > 0 || (long)(%s) != %dL) {' % (
name, name, value))
else:
prnt(' if ((%s) <= 0 || (unsigned long)(%s) != %dUL) {' % (
name, name, value))
prnt(' char buf[64];')
prnt(' if ((%s) <= 0)' % name)
prnt(' sprintf(buf, "%%ld", (long)(%s));' % name)
prnt(' else')
prnt(' sprintf(buf, "%%lu", (unsigned long)(%s));' %
name)
prnt(' sprintf(out_error, "%s has the real value %s, not %s",')
prnt(' "%s", buf, "%d");' % (name[:100], value))
prnt(' return -1;')
prnt(' }')
def _load_known_int_constant(self, module, funcname):
BType = self.ffi._typeof_locked("char[]")[0]
BFunc = self.ffi._typeof_locked("int(*)(char*)")[0]
function = module.load_function(BFunc, funcname)
p = self.ffi.new(BType, 256)
if function(p) < 0:
error = self.ffi.string(p)
if sys.version_info >= (3,):
error = str(error, 'utf-8')
raise VerificationError(error)
def _enum_funcname(self, prefix, name):
# "$enum_$1" => "___D_enum____D_1"
name = name.replace('$', '___D_')
return '_cffi_e_%s_%s' % (prefix, name)
def _generate_gen_enum_decl(self, tp, name, prefix='enum'):
if tp.partial:
for enumerator in tp.enumerators:
self._generate_gen_const(True, enumerator)
return
#
funcname = self._enum_funcname(prefix, name)
self.export_symbols.append(funcname)
prnt = self._prnt
prnt('int %s(char *out_error)' % funcname)
prnt('{')
for enumerator, enumvalue in zip(tp.enumerators, tp.enumvalues):
self._check_int_constant_value(enumerator, enumvalue)
prnt(' return 0;')
prnt('}')
prnt()
def _loading_gen_enum(self, tp, name, module, prefix='enum'):
if tp.partial:
enumvalues = [self._load_constant(True, tp, enumerator, module)
for enumerator in tp.enumerators]
tp.enumvalues = tuple(enumvalues)
tp.partial_resolved = True
else:
funcname = self._enum_funcname(prefix, name)
self._load_known_int_constant(module, funcname)
def _loaded_gen_enum(self, tp, name, module, library):
for enumerator, enumvalue in zip(tp.enumerators, tp.enumvalues):
setattr(library, enumerator, enumvalue)
type(library)._cffi_dir.append(enumerator)
# ----------
# macros: for now only for integers
def _generate_gen_macro_decl(self, tp, name):
if tp == '...':
check_value = None
else:
check_value = tp # an integer
self._generate_gen_const(True, name, check_value=check_value)
_loading_gen_macro = _loaded_noop
def _loaded_gen_macro(self, tp, name, module, library):
if tp == '...':
check_value = None
else:
check_value = tp # an integer
value = self._load_constant(True, tp, name, module,
check_value=check_value)
setattr(library, name, value)
type(library)._cffi_dir.append(name)
# ----------
# global variables
def _generate_gen_variable_decl(self, tp, name):
if isinstance(tp, model.ArrayType):
if tp.length_is_unknown():
prnt = self._prnt
funcname = '_cffi_sizeof_%s' % (name,)
self.export_symbols.append(funcname)
prnt("size_t %s(void)" % funcname)
prnt("{")
prnt(" return sizeof(%s);" % (name,))
prnt("}")
tp_ptr = model.PointerType(tp.item)
self._generate_gen_const(False, name, tp_ptr)
else:
tp_ptr = model.PointerType(tp)
self._generate_gen_const(False, name, tp_ptr, category='var')
_loading_gen_variable = _loaded_noop
def _loaded_gen_variable(self, tp, name, module, library):
if isinstance(tp, model.ArrayType): # int a[5] is "constant" in the
# sense that "a=..." is forbidden
if tp.length_is_unknown():
funcname = '_cffi_sizeof_%s' % (name,)
BFunc = self.ffi._typeof_locked('size_t(*)(void)')[0]
function = module.load_function(BFunc, funcname)
size = function()
BItemType = self.ffi._get_cached_btype(tp.item)
length, rest = divmod(size, self.ffi.sizeof(BItemType))
if rest != 0:
raise VerificationError(
"bad size: %r does not seem to be an array of %s" %
(name, tp.item))
tp = tp.resolve_length(length)
tp_ptr = model.PointerType(tp.item)
value = self._load_constant(False, tp_ptr, name, module)
# 'value' is a <cdata 'type *'> which we have to replace with
# a <cdata 'type[N]'> if the N is actually known
if tp.length is not None:
BArray = self.ffi._get_cached_btype(tp)
value = self.ffi.cast(BArray, value)
setattr(library, name, value)
type(library)._cffi_dir.append(name)
return
# remove ptr=<cdata 'int *'> from the library instance, and replace
# it by a property on the class, which reads/writes into ptr[0].
funcname = '_cffi_var_%s' % name
BFunc = self.ffi._typeof_locked(tp.get_c_name('*(*)(void)', name))[0]
function = module.load_function(BFunc, funcname)
ptr = function()
def getter(library):
return ptr[0]
def setter(library, value):
ptr[0] = value
setattr(type(library), name, property(getter, setter))
type(library)._cffi_dir.append(name)
cffimod_header = r'''
#include <stdio.h>
#include <stddef.h>
#include <stdarg.h>
#include <errno.h>
#include <sys/types.h> /* XXX for ssize_t on some platforms */
/* this block of #ifs should be kept exactly identical between
c/_cffi_backend.c, cffi/vengine_cpy.py, cffi/vengine_gen.py
and cffi/_cffi_include.h */
#if defined(_MSC_VER)
# include <malloc.h> /* for alloca() */
# if _MSC_VER < 1600 /* MSVC < 2010 */
typedef __int8 int8_t;
typedef __int16 int16_t;
typedef __int32 int32_t;
typedef __int64 int64_t;
typedef unsigned __int8 uint8_t;
typedef unsigned __int16 uint16_t;
typedef unsigned __int32 uint32_t;
typedef unsigned __int64 uint64_t;
typedef __int8 int_least8_t;
typedef __int16 int_least16_t;
typedef __int32 int_least32_t;
typedef __int64 int_least64_t;
typedef unsigned __int8 uint_least8_t;
typedef unsigned __int16 uint_least16_t;
typedef unsigned __int32 uint_least32_t;
typedef unsigned __int64 uint_least64_t;
typedef __int8 int_fast8_t;
typedef __int16 int_fast16_t;
typedef __int32 int_fast32_t;
typedef __int64 int_fast64_t;
typedef unsigned __int8 uint_fast8_t;
typedef unsigned __int16 uint_fast16_t;
typedef unsigned __int32 uint_fast32_t;
typedef unsigned __int64 uint_fast64_t;
typedef __int64 intmax_t;
typedef unsigned __int64 uintmax_t;
# else
# include <stdint.h>
# endif
# if _MSC_VER < 1800 /* MSVC < 2013 */
# ifndef __cplusplus
typedef unsigned char _Bool;
# endif
# endif
# define _cffi_float_complex_t _Fcomplex /* include <complex.h> for it */
# define _cffi_double_complex_t _Dcomplex /* include <complex.h> for it */
#else
# include <stdint.h>
# if (defined (__SVR4) && defined (__sun)) || defined(_AIX) || defined(__hpux)
# include <alloca.h>
# endif
# define _cffi_float_complex_t float _Complex
# define _cffi_double_complex_t double _Complex
#endif
'''

Some files were not shown because too many files have changed in this diff Show More