Bug#1056483: python-memory-profiler's autopkg tests fail with Python 3.12

2024-01-20 Thread Stefano Rivera
Note: This package is no longer maintained upstream:
https://github.com/pythonprofilers/memory_profiler/issues/383

Stefano

-- 
Stefano Rivera
  http://tumbleweed.org.za/
  +1 415 683 3272

-- 
debian-science-maintainers mailing list
debian-science-maintainers@alioth-lists.debian.net
https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/debian-science-maintainers


Bug#1025183: silx: (autopkgtest) needs update for python3.11: Segmentation fault

2023-01-27 Thread Stefano Rivera
Hi Simon (2023.01.26_19:30:35_-0400)
> While I can't confirm the original error in the bug report, I can confirm
> that the autopkgtests fail with Python 3.11, and require some fixes.
> 
> I have uploaded a fix for this to Ubuntu, the delta is attached. I also
> uploaded this to DELAYED/2 (the fix is non-intrusive and only affects the
> test suite, so it should be safe). Please let me know if you would like this
> delayed further, cancelled, or expedited.

Thanks, I'll team-upload that now.

I had looked at the next upstream version, and got a whole slew of other
test failures. This looks like a much simpler solution...

SR

-- 
Stefano Rivera
  http://tumbleweed.org.za/
  +1 415 683 3272

-- 
debian-science-maintainers mailing list
debian-science-maintainers@alioth-lists.debian.net
https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/debian-science-maintainers


Bug#1025181: qiskit-ibmq-provider: (autopkgtest) needs update for python3.11: module 'asyncio' has no attribute 'coroutine'

2022-12-25 Thread Stefano Rivera
I see this is fixed upstream in around 0.11, by
https://github.com/Qiskit/qiskit-ibmq-provider/pull/779

Given #1008627 and the deprecation upstream, should we just drop the
package, entirely?

SR

-- 
Stefano Rivera
  http://tumbleweed.org.za/
  +1 415 683 3272

-- 
debian-science-maintainers mailing list
debian-science-maintainers@alioth-lists.debian.net
https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/debian-science-maintainers


Bug#1024047: python-line-profiler FTBFS with Python 3.11 as supported version

2022-11-23 Thread Stefano Rivera
Control: tag -1 + fixed-upstream

Upstream claims that version 4.0.0 supports Python 3.11.

I tried backporting a minimal patch to bring 3.11 support back, but
tests fail. So... time to stop kicking this can down the road and update
to the latest upstream version.

SR

-- 
debian-science-maintainers mailing list
debian-science-maintainers@alioth-lists.debian.net
https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/debian-science-maintainers


Bug#1024265: opengv: diff for NMU version 1.0+1git91f4b1-6.1

2022-11-18 Thread Stefano Rivera
Control: tags 1024265 + pending

Dear maintainer,

I've prepared an NMU for opengv (versioned as 1.0+1git91f4b1-6.1) and
uploaded it to DELAYED/5. Please feel free to tell me if I
should delay it longer.

The patch (thanks Tyler!) may not be ideal, it's building the whole of the big
package, twice. There may be a more efficient way to only rebuild the
python bits. But I can't see that.

Also filed as: https://salsa.debian.org/science-team/opengv/-/merge_requests/2

Regards.

SR
diff -Nru opengv-1.0+1git91f4b1/debian/changelog opengv-1.0+1git91f4b1/debian/changelog
--- opengv-1.0+1git91f4b1/debian/changelog	2022-11-02 19:34:56.0 +0200
+++ opengv-1.0+1git91f4b1/debian/changelog	2022-11-18 11:43:32.0 +0200
@@ -1,3 +1,11 @@
+opengv (1.0+1git91f4b1-6.1) unstable; urgency=medium
+
+  * Non-maintainer upload.
+  * Use pybuild to build for multiple Python versions. Thanks Tyler Weaver.
+(Closes: #1024265)
+
+ -- Stefano Rivera   Fri, 18 Nov 2022 11:43:32 +0200
+
 opengv (1.0+1git91f4b1-6) unstable; urgency=medium
 
   * Arch:all builds work (Closes: #1020995)
diff -Nru opengv-1.0+1git91f4b1/debian/rules opengv-1.0+1git91f4b1/debian/rules
--- opengv-1.0+1git91f4b1/debian/rules	2022-11-02 19:34:56.0 +0200
+++ opengv-1.0+1git91f4b1/debian/rules	2022-11-18 11:43:32.0 +0200
@@ -4,14 +4,11 @@
 # Uncomment this to turn on verbose mode.
 #export DH_VERBOSE=1
 
-%:
-	dh $@ --with python3
+export PYBUILD_CONFIGURE_ARGS=-DBUILD_TESTS=ON -DBUILD_PYTHON=ON -DBUILD_SHARED_LIBS=ON
+export PYBUILD_SYSTEM=cmake
 
-override_dh_auto_configure:
-	dh_auto_configure -- \
-		-DBUILD_TESTS=ON \
-		-DBUILD_PYTHON=ON \
-		-DBUILD_SHARED_LIBS=ON
+%:
+	dh $@ --with python3 --buildsystem=pybuild
 
 override_dh_auto_build-arch:
 	dh_auto_build -a
@@ -32,5 +29,3 @@
 override_dh_auto_clean:
 	dh_auto_clean
 	rm -rf doc/html doc/latex
-
-
-- 
debian-science-maintainers mailing list
debian-science-maintainers@alioth-lists.debian.net
https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/debian-science-maintainers


Bug#1024265: opengv: Build for all supported python3 versions or Build-Depend on python3-dev

2022-11-16 Thread Stefano Rivera
Source: opengv
Version: 1.0+1git91f4b1-6
Severity: normal
User: debian-pyt...@lists.debian.org
Usertags: python3.11 python3-all-dev

This package build-depends on python3-all-dev, but does not build
extensions/libraries for all supported python3 versions.  This is seen
on the transition tracker for adding python3.11 support [1] and
creates false positives.

Please either replace the b-d python3-all-dev with python3-dev, or
build for all supported Python3 versions.  With the former solution
you get later exposure to a new python3 version, with the latter
solution you get early exposure.

https://release.debian.org/transitions/html/python3.11-add.html

SR

-- 
debian-science-maintainers mailing list
debian-science-maintainers@alioth-lists.debian.net
https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/debian-science-maintainers


Bug#1022480: python-peachpy: diff for NMU version 0.0~git20211013.257881e-1.1

2022-11-13 Thread Stefano Rivera
Control: tags 1022480 + patch
Control: tags 1022480 + pending

Dear maintainer,

I've prepared an NMU for python-peachpy (versioned as 
0.0~git20211013.257881e-1.1) and
uploaded it to DELAYED/2. Please feel free to tell me if I
should delay it longer.

Regards,

SR
diff -Nru python-peachpy-0.0~git20211013.257881e/debian/changelog python-peachpy-0.0~git20211013.257881e/debian/changelog
--- python-peachpy-0.0~git20211013.257881e/debian/changelog	2022-01-10 07:28:35.0 +0200
+++ python-peachpy-0.0~git20211013.257881e/debian/changelog	2022-11-13 11:31:03.0 +0200
@@ -1,3 +1,10 @@
+python-peachpy (0.0~git20211013.257881e-1.1) unstable; urgency=medium
+
+  * Non-maintainer upload.
+  * Patch: Support setuptools >= 60 (Closes: #1022480)
+
+ -- Stefano Rivera   Sun, 13 Nov 2022 11:31:03 +0200
+
 python-peachpy (0.0~git20211013.257881e-1) unstable; urgency=medium
 
   * New upstream version 0.0~git20211013.257881e (Closes: #966250)
diff -Nru python-peachpy-0.0~git20211013.257881e/debian/patches/empty-package-dir.patch python-peachpy-0.0~git20211013.257881e/debian/patches/empty-package-dir.patch
--- python-peachpy-0.0~git20211013.257881e/debian/patches/empty-package-dir.patch	1970-01-01 02:00:00.0 +0200
+++ python-peachpy-0.0~git20211013.257881e/debian/patches/empty-package-dir.patch	2022-11-13 11:31:03.0 +0200
@@ -0,0 +1,27 @@
+From eaa2eaef5e4f00bc61d0bc9dafbfda13518c7c57 Mon Sep 17 00:00:00 2001
+From: Alfredo Tupone 
+Date: Sun, 15 May 2022 20:03:39 +0200
+Subject: [PATCH] Update setup.py
+
+Consider the case where self.distribution.package_dir is an empty dictionary
+Origin: upstream, https://github.com/Maratyszcza/PeachPy/commit/eaa2eaef5e4f00bc61d0bc9dafbfda13518c7c57
+---
+ setup.py | 2 +-
+ 1 file changed, 1 insertion(+), 1 deletion(-)
+
+diff --git a/setup.py b/setup.py
+index bd80bd6..337b310 100755
+--- a/setup.py
 b/setup.py
+@@ -35,7 +35,7 @@ class GenerateInstructions(Command):
+ def run(self):
+ # package_dir may be None, in that case use the current directory.
+ import os
+-if self.distribution.package_dir is None:
++if not self.distribution.package_dir:
+ src_dir = os.getcwd()
+ else:
+ src_dir = os.path.abspath(self.distribution.package_dir[""])
+-- 
+2.35.1
+
diff -Nru python-peachpy-0.0~git20211013.257881e/debian/patches/series python-peachpy-0.0~git20211013.257881e/debian/patches/series
--- python-peachpy-0.0~git20211013.257881e/debian/patches/series	2022-01-10 07:24:58.0 +0200
+++ python-peachpy-0.0~git20211013.257881e/debian/patches/series	2022-11-13 11:31:03.0 +0200
@@ -1 +1,3 @@
 reproducible-build.patch
+empty-package-dir.patch
+setuptools-60.patch
diff -Nru python-peachpy-0.0~git20211013.257881e/debian/patches/setuptools-60.patch python-peachpy-0.0~git20211013.257881e/debian/patches/setuptools-60.patch
--- python-peachpy-0.0~git20211013.257881e/debian/patches/setuptools-60.patch	1970-01-01 02:00:00.0 +0200
+++ python-peachpy-0.0~git20211013.257881e/debian/patches/setuptools-60.patch	2022-11-13 11:31:03.0 +0200
@@ -0,0 +1,34 @@
+From a1a3b0b88ea2802d2b0c48c43c73f95897ed914f Mon Sep 17 00:00:00 2001
+From: Stefano Rivera 
+Date: Sun, 13 Nov 2022 11:28:59 +0200
+Subject: [PATCH] Import setuptools before distutils
+
+setuptools 60 uses its own bundled version of distutils, by default. It
+injects this into sys.modules, at import time. So we need to make sure
+that it is imported, before anything else imports distutils, to ensure
+everything is using the same distutils version.
+
+This change in setuptools is to prepare for Python 3.12, which will drop
+distutils.
+
+Bug-Debian: https://bugs.debian.org/1022480
+Forwarded: https://github.com/Maratyszcza/PeachPy/pull/133
+---
+ setup.py | 4 ++--
+ 1 file changed, 2 insertions(+), 2 deletions(-)
+
+--- a/setup.py
 b/setup.py
+@@ -1,10 +1,10 @@
+ #!/usr/bin/env python
+ 
++from setuptools import setup
++from setuptools.command.develop import develop
+ import distutils.log
+ from distutils.command.build import build
+-from setuptools.command.develop import develop
+ from distutils.cmd import Command
+-from setuptools import setup
+ 
+ def read_text_file(path):
+ import os
-- 
debian-science-maintainers mailing list
debian-science-maintainers@alioth-lists.debian.net
https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/debian-science-maintainers


Bug#1009240: esys-particle: diff for NMU version 2.3.5+dfsg2-4.1

2022-04-13 Thread Stefano Rivera
Control: tags 1009240 + patch
Control: tags 1009240 + pending

Dear maintainer,

I've prepared an NMU for esys-particle (versioned as 2.3.5+dfsg2-4.1) and
uploaded it to DELAYED/5. Please feel free to tell me if I
should delay it longer.

Regards,

SR
diff -Nru esys-particle-2.3.5+dfsg2/debian/changelog esys-particle-2.3.5+dfsg2/debian/changelog
--- esys-particle-2.3.5+dfsg2/debian/changelog	2022-01-17 11:16:32.0 -0400
+++ esys-particle-2.3.5+dfsg2/debian/changelog	2022-04-13 21:24:02.0 -0400
@@ -1,3 +1,11 @@
+esys-particle (2.3.5+dfsg2-4.1) unstable; urgency=medium
+
+  * Non-maintainer upload.
+  * Patch: Don't call Py_Initialize() before Py_Main(), fixing sys.path on
+Python 3.10. (Closes: #1009240)
+
+ -- Stefano Rivera   Wed, 13 Apr 2022 21:24:02 -0400
+
 esys-particle (2.3.5+dfsg2-4) unstable; urgency=medium
 
   * Team upload
diff -Nru esys-particle-2.3.5+dfsg2/debian/patches/python-3.10-pyinitialize.patch esys-particle-2.3.5+dfsg2/debian/patches/python-3.10-pyinitialize.patch
--- esys-particle-2.3.5+dfsg2/debian/patches/python-3.10-pyinitialize.patch	1969-12-31 20:00:00.0 -0400
+++ esys-particle-2.3.5+dfsg2/debian/patches/python-3.10-pyinitialize.patch	2022-04-13 21:23:18.0 -0400
@@ -0,0 +1,18 @@
+Description: Don't call Py_Initialize before Py_Main
+ This is unsupported, and breaks sys.path on Python 3.10.
+ Since https://github.com/python/cpython/commit/af1d64d9f7a7cf673279725fdbaf4adcca51d41f
+ See: https://bugs.python.org/issue36204
+ See: https://bugs.python.org/issue34008
+Bug-Debian: https://bugs.debian.org/1009240
+Author: Stefano Rivera 
+
+--- a/Python/MpiPython/MpiPythonMain.cpp
 b/Python/MpiPython/MpiPythonMain.cpp
+@@ -48,7 +48,6 @@
+ if(myrank==0){// if rank==0 --> master
+   // start python
+   // std::cout << "master start\n";
+-  Py_Initialize();
+   
+ #if PY_VERSION_HEX >= 0x0300
+   wchar_t** wargv = new wchar_t*[argc+1];
diff -Nru esys-particle-2.3.5+dfsg2/debian/patches/series esys-particle-2.3.5+dfsg2/debian/patches/series
--- esys-particle-2.3.5+dfsg2/debian/patches/series	2022-01-17 11:16:10.0 -0400
+++ esys-particle-2.3.5+dfsg2/debian/patches/series	2022-04-13 21:21:00.0 -0400
@@ -7,3 +7,4 @@
 07_fix_FTBFS_perl5.26.patch
 09_fix_tk9.patch
 support-python-3.10.patch
+python-3.10-pyinitialize.patch
-- 
debian-science-maintainers mailing list
debian-science-maintainers@alioth-lists.debian.net
https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/debian-science-maintainers


Bug#1009240: esys-particle: autopkgtest regression since rebuild for python3.10 as default

2022-04-13 Thread Stefano Rivera
I did some digging:

The issue here is that sys.path is not being correctly configured

It is being initialized to:
['', '/tmp/cpython/lib/python310.zip', '/tmp/cpython/lib/python3.10', 
'/tmp/cpython/lib/python3.10/lib-dynload']

not:
['', '/tmp/cpython/lib/python310.zip', '/tmp/cpython/lib/python3.10', 
'/tmp/cpython/lib/python3.10/lib-dynload', 
'/tmp/cpython/lib/python3.10/site-packages']

You can verify this by running:
esysparticle -c 'import sys; print(sys.path)'

More digging showed that it is actually being correctly initialized in
Py_Initialize(), but then re-initialized to be incorrect in Py_Main().

A git bisect got me to
https://github.com/python/cpython/commit/af1d64d9f7a7cf673279725fdbaf4adcca51d41f
as being the commit that made that change.

I haven't got to the bottom of exactly why that reinitialization is
going wrong, but these bugs hint at this code as being unsupported:
https://bugs.python.org/issue36204
https://bugs.python.org/issue34008

The easy answer here is to stop calling Py_Initialize(), as it's not
necessary.

The "right" long-term approach would be to migrate to Py_RunMain().

SR

-- 
Stefano Rivera
  http://tumbleweed.org.za/
  +1 415 683 3272

-- 
debian-science-maintainers mailing list
debian-science-maintainers@alioth-lists.debian.net
https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/debian-science-maintainers


Bug#1001484: sfepy: (autopkgtest) needs update for python3.10: Sequence' from 'collections' is removed

2022-03-24 Thread Stefano Rivera
Control: tag -1 + pending

I see a commit fixing this in git, just waiting to be uploaded.
https://salsa.debian.org/science-team/sfepy/-/commit/f6a4f8d2435e8406a629a75ee89891a24631233a

SR

-- 
Stefano Rivera
  http://tumbleweed.org.za/
  +1 415 683 3272

-- 
debian-science-maintainers mailing list
debian-science-maintainers@alioth-lists.debian.net
https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/debian-science-maintainers


Bug#1006615: python3-pybind11: Handle Python 3.10's default sysconfig paths

2022-03-01 Thread Stefano Rivera
Hi Debian (2022.02.28_13:58:36_-0400)
> Attached are a pair of patches to address the issue. I've forwarded them
> upstream in https://github.com/pybind/pybind11/pull/3764

And it has been merged, upstream.

SR

-- 
Stefano Rivera
  http://tumbleweed.org.za/
  +1 415 683 3272

-- 
debian-science-maintainers mailing list
debian-science-maintainers@alioth-lists.debian.net
https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/debian-science-maintainers


Bug#1006615: python3-pybind11: Handle Python 3.10's default sysconfig paths

2022-02-28 Thread Stefano Rivera
Package: python3-pybind11
Version: 2.9.1-1
Severity: normal
Tags: patch
Forwarded: https://github.com/pybind/pybind11/pull/3764

Until recently, the Python sysconfig module didn't know about Debian's
deb_system and posix_local schemes that were patched into
distutils.sysconfig. This changed in 3.10.2-3, to allow the deprecation
of distutils, upstream.

See https://bugs.launchpad.net/ubuntu/+source/python3.10/+bug/1960608
for some of the back-story here.

The effective change here is that with Python 3.10 as default, packages
using pybind11 are FTBFS because they can't find Python.h. e.g.
https://launchpad.net/ubuntu/+source/avogadrolibs/1.95.1-8

> CMake Error in python/CMakeLists.txt:
>   Imported target "pybind11::module" includes non-existent path
> 
> "/usr/local/include/python3.10"
> 
>   in its INTERFACE_INCLUDE_DIRECTORIES.

I'm pointing to Ubuntu bugs here, because we're hitting these issues in
Ubuntu first, where Python 3.10 is the default Python 3. Debian hasn't
got there, yet.

The solutions here are two-fold:
1. Migrate from distutils.sysconfig to sysconfig. This needs to happen,
   upstream, before distutils is gone from stdlib.
2. Explicitly select the correct system scheme when looking for
   Python3's headers, rather than the default install-to-/usr/local
   scheme.

Attached are a pair of patches to address the issue. I've forwarded them
upstream in https://github.com/pybind/pybind11/pull/3764

SR
From 5262ff794498b45c12f23f64e9adef30d51e667f Mon Sep 17 00:00:00 2001
From: Stefano Rivera 
Date: Mon, 28 Feb 2022 12:49:51 -0400
Subject: [PATCH 1/2] Use sysconfig in Python >= 3.10

Rely on sysconfig for installation paths for Python >= 3.10. distutils
has been deprecated and will be removed.

Fixes: #3677
---
 tools/FindPythonLibsNew.cmake | 15 ---
 1 file changed, 12 insertions(+), 3 deletions(-)

diff --git a/tools/FindPythonLibsNew.cmake b/tools/FindPythonLibsNew.cmake
index 7aeffa44..31eeffa0 100644
--- a/tools/FindPythonLibsNew.cmake
+++ b/tools/FindPythonLibsNew.cmake
@@ -112,11 +112,20 @@ endif()
 # VERSION. VERSION will typically be like "2.7" on unix, and "27" on windows.
 execute_process(
   COMMAND
-"${PYTHON_EXECUTABLE}" "-c" "from distutils import sysconfig as s;import sys;import struct;
+"${PYTHON_EXECUTABLE}" "-c" "
+import sys;import struct;
+import sysconfig as s
+USE_SYSCONFIG = sys.version_info >= (3, 10)
+if not USE_SYSCONFIG:
+from distutils import sysconfig as ds
 print('.'.join(str(v) for v in sys.version_info));
 print(sys.prefix);
-print(s.get_python_inc(plat_specific=True));
-print(s.get_python_lib(plat_specific=True));
+if USE_SYSCONFIG:
+print(s.get_path('platinclude'))
+print(s.get_path('platlib'))
+else:
+print(ds.get_python_inc(plat_specific=True));
+print(ds.get_python_lib(plat_specific=True));
 print(s.get_config_var('EXT_SUFFIX') or s.get_config_var('SO'));
 print(hasattr(sys, 'gettotalrefcount')+0);
 print(struct.calcsize('@P'));
-- 
2.34.1

From 697e8e1fca67bb48e477a61a9bbfa049c653ecd2 Mon Sep 17 00:00:00 2001
From: Stefano Rivera 
Date: Mon, 28 Feb 2022 13:11:45 -0400
Subject: [PATCH 2/2] Explicitly select the posix_prefix scheme for platinclude
 on Debian

Debian's default scheme is posix_local, for installing locally-built
packages to /usr/local/.  We want to find the Python headers in /usr/,
so search posix_prefix.
---
 tools/FindPythonLibsNew.cmake | 6 +-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/tools/FindPythonLibsNew.cmake b/tools/FindPythonLibsNew.cmake
index 31eeffa0..6ad5fe15 100644
--- a/tools/FindPythonLibsNew.cmake
+++ b/tools/FindPythonLibsNew.cmake
@@ -121,7 +121,11 @@ if not USE_SYSCONFIG:
 print('.'.join(str(v) for v in sys.version_info));
 print(sys.prefix);
 if USE_SYSCONFIG:
-print(s.get_path('platinclude'))
+scheme = s.get_default_scheme()
+if scheme == 'posix_local':
+# Debian's default scheme installs to /usr/local/ but we want to find headers in /usr/
+scheme = 'posix_prefix'
+print(s.get_path('platinclude', scheme))
 print(s.get_path('platlib'))
 else:
 print(ds.get_python_inc(plat_specific=True));
-- 
2.34.1

-- 
debian-science-maintainers mailing list
debian-science-maintainers@alioth-lists.debian.net
https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/debian-science-maintainers


Bug#1000204: pyfai: diff for NMU version 0.20.0+dfsg1-4.1

2021-11-22 Thread Stefano Rivera
Control: tags 1000204 + pending

Dear maintainer,

I've prepared an NMU for pyfai (versioned as 0.20.0+dfsg1-4.1) and
uploaded it to DELAYED/5. Please feel free to tell me if I
should delay it longer.

Also filed as https://salsa.debian.org/science-team/pyfai/-/merge_requests/4

Regards,

SR
diff -Nru pyfai-0.20.0+dfsg1/debian/changelog pyfai-0.20.0+dfsg1/debian/changelog
--- pyfai-0.20.0+dfsg1/debian/changelog	2021-09-19 07:57:57.0 -0400
+++ pyfai-0.20.0+dfsg1/debian/changelog	2021-11-22 21:29:59.0 -0400
@@ -1,3 +1,10 @@
+pyfai (0.20.0+dfsg1-4.1) UNRELEASED; urgency=medium
+
+  * Non-maintainer upload.
+  * Patch: Python 3.10 support, thanks Lukas Märdian. (Closes: #1000204)
+
+ -- Stefano Rivera   Mon, 22 Nov 2021 21:29:59 -0400
+
 pyfai (0.20.0+dfsg1-4) unstable; urgency=medium
 
   * Bug fix: "Removal of the python3-*-dbg packages in sid/bookworm",
diff -Nru pyfai-0.20.0+dfsg1/debian/patches/0009-Fix-collections.MutableSets-compat-with-python3.10.patch pyfai-0.20.0+dfsg1/debian/patches/0009-Fix-collections.MutableSets-compat-with-python3.10.patch
--- pyfai-0.20.0+dfsg1/debian/patches/0009-Fix-collections.MutableSets-compat-with-python3.10.patch	1969-12-31 20:00:00.0 -0400
+++ pyfai-0.20.0+dfsg1/debian/patches/0009-Fix-collections.MutableSets-compat-with-python3.10.patch	2021-11-22 21:29:59.0 -0400
@@ -0,0 +1,45 @@
+From: Lukas Märdian 
+Date: Mon, 22 Nov 2021 21:30:44 -0400
+Subject: Fix collections.MutableSets compat with python3.10
+
+collections.MutableSet was deprecated a long time ago and removed with python3.10.
+https://docs.python.org/3/whatsnew/3.10.html#removed
+https://bugs.python.org/issue37324
+
+Using the old class name leads to test failures such as this one:
+==
+ERROR: test_import_all_modules (pyFAI.test.test_bug_regression.TestBugRegression)
+Try to import every single module in the package
+--
+Traceback (most recent call last):
+  File "/<>/pyfai-0.20.0+dfsg1/.pybuild/cpython3_3.10_pyfai/build/pyFAI/test/test_bug_regression.py", line 286, in test_import_all_modules
+raise err
+  File "/<>/pyfai-0.20.0+dfsg1/.pybuild/cpython3_3.10_pyfai/build/pyFAI/test/test_bug_regression.py", line 271, in test_import_all_modules
+load_source(fqn, path)
+  File "/<>/pyfai-0.20.0+dfsg1/.pybuild/cpython3_3.10_pyfai/build/pyFAI/test/test_bug_regression.py", line 78, in load_source
+spec.loader.exec_module(module)
+  File "", line 883, in exec_module
+  File "", line 241, in _call_with_frames_removed
+  File "/<>/pyfai-0.20.0+dfsg1/.pybuild/cpython3_3.10_pyfai/build/pyFAI/utils/orderedset.py", line 30, in 
+class OrderedSet(collections.MutableSet):
+AttributeError: module 'collections' has no attribute 'MutableSet'
+
+Origin: vendor, Ubuntu
+Forwarded: https://github.com/silx-kit/pyFAI/pull/1588
+---
+ pyFAI/utils/orderedset.py | 2 +-
+ 1 file changed, 1 insertion(+), 1 deletion(-)
+
+diff --git a/pyFAI/utils/orderedset.py b/pyFAI/utils/orderedset.py
+index de097cf..417490f 100644
+--- a/pyFAI/utils/orderedset.py
 b/pyFAI/utils/orderedset.py
+@@ -27,7 +27,7 @@
+ import collections
+ 
+ 
+-class OrderedSet(collections.MutableSet):
++class OrderedSet(collections.abc.MutableSet):
+ 
+ def __init__(self, iterable=None):
+ self.end = end = []
diff -Nru pyfai-0.20.0+dfsg1/debian/patches/series pyfai-0.20.0+dfsg1/debian/patches/series
--- pyfai-0.20.0+dfsg1/debian/patches/series	2021-09-19 07:57:57.0 -0400
+++ pyfai-0.20.0+dfsg1/debian/patches/series	2021-11-22 21:29:59.0 -0400
@@ -6,3 +6,4 @@
 0006-fix-a-couple-of-test-failing-with-i386.patch
 0007-cleaner-test.patch
 0008-tunning.patch
+0009-Fix-collections.MutableSets-compat-with-python3.10.patch
-- 
debian-science-maintainers mailing list
debian-science-maintainers@alioth-lists.debian.net
https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/debian-science-maintainers


Bug#999519: xrayutilities: diff for NMU version 1.7.1-2.1

2021-11-22 Thread Stefano Rivera
Control: tags 999519 + patch
Control: tags 999519 + pending

Dear maintainer,

I've prepared an NMU for xrayutilities (versioned as 1.7.1-2.1) and
uploaded it to DELAYED/2. Please feel free to tell me if I
should delay it longer.

Also filed as: 
https://salsa.debian.org/science-team/python-xrayutilities/-/merge_requests/1

Regards,

SR
diff -Nru xrayutilities-1.7.1/debian/changelog xrayutilities-1.7.1/debian/changelog
--- xrayutilities-1.7.1/debian/changelog	2021-09-19 08:54:46.0 -0400
+++ xrayutilities-1.7.1/debian/changelog	2021-11-22 21:22:36.0 -0400
@@ -1,3 +1,10 @@
+xrayutilities (1.7.1-2.1) unstable; urgency=medium
+
+  * Non-maintainer upload.
+  * Patch: matplotlib 3.5 support (Closes: #999519)
+
+ -- Stefano Rivera   Mon, 22 Nov 2021 21:22:36 -0400
+
 xrayutilities (1.7.1-2) unstable; urgency=medium
 
   * Corrected Profile nodocs -> nodoc
diff -Nru xrayutilities-1.7.1/debian/patches/0005-Fix-matplotlib-3.5-issue.patch xrayutilities-1.7.1/debian/patches/0005-Fix-matplotlib-3.5-issue.patch
--- xrayutilities-1.7.1/debian/patches/0005-Fix-matplotlib-3.5-issue.patch	1969-12-31 20:00:00.0 -0400
+++ xrayutilities-1.7.1/debian/patches/0005-Fix-matplotlib-3.5-issue.patch	2021-11-22 21:22:36.0 -0400
@@ -0,0 +1,35 @@
+From: Dominik Kriegner 
+Date: Fri, 12 Nov 2021 12:30:16 +0100
+Subject: Fix matplotlib-3.5 issue.
+
+reported by Debian. The best solution seems to be to avoid the
+problematic keyword argument since we anyhow specified the default
+value.
+
+Bug-Debian: https://bugs.debian.org/999519
+---
+ lib/xrayutilities/simpack/fit.py | 4 ++--
+ 1 file changed, 2 insertions(+), 2 deletions(-)
+
+diff --git a/lib/xrayutilities/simpack/fit.py b/lib/xrayutilities/simpack/fit.py
+index 8e71783..5d49dcf 100644
+--- a/lib/xrayutilities/simpack/fit.py
 b/lib/xrayutilities/simpack/fit.py
+@@ -13,7 +13,7 @@
+ # You should have received a copy of the GNU General Public License
+ # along with this program; if not, see <http://www.gnu.org/licenses/>.
+ #
+-# Copyright (C) 2016-2020 Dominik Kriegner 
++# Copyright (C) 2016-2021 Dominik Kriegner 
+ 
+ import warnings
+ 
+@@ -208,7 +208,7 @@ class FitModel(object):
+ else:
+ self.zord = 1
+ if self.logscale:
+-self.ax.set_yscale("log", nonposy='clip')
++self.ax.set_yscale("log")
+ self.fline = None
+ 
+ def showplot(self, xlab=self.lmodel.xlabelstr,
diff -Nru xrayutilities-1.7.1/debian/patches/series xrayutilities-1.7.1/debian/patches/series
--- xrayutilities-1.7.1/debian/patches/series	2021-09-19 08:54:46.0 -0400
+++ xrayutilities-1.7.1/debian/patches/series	2021-11-22 21:22:36.0 -0400
@@ -2,3 +2,4 @@
 0002-Desactivated-pdf-generation-for-the-doc.patch
 0003-Do-not-change-PYTHONPATH-during-the-build.patch
 0004-Used-the-system-Mathjax.patch
+0005-Fix-matplotlib-3.5-issue.patch
-- 
debian-science-maintainers mailing list
debian-science-maintainers@alioth-lists.debian.net
https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/debian-science-maintainers


Bug#1000422: pandas: FTBFS with Python 3.10 - Test failures

2021-11-22 Thread Stefano Rivera
Source: pandas
Version: 1.1.5+dfsg-2
Severity: serious
Justification: FTBFS

Pandas fails to build with Python 3.10.
See: 
https://buildd.debian.org/status/fetch.php?pkg=pandas=amd64=1.1.5%2Bdfsg-2%2Bb1=1637461771=0

=== FAILURES ===
 TestDatetime64SeriesComparison.test_comparison_invalid[None-DataFrame] 

left =0  1  2  3  4
0  0  1  2  3  4
right =0  1  2  3  4
0 2001-01-01 2001-01-02 2001-01-03 2001-01-04 2001-01-05
box = 

def assert_invalid_comparison(left, right, box):
"""
Assert that comparison operations with mismatched types behave 
correctly.

Parameters
--
left : np.ndarray, ExtensionArray, Index, or Series
right : object
box : {pd.DataFrame, pd.Series, pd.Index, tm.to_array}
"""
# Not for tznaive-tzaware comparison

# Note: not quite the same as how we do this for tm.box_expected
xbox = box if box is not Index else np.array

result = left == right
expected = xbox(np.zeros(result.shape, dtype=np.bool_))

tm.assert_equal(result, expected)

result = right == left
tm.assert_equal(result, expected)

result = left != right
tm.assert_equal(result, ~expected)

result = right != left
tm.assert_equal(result, ~expected)

msg = "|".join(
[
"Invalid comparison between",
"Cannot compare type",
"not supported between",
"invalid type promotion",
(
# GH#36706 npdev 1.20.0 2020-09-28
r"The DTypes  and "
r" do not have a common 
DType. "
"For example they cannot be stored in a single array unless 
the "
"dtype is `object`."
),
]
)
with pytest.raises(TypeError, match=msg):
>   left < right

pandas/tests/arithmetic/common.py:89:

Looks like upstream has got there, but a quick search didn't find the
relevant patches, yet:
https://github.com/pandas-dev/pandas/commit/e7efd02c71c48c7968b2f8fdb845a0b0bf61a3fa

SR

-- 
debian-science-maintainers mailing list
debian-science-maintainers@alioth-lists.debian.net
https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/debian-science-maintainers


Bug#1000360: segyio: diff for NMU version 1.8.3-1.1

2021-11-21 Thread Stefano Rivera
Control: tags 1000360 + pending

Dear maintainer,

I've prepared an NMU for segyio (versioned as 1.8.3-1.1) and
uploaded it to DELAYED/5. Please feel free to tell me if I
should delay it longer.

Also filed as:
https://salsa.debian.org/science-team/segyio/-/merge_requests/1

I'd really suggest a new upstream release, instead, though...

Regards,

SR
diff -Nru segyio-1.8.3/debian/changelog segyio-1.8.3/debian/changelog
--- segyio-1.8.3/debian/changelog	2018-12-23 08:17:03.0 -0400
+++ segyio-1.8.3/debian/changelog	2021-11-21 22:51:13.0 -0400
@@ -1,3 +1,12 @@
+segyio (1.8.3-1.1) unstable; urgency=medium
+
+  * Non-maintainer upload.
+  * Python 3.10 support. (Closes: #1000360)
+- Patch: Import from collections.abc.
+- Patch: PY_SSIZE_T_CLEAN.
+
+ -- Stefano Rivera   Sun, 21 Nov 2021 22:51:13 -0400
+
 segyio (1.8.3-1) unstable; urgency=medium
 
   * New upstream release
diff -Nru segyio-1.8.3/debian/patches/python-3.10-collections.abc.patch segyio-1.8.3/debian/patches/python-3.10-collections.abc.patch
--- segyio-1.8.3/debian/patches/python-3.10-collections.abc.patch	1969-12-31 20:00:00.0 -0400
+++ segyio-1.8.3/debian/patches/python-3.10-collections.abc.patch	2021-11-21 22:51:13.0 -0400
@@ -0,0 +1,210 @@
+From: =?utf-8?q?J=C3=B8rgen_Kvalsvik?= 
+Date: Mon, 7 Oct 2019 13:42:35 +0200
+Subject: Try-import collections.abc
+
+In python 3.8 [1], importing the abstract base classes from collections
+will break. Protecting the import in try is both backwards- and forward
+compatible, as collections.abc is not valid in python2.
+
+[1] DeprecationWarning: Using or importing the ABCs from 'collections'
+instead of from 'collections.abc' is deprecated, and in 3.8 it will
+stop working
+
+Origin: upstream, https://github.com/equinor/segyio/commit/b6378fb2a283c504209b55e1d1af402d4d765e11
+Bug-Debian: https://bugs.debian.org/1000360
+---
+ python/segyio/depth.py |  7 ++-
+ python/segyio/field.py | 17 +++--
+ python/segyio/line.py  | 14 +-
+ python/segyio/trace.py | 18 +++---
+ 4 files changed, 37 insertions(+), 19 deletions(-)
+
+diff --git a/python/segyio/depth.py b/python/segyio/depth.py
+index 948e1e5..7346cfa 100644
+--- a/python/segyio/depth.py
 b/python/segyio/depth.py
+@@ -1,3 +1,8 @@
++try:
++from collections.abc import Sequence # noqa
++except ImportError:
++from collections import Sequence # noqa
++
+ import numpy as np
+ try: from future_builtins import zip
+ except ImportError: pass
+@@ -23,7 +28,7 @@ class Depth(Sequence):
+ .. versionadded:: 1.1
+ 
+ .. versionchanged:: 1.6
+-common list operations (collections.Sequence)
++common list operations (Sequence)
+ 
+ .. versionchanged:: 1.7.1
+enabled for unstructured files
+diff --git a/python/segyio/field.py b/python/segyio/field.py
+index f6e8a46..8389935 100644
+--- a/python/segyio/field.py
 b/python/segyio/field.py
+@@ -1,10 +1,15 @@
+-import collections
++try:
++from collections.abc import Mapping # noqa
++from collections.abc import MutableMapping # noqa
++except ImportError:
++from collections import Mapping # noqa
++from collections import MutableMapping # noqa
+ 
+ import segyio
+ from .binfield import BinField
+ from .tracefield import TraceField
+ 
+-class Field(collections.MutableMapping):
++class Field(MutableMapping):
+ """
+ The Field implements the dict interface, with a fixed set of keys. It's
+ used for both binary- and trace headers. Any modifications to this
+@@ -22,7 +27,7 @@ class Field(collections.MutableMapping):
+ common dict operations (update, keys, values)
+ 
+ .. versionchanged:: 1.6
+-more common dict operations (collections.MutableMapping)
++more common dict operations (MutableMapping)
+ """
+ _bin_keys = [x for x in BinField.enums()
+  if  x != BinField.Unassigned1
+@@ -438,7 +443,7 @@ class Field(collections.MutableMapping):
+ def __eq__(self, other):
+ """x.__eq__(y) <==> x == y"""
+ 
+-if not isinstance(other, collections.Mapping):
++if not isinstance(other, Mapping):
+ return NotImplemented
+ 
+ if len(self) != len(other):
+@@ -492,13 +497,13 @@ class Field(collections.MutableMapping):
+ 
+ buf = bytearray(self.buf)
+ 
+-# Implementation largely borrowed from collections.mapping
++# Implementation largely borrowed from Mapping
+ # If E present and has a .keys() method: for k in E: D[k] = E[k]
+ # If E present and lacks .keys() method: for (k, v) in E: D[k] = v
+ # In either case, this is followed by: for k, v in F.items(): D[k] = v
+ if len(args) == 1:
+ other = args[0]
+-if isinstance(other, collections.Mapping):
++if isinstance(other, Mapping):
+ for key in other:
+ self.pu

Bug#1000360: segyio: FTBFS with Python 3.10

2021-11-21 Thread Stefano Rivera
Source: segyio
Version: 1.8.3-1
Severity: serious
Tags: upstream patch
Justification: FTBFS

FTBFS with Python 3.10:

Tests fail with:
E   AttributeError: module 'collections' has no attribute 'Mapping'

See: 
https://buildd.debian.org/status/fetch.php?pkg=segyio=amd64=1.8.3-1%2Bb5=1637491602=0

Looks like upstream fixed this issue years ago, updating to a new
upstream version would probably be good.

But, in the meantime, we can probably patch it.

SR

-- 
debian-science-maintainers mailing list
debian-science-maintainers@alioth-lists.debian.net
https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/debian-science-maintainers


Bug#1000336: numba: FTBFS with Python 3.10

2021-11-21 Thread Stefano Rivera
Source: numba
Version: 0.52.0-5
Severity: serious
Tags: ftbfs upstream
Justification: FTBFS
Forwarded: https://github.com/numba/numba/issues/7562

numba isn't compatible with Python 3.10, yet.
Upstream is working on it, see https://github.com/numba/numba/issues/7562

setup.py immediately fails with:
RuntimeError: Cannot install on Python version 3.10.0; only versions 
>=3.6,<3.10 are supported.

See e.g. 
https://buildd.debian.org/status/fetch.php?pkg=numba=amd64=0.52.0-5%2Bb1=1637478629=0

SR

-- 
debian-science-maintainers mailing list
debian-science-maintainers@alioth-lists.debian.net
https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/debian-science-maintainers


Bug#997448: mlpy: diff for NMU version 3.5.0+ds-1.3

2021-11-21 Thread Stefano Rivera
Control: tags 997448 + patch

Dear maintainer,

I've prepared an NMU for mlpy (versioned as 3.5.0+ds-1.3). The diff
is attached to this message.

Also forwarded as https://salsa.debian.org/science-team/mlpy/-/merge_requests/3

Regards,

SR
diff -Nru mlpy-3.5.0+ds/debian/changelog mlpy-3.5.0+ds/debian/changelog
--- mlpy-3.5.0+ds/debian/changelog	2020-06-12 10:20:54.0 -0400
+++ mlpy-3.5.0+ds/debian/changelog	2021-11-21 10:46:25.0 -0400
@@ -1,3 +1,10 @@
+mlpy (3.5.0+ds-1.3) unstable; urgency=medium
+
+  * Non-maintainer upload.
+  * Add Build-Dependency on tex-gyre, for docs (Closes: #997448)
+
+ -- Stefano Rivera   Sun, 21 Nov 2021 10:46:25 -0400
+
 mlpy (3.5.0+ds-1.2) unstable; urgency=medium
 
   * Non-maintainer upload.
diff -Nru mlpy-3.5.0+ds/debian/control mlpy-3.5.0+ds/debian/control
--- mlpy-3.5.0+ds/debian/control	2020-06-12 10:20:54.0 -0400
+++ mlpy-3.5.0+ds/debian/control	2021-11-21 10:46:25.0 -0400
@@ -13,6 +13,7 @@
python3-numpy,
python3-sphinx,
python3-scipy ,
+   tex-gyre,
texlive,
texlive-latex-extra,
latexmk,
-- 
debian-science-maintainers mailing list
debian-science-maintainers@alioth-lists.debian.net
https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/debian-science-maintainers


Bug#999538: pytables: diff for NMU version 3.6.1-5.1

2021-11-20 Thread Stefano Rivera
Control: tags 999538 + patch
Control: tags 999538 + pending

Dear maintainer,

I've prepared an NMU for pytables (versioned as 3.6.1-5.1) and
uploaded it to DELAYED/2. Please feel free to tell me if I
should delay it longer.

Also available as an MR: 
https://salsa.debian.org/science-team/pytables/-/merge_requests/1

Regards,

SR
diff -Nru pytables-3.6.1/debian/changelog pytables-3.6.1/debian/changelog
--- pytables-3.6.1/debian/changelog	2021-09-28 02:05:13.0 -0400
+++ pytables-3.6.1/debian/changelog	2021-11-20 23:32:58.0 -0400
@@ -1,3 +1,12 @@
+pytables (3.6.1-5.1) unstable; urgency=medium
+
+  * Non-maintainer upload.
+  * Patch: Fix FTBFS with numpy >= 1.20 (Closes: #999538)
+  * Patch: Fix FTBFS with Python 3.10.
+  * Install Python 3.10 C Extensions (2-digit minor version).
+
+ -- Stefano Rivera   Sat, 20 Nov 2021 23:32:58 -0400
+
 pytables (3.6.1-5) unstable; urgency=medium
 
   * debian/control:
diff -Nru pytables-3.6.1/debian/patches/0010-Use-lowercase-float-int-as-numpy-dtype.patch pytables-3.6.1/debian/patches/0010-Use-lowercase-float-int-as-numpy-dtype.patch
--- pytables-3.6.1/debian/patches/0010-Use-lowercase-float-int-as-numpy-dtype.patch	1969-12-31 20:00:00.0 -0400
+++ pytables-3.6.1/debian/patches/0010-Use-lowercase-float-int-as-numpy-dtype.patch	2021-11-20 23:32:58.0 -0400
@@ -0,0 +1,176 @@
+From: =?utf-8?q?Zbigniew_J=C4=99drzejewski-Szmek?= 
+Date: Sun, 24 Jan 2021 16:36:21 +0100
+Subject: Use lowercase float/int as numpy dtype
+
+Float64 is gone with numpy 1.20, which causes doctests to fail
+(https://bugzilla.redhat.com/show_bug.cgi?id=1914335).
+
+Similarly all uses of Float32, Int32 should be replaced by float32 and int32.
+
+>>> numpy.__version__
+'1.19.4'
+>>> [k for k in numpy.sctypeDict.keys() if str(k).lower().startswith('float')]
+['float16', 'Float16', 'float32', 'Float32', 'float64', 'Float64', 'float128', 'Float128', 'float_', 'float']
+
+>>> numpy.__version__
+'1.20.0rc2'
+>>> [k for k in numpy.sctypeDict.keys() if str(k).lower().startswith('float')]
+['float16', 'float32', 'float64', 'float128', 'float_', 'float']
+
+Bug-Upstream: https://github.com/PyTables/PyTables/issues/875
+Bug-Debian: https://bugs.debian.org/999538
+Origin: upstream, https://github.com/PyTables/PyTables/pull/862
+---
+ bench/bsddb-table-bench.py | 10 +-
+ bench/postgres-search-bench.py |  4 ++--
+ bench/pytables-search-bench.py |  6 +++---
+ bench/recarray2-test.py|  2 +-
+ bench/shelve-bench.py  | 10 +-
+ bench/sqlite-search-bench.py   |  4 ++--
+ tables/atom.py |  2 +-
+ 7 files changed, 19 insertions(+), 19 deletions(-)
+
+diff --git a/bench/bsddb-table-bench.py b/bench/bsddb-table-bench.py
+index dd9f875..a2e0236 100644
+--- a/bench/bsddb-table-bench.py
 b/bench/bsddb-table-bench.py
+@@ -83,11 +83,11 @@ def createFile(filename, totalrows, recsize, verbose):
+ # Get the record object associated with the new table
+ if recsize == "big":
+ isrec = Big()
+-arr = np.array(np.arange(32), type=np.Float64)
+-arr2 = np.array(np.arange(32), type=np.Float64)
++arr = np.array(np.arange(32), type=np.float64)
++arr2 = np.array(np.arange(32), type=np.float64)
+ elif recsize == "medium":
+ isrec = Medium()
+-arr = np.array(np.arange(2), type=np.Float64)
++arr = np.array(np.arange(2), type=np.float64)
+ else:
+ isrec = Small()
+ # print d
+@@ -107,8 +107,8 @@ def createFile(filename, totalrows, recsize, verbose):
+ #d['TDCcount'] = i % 256
+ d['ADCcount'] = (i * 256) % (1 << 16)
+ if recsize == "big":
+-#d.float1 = np.array([i]*32, np.Float64)
+-#d.float2 = np.array([i**2]*32, np.Float64)
++#d.float1 = np.array([i]*32, np.float64)
++#d.float2 = np.array([i**2]*32, np.float64)
+ arr[0] = 1.1
+ d['float1'] = arr
+ arr2[0] = 2.2
+diff --git a/bench/postgres-search-bench.py b/bench/postgres-search-bench.py
+index d2c9f4f..7fe83f6 100644
+--- a/bench/postgres-search-bench.py
 b/bench/postgres-search-bench.py
+@@ -15,11 +15,11 @@ def flatten(l):
+ 
+ 
+ def fill_arrays(start, stop):
+-col_i = numpy.arange(start, stop, type=numpy.Int32)
++col_i = numpy.arange(start, stop, type=numpy.int32)
+ if userandom:
+ col_j = numpy.random.uniform(0, nrows, size=[stop - start])
+ else:
+-col_j = numpy.array(col_i, type=numpy.Float64)
++col_j = numpy.array(col_i, type=numpy.float64)
+ return col_i, col_j
+ 
+ # Generator for ensure pytables benchmark compatibility
+diff --git a/bench/pytables-search-bench.py b/bench/pytables-search-bench.py
+index 726d30b..6417186 100644
+--- a/bench/pytables-search-bench.py
 b/bench/pytables-search-bench.py
+@@ -37,11 +37,11 @@ def create_db(

Bug#972372: sfepy: diff for NMU version 2019.4-2.1

2020-10-17 Thread Stefano Rivera
Control: tags 972372 + patch
Control: tags 972372 + pending

Dear maintainer,

I've prepared an NMU for sfepy (versioned as 2019.4-2.1) and
uploaded it to DELAYED/3. Please feel free to tell me if I
should delay it longer.

Regards.

SR
diff -Nru sfepy-2019.4/debian/changelog sfepy-2019.4/debian/changelog
--- sfepy-2019.4/debian/changelog	2020-01-21 11:29:55.0 -0800
+++ sfepy-2019.4/debian/changelog	2020-10-17 13:02:43.0 -0700
@@ -1,3 +1,11 @@
+sfepy (2019.4-2.1) unstable; urgency=medium
+
+  * Non-maintainer upload.
+  * Declare internal inline functions "static inline", fixing FTBFS on 32-bit
+archs (Closes: #972372).
+
+ -- Stefano Rivera   Sat, 17 Oct 2020 13:02:43 -0700
+
 sfepy (2019.4-2) unstable; urgency=medium
 
   * [40073a3] Trim trailing whitespace.
diff -Nru sfepy-2019.4/debian/patches/extern-inline.patch sfepy-2019.4/debian/patches/extern-inline.patch
--- sfepy-2019.4/debian/patches/extern-inline.patch	1969-12-31 16:00:00.0 -0800
+++ sfepy-2019.4/debian/patches/extern-inline.patch	2020-10-17 12:50:17.0 -0700
@@ -0,0 +1,35 @@
+Declare internal inline functions "static inline"
+
+Avoids FTBFS on 32-bit platforms where an undefined symbol _aux_hex was
+referenced from mesh.o.
+
+Author: Stefano Rivera 
+Bug-Debian: https://bugs.debian.org/972372
+Bug-Upstream: https://github.com/sfepy/sfepy/issues/651
+--- a/sfepy/discrete/common/extmods/mesh.c
 b/sfepy/discrete/common/extmods/mesh.c
+@@ -910,13 +910,13 @@
+   return(RET_OK);
+ }
+ 
+-inline float64 _det3x3(float64 j[9])
++static inline float64 _det3x3(float64 j[9])
+ {
+   return (j[0]*j[4]*j[8] + j[3]*j[7]*j[2] + j[1]*j[5]*j[6]
+   - j[2]*j[4]*j[6] - j[5]*j[7]*j[0] - j[1]*j[3]*j[8]);
+ }
+ 
+-inline float64 _tri_area(float64 *coors, uint32 *indices, uint32 nc)
++static inline float64 _tri_area(float64 *coors, uint32 *indices, uint32 nc)
+ {
+ #define VS(ic, id) (coors[nc*indices[ic] + id])
+ 
+@@ -945,7 +945,7 @@
+ #undef VS
+ }
+ 
+-inline float64 _aux_hex(float64 *coors, uint32 *indices, uint32 nc,
++static inline float64 _aux_hex(float64 *coors, uint32 *indices, uint32 nc,
+ int32 h, int32 i, int32 j, int32 k)
+ {
+ #define VS(ic, id) (coors[nc*indices[ic] + id])
diff -Nru sfepy-2019.4/debian/patches/series sfepy-2019.4/debian/patches/series
--- sfepy-2019.4/debian/patches/series	2020-01-21 11:29:55.0 -0800
+++ sfepy-2019.4/debian/patches/series	2020-10-17 12:46:14.0 -0700
@@ -1 +1,2 @@
 20_reproducible_build.patch
+extern-inline.patch
-- 
debian-science-maintainers mailing list
debian-science-maintainers@alioth-lists.debian.net
https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/debian-science-maintainers

Bug#972374: qiskit-terra: diff for NMU version 0.12.0-2.1

2020-10-17 Thread Stefano Rivera
Control: tags 972374 + patch
Control: tags 972374 + pending

Dear maintainer,

I've prepared an NMU for qiskit-terra (versioned as 0.12.0-2.1) and
uploaded it to DELAYED/3. Please feel free to tell me if I
should delay it longer.

Also, FYI, your git repo is missing on Salsa.

Regards.

SR
diff -Nru qiskit-terra-0.12.0/debian/changelog qiskit-terra-0.12.0/debian/changelog
--- qiskit-terra-0.12.0/debian/changelog	2020-08-15 03:26:42.0 -0700
+++ qiskit-terra-0.12.0/debian/changelog	2020-10-17 11:53:16.0 -0700
@@ -1,3 +1,10 @@
+qiskit-terra (0.12.0-2.1) unstable; urgency=medium
+
+  * Non-maintainer upload.
+  * Patch: Support Python 3.9, in test infrastructure. Closes: #972374
+
+ -- Stefano Rivera   Sat, 17 Oct 2020 11:53:16 -0700
+
 qiskit-terra (0.12.0-2) unstable; urgency=medium
 
   * d/patches: backport upstream fix. Closes: #964669
diff -Nru qiskit-terra-0.12.0/debian/patches/python3.9.patch qiskit-terra-0.12.0/debian/patches/python3.9.patch
--- qiskit-terra-0.12.0/debian/patches/python3.9.patch	1969-12-31 16:00:00.0 -0800
+++ qiskit-terra-0.12.0/debian/patches/python3.9.patch	2020-10-17 11:37:15.0 -0700
@@ -0,0 +1,33 @@
+Adapt to the new location of the private _AssertLogsContext class in Python 3.9
+
+Upstream has started to look at 3.9 support in
+https://github.com/Qiskit/qiskit-terra/pull/5189 but haven't fixed this, yet.
+
+Author: Stefano Rivera 
+Bug-Debian: https://bugs.debian.org/972374
+Forwarded: no
+--- a/qiskit/test/utils.py
 b/qiskit/test/utils.py
+@@ -16,9 +16,12 @@
+ 
+ import logging
+ import os
+-import unittest
+ from enum import Enum
+ from itertools import product
++try:
++from unittest._log import _AssertLogsContext
++except ImportError:
++from unittest.case import _AssertLogsContext
+ 
+ from qiskit import __path__ as qiskit_path
+ 
+@@ -68,7 +71,7 @@
+ logger.setLevel(level)
+ 
+ 
+-class _AssertNoLogsContext(unittest.case._AssertLogsContext):
++class _AssertNoLogsContext(_AssertLogsContext):
+ """A context manager used to implement TestCase.assertNoLogs()."""
+ 
+ # pylint: disable=inconsistent-return-statements
diff -Nru qiskit-terra-0.12.0/debian/patches/series qiskit-terra-0.12.0/debian/patches/series
--- qiskit-terra-0.12.0/debian/patches/series	2020-08-15 03:05:52.0 -0700
+++ qiskit-terra-0.12.0/debian/patches/series	2020-10-17 11:32:33.0 -0700
@@ -2,3 +2,4 @@
 0002-Simplify-networkx-dependency-declaration.patch
 0003-Trim-included-package_data-in-setup.py.patch
 0004-Fix-numpy-1.19.0-pulse-error-upstream-backport.patch
+python3.9.patch
-- 
debian-science-maintainers mailing list
debian-science-maintainers@alioth-lists.debian.net
https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/debian-science-maintainers

Bug#972246: numba ftbfs in unstable with python3.9 as supported python3 version

2020-10-16 Thread Stefano Rivera
Looks like it's going to take a while to get Python 3.9 support in the
numba stack:
https://github.com/numba/numba/issues/6345#issuecomment-706722002

This isn't just a little python library, it's a Python JIT for Python,
so I'd believe that supporting 3.9 is non-trivial.

SR

-- 
Stefano Rivera
  http://tumbleweed.org.za/
  +1 415 683 3272

-- 
debian-science-maintainers mailing list
debian-science-maintainers@alioth-lists.debian.net
https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/debian-science-maintainers

Bug#972251: python-line-profiler: diff for NMU version 2.1-2.1

2020-10-16 Thread Stefano Rivera
Control: tags 972251 + patch
Control: tags 972251 + pending

Dear maintainer,

I've prepared an NMU for python-line-profiler (versioned as 2.1-2.1) and
uploaded it to DELAYED/5. Please feel free to tell me if I
should delay it longer.

Regards,

SR
diff -Nru python-line-profiler-2.1/debian/changelog python-line-profiler-2.1/debian/changelog
--- python-line-profiler-2.1/debian/changelog	2018-11-01 02:25:39.0 -0700
+++ python-line-profiler-2.1/debian/changelog	2020-10-15 23:00:41.0 -0700
@@ -1,3 +1,11 @@
+python-line-profiler (2.1-2.1) unstable; urgency=medium
+
+  * Non-maintainer upload.
+  * Patch: Python 3.9 compatibility, assume gettimeofday() exists and takes 2
+arguments. (Closes: #972251)
+
+ -- Stefano Rivera   Thu, 15 Oct 2020 23:00:41 -0700
+
 python-line-profiler (2.1-2) unstable; urgency=medium
 
   * Team upload
diff -Nru python-line-profiler-2.1/debian/patches/py39-compat.diff python-line-profiler-2.1/debian/patches/py39-compat.diff
--- python-line-profiler-2.1/debian/patches/py39-compat.diff	1969-12-31 16:00:00.0 -0800
+++ python-line-profiler-2.1/debian/patches/py39-compat.diff	2020-10-15 23:00:41.0 -0700
@@ -0,0 +1,32 @@
+Description: Python 3.9 dropped gettimeofday configure checks
+ In bpo-38068, cPython assumes gettimeofday exists and takes two arguments.
+ That's a reasonable assumption that we can repeat.
+Author: Stefano Rivera 
+Forwarded: https://github.com/pyutils/line_profiler/pull/31
+Bug-Debian: https://bugs.debian.org/972251
+
+--- a/timers.c
 b/timers.c
+@@ -32,10 +32,6 @@
+ 
+ #else  /* !MS_WINDOWS */
+ 
+-#ifndef HAVE_GETTIMEOFDAY
+-#error "This module requires gettimeofday() on non-Windows platforms!"
+-#endif
+-
+ #if (defined(PYOS_OS2) && defined(PYCC_GCC))
+ #include 
+ #else
+@@ -48,11 +44,7 @@
+ {
+ struct timeval tv;
+ PY_LONG_LONG ret;
+-#ifdef GETTIMEOFDAY_NO_TZ
+-gettimeofday();
+-#else
+ gettimeofday(, (struct timezone *)NULL);
+-#endif
+ ret = tv.tv_sec;
+ ret = ret * 100 + tv.tv_usec;
+ return ret;
diff -Nru python-line-profiler-2.1/debian/patches/series python-line-profiler-2.1/debian/patches/series
--- python-line-profiler-2.1/debian/patches/series	2018-11-01 02:21:39.0 -0700
+++ python-line-profiler-2.1/debian/patches/series	2020-10-15 22:58:51.0 -0700
@@ -1 +1,2 @@
 py37-compat.diff
+py39-compat.diff
-- 
debian-science-maintainers mailing list
debian-science-maintainers@alioth-lists.debian.net
https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/debian-science-maintainers

Bug#972319: python-line-profiler: New upstream releases in new GitHub repo

2020-10-16 Thread Stefano Rivera
Source: python-line-profiler
Severity: normal

The canonical upstream source seems to be
https://github.com/pyutils/line_profiler these days.

You probably want to update your watch file.

3.0.2 is the current latest version.

SR

-- 
debian-science-maintainers mailing list
debian-science-maintainers@alioth-lists.debian.net
https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/debian-science-maintainers

Bug#960010: freefem++: diff for NMU version 3.61.1+dfsg1-5.1

2020-05-17 Thread Stefano Rivera
Control: tags 960010 + patch

Dear maintainer,

I've prepared an NMU for freefem++ (versioned as 3.61.1+dfsg1-5.1). The diff
is attached to this message.

Regards,

SR
diff -Nru freefem++-3.61.1+dfsg1/debian/changelog freefem++-3.61.1+dfsg1/debian/changelog
--- freefem++-3.61.1+dfsg1/debian/changelog	2019-10-05 00:21:24.0 -0700
+++ freefem++-3.61.1+dfsg1/debian/changelog	2020-05-17 13:41:04.0 -0700
@@ -1,3 +1,10 @@
+freefem++ (3.61.1+dfsg1-5.1) unstable; urgency=medium
+
+  * Non-maintainer upload.
+  * Fix FTBFS with gsl 2.6 (Closes: #960010)
+
+ -- Stefano Rivera   Sun, 17 May 2020 13:41:04 -0700
+
 freefem++ (3.61.1+dfsg1-5) unstable; urgency=medium
 
   * Team upload.
diff -Nru freefem++-3.61.1+dfsg1/debian/patches/double-cblas-import.patch freefem++-3.61.1+dfsg1/debian/patches/double-cblas-import.patch
--- freefem++-3.61.1+dfsg1/debian/patches/double-cblas-import.patch	1969-12-31 16:00:00.0 -0800
+++ freefem++-3.61.1+dfsg1/debian/patches/double-cblas-import.patch	2020-05-17 13:21:25.0 -0700
@@ -0,0 +1,25 @@
+Description: Avoid FTBFS with gsl 2.6 by including 2 incompatible cblas headers
+Bug-Debian: https://bugs.debian.org/960010
+Author: Frederic Hecht 
+Origin: upstream, https://github.com/FreeFem/FreeFem-sources/commit/3bfe3eb669c580583e9290474614b45cee52a96c
+
+--- a/src/femlib/MatriceCreuse_tpl.hpp
 b/src/femlib/MatriceCreuse_tpl.hpp
+@@ -12,7 +12,7 @@
+ // test blas 
+ //  on MacOS9 under MWERKS
+ //  cblas_ddot macos-9 is not 
+-#ifdef HAVE_CBLAS_H
++#ifdef HAVE_CBLAS_H_BUG
+ extern "C" {
+ #define FF_VERSION VERSION
+ #undef VERSION
+@@ -21,7 +21,7 @@
+ #define VERSION VERSION
+ }
+ #define WITHBLAS 1
+-#elif HAVE_VECLIB_CBLAS_H
++#elif HAVE_VECLIB_CBLAS_BUG
+ #include  
+ #define WITHBLAS 1
+ #endif  
diff -Nru freefem++-3.61.1+dfsg1/debian/patches/series freefem++-3.61.1+dfsg1/debian/patches/series
--- freefem++-3.61.1+dfsg1/debian/patches/series	2019-10-05 00:21:24.0 -0700
+++ freefem++-3.61.1+dfsg1/debian/patches/series	2020-05-17 13:17:38.0 -0700
@@ -3,3 +3,4 @@
 examples++-load.patch
 src_fflib.patch
 fix_all.edp.patch
+double-cblas-import.patch
-- 
debian-science-maintainers mailing list
debian-science-maintainers@alioth-lists.debian.net
https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/debian-science-maintainers