summaryrefslogtreecommitdiff
path: root/tools
Commit message (Collapse)AuthorAgeFilesLines
* ENH: Vendorize tempita.Charles Harris2016-09-305-16/+1560
| | | | | | | | | | | Copy tempita from https://github.com/gjhiggins/tempita with a few changes to remove the six dependency. This removes a dependency on Cython's tempita, which is not to be relied on as it is considered by the Cython folks to be an implementation detail. The package is named npy_tempita so that it can always be distinguished from an installed tempita. The cythonize.py script is changed to use the vendorized tempita.
* MAINT: Add Tempita to randint helpersgfyoung2016-09-201-12/+45
| | | | | | Refactors the randint helpers to use a Tempita template. This will reduce technical debt in the long run.
* BUG: Assert fromfile ending earlier in pyx processinggfyoung2016-09-191-1/+1
|
* DOC: change Numpy to NumPy in remaining filesPierre de Buyl2016-09-064-8/+8
| | | | the files in doc/ and numpy/ were covered in previous commits
* TST: Make one 3.5 test run the full test suit.Sebastian Berg2016-09-021-1/+5
|
* Avoid NPY_ARRAY_F_CONTIGUOUS for numpy < 1.7Jan David Mol2016-06-181-1/+6
| | | numpy API < 1.7 does not support NPY_ARRAY_F_CONTIGUOUS.or PyArray_IS_F_CONTIGUOUS. Supporting old APIs is useful as users tend to ship modern numpy.i files to compile against possibly older numpy installations.
* Fix a false positive OverflowError in Python 3.x when value above 0x7FFFFFF ↵Mathieu Lamarre2016-04-221-7/+14
| | | | | | | | | | | | are passed to a function accepting "unsigned int". This a port of a fix in pyprimtype.swg from which several code snippets where copy pasted into swig/pyfragments.swg. Please see SWIG changes log (2015-12-23) for more details: http://www.swig.org/Release/CHANGES.current 2015-12-23: ahnolds [Python] Fixes for conversion of signed and unsigned integer types ...
* DOC: Fix more typos in docs and comments.Dongjoon Hyun2016-02-252-3/+3
|
* TST: test installing from sdist on TravisCI.Ralf Gommers2016-01-271-1/+14
|
* BUG: fix TravisCI test issues when using setuptools unconditionally.Ralf Gommers2016-01-161-0/+1
| | | | Also remove all mentions of setupegg.py from the documentation.
* TST: make pointer-arith error fatalJulian Taylor2016-01-041-1/+2
| | | | matches windows compiler behavior
* TST: Increase the allowable warning count to 1 for i386 debug test.Charles Harris2016-01-021-2/+4
| | | | | | | | | | | | Cython generated C code contains the number '-2147483648L', which leads to a warning on 32 bit platforms: "Warning: this decimal constant is unsigned only in ISO C90" See the discussion at http://stackoverflow.com/questions/9941261/ The compiled code seems to run correctly despite the warning and if there are problems, they should turn up in the nose testing.
* MAINT: Remove commented out code blocksgfyoung2015-12-202-19/+0
|
* DOC: Use print only as function when print_function is imported from __future__gfyoung2015-12-192-4/+4
| | | | Closes gh-6863.
* STY: Style fixes for .travis.yml and travis-upload-wheel.shCharles Harris2015-12-161-5/+7
| | | | | | | * shell script style fixes inspired by google shell style guide https://google.github.io/styleguide/shell.xml * .travis.yml longline breaking tested with http://yaml-online-parser.appspot.com/
* TST: Clean up travis-test and make it work with current travis ci.Charles Harris2015-12-161-52/+79
| | | | | | | | | | Travis ci is migrating to GCI and the 32 bit tests broke in the process. This cleans up the tools/travis-test script, fixes it for current travis, and changes the 32 bit test to use python 2.7 in order to turn up errors involving python long integers. In preparation for dropping Python 3.2 and 3.3, the USE_DEBUG test is run in the travis ci trusty beta so that python3 defaults to 3.4.
* CI: run benchmark suite in travis-CIPauli Virtanen2015-12-141-0/+11
| | | | | This should ensure the suite stays in working condition, not to produce reliable timing information.
* ENH deploy dev wheels to rackspaceOlivier Grisel2015-12-102-3/+16
|
* MAINT: Include from __future__ boilerplate in some files missing it.Charles Harris2015-12-051-0/+2
| | | | | Some newer *.py files are missing the `from __future__` boilerplate that helps assure Python2 and Python3 compatibility.
* Merge pull request #6726 from ↵Charles Harris2015-11-261-1/+1
|\ | | | | | | | | charris/fix-bugs-exposed-by-relaxed-stride-rollback Fix bugs exposed by relaxed stride rollback
| * TST: Fix travis-ci test for numpy wheels.Charles Harris2015-11-251-1/+1
| |
* | MAINT: enable Werror=vla in travisJulian Taylor2015-11-241-4/+4
|/ | | | avoids issues with stone age compilers like MSVC
* BUG: Fix swig make_fortran function.Charles Harris2015-11-041-1/+1
| | | | | | | | The function was calling PyArray_FromArray with NPY_FORTRANORDER instead of NPY_ARRAY_F_CONTIGUOUS. The first is of type NPY_ORDER and the second is a flag. Closes #6618. [ci skip]
* BUG: Fix use of PyArray_ISFORTRAN in numpy.i.Charles Harris2015-11-011-1/+1
| | | | | | | | | | PyArray_ISFORTRAN was used to implement array_is_fortran in numpy.i when what was wanted was PyArray_IS_F_CONTIGUOUS. The difference is that PyArray_ISFORTRAN will return False if the array is c_contiguous. Previous to relaxed stride checking this did not matter, but currently arrays with ndim > 1 may be both C and Fortran contiguous and that results in errors when PyArray_ISFORTRAN is mistakenly used to check for Fortran contiguity.
* typo in numpy.ijschueller2015-09-021-1/+1
|
* MAINT,TST: Remove Bento from the travis ci testing.Charles Harris2015-08-191-35/+1
|
* BLD: Speed up up Travis jobs by running on container-based infrastructureAlex Willmer2015-08-051-8/+1
| | | | | | | | | | | All jobs currently run on Travis's legacy infrastructure - which supports sudo. The newer container-based infrastructure is faster, but doesn't allow sudo. This patch - sets sudo=false for all jobs, except the chroot job - uses Travis's apt addon to install all packages - installs eatmydata for all jobs to reduce disk IO - removes the tmpfs workaround for chroot builds
* TST: fail tests on compiler warningsJulian Taylor2015-07-031-2/+11
|
* New typemap for in-place arrays of arbitrary number of dimensions:Tom Krauss2015-05-258-26/+394
| | | | | (DATA_TYPE* INPLACE_ARRAY_FLAT, DIM_TYPE DIM_FLAT) Added unittests, updated documentation.
* MAINT: switch waf download to https://github.com/numpy/numpy-vendorRalf Gommers2015-05-071-1/+1
| | | | Follow-up of gh-5829, which got merged too early.
* Merge pull request #5829 from rgommers/bento-ciCharles Harris2015-05-031-1/+1
|\ | | | | TST: re-enable TravisCI testing with Bento.
| * TST: re-enable TravisCI testing with Bento.Ralf Gommers2015-05-031-1/+1
| | | | | | | | | | Disabling was done in gh-5708, due to the Waf download site being down for a while.
* | TEST: add module to test installed scriptsMatthew Brett2015-04-301-3/+10
|/ | | | | | | | | Module tests whether we can run f2py and return correct version. Skip this test when running in-place (we don't install f2py in that case). Use our own virtualenvs in travis-ci to avoid picking up travis' numpy.
* DOC: put a copy of the text in LICENSE.txt into tools/swig/numpy.iRalf Gommers2015-03-181-0/+34
| | | | | | | | | Reason: numpy.i is supposed to be copied, not used from within an installed Numpy version. Closes gh-5690 [ci skip]
* Correct exit code if number of errors % 256 == 0Chris Lamb2015-02-157-7/+7
| | | | | | | For example, we had 256 errors (etc.) our process will exit with a successful error code which is incorrect and/or misleading. Signed-off-by: Chris Lamb <chris@chris-lamb.co.uk>
* Use correct #ifdef __cplusplus guard to include std_complex.iMichele Vallisneri2015-02-131-2/+2
|
* ENH: added std::complex support to numpy.iGlen Mabey2014-11-077-12/+333
| | | | | | numpy.i now includes ready-made typemaps for std::complex<float> and std::complex<double> . Tests were added to testArray using a newly defined ArrayZ class.
* ENH: support parallel compilation of extensionsJulian Taylor2014-10-281-0/+3
| | | | | | | | | | | | | | Allow extensions using numpy.distutils to compile in parallel. By passing `--jobs=n` or `-j n` to `setup.py build` the compilation of extensions is now performed in `n` parallel processes. Additionally the environment variable NPY_NUM_BUILD_JOBS is used as the default value, if its unset the default is serial compilation. The parallelization is limited to within the files of an extension, so only numpy multiarraymodule really profits but its still a nice improvement when you have 2-4 cores. Unfortunately Cython will not profit at all as it tends to build one module per file.
* Merge pull request #5149 from charris/fix-ma-docs-for-OO-optionJulian Taylor2014-10-041-0/+5
|\ | | | | | | BUG: Make numpy import when run with Python flag '-OO
| * TST: Add test with python flag -OO.Charles Harris2014-10-041-0/+5
| |
* | MAINT: Update waf to 1.7.16Charles Harris2014-09-041-3/+3
| |
* | Merge pull request #4671 from charris/prepare-1.10-develCharles Harris2014-07-061-1/+1
|\ \ | | | | | | MAINT: start 1.10-devel.
| * | MAINT: Make NPY_RELAXED_STRIDE_CHECKING true by default.Charles Harris2014-05-061-1/+1
| | |
* | | BUG: disable garbage collector during memory allocation hookJulian Taylor2014-07-051-0/+10
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When a cython object contains numpy arrays the pure python allocation hook can trigger during cythons __dealloc__ method and trigger a second deletion of the object currently being deleted. To minimize the probabily that that happens disable the garbage collector during the hook. As this involves python calls it is still possible that a double delete occurs but chances are lowered, a proper solution would be C only hook like python 3.4 tracemalloc module. Closes gh-4834
* | | Fix typemap for Fortran ordered array inputAdam Reeve2014-06-161-3/+3
| |/ |/| | | | | | | | | The typemaps with size parameters after the array pointer were correct, but the typemaps with size parameters before the array pointer created arrays with C ordering.
* | BLD: add a simple python file to build cpucaps.dllJulian Taylor2014-06-042-0/+17
| | | | | | | | | | | | using scons seems to fail with wine 1.6, but one only needs to run three commands so its simpler to just put these into a script instead of trying to debug scons.
* | BUG: Fix missing fragment in ARGOUTVIEWM typemaps.David Froger2014-05-261-18/+18
| | | | | | | | | | | | | | | | | | | | The argout typemaps of all the ARGOUTVIEWM typemap suites make use of the free_cap function, defined in the NumPy_Utilities fragment, which was not listed in the fragment argument of the typemaps. As a result, the free_cap function is not included in the source generated code by SWIG, and the code do not compile (unless another typemap including the missing fragment has been used).
* | Merge pull request #4662 from matthew-brett/cythonize-for-wheelsRalf Gommers2014-05-111-1/+7
|\ \ | | | | | | BUG: cythonize sources for wheel build
| * | TST: add travis test for wheel build and installMatthew Brett2014-05-041-1/+7
| | | | | | | | | | | | | | | Add travis build that builds and installs wheel, runs tests from installed wheel.
* | | TST: update package list for chroot buildJulian Taylor2014-05-091-1/+2
| |/ |/| | | | | travis boxes are again badly out of date