summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* Fix test that depended on old-style classes in Py2.ll3_new_style_classesStefan Behnel2020-04-191-0/+1
|
* Update migration guide regarding old-style classes with language_level=3.Stefan Behnel2020-04-191-0/+3
|
* Turn plain classes without bases into new-style classes with language_level=3.Stefan Behnel2020-04-193-1/+19
|
* Prepare release of 3.0a2.Stefan Behnel2020-04-191-1/+1
|
* Update changelog.Stefan Behnel2020-04-181-5/+11
|
* Increase warning level for legacy "dict in annotations" typing.Stefan Behnel2020-04-181-2/+2
|
* Make "cimport numpy" without import_array() safer by automatically calling ↵da-woods2020-04-1812-8/+200
| | | | it (GH-3524)
* Make C++ typeid accept specializations of fused types (#3205)Stefan Behnel2020-04-183-1/+72
|\ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | * Potential fix for GH issue #3203 Gets the specialized type if possible from NameNode.analyse_as_type This does introduce a potential new bug: ``` cimport cython just_float = cython.fused_type(float) cdef OK1(just_float x): return just_float in floating cdef fail1(just_float x, floating y): return just_float in floating cdef fail2(floating x): return floating in floating def show(): """ >>> show() True True True True """ print(OK1(1.0)) print(fail1(1.0, 2.0)) print(fail1[float, double](1.0, 2.0)) print(fail2[float](1.0)) ``` fail1 and fail2 work before this patch but fail with it. It isn't clear to me if this should actually be considered a bug. It works in both versions with `cython.floating`, which possibly suggests analyse_as_type in AttributeNode should also be changed * Bring attribute.fused types in line * Removed try-catch * Fix and test "type in fused_type" special-case * Added "analyse_as_specialized_type" * Fixed cpp_operators (handle type is None)
| * Fixed cpp_operators (handle type is None)da-woods2020-04-181-2/+2
| |
| * Added "analyse_as_specialized_type"da-woods2020-04-182-20/+18
| |
| * Fix and test "type in fused_type" special-caseda-woods2020-04-182-0/+36
| |
| * Removed try-catchda-woods2019-10-261-7/+5
| |
| * Bring attribute.fused types in lineda-woods2019-10-242-4/+21
| |
| * Potential fix for GH issue #3203da-woods2019-10-242-0/+22
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Gets the specialized type if possible from NameNode.analyse_as_type This does introduce a potential new bug: ``` cimport cython just_float = cython.fused_type(float) cdef OK1(just_float x): return just_float in floating cdef fail1(just_float x, floating y): return just_float in floating cdef fail2(floating x): return floating in floating def show(): """ >>> show() True True True True """ print(OK1(1.0)) print(fail1(1.0, 2.0)) print(fail1[float, double](1.0, 2.0)) print(fail2[float](1.0)) ``` fail1 and fail2 work before this patch but fail with it. It isn't clear to me if this should actually be considered a bug. It works in both versions with `cython.floating`, which possibly suggests analyse_as_type in AttributeNode should also be changed
* | Make CodeWriter inherit from ExpressionWriter in order to support all kinds ↵Stefan Behnel2020-04-183-88/+63
| | | | | | | | of expressions without duplicating code. (GH-3528)
* | Start writing up some help for users who want to migrate from Cython 0.29.x ↵Stefan Behnel2020-04-182-0/+121
| | | | | | | | to Cython 3.0.
* | Add some missing branch hints.Stefan Behnel2020-04-181-6/+6
| |
* | Merge branch '0.29.x'Stefan Behnel2020-04-171-4/+5
|\ \
| * | Update changelog.Stefan Behnel2020-04-171-0/+4
| | |
* | | Merge branch '0.29.x'Stefan Behnel2020-04-171-0/+1
|\ \ \ | |/ /
| * | Clear "self.index_temps" after cleanup to avoid keeping dangling temporary ↵Stefan Behnel2020-04-171-0/+1
| | | | | | | | | | | | state.
| * | Cleanup more generator temps (GH-3522)Stefan Behnel2020-04-171-11/+12
| | | | | | | | | | | | | | | * Fixed indexing temps for non-python objects. * Moved cleanup into release_temps since the temps survived into the result_code.
* | | Add comment and make sure the type's (empty) list of index temps cannot be ↵Stefan Behnel2020-04-171-1/+2
| | | | | | | | | | | | changed by accident.
* | | Simplify some redundant code by calling the obvious helper function instead.Stefan Behnel2020-04-171-6/+4
| | |
* | | Cleanup more generator temps (GH-3522)da-woods2020-04-171-11/+12
| | | | | | | | | | | | | | | | | | * Fixed indexing temps for non-python objects. * Release another temp in PyMethodCallNode. Was causing warnings in generators_py. * Moved cleanup into release_temps since the temps survived into the result_code.
* | | Merge branch '0.29.x'Stefan Behnel2020-04-170-0/+0
|\ \ \ | |/ /
| * | Release temps used for buffer indexing after use (GH-3517)da-woods2020-04-151-3/+10
| | | | | | | | | Closes https://github.com/cython/cython/issues/3430
* | | Update changelog.Stefan Behnel2020-04-171-0/+4
| | |
* | | Convert Py2 str to unicode directly when building fused function signature.Stefan Behnel2020-04-171-6/+6
| | |
* | | Remove accidentally duplicated code line.Stefan Behnel2020-04-171-1/+0
| | |
* | | Avoid an unnecessary conditional branch.Stefan Behnel2020-04-161-4/+4
| | |
* | | Prevent fused dispatcher optional arguments being overwritten (GH-3519)da-woods2020-04-153-17/+26
| | | | | | | | | | | | | | | | | | | | | | | | | | | * Prevent fused dispatcher optional arguments being overwritten If the 5th argument of the regular functions was optional it'd overwrite the _fused_sigindex default of the dispatcher, causing type errors at runtime. closes https://github.com/cython/cython/issues/3511
* | | Release temps used for buffer indexing after use (GH-3517)da-woods2020-04-151-3/+10
| | | | | | | | | Closes https://github.com/cython/cython/issues/3430
* | | Mention "cython_freeze" in the embedding docs page.Stefan Behnel2020-04-141-1/+3
| | |
* | | Add a stub documentation page about embedding Cython modules in C/C++ ↵Stefan Behnel2020-04-142-1/+28
| | | | | | | | | | | | applications.
* | | Update changelog.Stefan Behnel2020-04-131-0/+5
| | |
* | | Drop unused code-paths associated with "if cython.compiled" early (GH-3507)da-woods2020-04-133-3/+54
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This allows things to work like: # in pxd file from libc.math cimport sin # in py file if not cython.compiled: from math import sin # previously failed with cython compile # error because it was assigning to a cdef name This seems worthwhile because it makes it easier to write code that re-assigns cdef names and so works in both modes.
* | | Reduce some overhead in the fused function selection code.Stefan Behnel2020-04-131-13/+12
| | |
* | | Merge branch '0.29.x'Stefan Behnel2020-04-132-11/+11
|\ \ \ | |/ /
| * | Prevent calling PyObject_GC_Track() in __Pyx_CyFunction_New() when ↵Stefan Behnel2020-04-133-34/+60
| | | | | | | | | | | | | | | | | | | | | | | | instantiating a fused function before the object fields of the fused function subtype are fully initialised. See https://bugs.python.org/issue38392 Closes GH-3215. Supersedes GH-3216.
| * | Fix a reference leak when PyList_Append() fails.Stefan Behnel2020-04-131-6/+7
| | |
| * | Fix error label name.Stefan Behnel2020-04-131-1/+1
| | |
| * | Update changelog.Stefan Behnel2020-04-121-0/+1
| | |
* | | Merge branch '0.29.x'Stefan Behnel2020-04-122-3/+5
|\ \ \ | |/ /
| * | Make travis install script fail on intermediate errors.Stefan Behnel2020-04-121-6/+6
| | |
| * | Debug conda issues on travis.Stefan Behnel2020-04-121-3/+5
| | |
| * | Add missing "@cython.binding" decorator for Python code.Stefan Behnel2020-04-122-0/+4
| | |
| * | Fix missing error handling in non-CPython code.Stefan Behnel2020-04-121-1/+1
| | |
* | | Sprinkle lots of branch prediction hints in places where exceptions are ↵Stefan Behnel2020-04-125-39/+40
| | | | | | | | | | | | being handled or otherwise exceptional cases occur.
* | | Simplify some decref-set patterns using the __Pyx_Py_XDECREF_SET() macro.Stefan Behnel2020-04-123-44/+15
| | |