pybind/pybind11
1
0
Fork 0
mirror of https://github.com/pybind/pybind11.git synced 2026-03-14 20:27:47 +00:00
Code Issues Packages Projects Releases Wiki Activity

Expand float and complex strict mode to allow ints and ints/float (for PEP 484 compatibility). (#5879)

Browse Source 
* init

Signed-off-by: Michael Carlstrom <rmc@carlstrom.com>

* Add constexpr to is_floating_point check

This is known at compile time so it can be constexpr

* Allow noconvert float to accept int

* Update noconvert documentation

* Allow noconvert complex to accept int and float

* Add complex strict test

* style: pre-commit fixes

* Update unit tests so int, becomes double.

* style: pre-commit fixes

* remove if (constexpr)

Signed-off-by: Michael Carlstrom <rmc@carlstrom.com>

* fix spelling error

* bump order in #else

* Switch order in c++11 only section

Signed-off-by: Michael Carlstrom <rmc@carlstrom.com>

* ci: trigger build

* ci: trigger build

* Allow casting from float to int

The int type caster allows anything that implements __int__ with explicit exception of the python float. I can't see any reason for this.
This modifies the int casting behaviour to accept a float.
If the argument is marked as noconvert() it will only accept int.

* tests for py::float into int

* Update complex_cast tests

* Add SupportsIndex to int and float

* style: pre-commit fixes

* fix assert

* Update docs to mention other conversions

* fix pypy __index__ problems

* style: pre-commit fixes

* extract out PyLong_AsLong __index__ deprecation

Signed-off-by: Michael Carlstrom <rmc@carlstrom.com>

* style: pre-commit fixes

* Add back env.deprecated_call

Signed-off-by: Michael Carlstrom <rmc@carlstrom.com>

* remove note

Signed-off-by: Michael Carlstrom <rmc@carlstrom.com>

* remove untrue comment

Signed-off-by: Michael Carlstrom <rmc@carlstrom.com>

* fix noconvert_args

Signed-off-by: Michael Carlstrom <rmc@carlstrom.com>

* resolve error

Signed-off-by: Michael Carlstrom <rmc@carlstrom.com>

* Add comment

Signed-off-by: Michael Carlstrom <rmc@carlstrom.com>

* [skip ci]

tests: Add overload resolution test for float/int breaking change

Add test_overload_resolution_float_int() to explicitly test the breaking
change where int arguments now match float overloads when registered first.

The existing tests verify conversion behavior (int -> float, int/float -> complex)
but do not test overload resolution when both float and int overloads exist.
This test fills that gap by:

- Testing that float overload registered before int overload matches int(42)
- Testing strict mode (noconvert) overload resolution breaking change
- Testing complex overload resolution with int/float/complex overloads
- Documenting the breaking change explicitly

This complements existing tests which verify 'can it convert?' by testing
'which overload wins when multiple can convert?'

* Add test to verify that custom __index__ objects (not PyLong) work correctly with complex conversion. These should be consistent across CPython, PyPy, and GraalPy.

* Improve comment clarity for PyPy __index__ handling

Replace cryptic 'So: PYBIND11_INDEX_CHECK(src.ptr())' comment with
clearer explanation of the logic:

- Explains that we need to call PyNumber_Index explicitly on PyPy
  for non-PyLong objects
- Clarifies the relationship to the outer condition: when convert
  is false, we only reach this point if PYBIND11_INDEX_CHECK passed
  above

This makes the code more maintainable and easier to understand
during review.

* Undo inconsequential change to regex in test_enum.py

During merge, HEAD's regex pattern was kept, but master's version is preferred.
The order of ` ` and `\|` in the character class is arbitrary. Keep master's order
(already fixed in PR #5891; sorry I missed looking back here when working on 5891).

* test_methods_and_attributes.py: Restore existing `m.overload_order(1.1)` call and clearly explain the behavior change.

* Reject float → int conversion even in convert mode

Enabling implicit float → int conversion in convert mode causes
silent truncation (e.g., 1.9 → 1). This is dangerous because:

1. It's implicit - users don't expect truncation when calling functions
2. It's silent - no warning or error
3. It can hide bugs - precision loss is hard to detect

This change restores the explicit rejection of PyFloat_Check for integer
casters, even in convert mode. This is more in line with Python's behavior
where int(1.9) must be explicit.

Note that the int → float conversion in noconvert mode is preserved,
as that's a safe widening conversion.

* Revert test changes that sidestepped implicit float→int conversion

This reverts all test modifications that were made to accommodate
implicit float→int conversion in convert mode. With the production
code change that explicitly rejects float→int conversion even in
convert mode, these test workarounds are no longer needed.

Changes reverted:
- test_builtin_casters.py: Restored cant_convert(3.14159) and
  np.float32 conversion with deprecated_call wrapper
- test_custom_type_casters.py: Restored TypeError expectation for
  m.ints_preferred(4.0)
- test_methods_and_attributes.py: Restored TypeError expectation
  for m.overload_order(1.1)
- test_stl.py: Restored float literals (2.0) that were replaced with
  strings to avoid conversion
- test_factory_constructors.py: Restored original constructor calls
  that were modified to avoid float→int conversion

Also removes the unused avoid_PyLong_AsLong_deprecation fixture
and related TypeVar imports, as all uses were removed.

* Replace env.deprecated_call() with pytest.deprecated_call()

The env.deprecated_call() function was removed, but two test cases
still reference it. Replace with pytest.deprecated_call(), which is
the standard pytest context manager for handling deprecation warnings.

Since we already require pytest>=6 (see tests/requirements.txt), the
compatibility function is obsolete and pytest.deprecated_call() is
available.

* Update test expectations for swapped NoisyAlloc overloads

PR 5879 swapped the order of NoisyAlloc constructor overloads:
- (int i, double) is now placement new (comes first)
- (double d, double) is now factory pointer (comes second)

This swap is necessary because pybind11 tries overloads in order
until one matches. With int → float conversion now allowed:

- create_and_destroy(4, 0.5): Without the swap, (double d, double)
  would match first (since int → double conversion is allowed),
  bypassing the more specific (int i, double) overload. With the
  swap, (int i, double) matches first (exact match), which is
  correct.

- create_and_destroy(3.5, 4.5): (int i, double) fails (float → int
  is rejected), then (double d, double) matches, which is correct.

The swap ensures exact int matches are preferred over double matches
when an int is provided, which is the expected overload resolution
behavior.

Update the test expectations to match the new overload resolution
order.

* Resolve clang-tidy error:

/__w/pybind11/pybind11/include/pybind11/cast.h:253:46: error: repeated branch body in conditional chain [bugprone-branch-clone,-warnings-as-errors]
  253 |         } else if (PyFloat_Check(src.ptr())) {
      |                                              ^
/__w/pybind11/pybind11/include/pybind11/cast.h:258:10: note: end of the original
  258 |         } else if (convert || PYBIND11_LONG_CHECK(src.ptr()) || PYBIND11_INDEX_CHECK(src.ptr())) {
      |          ^
/__w/pybind11/pybind11/include/pybind11/cast.h:283:16: note: clone 1 starts here
  283 |         } else {
      |                ^

* Add test coverage for __index__ and __int__ edge cases: incorrectly returning float

These tests ensure that:
- Invalid return types (floats) are properly rejected
- The fallback from __index__ to __int__ works correctly in convert mode
- noconvert mode correctly prevents fallback when __index__ fails

* Minor comment-only changes: add PR number, for easy future reference

* Ensure we are not leaking a Python error is something is wrong elsewhere (e.g. UB, or bug in Python beta testing).

See also: https://github.com/pybind/pybind11/pull/5879#issuecomment-3521099331

* [skip ci] Bump PYBIND11_INTERNALS_VERSION to 12 (for PRs 5879, 5887, 5960)

---------

Signed-off-by: Michael Carlstrom <rmc@carlstrom.com>
Co-authored-by: gentlegiantJGC <gentlegiantJGC@users.noreply.github.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Ralf W. Grosse-Kunstleve <rgrossekunst@nvidia.com>
This commit is contained in:
Michael Carlstrom
2026-02-16 23:00:01 -08:00
committed by GitHub
parent 2448bc5853
commit e8e8d6ab22
11 changed files with 256 additions and 63 deletions
Download Patch File Download Diff File Expand all files Collapse all files

39
docs/advanced/functions.rst
View File

@@ -437,10 +437,10 @@ Certain argument types may support conversion from one type to another. Some
examples of conversions are:
* :ref:`implicit_conversions` declared using ``py::implicitly_convertible<A,B>()``
* Calling a method accepting a double with an integer argument
* Calling a ``std::complex<float>`` argument with a non-complex python type
(for example, with a float). (Requires the optional ``pybind11/complex.h``
header).
* Passing an argument that implements ``__float__`` or ``__index__`` to ``float`` or ``double``.
* Passing an argument that implements ``__int__`` or ``__index__`` to ``int``.
* Passing an argument that implements ``__complex__``, ``__float__``, or ``__index__`` to ``std::complex<float>``.
(Requires the optional ``pybind11/complex.h`` header).
* Calling a function taking an Eigen matrix reference with a numpy array of the
wrong type or of an incompatible data layout. (Requires the optional
``pybind11/eigen.h`` header).
@@ -452,24 +452,37 @@ object, such as:
.. code-block:: cpp
m.def("floats_only", [](double f) { return 0.5 * f; }, py::arg("f").noconvert());
m.def("floats_preferred", [](double f) { return 0.5 * f; }, py::arg("f"));
m.def("supports_float", [](double f) { return 0.5 * f; }, py::arg("f"));
m.def("only_float", [](double f) { return 0.5 * f; }, py::arg("f").noconvert());
Attempting the call the second function (the one without ``.noconvert()``) with
an integer will succeed, but attempting to call the ``.noconvert()`` version
will fail with a ``TypeError``:
``supports_float`` will accept any argument that implements ``__float__`` or ``__index__``.
``only_float`` will only accept a float or int argument. Anything else will fail with a ``TypeError``:
.. note::
The noconvert behaviour of float, double and complex has changed to match PEP 484.
A float/double argument marked noconvert will accept float or int.
A std::complex<float> argument will accept complex, float or int.
.. code-block:: pycon
>>> floats_preferred(4)
class MyFloat:
def __init__(self, value: float) -> None:
self._value = float(value)
def __repr__(self) -> str:
return f"MyFloat({self._value})"
def __float__(self) -> float:
return self._value
>>> supports_float(MyFloat(4))
2.0
>>> floats_only(4)
>>> only_float(MyFloat(4))
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
TypeError: floats_only(): incompatible function arguments. The following argument types are supported:
TypeError: only_float(): incompatible function arguments. The following argument types are supported:
1. (f: float) -> float
Invoked with: 4
Invoked with: MyFloat(4)
You may, of course, combine this with the :var:`_a` shorthand notation (see
:ref:`keyword_args`) and/or :ref:`default_args`. It is also permitted to omit

27
include/pybind11/cast.h
View File

@@ -244,29 +244,28 @@ public:
return false;
}
#if !defined(PYPY_VERSION)
auto index_check = [](PyObject *o) { return PyIndex_Check(o); };
#else
// In PyPy 7.3.3, `PyIndex_Check` is implemented by calling `__index__`,
// while CPython only considers the existence of `nb_index`/`__index__`.
auto index_check = [](PyObject *o) { return hasattr(o, "__index__"); };
#endif
if (std::is_floating_point<T>::value) {
if (convert || PyFloat_Check(src.ptr())) {
if (convert || PyFloat_Check(src.ptr()) || PYBIND11_LONG_CHECK(src.ptr())) {
py_value = (py_type) PyFloat_AsDouble(src.ptr());
} else {
return false;
}
} else if (PyFloat_Check(src.ptr())
|| (!convert && !PYBIND11_LONG_CHECK(src.ptr()) && !index_check(src.ptr()))) {
|| !(convert || PYBIND11_LONG_CHECK(src.ptr())
|| PYBIND11_INDEX_CHECK(src.ptr()))) {
// Explicitly reject float → int conversion even in convert mode.
// This prevents silent truncation (e.g., 1.9 → 1).
// Only int → float conversion is allowed (widening, no precision loss).
// Also reject if none of the conversion conditions are met.
return false;
} else {
handle src_or_index = src;
// PyPy: 7.3.7's 3.8 does not implement PyLong_*'s __index__ calls.
#if defined(PYPY_VERSION)
object index;
if (!PYBIND11_LONG_CHECK(src.ptr())) { // So: index_check(src.ptr())
// If not a PyLong, we need to call PyNumber_Index explicitly on PyPy.
// When convert is false, we only reach here if PYBIND11_INDEX_CHECK passed above.
if (!PYBIND11_LONG_CHECK(src.ptr())) {
index = reinterpret_steal<object>(PyNumber_Index(src.ptr()));
if (!index) {
PyErr_Clear();
@@ -286,8 +285,10 @@ public:
}
}
// Python API reported an error
bool py_err = py_value == (py_type) -1 && PyErr_Occurred();
bool py_err = (PyErr_Occurred() != nullptr);
if (py_err) {
assert(py_value == static_cast<py_type>(-1));
}
// Check to see if the conversion is valid (integers should match exactly)
// Signed/unsigned checks happen elsewhere

4
include/pybind11/complex.h
View File

@@ -51,7 +51,9 @@ public:
if (!src) {
return false;
}
if (!convert && !PyComplex_Check(src.ptr())) {
if (!convert
&& !(PyComplex_Check(src.ptr()) || PyFloat_Check(src.ptr())
|| PYBIND11_LONG_CHECK(src.ptr()))) {
return false;
}
handle src_or_index = src;

6
include/pybind11/detail/internals.h
View File

@@ -39,11 +39,11 @@
/// further ABI-incompatible changes may be made before the ABI is officially
/// changed to the new version.
#ifndef PYBIND11_INTERNALS_VERSION
# define PYBIND11_INTERNALS_VERSION 11
# define PYBIND11_INTERNALS_VERSION 12
#endif
#if PYBIND11_INTERNALS_VERSION < 11
# error "PYBIND11_INTERNALS_VERSION 11 is the minimum for all platforms for pybind11v3."
#if PYBIND11_INTERNALS_VERSION < 12
# error "PYBIND11_INTERNALS_VERSION 12 is the minimum for all platforms for pybind11 v3.1.0"
#endif
PYBIND11_NAMESPACE_BEGIN(PYBIND11_NAMESPACE)

25
tests/test_builtin_casters.cpp
View File

@@ -363,9 +363,34 @@ TEST_SUBMODULE(builtin_casters, m) {
m.def("complex_cast", [](float x) { return "{}"_s.format(x); });
m.def("complex_cast",
[](std::complex<float> x) { return "({}, {})"_s.format(x.real(), x.imag()); });
m.def(
"complex_cast_strict",
[](std::complex<float> x) { return "({}, {})"_s.format(x.real(), x.imag()); },
py::arg{}.noconvert());
m.def("complex_convert", [](std::complex<float> x) { return x; });
m.def("complex_noconvert", [](std::complex<float> x) { return x; }, py::arg{}.noconvert());
// test_overload_resolution_float_int
// Test that float overload registered before int overload gets selected when passing int
// This documents the breaking change: int can now match float in strict mode
m.def("overload_resolution_test", [](float x) { return "float: " + std::to_string(x); });
m.def("overload_resolution_test", [](int x) { return "int: " + std::to_string(x); });
// Test with noconvert (strict mode) - this is the key breaking change
m.def(
"overload_resolution_strict",
[](float x) { return "float_strict: " + std::to_string(x); },
py::arg{}.noconvert());
m.def("overload_resolution_strict", [](int x) { return "int_strict: " + std::to_string(x); });
// Test complex overload resolution: complex registered before float/int
m.def("overload_resolution_complex", [](std::complex<float> x) {
return "complex: (" + std::to_string(x.real()) + ", " + std::to_string(x.imag()) + ")";
});
m.def("overload_resolution_complex", [](float x) { return "float: " + std::to_string(x); });
m.def("overload_resolution_complex", [](int x) { return "int: " + std::to_string(x); });
// test int vs. long (Python 2)
m.def("int_cast", []() { return 42; });
m.def("long_cast", []() { return (long) 42; });

169
tests/test_builtin_casters.py
View File

@@ -315,6 +315,7 @@ def test_int_convert(doc):
# Before Python 3.8, `PyLong_AsLong` does not pick up on `obj.__index__`,
# but pybind11 "backports" this behavior.
assert convert(Index()) == 42
assert isinstance(convert(Index()), int)
assert noconvert(Index()) == 42
assert convert(IntAndIndex()) == 0 # Fishy; `int(DoubleThought)` == 42
assert noconvert(IntAndIndex()) == 0
@@ -323,6 +324,50 @@ def test_int_convert(doc):
assert convert(RaisingValueErrorOnIndex()) == 42
requires_conversion(RaisingValueErrorOnIndex())
class IndexReturnsFloat:
def __index__(self):
return 3.14 # noqa: PLE0305 Wrong: should return int
class IntReturnsFloat:
def __int__(self):
return 3.14 # Wrong: should return int
class IndexFloatIntInt:
def __index__(self):
return 3.14 # noqa: PLE0305 Wrong: should return int
def __int__(self):
return 42 # Correct: returns int
class IndexIntIntFloat:
def __index__(self):
return 42 # Correct: returns int
def __int__(self):
return 3.14 # Wrong: should return int
class IndexFloatIntFloat:
def __index__(self):
return 3.14 # noqa: PLE0305 Wrong: should return int
def __int__(self):
return 2.71 # Wrong: should return int
cant_convert(IndexReturnsFloat())
requires_conversion(IndexReturnsFloat())
cant_convert(IntReturnsFloat())
requires_conversion(IntReturnsFloat())
assert convert(IndexFloatIntInt()) == 42 # convert: __index__ fails, uses __int__
requires_conversion(IndexFloatIntInt()) # noconvert: __index__ fails, no fallback
assert convert(IndexIntIntFloat()) == 42 # convert: __index__ succeeds
assert noconvert(IndexIntIntFloat()) == 42 # noconvert: __index__ succeeds
cant_convert(IndexFloatIntFloat()) # convert mode rejects (both fail)
requires_conversion(IndexFloatIntFloat()) # noconvert mode also rejects
def test_float_convert(doc):
class Int:
@@ -356,7 +401,7 @@ def test_float_convert(doc):
assert pytest.approx(convert(Index())) == -7.0
assert isinstance(convert(Float()), float)
assert pytest.approx(convert(3)) == 3.0
requires_conversion(3)
assert pytest.approx(noconvert(3)) == 3.0
cant_convert(Int())
@@ -505,6 +550,11 @@ def test_complex_cast(doc):
assert m.complex_cast(Complex()) == "(5.0, 4.0)"
assert m.complex_cast(2j) == "(0.0, 2.0)"
assert m.complex_cast_strict(1) == "(1.0, 0.0)"
assert m.complex_cast_strict(3.0) == "(3.0, 0.0)"
assert m.complex_cast_strict(complex(5, 4)) == "(5.0, 4.0)"
assert m.complex_cast_strict(2j) == "(0.0, 2.0)"
convert, noconvert = m.complex_convert, m.complex_noconvert
def requires_conversion(v):
@@ -529,14 +579,127 @@ def test_complex_cast(doc):
assert convert(Index()) == 1
assert isinstance(convert(Index()), complex)
requires_conversion(1)
requires_conversion(2.0)
assert noconvert(1) == 1.0
assert noconvert(2.0) == 2.0
assert noconvert(1 + 5j) == 1.0 + 5.0j
requires_conversion(Complex())
requires_conversion(Float())
requires_conversion(Index())
def test_complex_index_handling():
"""
Test __index__ handling in complex caster (added with PR #5879).
This test verifies that custom __index__ objects (not PyLong) work correctly
with complex conversion. The behavior should be consistent across CPython,
PyPy, and GraalPy.
- Custom __index__ objects work with convert (non-strict mode)
- Custom __index__ objects do NOT work with noconvert (strict mode)
- Regular int (PyLong) works with both convert and noconvert
"""
class CustomIndex:
"""Custom class with __index__ but not __int__ or __float__"""
def __index__(self) -> int:
return 42
class CustomIndexNegative:
"""Custom class with negative __index__"""
def __index__(self) -> int:
return -17
convert, noconvert = m.complex_convert, m.complex_noconvert
# Test that regular int (PyLong) works
assert convert(5) == 5.0 + 0j
assert noconvert(5) == 5.0 + 0j
# Test that custom __index__ objects work with convert (non-strict mode)
# This exercises the PyPy-specific path in complex.h
assert convert(CustomIndex()) == 42.0 + 0j
assert convert(CustomIndexNegative()) == -17.0 + 0j
# With noconvert (strict mode), custom __index__ objects are NOT accepted
# Strict mode only accepts complex, float, or int (PyLong), not custom __index__ objects
def requires_conversion(v):
pytest.raises(TypeError, noconvert, v)
requires_conversion(CustomIndex())
requires_conversion(CustomIndexNegative())
# Verify the result is actually a complex
result = convert(CustomIndex())
assert isinstance(result, complex)
assert result.real == 42.0
assert result.imag == 0.0
def test_overload_resolution_float_int():
"""
Test overload resolution behavior when int can match float (added with PR #5879).
This test documents the breaking change in PR #5879: when a float overload is
registered before an int overload, passing a Python int will now match the float
overload (because int can be converted to float in strict mode per PEP 484).
Before PR #5879: int(42) would match int overload (if both existed)
After PR #5879: int(42) matches float overload (if registered first)
This is a breaking change because existing code that relied on int matching
int overloads may now match float overloads instead.
"""
# Test 1: float overload registered first, int second
# When passing int(42), pybind11 tries overloads in order:
# 1. float overload - can int(42) be converted? Yes (with PR #5879 changes)
# 2. Match! Use float overload (int overload never checked)
result = m.overload_resolution_test(42)
assert result == "float: 42.000000", (
f"Expected int(42) to match float overload, got: {result}. "
"This documents the breaking change: int now matches float overloads."
)
assert m.overload_resolution_test(42.0) == "float: 42.000000"
# Test 2: With noconvert (strict mode) - this is the KEY breaking change
# Before PR #5879: int(42) would NOT match float overload with noconvert, would match int overload
# After PR #5879: int(42) DOES match float overload with noconvert (because int->float is now allowed)
result_strict = m.overload_resolution_strict(42)
assert result_strict == "float_strict: 42.000000", (
f"Expected int(42) to match float overload with noconvert, got: {result_strict}. "
"This is the key breaking change: int now matches float even in strict mode."
)
assert m.overload_resolution_strict(42.0) == "float_strict: 42.000000"
# Test 3: complex overload registered first, then float, then int
# When passing int(5), pybind11 tries overloads in order:
# 1. complex overload - can int(5) be converted? Yes (with PR #5879 changes)
# 2. Match! Use complex overload
assert m.overload_resolution_complex(5) == "complex: (5.000000, 0.000000)"
assert m.overload_resolution_complex(5.0) == "complex: (5.000000, 0.000000)"
assert (
m.overload_resolution_complex(complex(3, 4)) == "complex: (3.000000, 4.000000)"
)
# Verify that the overloads are registered in the expected order
# The docstring should show float overload before int overload
doc = m.overload_resolution_test.__doc__
assert doc is not None
# Check that float overload appears before int overload in docstring
# The docstring uses "typing.SupportsFloat" and "typing.SupportsInt"
float_pos = doc.find("SupportsFloat")
int_pos = doc.find("SupportsInt")
assert float_pos != -1, f"Could not find 'SupportsFloat' in docstring: {doc}"
assert int_pos != -1, f"Could not find 'SupportsInt' in docstring: {doc}"
assert float_pos < int_pos, (
f"Float overload should appear before int overload in docstring. "
f"Found 'SupportsFloat' at {float_pos}, 'SupportsInt' at {int_pos}. "
f"Docstring: {doc}"
)
def test_bool_caster():
"""Test bool caster implicit conversions."""
convert, noconvert = m.bool_passthrough, m.bool_passthrough_noconvert

12
tests/test_custom_type_casters.py
View File

@@ -55,17 +55,7 @@ def test_noconvert_args(msg):
assert m.floats_preferred(4) == 2.0
assert m.floats_only(4.0) == 2.0
with pytest.raises(TypeError) as excinfo:
m.floats_only(4)
assert (
msg(excinfo.value)
== """
floats_only(): incompatible function arguments. The following argument types are supported:
1. (f: float) -> float
Invoked with: 4
"""
)
assert m.floats_only(4) == 2.0
assert m.ints_preferred(4) == 2
assert m.ints_preferred(True) == 0

5
tests/test_factory_constructors.cpp
View File

@@ -405,11 +405,10 @@ TEST_SUBMODULE(factory_constructors, m) {
pyNoisyAlloc.def(py::init([](double d, int) { return NoisyAlloc(d); }));
// Old-style placement new init; requires preallocation
ignoreOldStyleInitWarnings([&pyNoisyAlloc]() {
pyNoisyAlloc.def("__init__",
[](NoisyAlloc &a, double d, double) { new (&a) NoisyAlloc(d); });
pyNoisyAlloc.def("__init__", [](NoisyAlloc &a, int i, double) { new (&a) NoisyAlloc(i); });
});
// Requires deallocation of previous overload preallocated value:
pyNoisyAlloc.def(py::init([](int i, double) { return new NoisyAlloc(i); }));
pyNoisyAlloc.def(py::init([](double d, double) { return new NoisyAlloc(d); }));
// Regular again: requires yet another preallocation
ignoreOldStyleInitWarnings([&pyNoisyAlloc]() {
pyNoisyAlloc.def(

12
tests/test_factory_constructors.py
View File

@@ -433,9 +433,10 @@ def test_reallocation_e(capture, msg):
create_and_destroy(3.5, 4.5)
assert msg(capture) == strip_comments(
"""
noisy new # preallocation needed before invoking placement-new overload
noisy placement new # Placement new
NoisyAlloc(double 3.5) # construction
noisy new # preallocation needed before invoking factory pointer overload
noisy delete # deallocation of preallocated storage
noisy new # Factory pointer allocation
NoisyAlloc(double 3.5) # factory pointer construction
---
~NoisyAlloc() # Destructor
noisy delete # operator delete
@@ -450,9 +451,8 @@ def test_reallocation_f(capture, msg):
assert msg(capture) == strip_comments(
"""
noisy new # preallocation needed before invoking placement-new overload
noisy delete # deallocation of preallocated storage
noisy new # Factory pointer allocation
NoisyAlloc(int 4) # factory pointer construction
noisy placement new # Placement new
NoisyAlloc(int 4) # construction
---
~NoisyAlloc() # Destructor
noisy delete # operator delete

18
tests/test_methods_and_attributes.cpp
View File

@@ -240,33 +240,33 @@ TEST_SUBMODULE(methods_and_attributes, m) {
#if defined(PYBIND11_OVERLOAD_CAST)
.def("overloaded", py::overload_cast<>(&ExampleMandA::overloaded))
.def("overloaded", py::overload_cast<int>(&ExampleMandA::overloaded))
.def("overloaded", py::overload_cast<int, int>(&ExampleMandA::overloaded))
.def("overloaded", py::overload_cast<int, float>(&ExampleMandA::overloaded))
.def("overloaded", py::overload_cast<float, int>(&ExampleMandA::overloaded))
.def("overloaded", py::overload_cast<int, int>(&ExampleMandA::overloaded))
.def("overloaded", py::overload_cast<float, float>(&ExampleMandA::overloaded))
.def("overloaded_float", py::overload_cast<float, float>(&ExampleMandA::overloaded))
.def("overloaded_const", py::overload_cast<int>(&ExampleMandA::overloaded, py::const_))
.def("overloaded_const",
py::overload_cast<int, int>(&ExampleMandA::overloaded, py::const_))
.def("overloaded_const",
py::overload_cast<int, float>(&ExampleMandA::overloaded, py::const_))
.def("overloaded_const",
py::overload_cast<float, int>(&ExampleMandA::overloaded, py::const_))
.def("overloaded_const",
py::overload_cast<int, int>(&ExampleMandA::overloaded, py::const_))
.def("overloaded_const",
py::overload_cast<float, float>(&ExampleMandA::overloaded, py::const_))
#else
// Use both the traditional static_cast method and the C++11 compatible overload_cast_
.def("overloaded", overload_cast_<>()(&ExampleMandA::overloaded))
.def("overloaded", overload_cast_<int>()(&ExampleMandA::overloaded))
.def("overloaded", overload_cast_<int, float>()(&ExampleMandA::overloaded))
.def("overloaded", static_cast<py::str (ExampleMandA::*)(float, int)>(&ExampleMandA::overloaded))
.def("overloaded", static_cast<py::str (ExampleMandA::*)(int, int)>(&ExampleMandA::overloaded))
.def("overloaded", overload_cast_<int, int>()(&ExampleMandA::overloaded))
.def("overloaded", static_cast<py::str (ExampleMandA::*)(int, float)>(&ExampleMandA::overloaded))
.def("overloaded", static_cast<py::str (ExampleMandA::*)(float, int)>(&ExampleMandA::overloaded))
.def("overloaded", static_cast<py::str (ExampleMandA::*)(float, float)>(&ExampleMandA::overloaded))
.def("overloaded_float", overload_cast_<float, float>()(&ExampleMandA::overloaded))
.def("overloaded_const", overload_cast_<int >()(&ExampleMandA::overloaded, py::const_))
.def("overloaded_const", overload_cast_<int, float>()(&ExampleMandA::overloaded, py::const_))
.def("overloaded_const", static_cast<py::str (ExampleMandA::*)(float, int) const>(&ExampleMandA::overloaded))
.def("overloaded_const", static_cast<py::str (ExampleMandA::*)(int, int) const>(&ExampleMandA::overloaded))
.def("overloaded_const", overload_cast_<int, int>()(&ExampleMandA::overloaded, py::const_))
.def("overloaded_const", static_cast<py::str (ExampleMandA::*)(int, float) const>(&ExampleMandA::overloaded))
.def("overloaded_const", static_cast<py::str (ExampleMandA::*)(float, int) const>(&ExampleMandA::overloaded))
.def("overloaded_const", static_cast<py::str (ExampleMandA::*)(float, float) const>(&ExampleMandA::overloaded))
#endif
// test_no_mixed_overloads

2
tests/test_stl.cpp
View File

@@ -528,7 +528,7 @@ TEST_SUBMODULE(stl, m) {
m.def("load_variant", [](const variant<int, std::string, double, std::nullptr_t> &v) {
return py::detail::visit_helper<variant>::call(visitor(), v);
});
m.def("load_variant_2pass", [](variant<double, int> v) {
m.def("load_variant_2pass", [](variant<int, double> v) {
return py::detail::visit_helper<variant>::call(visitor(), v);
});
m.def("cast_variant", []() {