* feat: type<T>()

* refactor: using py::type as class

* refactor: py::object as base

* wip: tigher api

* refactor: fix conversion and limit API further

* docs: some added notes from @EricCousineau-TRI

* refactor: use py::type::of

Why only for pointers? Because C casts are hard to grep for.

* Added guards to the includes

Added new CI config

Added new trigger

Changed CI workflow name

Debug CI

Debug CI

Debug CI

Debug CI

Added flags fro PGI

Disable Eigen

Removed tests that fail

Uncomment lines

* fix: missing include

fix: minor style cleanup

tests: support skipping

ci: remove and tighten a bit

fix: try msvc workaround for pgic

* tests: split up prealoc tests

* fix: PGI compiler fix

* fix: PGI void_t only

* fix: try to appease nvcc

* ci: better ordering for slow tests

* ci: minor improvements to testing

* ci: Add NumPy to testing

* ci: Eigen generates CUDA warnings / PGI errors

* Added CentOS7 back for a moment

* Fix YAML

* ci: runs-on missing

* centos7 is missing pytest

* ci: use C++11 on CentOS 7

* ci: test something else

* Try just adding flags on CentOS 7

* fix: CentOS 7

* refactor: move include to shared location

* Added verbose flag

* Try to use system cmake3 on CI

* Try to use system cmake3 on CI, attempt2

* Try to use system cmake3 on CI, attempt3

* tests: not finding pytest should be a warning, not a fatal error

* tests: cleanup

* Weird issue?

* fix: final polish

Co-authored-by: Andrii Verbytskyi <andrii.verbytskyi@mpp.mpg.de>
Co-authored-by: Henry Schreiner <henryschreineriii@gmail.com>
Co-authored-by: Andrii Verbytskyi <averbyts@cern.ch>

Add check if `str(handle)` correctly converted the object, and throw py::error_already_set if not (bis) (#2477)

* Add check if `str(handle)` correctly converted the object, and throw py::error_already_set if not

* Fix tests on Python 3

* Apply @rwgk's fixes to cherry-picked commits from #2392

The variables PYBIND11_HAS_OPTIONAL, PYBIND11_HAS_EXP_OPTIONAL, PYBIND11_HAS_VARIANT,
__clang__, __APPLE__ were not checked for defined in a minortity of instances.

If the project using pybind11 sets -Wundef, the warnings will show.

The test build is also modified to catch the problem.

* fix: support nvcc and test

* fixup! fix: support nvcc and test

* docs: mention what compilers fail

* fix: much simpler logic

* refactor: slightly faster / clearer

* Add py::object casting example to embedding docs

* Move implicit cast example to object.rst

* Move to bottom and improve implicit casting text

* Fix xref

* Improve wording as per @bstaletic's suggestion

* ci: only annotate linux for now

* style: block some common mistakes

* Add note about specifying custom base.

* Update exception docs based on PR feedback.

* Fix trailing whitespace.

Co-authored-by: Michal Sustr <michal.sustr@aic.fel.cvut.cz>

* fix: use classic extension handling unless otherwise requested

* fix: variable must be cached to be used externally

* feat: py::pos_only

* fix: review points from @YannickJadoul

* fix: review points from @bstaletic

* refactor: kwonly -> kw_only

* doc: avoid C++ types in docstrings

* A bit of rewording

* Another bit of rewording

* Third rewording

* ci: harden chrono test, mark another macos 4.9 dev failure

This should help with a little of the flakiness seen with the timing test

* Update tests/test_chrono.py

* Can also fail

Adding missing `bytes` type to `test_constructors()`, to exercise the code change.

The changes in the PR were cherry-picked from PR #2409 (with a very minor
modification in test_pytypes.py related to flake8). Via PR #2409, these
changes were extensively tested in the Google environment, as summarized here:
https://docs.google.com/document/d/1TPL-J__mph_yHa1quDvsO12E_F5OZnvBaZlW9IIrz8M/
The changes in this PR did not cause an issues at all.

Note that `test_constructors()` before this PR passes for Python 2 only
because `pybind11::str` can hold `PyUnicodeObject` or `PyBytesObject`. As a
side-effect of this PR, `test_constructors()` no longer relies on this
permissive `pybind11::str` behavior. However, the permissive behavior is still
exercised/exposed via the existing `test_pybind11_str_raw_str()`.

The test code change is designed to enable easy removal later, when Python 2
support is dropped.

For completeness: confusingly, the non-test code changes travelled through PR

Example `ambiguous conversion` error fixed by this PR:
```
pybind11/tests/test_pytypes.cpp:214:23: error: ambiguous conversion for functional-style cast from 'pybind11::detail::item_accessor' (aka 'accessor<accessor_policies::generic_item>') to 'py::bytes'
            "bytes"_a=py::bytes(d["bytes"]),
                      ^~~~~~~~~~~~~~~~~~~~
pybind11/include/pybind11/detail/../pytypes.h:957:21: note: candidate constructor
    PYBIND11_OBJECT(bytes, object, PYBIND11_BYTES_CHECK)
                    ^
pybind11/include/pybind11/detail/../pytypes.h:957:21: note: candidate constructor
pybind11/include/pybind11/detail/../pytypes.h:987:15: note: candidate constructor
inline bytes::bytes(const pybind11::str &s) {
              ^
1 error generated.
```

Fix bug roundtripping datetime.time objects after midnight in eastern hemisphere timezones (#2417) (#2438)

* Fix bug roundtripping datetime.time objects after midnight in eastern hemisphere timezones (#2417)

* tests: check more timezones

* Fix review remarks: remove useless comment and skip setting TZ environment variable on Windows

* fix: respect PYTHON_VERSION if set in classic mode

* fix: add warning when using PYTHON_VERSION

tests: fix CI by including <algorithm> to stop MSVC from complaining about std::count_if in tests/test_sequences_and_iterators.cpp (#2435)

Throw exception on returning a unique_ptr or shared_ptr nullptr (or any other holder type) from py::init, rather than crashing (#2430)

* Add note that VS2017 requires /permissive- to build in C++17 mode

* ci: test C++17 on MSVC 2017

* ci: args1/2, use args to override max cxx

Co-authored-by: Henry Schreiner <henryschreineriii@gmail.com>

Adapt code example in advanced/classes.rst to new handling of forgetting to call the superclass __init__ (#2429)

The main change is to treat error_already_set as a separate category
of exception that arises in different circumstances and needs to be
handled differently. The asymmetry between Python and C++ exceptions
is further emphasized.

* fix: reduce target collision in add_submodule mode

Closes #2420

* fix: update CMakeLists.txt

CPython configured with `--with-pydebug` could not use `pybind11_add_module`

* docs: move CONTRIBUTING

* docs: clarify PyPy promise

* tests: keep source dir clean

* ci: make first build inplace

* ci: drop dev setting (wasn't doing anything)

* tests: warn if source directory is dirty

* feat: FindPython support

* refactor: rename to PYBIND11_FINDPYTHON

* docs: Caps fixes

* feat: NOPYTHON mode

* test: check simple call

* docs: add changelog/upgrade guide

* feat: Support Python3 and Python2

* refactor: Use targets in tests

* fix: support CMake 3.4+

* feat: classic search also finds virtual environments

* docs: some updates from @wjakob's review

* fix: wrong name for QUIET mode variable, reported by @skoslowski

* refactor: cleaner output messaging

* fix: support debug Python's in FindPython mode too

* fixup! refactor: cleaner output messaging

* fix: missing pybind11_FOUND and pybind11_INCLUDE_DIR restored to subdir mode

* fix: nicer reporting of Python / PyPy

* fix: out-of-order variable fix

* docs: minor last-minute cleanup