* Added c++17 test configurations for clang5.0 and clang6.0
* Fix#739 correct cppcorecheck warnings for clang-cl
* Add clang-cl configurations
* Corrections Appveyor;
Temporarily disable msvc compilation for faster testing (#741)
* Add path to clang-cl.exe (#741)
* Escape backslash in path (#741)
* Update vcpkg (#741)
* Check vcpkg version; try without building vcpkg; use latest clang-cl from path (#741)
* Fix blocks in ps script (#741)
* Try accessing APPVEYOR_BUILD_FOLDER variable (#471)
* Update span size() bug confirmation test for GCC 6.5 (#741)
* MSVC flags to Clang-cl; disable c++98-compat and undefined macro warnings (#741)
* Suppress clang warning on tests (missing-prototypes) (#741)
* Fix clang warning -Wnewline-eof (#741)
* Fix clang warning -Wdouble-promotion (#741)
* Set linker explicitly
* Clean condition statement
* For Clang, fallback to the AppVeyor installed version of CMake
* Fix clang warning -Wmissing-variable-declarations
* Fallback to the MSVC linker until vcpkg has been updated
* Revert "Fallback to the MSVC linker until vcpkg has been updated"
This reverts commit 7263f3289e.
* Fix clang warning -Wunused-member-function
* Fix clang warning -Wmissing-noreturn
* Fix clang warning -Winvalid-noreturn on Windows
* Add macro block end comment on large #if blocks
* Workaround: fallback to mscv link.exe
* Workaround: get msvc paths into PowerShell through intermediate file
* Workaround: fix, remove "PATH=" from text
* Workaround: try with full-path; and return user PATH
* Workaround: fix, escape backslashes
* Revert all since "Workaround: fallback to mscv link.exe" did not work on AppVeyor
This reverts the commits:
bda3d6a42897062933ac0f4fb04bac1b0c19afd1a5739ea5f0
* Suppress output of git pull; remove vcpkg from cache
* Re-enable AppVeyor builds for all platforms
* Correct typo
Co-Authored-By: Farwaykorse <Farwaykorse@users.noreply.github.com>
* Add Clang-cl 7.0.0 to the supported platforms
* Revert "Fix clang warning -Wunused-member-function"
This reverts commit 6fe1a42035.
* Fix or locally suppress clang warning -Wunused-member-function
* format touched code and correct comment
* git pull --quiet
* fix logic error in workaround
* fix missing bracket
* Suppress output of mkdir
* Replace MSBuild with Ninja
* Suppress output of 7z
* Add architecture flags for Clang
* Drop workaround for lld-link
* 7-zip Overwrite and Alternative output suppression without suppressing errors
Replaces 3c1c0794dd
* AppVeyor setup and CMake before build
* reorder compiler configuration
* remove unnecessary
* remove -fno-strict-aliasing
* remove -Wsign-conversion, since before v4.0 part of -Wconversion
* -Wctor-dtor-privacy is GCC only
* remove -Woverloaded-virtual part of -Wmost, part of -Wall
* add -Wmissing-noreturn
* remove the pragmas for -Wunused-member-function
* Re-add MSBuild generator on AppVeyor
* Print CMake commands
* Add MSBuild toolset selection
* Separate Architecture setting
* clang-cl: add -Weverything
* clang-cl -Wno-c++98-compat
* clang-cl -Wno-c++98-compat-pedantic
* clang-cl -Wno-missing-prototypes
* clang-cl C++14 -Wno-unused-member-function
* clang-cl -Wundef __GNUC__
* clang++: add -Weverything
* clang++ -Wno-c++98-compat
* clang++ -Wno-c++98-compat-pedantic
* clang++ -Wno-missing-prototypes
* clang++ -Wno-weak-vtables
* clang++ C++14 -Wno-unused-member-function
* clang++ fix -Wundef _MSC_VER
* clang++ -Wno-padded
* clang++ solve -Wdeprecated
* Add AppleClang compiler target
Since CMake v3.0 use of Clang for both is deprecated
* clang++ v5.0 C++17 -Wno-undefined-func-template
* Add VS2015 + LLVM/clang-cl to AppVeyor
* Do not disable constexpr when compiling with clang-cl on Windows
* Clean-up clang-only warnings (now under -Weverything)
* Revert "Fix clang warning -Winvalid-noreturn on Windows"
This reverts commit 2238c4760e.
* Suppress -Winvalid-noreturn for the MS STL noexception workaround
* CMake: put preprocessor definition in target_compile_definitions
* Solve compiler warning C4668: __GNUC__ not defined
* Added c++17 test configurations for clang5.0 and clang6.0
* Removed explicit not_null constructor, sloppy_not_null, added strict_not_null
We added explicit not_null constructor in version 2.0.0.
It proved very difficult to switch to the new version for
large code bases that adopted previous versions of gsl,
due to not_null used extensively in the code. Still, using
explicit constructor is very benefitial for new code, since
it encorages better API design and make null checks intentional.
To resolve the issue, this change:
- removes explicit keyword from not_null constructor
- removes unneded sloppy_not_null type
- adds strict_not_null type to behave the same way as v2 not_null
- updates tests
* Removed explicit not_null constructor, sloppy_not_null, added strict_not_null
We added explicit not_null constructor in version 2.0.0.
It proved very difficult to switch to the new version for
large code bases that adopted previous versions of gsl,
due to not_null used extensively in the code. Still, using
explicit constructor is very benefitial for new code, since
it encorages better API design and make null checks intentional.
To resolve the issue, this change:
- removes explicit keyword from not_null constructor
- removes unneded sloppy_not_null type
- adds strict_not_null type to behave the same way as v2 not_null
- updates tests
* fixed build break for gcc7
* added more tests
* added more non-compiling tests
* Addressed PR comments and suppressed a code analysis warning
* Fixed test failure in not_null tests
* Span std::array c'tor uses arr.data() instead of &arr[0]
- Fixes runtime issues when constructing from an empty std::array
* Construct span with std::data if C++17 detected
* Specialize span c'tor for std::array of length 0, set storage to nullptr
* Changed &arr[0] to std::array<T, N>::data and std::address_of to protect against overloaded operator&.
* Removed the usage of `std::addressof` because it is a C++ 17 feature. Using decay for C arrays instead.
* Add unit tests for #662.
* Added c++17 test configurations for clang5.0 and clang6.0
* fixed CppCoreCheck pointer decay warning
* Added c++17 test configurations for clang5.0 and clang6.0
* Added optimization that removes a branch from span::operator[]
* minor beauty fix
* added a better message for the optimization, fixed signed/unsigned warning
* better check fir wrap-around possibility
* Added c++17 test configurations for clang5.0 and clang6.0
* Added make_not_null helper to create a not_null
Introduction of explicit not_null constructor made it cumbersome to create not_nulls
in c++14. Adding make_not_null helper. Usage (see tests):
int i = 42;
auto x = make_not_null(&i);
helper(make_not_null(&i));
helper_const(make_not_null(&i));
* Added std::forward to make_not_null, fixed some code analysis warnings
* Fix build break in VS2015 Release configuration
* Fix build break in VS2015 Release configuration
* Added c++17 test configurations for clang5.0 and clang6.0
* Fixed CppCoreCheck warnings in GSL and tests
- Added CMakeSettings.json for VS Open Folder configuration
- So we can easily run CppCoreCheck in VS
- Fixed CppCorecheck warnings where it made sense
- Suppressed the rest
- Some suppression does not work due to compiler/tool bugs,
so replaced by #pragma disable
- CppCoreCheck has noise, suppressed those with comments
- Catch produces many warnings, blanket-supressed them all
- Had to fix clang formatting to keep attributes in place
- clang-format does not support attributes, so I am using
- "CommentPragmas: '^ NO-FORMAT:'" to skip formatiting on them
- Removed GSL_NOEXCEPT macro, removed incorred noexcepts
* Ignore unknown attributes
* ignore unknown attributes in noexception mode tests
* fixed C26472 in at()
* created GSL_SUPPRESS macro to allow all compilers to parse suppression attributes
* try to fix gcc compilation problems with attributes
* ignore gsl::suppress for gcc
* move suppression to function level on return statements
clang5.0 and up does not allow attributes on return statemets in constexpr functions
* move suppression to function level on return statements
* use GSL_SUPPRESS in algorithm_tests
* Addressed PR comments
* Added c++17 test configurations for clang5.0 and clang6.0
* removed GSL_NOEXCEPT
- Removed GSL_NOEXCEPT macro
- Replaced by noexcept keyword when needed
- removed noexcept where a function could throw
* remove unneded undef
* fixed replace errors
* removed noexcept where function can throw
* Explicitly check for availbility of std::byte
GCC > 7.3 defines __cpp_lib_byte >= 201603 when std::byte is available.
On libc++ std::byte is available since version 5.0. This can be checked
with _LIBCPP_VERSION >= 5000.
This fixes 713.
* Add missing \ in preprocessor check
- gsl::narrow, gsl::narrow_cast and gsl::at are the safe variants suggested by CppCoreGuideline. It does not make sense to let VS warn inside the implementation of these functions that unsafe static_cast is used and that the safe variants shall be used.
- Suppress warning that throw_exception can be declared noexcept (for the GSL_TERMINATE_ON_CONTRACT_VIOLATION case)
* Added template argument deduction for not_null
This allows compilers with c++17 support to infer template
instantiation types when calling not_null constructor:
int foo(not_null<const int*> x);
int main()
{
int t = 0;
not_null x{ &t };
return foo(not_null{ &t });
}
* replaced deduction guides by a simple constructor
* Updated tests
* fixed check for availability of std::byte
* testing c++1z on clang
* fixed cmakelists extra endif
* include cstddef header for clang and gcc in pointers
* fixed a typo
* fix missing nullptr_t type
* fixed typo in CMakeLists.tst
* change approach to c++17 testing, step one: revert cmake testing, update clang5.0 package
removed using latest c++ due to clang5.0 failing,
update clang5.0 travis config to use llvm-toolchain-trusty-5.0
* addressed comments
* Add usage for check-and-unwrap of MS STL iterators
This is Billy ONeal's PR #682 with a typo fixed.
See corresponding change here:
ca77129308
That change officially exposes the STL's range checking machinery,
available in MSVC++ "15.8"+
This change augments GSL span::iterator to call into that newly exposed
machinery.
_Verify_range(cit, cit)
Requests that the iterator type check that the parameters form a valid
[First, Last) range iterator pair. This replaces _DEBUG_RANGE and
supporting machinery. The standard library provides a version for
pointers under _IDL != 0; otherwise they are normally provided via
hidden friend functions. Note that declaring some of these hidden
friends for "wrapper" iterators like move_iterator and reverse_iterator
triggers VSO#610735.
cit._Verify_offset(difference_type _Off)
For random-access iterators, performs any asserts that would be
performed by i += _Off; except with possibly a better error message and
without moving the iterator.
cit._Unwrapped()
Returns an "unchecked" or "unwrapped" iterator which has previously been
validated. The iterator may have been validated by a call to
_Verify_range or _Verify_offset (above), or by seeking a checked
iterator to a "high water mark" point.
it._Seek_to(cit) / it._Seek_to(return value from _Unwrapped())
Moves the position of the checked iterator it to the position of the
unchecked iterator supplied. Generally not intended to perform range
checks.
* Fixed build break in VS2015
* fixed constexpr build break when using VS2015
* [span] implement non-member + as a hidden friend
Drive-by: remove garbage `operator-(difference_type, span_iterator)`.
* [span] Use pointers for range-based-for on MSVC
C++17 defines `std::byte*` to have the same aliasing properties as
`char*`. Hence, we mark it with the `__may_alias__` attribute under gcc
and clang.
Based on the fix by Martin Moene in byte-lite.
Fixes#663
This fixes the build with clang 7 which introduces a new warning
`-Wreturn-std-move` which warns about needless copies when a move
operation is available (see https://reviews.llvm.org/rL329914).
We take the exception object by uref so we should throw the forwarded
version.
