b33cfe23d9
Switched from sprintf to snprintf
2026-01-05 13:56:27 +01:00
8dc4f659a1
Fix cases when png reader ended up in an infinite loop. #774
...
Additionally, throw exceptions directly, instead of handling
errors by longjumps.
Longjumps lead to segfaults under compiler optimizations.
Besides, setjmp setups always ended up throwing
exceptions, so why not throwing them right away instead?
2025-12-03 22:36:21 +03:00
2956a0e9ef
fix compile error on clang
2025-07-03 01:11:11 +08:00
dcce96467a
Propose repair for issue 763
2025-04-18 19:41:10 +02:00
dc07bcadcd
fix is_floating_point: symbol is not a class template
2025-04-04 16:33:23 +02:00
e5f5bce2b9
Fix dynamic transposed_view ( #756 )
...
Just a little typo
2024-09-18 20:57:06 +02:00
32eabbc553
chore: fix more typos ( #753 )
2024-07-23 09:40:16 +02:00
74ffdcb047
chore: fix some typos ( #751 )
2024-07-08 10:40:31 +02:00
481b6e7c14
feat: add tell() and error() functions to istream + ostream devices ( #747 )
...
The class file_stream_device already has such functions, and now
both istream_device and ostream_device have them, too, making the
device classes more consistent.
Closes #724
2024-05-17 11:43:44 +02:00
abb561a226
fix: Wrong buffer size in path string conversion functions ( #746 )
...
Fixes #733
2024-05-13 09:58:59 +02:00
2c3ee866b6
Portable minmax Win32-API workaround #744 ( #745 )
...
Co-authored-by: Christopher Kormanyos <ckormanyos@yahoo.com >
2024-04-18 10:21:09 +02:00
ecea33a905
fix: Make locator.hpp self contained ( #729 )
2023-02-22 11:59:12 +01:00
8f4cdcbcce
fix: Fixed custom color converter in dynamic_factory and added corresponding ( #726 )
2023-02-21 14:38:03 +01:00
d6e67f38c5
fix: Normalize Gaussian 2D kernel. ( #725 )
2023-02-07 14:11:51 +01:00
712b827edb
fix: Convolution in convolve_2d ( #723 )
...
Fixes #722
2023-02-07 14:10:38 +01:00
573ba132cd
fix: broken build for midpoint_ellipse_rasterizer::draw_curve ( #705 )
...
* fix: ellipse_rasterizer draw_curve()
* moved ostream operator overload of point into test fixture file
2022-07-04 21:09:21 +02:00
233dc53ff7
fix: Unused variable warning ( #704 )
2022-07-02 11:00:09 +02:00
4dbf35a510
Add pmr image typedefs ( #529 )
...
* Add pmr image typedefs
* Swap allocators only if it propagate on container
* Do not call propagate_on_container_swap for C++14
Co-authored-by: Mateusz Łoskot <mateusz@loskot.net >
2022-06-28 07:35:27 +01:00
8d7034c71c
fix: Memory leak in image class for empty dimensions ( #649 )
2022-06-27 22:59:52 +02:00
fe63aa2a10
refactor: Switch to trailing return types ( #599 )
...
- Trailing return types everywhere
- Optionally, return type deduction where sensible (simple and short functions)
This is related to introduction of common .clang-format,
see https://github.com/boostorg/gil/pull/596#issuecomment-822681523
2022-06-27 18:01:49 +02:00
d5492e1ace
feat: Added apply_rasterizer() free function ( #695 )
...
This PR implements the gil::apply_rasterizer()
free function mentioned in #680 .
2022-06-27 09:54:58 +02:00
adddbec896
refactor: Ellipse rasterizer according to the comment at ( #692 )
...
https://github.com/boostorg/gil/pull/585#issuecomment-1144412220
2022-06-26 13:33:53 +02:00
bfed3de004
refactor: Deprecate apply_operation in favor of variant2::visit for any_image ( #656 )
...
build: Add /bigobj to MSVC flags
Avoid test\extension\dynamic_image\algorithm\copy_and_convert_pixels.cpp
compilation error:
fatal error C1128: number of sections exceeded object file format limit: compile with /bigobj
Co-authored-by: Mateusz Łoskot <mateusz@loskot.net >
2022-06-26 13:21:18 +02:00
151fd9c8a3
refactor: Replace deprecated libtiff v4.3 typedefs with C99 fixed-size integers ( #685 )
2022-06-26 00:51:24 +02:00
b9e652dce8
fix: Automatic detection of <filesystem> header ( #684 )
2022-06-25 23:25:12 +02:00
c436ea3e67
fix: Add missing #include <array>
2022-06-24 17:21:52 +02:00
5273678bfc
fix: Wrong RGB -> HSL convertion ( #505 )
...
Tests will follow as part of https://github.com/boostorg/gil/issues/690
2022-06-24 08:44:28 +02:00
1c6c427f53
fix: Added missing include guard to line rasterizer ( #687 )
2022-06-12 15:05:31 +02:00
9ecdb876b3
Fix deprecation warning bug for gcc 10.2 ( #683 )
...
Related to #660 ,
see https://github.com/boostorg/gil/pull/660#issuecomment-1147735559
2022-06-07 09:24:30 +02:00
48d7ebffe0
Move diffusion and Hough transform
...
Move anisotropic diffusion and
Hough transform into extension.
Adjust tests and cmakelists
2022-05-31 02:37:56 +06:00
cc39f1a007
Change "" includes into <>
2022-05-31 02:37:56 +06:00
bab2a370ff
Fix Hough transform and move rasterization
...
Hough line transform had incorrect
rounding which lead to misleading message
2022-05-31 02:37:56 +06:00
98f49711dd
refactor: Make packed_pixel trivially copyable and assignable ( #679 )
2022-05-28 11:07:54 +02:00
6d312af51a
chore!: deprecate any_color_converted_view ( #678 )
...
color_converted_view has the same implementation and can be used
as a replacement, providing the same functionality.
This builds on #660 by @marco-langer (Thank you!)
Due to the fact that the switch to C++14 has been announced in #677 ,
we can now officially use the [[deprecated]] attribute which was standardized
in C++14.
This is initial part of deprecating the `any_color_converted_view` and
as @marco-langer pointed in
https://github.com/boostorg/gil/pull/660#issuecomment-1138183545
there are more changes to follow.
2022-05-26 08:58:19 +02:00
1be8c87c10
chore: do not use deprecated header in test or elsewhere ( #675 )
2022-05-18 23:25:36 +02:00
3090f866f2
fix: Re-allow devicen_t with two color components ( #654 )
...
devicen_t<2> was possible in Boost 1.71, but broke in Boost 1.72.
The static_assert was introduced in commit
5611bd5807#274
Now it's possible to use it again.
Fixes #519
2022-05-18 09:30:06 +02:00
f839504f20
Fixed io extension build error ( #671 )
...
The error was due to missing dynamic.hpp (renamed in #653 )
header that contained dynamic_io_fnobj.
Adding the header back in to write.hpp
headers of io extensions fixed the problem.
2022-05-13 14:13:00 +02:00
9666a95bc5
chore: Remove unused variable in diffusion.hpp ( #666 )
...
The variable `dims` is not used and causes a compiler warning.
2022-05-12 14:23:49 +02:00
c063d1c185
feat: Added for_each_pixel overload for any_image ( #648 )
...
Take functor by value instead of reference.
Test cases to use the same fixture as the other any_image tests.
Fixes #579
2022-04-30 23:57:00 +02:00
3289fe0bc4
refactor: Unified operation names for pixel and channel operations ( #655 )
...
Renamed
- gil::pixel_multiply_t to gil::pixel_multiplies_t
- gil::pixel_divide_t to gil::pixel_divides_t
Closes #368
2022-04-30 23:53:41 +02:00
9d526ed706
refactor: Renamed io/dynamic_io_new.hpp to io/detail/dynamic.hpp ( #653 )
...
Closes #189
2022-04-27 13:27:12 +02:00
791148660e
fix: is_equal_to_sixteen in PNG I/O was less-than test ( #650 )
...
While is_equal_to_sixteen should (as the name suggests) check for
equality to 16, it actually checked whether the bit depth was
less than 16.
2022-04-26 21:27:22 +02:00
4f83beb735
style: Remove trailing space characters ( #651 )
2022-04-26 20:52:56 +02:00
63f5837bd3
Removed deprecated.hpp ( #627 )
...
Closes #625 too
2022-04-06 11:42:56 +02:00
da0655fb66
Merge pull request #621 from sdebionne/fix-620
...
Fix for_each_pixel for non 1d iterable views
2022-03-01 14:35:20 +01:00
f4c70a8606
Add support for <filesystem> to IO ( #636 )
...
If neither <filesystem> nor <experimental/filesystem> is detected,
then require <boost/filesystem.hpp>.
If user defines BOOST_GIL_IO_USE_BOOST_FILESYSTEM macro,
then <boost/filesystem.hpp> is pre-selected and required,
and search for any of the C++ standard implementation is skipped.
Remove end-user macro BOOST_GIL_IO_ADD_FS_PATH_SUPPORT
Require tests to always build with support of either
detected C++ filesystem or pre-selected Boost.Filesystem.
Closes ##222
2022-02-22 19:18:55 +01:00
9bd8642f69
numeric extension move into core part 5
...
moved numeric/kernel.hpp into core
2022-02-18 14:25:59 +05:30
87a3157c4b
numeric extension move into core part 4
...
moved numeric/pixel_numeric_operations.hpp into core
2022-02-18 14:25:59 +05:30
483915cb20
numeric extension move into core part 3
...
moved numeric/convolve.hpp into core
2022-02-18 14:25:59 +05:30
baf246a726
numeric extension move into core part 2
...
moved numeric/channel_numeric_operations.hpp into core
2022-02-18 14:25:59 +05:30
Amsakan Bavan