You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

1153 lines
53 KiB

/*
This file is part of Magnum.
Copyright © 2010, 2011, 2012, 2013, 2014, 2015, 2016, 2017, 2018, 2019,
2020, 2021 Vladimír Vondruš <mosra@centrum.cz>
Permission is hereby granted, free of charge, to any person obtaining a
copy of this software and associated documentation files (the "Software"),
to deal in the Software without restriction, including without limitation
the rights to use, copy, modify, merge, publish, distribute, sublicense,
and/or sell copies of the Software, and to permit persons to whom the
Software is furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included
in all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
DEALINGS IN THE SOFTWARE.
*/
#include "AbstractImageConverter.h"
#include <Corrade/Containers/Array.h>
#include <Corrade/Containers/EnumSet.hpp>
#include <Corrade/Containers/Optional.h>
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
#include <Corrade/Containers/StringView.h>
#include <Corrade/Containers/StringStl.h> /* for Directory */
#include <Corrade/Utility/Assert.h>
#include <Corrade/Utility/Directory.h>
#include <Corrade/Utility/DebugStl.h>
#include "Magnum/Image.h"
#include "Magnum/ImageView.h"
#include "Magnum/PixelFormat.h"
#include "Magnum/Trade/ImageData.h"
#ifndef CORRADE_PLUGINMANAGER_NO_DYNAMIC_PLUGIN_SUPPORT
#include "Magnum/Trade/configure.h"
#endif
namespace Magnum { namespace Trade {
std::string AbstractImageConverter::pluginInterface() {
return
/* [interface] */
"cz.mosra.magnum.Trade.AbstractImageConverter/0.3.1"
/* [interface] */
;
}
#ifndef CORRADE_PLUGINMANAGER_NO_DYNAMIC_PLUGIN_SUPPORT
std::vector<std::string> AbstractImageConverter::pluginSearchPaths() {
return PluginManager::implicitPluginSearchPaths(
#ifndef MAGNUM_BUILD_STATIC
Utility::Directory::libraryLocation(&pluginInterface),
#else
{},
#endif
#ifdef CORRADE_IS_DEBUG_BUILD
MAGNUM_PLUGINS_IMAGECONVERTER_DEBUG_DIR,
#else
MAGNUM_PLUGINS_IMAGECONVERTER_DIR,
#endif
#ifdef CORRADE_IS_DEBUG_BUILD
"magnum-d/"
#else
"magnum/"
#endif
"imageconverters");
}
#endif
AbstractImageConverter::AbstractImageConverter() = default;
AbstractImageConverter::AbstractImageConverter(PluginManager::Manager<AbstractImageConverter>& manager): PluginManager::AbstractManagingPlugin<AbstractImageConverter>{manager} {}
AbstractImageConverter::AbstractImageConverter(PluginManager::AbstractManager& manager, const std::string& plugin): PluginManager::AbstractManagingPlugin<AbstractImageConverter>{manager, plugin} {}
void AbstractImageConverter::setFlags(ImageConverterFlags flags) {
_flags = flags;
doSetFlags(flags);
}
void AbstractImageConverter::doSetFlags(ImageConverterFlags) {}
void AbstractImageConverter::addFlags(ImageConverterFlags flags) {
setFlags(_flags|flags);
}
void AbstractImageConverter::clearFlags(ImageConverterFlags flags) {
setFlags(_flags & ~flags);
}
Containers::Optional<ImageData1D> AbstractImageConverter::convert(const ImageView1D& image) {
CORRADE_ASSERT(features() & ImageConverterFeature::Convert1D,
"Trade::AbstractImageConverter::convert(): 1D image conversion not supported", {});
/* Unlike with convertToData() / convertToFile(), where images are checked
for having a non-zero size and being non-null, here it's explicitly
allowed:
- Usual format conversions / resamplings are a loop that "just works"
with zero-sized dimensions.
- Nullptr views could be potentially useful for converters that fill
an image of desired format and size with data, such as generating a
noise image or some test pattern. While it could be implemented as
an importer plugin as well, the size/format would have to be passed
manually through plugin-specific configuration there, which is
rather annoying compared to having an API prepared directly for
that.
Plugin implementations that can't work with these have to do this check
themselves, and since that's then a plugin-specific behavior, it should
be a runtime error, not an assert. */
Containers::Optional<ImageData1D> out = doConvert(image);
CORRADE_ASSERT(!out || !out->_data.deleter(), "Trade::AbstractImageConverter::convert(): implementation is not allowed to use a custom Array deleter", {});
return out;
}
Containers::Optional<ImageData1D> AbstractImageConverter::doConvert(const ImageView1D&) {
CORRADE_ASSERT_UNREACHABLE("Trade::AbstractImageConverter::convert(): 1D image conversion advertised but not implemented", {});
}
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
Containers::Optional<ImageData2D> AbstractImageConverter::convert(const ImageView2D& image) {
CORRADE_ASSERT(features() & ImageConverterFeature::Convert2D,
"Trade::AbstractImageConverter::convert(): 2D image conversion not supported", {});
/* No zero size / nullptr checks here, see convert(const ImageView1D&) for
reasons why */
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
Containers::Optional<ImageData2D> out = doConvert(image);
CORRADE_ASSERT(!out || !out->_data.deleter(), "Trade::AbstractImageConverter::convert(): implementation is not allowed to use a custom Array deleter", {});
return out;
}
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
Containers::Optional<ImageData2D> AbstractImageConverter::doConvert(const ImageView2D&) {
CORRADE_ASSERT_UNREACHABLE("Trade::AbstractImageConverter::convert(): 2D image conversion advertised but not implemented", {});
}
#ifdef MAGNUM_BUILD_DEPRECATED
Containers::Optional<Image2D> AbstractImageConverter::exportToImage(const ImageView2D& image) {
Containers::Optional<ImageData2D> out = convert(image);
if(!out) return {};
if(out->isCompressed()) {
Error{} << "Trade::AbstractImageConverter::exportToImage(): implementation returned a compressed image";
return {};
}
const PixelStorage storage = out->storage();
const PixelFormat format = out->format();
const UnsignedInt formatExtra = out->formatExtra();
const UnsignedInt pixelSize = out->pixelSize();
const Vector2i size = out->size();
return Image2D{storage, format, formatExtra, pixelSize, size, out->release()};
}
Containers::Optional<CompressedImage2D> AbstractImageConverter::exportToCompressedImage(const ImageView2D& image) {
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
Containers::Optional<ImageData2D> out = convert(image);
if(!out) return {};
if(!out->isCompressed()) {
Error{} << "Trade::AbstractImageConverter::exportToCompressedImage(): implementation returned an uncompressed image";
return {};
}
const CompressedPixelStorage storage = out->compressedStorage();
const CompressedPixelFormat format = out->compressedFormat();
const Vector2i size = out->size();
return CompressedImage2D{storage, format, size, out->release()};
}
#endif
Containers::Optional<ImageData3D> AbstractImageConverter::convert(const ImageView3D& image) {
CORRADE_ASSERT(features() & ImageConverterFeature::Convert3D,
"Trade::AbstractImageConverter::convert(): 3D image conversion not supported", {});
/* No zero size / nullptr checks here, see convert(const ImageView1D&) for
reasons why */
Containers::Optional<ImageData3D> out = doConvert(image);
CORRADE_ASSERT(!out || !out->_data.deleter(), "Trade::AbstractImageConverter::convert(): implementation is not allowed to use a custom Array deleter", {});
return out;
}
Containers::Optional<ImageData3D> AbstractImageConverter::doConvert(const ImageView3D&) {
CORRADE_ASSERT_UNREACHABLE("Trade::AbstractImageConverter::convert(): 3D image conversion advertised but not implemented", {});
}
Containers::Optional<ImageData1D> AbstractImageConverter::convert(const CompressedImageView1D& image) {
CORRADE_ASSERT(features() & ImageConverterFeature::ConvertCompressed1D,
"Trade::AbstractImageConverter::convert(): compressed 1D image conversion not supported", {});
/* No zero size / nullptr checks here, see convert(const ImageView1D&) for
reasons why */
Containers::Optional<ImageData1D> out = doConvert(image);
CORRADE_ASSERT(!out || !out->_data.deleter(), "Trade::AbstractImageConverter::convert(): implementation is not allowed to use a custom Array deleter", {});
return out;
}
Containers::Optional<ImageData1D> AbstractImageConverter::doConvert(const CompressedImageView1D&) {
CORRADE_ASSERT_UNREACHABLE("Trade::AbstractImageConverter::convert(): compressed 1D image conversion advertised but not implemented", {});
}
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
Containers::Optional<ImageData2D> AbstractImageConverter::convert(const CompressedImageView2D& image) {
CORRADE_ASSERT(features() & ImageConverterFeature::ConvertCompressed2D,
"Trade::AbstractImageConverter::convert(): compressed 2D image conversion not supported", {});
/* No zero size / nullptr checks here, see convert(const ImageView1D&) for
reasons why */
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
Containers::Optional<ImageData2D> out = doConvert(image);
CORRADE_ASSERT(!out || !out->_data.deleter(), "Trade::AbstractImageConverter::convert(): implementation is not allowed to use a custom Array deleter", {});
return out;
}
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
Containers::Optional<ImageData2D> AbstractImageConverter::doConvert(const CompressedImageView2D&) {
CORRADE_ASSERT_UNREACHABLE("Trade::AbstractImageConverter::convert(): compressed 2D image conversion advertised but not implemented", {});
}
Containers::Optional<ImageData3D> AbstractImageConverter::convert(const CompressedImageView3D& image) {
CORRADE_ASSERT(features() & ImageConverterFeature::ConvertCompressed3D,
"Trade::AbstractImageConverter::convert(): compressed 3D image conversion not supported", {});
/* No zero size / nullptr checks here, see convert(const ImageView1D&) for
reasons why */
Containers::Optional<ImageData3D> out = doConvert(image);
CORRADE_ASSERT(!out || !out->_data.deleter(), "Trade::AbstractImageConverter::convert(): implementation is not allowed to use a custom Array deleter", {});
return out;
}
Containers::Optional<ImageData3D> AbstractImageConverter::doConvert(const CompressedImageView3D&) {
CORRADE_ASSERT_UNREACHABLE("Trade::AbstractImageConverter::convert(): compressed 3D image conversion advertised but not implemented", {});
}
Containers::Optional<ImageData1D> AbstractImageConverter::convert(const ImageData1D& image) {
return image.isCompressed() ? convert(CompressedImageView1D(image)) : convert(ImageView1D(image));
}
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
Containers::Optional<ImageData2D> AbstractImageConverter::convert(const ImageData2D& image) {
return image.isCompressed() ? convert(CompressedImageView2D(image)) : convert(ImageView2D(image));
}
Containers::Optional<ImageData3D> AbstractImageConverter::convert(const ImageData3D& image) {
return image.isCompressed() ? convert(CompressedImageView3D(image)) : convert(ImageView3D(image));
}
#ifndef CORRADE_NO_ASSERT
namespace {
template<UnsignedInt dimensions, template<UnsignedInt, class> class View> bool checkImageValidity(const char* const messagePrefix, const View<dimensions, const char>& image) {
/* At some point there might be a file format that allows zero-sized
images, but so far I don't know about any. When such format appears,
this check will get moved to plugin implementations that can't work with
zero-sized images.
Also note that this check isn't done for the Image->Image conversion
above, there zero-sized images and nullptr *could* make sense. */
CORRADE_ASSERT(image.size().product(),
messagePrefix << "can't convert image with a zero size:" << image.size(), false);
CORRADE_ASSERT(image.data(),
messagePrefix << "can't convert image with a nullptr view", false);
return true;
}
}
#endif
Containers::Array<char> AbstractImageConverter::convertToData(const ImageView1D& image) {
CORRADE_ASSERT(features() >= ImageConverterFeature::Convert1DToData,
"Trade::AbstractImageConverter::convertToData(): 1D image conversion not supported", nullptr);
#ifndef CORRADE_NO_ASSERT
/* Explicitly return if checks fail for CORRADE_GRACEFUL_ASSERT builds */
if(!checkImageValidity("Trade::AbstractImageConverter::convertToData():", image))
return {};
#endif
Containers::Array<char> out = doConvertToData(image);
CORRADE_ASSERT(!out.deleter(), "Trade::AbstractImageConverter::convertToData(): implementation is not allowed to use a custom Array deleter", {});
return out;
}
Containers::Array<char> AbstractImageConverter::doConvertToData(const ImageView1D& image) {
CORRADE_ASSERT(features() >= ImageConverterFeature::ConvertLevels1DToData,
"Trade::AbstractImageConverter::convertToData(): 1D image conversion advertised but not implemented", nullptr);
return doConvertToData(Containers::arrayView({image}));
}
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
Containers::Array<char> AbstractImageConverter::convertToData(const ImageView2D& image) {
CORRADE_ASSERT(features() >= ImageConverterFeature::Convert2DToData,
"Trade::AbstractImageConverter::convertToData(): 2D image conversion not supported", nullptr);
#ifndef CORRADE_NO_ASSERT
/* Explicitly return if checks fail for CORRADE_GRACEFUL_ASSERT builds */
if(!checkImageValidity("Trade::AbstractImageConverter::convertToData():", image))
return {};
#endif
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
Containers::Array<char> out = doConvertToData(image);
CORRADE_ASSERT(!out.deleter(), "Trade::AbstractImageConverter::convertToData(): implementation is not allowed to use a custom Array deleter", {});
return out;
}
Containers::Array<char> AbstractImageConverter::doConvertToData(const ImageView2D& image) {
CORRADE_ASSERT(features() >= ImageConverterFeature::ConvertLevels2DToData,
"Trade::AbstractImageConverter::convertToData(): 2D image conversion advertised but not implemented", nullptr);
return doConvertToData(Containers::arrayView({image}));
}
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
#ifdef MAGNUM_BUILD_DEPRECATED
Containers::Array<char> AbstractImageConverter::exportToData(const ImageView2D& image) {
return convertToData(image);
}
#endif
Containers::Array<char> AbstractImageConverter::convertToData(const ImageView3D& image) {
CORRADE_ASSERT(features() >= ImageConverterFeature::Convert3DToData,
"Trade::AbstractImageConverter::convertToData(): 3D image conversion not supported", nullptr);
#ifndef CORRADE_NO_ASSERT
/* Explicitly return if checks fail for CORRADE_GRACEFUL_ASSERT builds */
if(!checkImageValidity("Trade::AbstractImageConverter::convertToData():", image))
return {};
#endif
Containers::Array<char> out = doConvertToData(image);
CORRADE_ASSERT(!out.deleter(), "Trade::AbstractImageConverter::convertToData(): implementation is not allowed to use a custom Array deleter", {});
return out;
}
Containers::Array<char> AbstractImageConverter::doConvertToData(const ImageView3D& image) {
CORRADE_ASSERT(features() >= ImageConverterFeature::ConvertLevels3DToData,
"Trade::AbstractImageConverter::convertToData(): 3D image conversion advertised but not implemented", nullptr);
return doConvertToData(Containers::arrayView({image}));
}
Containers::Array<char> AbstractImageConverter::convertToData(const CompressedImageView1D& image) {
CORRADE_ASSERT(features() >= ImageConverterFeature::ConvertCompressed1DToData,
"Trade::AbstractImageConverter::convertToData(): compressed 1D image conversion not supported", nullptr);
#ifndef CORRADE_NO_ASSERT
/* Explicitly return if checks fail for CORRADE_GRACEFUL_ASSERT builds */
if(!checkImageValidity("Trade::AbstractImageConverter::convertToData():", image))
return {};
#endif
Containers::Array<char> out = doConvertToData(image);
CORRADE_ASSERT(!out.deleter(), "Trade::AbstractImageConverter::convertToData(): implementation is not allowed to use a custom Array deleter", {});
return out;
}
Containers::Array<char> AbstractImageConverter::doConvertToData(const CompressedImageView1D& image) {
CORRADE_ASSERT(features() >= ImageConverterFeature::ConvertCompressedLevels1DToData,
"Trade::AbstractImageConverter::convertToData(): compressed 1D image conversion advertised but not implemented", nullptr);
return doConvertToData(Containers::arrayView({image}));
}
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
Containers::Array<char> AbstractImageConverter::convertToData(const CompressedImageView2D& image) {
CORRADE_ASSERT(features() >= ImageConverterFeature::ConvertCompressed2DToData,
"Trade::AbstractImageConverter::convertToData(): compressed 2D image conversion not supported", nullptr);
#ifndef CORRADE_NO_ASSERT
/* Explicitly return if checks fail for CORRADE_GRACEFUL_ASSERT builds */
if(!checkImageValidity("Trade::AbstractImageConverter::convertToData():", image))
return {};
#endif
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
Containers::Array<char> out = doConvertToData(image);
CORRADE_ASSERT(!out.deleter(), "Trade::AbstractImageConverter::convertToData(): implementation is not allowed to use a custom Array deleter", {});
return out;
}
Containers::Array<char> AbstractImageConverter::doConvertToData(const CompressedImageView2D& image) {
CORRADE_ASSERT(features() >= ImageConverterFeature::ConvertCompressedLevels2DToData,
"Trade::AbstractImageConverter::convertToData(): compressed 2D image conversion advertised but not implemented", nullptr);
return doConvertToData(Containers::arrayView({image}));
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
}
#ifdef MAGNUM_BUILD_DEPRECATED
Containers::Array<char> AbstractImageConverter::exportToData(const CompressedImageView2D& image) {
return convertToData(image);
}
#endif
Containers::Array<char> AbstractImageConverter::convertToData(const CompressedImageView3D& image) {
CORRADE_ASSERT(features() >= ImageConverterFeature::ConvertCompressed3DToData,
"Trade::AbstractImageConverter::convertToData(): compressed 3D image conversion not supported", nullptr);
#ifndef CORRADE_NO_ASSERT
/* Explicitly return if checks fail for CORRADE_GRACEFUL_ASSERT builds */
if(!checkImageValidity("Trade::AbstractImageConverter::convertToData():", image))
return {};
#endif
Containers::Array<char> out = doConvertToData(image);
CORRADE_ASSERT(!out.deleter(), "Trade::AbstractImageConverter::convertToData(): implementation is not allowed to use a custom Array deleter", {});
return out;
}
Containers::Array<char> AbstractImageConverter::doConvertToData(const CompressedImageView3D& image) {
CORRADE_ASSERT(features() >= ImageConverterFeature::ConvertCompressedLevels3DToData,
"Trade::AbstractImageConverter::convertToData(): compressed 3D image conversion advertised but not implemented", nullptr);
return doConvertToData(Containers::arrayView({image}));
}
Containers::Array<char> AbstractImageConverter::convertToData(const ImageData1D& image) {
return image.isCompressed() ? convertToData(CompressedImageView1D(image)) : convertToData(ImageView1D(image));
}
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
Containers::Array<char> AbstractImageConverter::convertToData(const ImageData2D& image) {
return image.isCompressed() ? convertToData(CompressedImageView2D(image)) : convertToData(ImageView2D(image));
}
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
#ifdef MAGNUM_BUILD_DEPRECATED
Containers::Array<char> AbstractImageConverter::exportToData(const ImageData2D& image) {
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
return convertToData(image);
}
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
#endif
Containers::Array<char> AbstractImageConverter::convertToData(const ImageData3D& image) {
return image.isCompressed() ? convertToData(CompressedImageView3D(image)) : convertToData(ImageView3D(image));
}
#ifndef CORRADE_NO_ASSERT
namespace {
template<UnsignedInt dimensions> bool checkImageValidity(const char* const messagePrefix, const Containers::ArrayView<const BasicImageView<dimensions>> imageLevels) {
CORRADE_ASSERT(!imageLevels.empty(),
messagePrefix << "at least one image has to be specified", false);
const PixelFormat format = imageLevels[0].format();
const UnsignedInt formatExtra = imageLevels[0].formatExtra();
/* Going through *all* levels although the format assertion is never fired
in the first iteration in order to properly check also the first one for
zero size / nullptr. */
for(std::size_t i = 0; i != imageLevels.size(); ++i) {
CORRADE_ASSERT(imageLevels[i].size().product(),
messagePrefix << "can't convert image" << i << "with a zero size:" << imageLevels[i].size(), false);
CORRADE_ASSERT(imageLevels[i].data(),
messagePrefix << "can't convert image" << i << "with a nullptr view", false);
CORRADE_ASSERT(imageLevels[i].format() == format,
messagePrefix << "levels don't have the same format, expected" << format << "but got" << imageLevels[i].format() << "for image" << i, false);
CORRADE_ASSERT(imageLevels[i].formatExtra() == formatExtra,
messagePrefix << "levels don't have the same extra format field, expected" << formatExtra << "but got" << imageLevels[i].formatExtra() << "for image" << i, false);
}
return true;
}
template<UnsignedInt dimensions> bool checkImageValidity(const char* const messagePrefix, const Containers::ArrayView<const BasicCompressedImageView<dimensions>> imageLevels) {
CORRADE_ASSERT(!imageLevels.empty(),
messagePrefix << "at least one image has to be specified", false);
const CompressedPixelFormat format = imageLevels[0].format();
/* Going through *all* levels although the format assertion is never fired
in the first iteration in order to properly check also the first one for
zero size / nullptr. */
for(std::size_t i = 0; i != imageLevels.size(); ++i) {
CORRADE_ASSERT(imageLevels[i].size().product(),
messagePrefix << "can't convert image" << i << "with a zero size:" << imageLevels[i].size(), false);
CORRADE_ASSERT(imageLevels[i].data(),
messagePrefix << "can't convert image" << i << "with a nullptr view", false);
CORRADE_ASSERT(imageLevels[i].format() == format,
messagePrefix << "levels don't have the same format, expected" << format << "but got" << imageLevels[i].format() << "for image" << i, false);
}
return true;
}
}
#endif
Containers::Array<char> AbstractImageConverter::convertToData(const Containers::ArrayView<const ImageView1D> imageLevels) {
CORRADE_ASSERT(features() >= ImageConverterFeature::ConvertLevels1DToData,
"Trade::AbstractImageConverter::convertToData(): multi-level 1D image conversion not supported", nullptr);
#ifndef CORRADE_NO_ASSERT
/* Explicitly return if checks fail for CORRADE_GRACEFUL_ASSERT builds */
if(!checkImageValidity("Trade::AbstractImageConverter::convertToData():", imageLevels))
return {};
#endif
Containers::Array<char> out = doConvertToData(imageLevels);
CORRADE_ASSERT(!out.deleter(), "Trade::AbstractImageConverter::convertToData(): implementation is not allowed to use a custom Array deleter", {});
return out;
}
Containers::Array<char> AbstractImageConverter::convertToData(const std::initializer_list<ImageView1D> imageLevels) {
return convertToData(Containers::arrayView(imageLevels));
}
Containers::Array<char> AbstractImageConverter::doConvertToData(Containers::ArrayView<const ImageView1D>) {
CORRADE_ASSERT_UNREACHABLE("Trade::AbstractImageConverter::convertToData(): multi-level 1D image conversion advertised but not implemented", nullptr);
}
Containers::Array<char> AbstractImageConverter::convertToData(const Containers::ArrayView<const ImageView2D> imageLevels) {
CORRADE_ASSERT(features() >= ImageConverterFeature::ConvertLevels2DToData,
"Trade::AbstractImageConverter::convertToData(): multi-level 2D image conversion not supported", nullptr);
#ifndef CORRADE_NO_ASSERT
/* Explicitly return if checks fail for CORRADE_GRACEFUL_ASSERT builds */
if(!checkImageValidity("Trade::AbstractImageConverter::convertToData():", imageLevels))
return {};
#endif
Containers::Array<char> out = doConvertToData(imageLevels);
CORRADE_ASSERT(!out.deleter(), "Trade::AbstractImageConverter::convertToData(): implementation is not allowed to use a custom Array deleter", {});
return out;
}
Containers::Array<char> AbstractImageConverter::convertToData(const std::initializer_list<ImageView2D> imageLevels) {
return convertToData(Containers::arrayView(imageLevels));
}
Containers::Array<char> AbstractImageConverter::doConvertToData(Containers::ArrayView<const ImageView2D>) {
CORRADE_ASSERT_UNREACHABLE("Trade::AbstractImageConverter::convertToData(): multi-level 2D image conversion advertised but not implemented", nullptr);
}
Containers::Array<char> AbstractImageConverter::convertToData(const Containers::ArrayView<const ImageView3D> imageLevels) {
CORRADE_ASSERT(features() >= ImageConverterFeature::ConvertLevels3DToData,
"Trade::AbstractImageConverter::convertToData(): multi-level 3D image conversion not supported", nullptr);
#ifndef CORRADE_NO_ASSERT
/* Explicitly return if checks fail for CORRADE_GRACEFUL_ASSERT builds */
if(!checkImageValidity("Trade::AbstractImageConverter::convertToData():", imageLevels))
return {};
#endif
Containers::Array<char> out = doConvertToData(imageLevels);
CORRADE_ASSERT(!out.deleter(), "Trade::AbstractImageConverter::convertToData(): implementation is not allowed to use a custom Array deleter", {});
return out;
}
Containers::Array<char> AbstractImageConverter::convertToData(const std::initializer_list<ImageView3D> imageLevels) {
return convertToData(Containers::arrayView(imageLevels));
}
Containers::Array<char> AbstractImageConverter::doConvertToData(Containers::ArrayView<const ImageView3D>) {
CORRADE_ASSERT_UNREACHABLE("Trade::AbstractImageConverter::convertToData(): multi-level 3D image conversion advertised but not implemented", nullptr);
}
Containers::Array<char> AbstractImageConverter::convertToData(const Containers::ArrayView<const CompressedImageView1D> imageLevels) {
CORRADE_ASSERT(features() >= ImageConverterFeature::ConvertCompressedLevels1DToData,
"Trade::AbstractImageConverter::convertToData(): multi-level compressed 1D image conversion not supported", nullptr);
#ifndef CORRADE_NO_ASSERT
/* Explicitly return if checks fail for CORRADE_GRACEFUL_ASSERT builds */
if(!checkImageValidity("Trade::AbstractImageConverter::convertToData():", imageLevels))
return {};
#endif
Containers::Array<char> out = doConvertToData(imageLevels);
CORRADE_ASSERT(!out.deleter(), "Trade::AbstractImageConverter::convertToData(): implementation is not allowed to use a custom Array deleter", {});
return out;
}
Containers::Array<char> AbstractImageConverter::convertToData(const std::initializer_list<CompressedImageView1D> imageLevels) {
return convertToData(Containers::arrayView(imageLevels));
}
Containers::Array<char> AbstractImageConverter::doConvertToData(Containers::ArrayView<const CompressedImageView1D>) {
CORRADE_ASSERT_UNREACHABLE("Trade::AbstractImageConverter::convertToData(): multi-level compressed 1D image conversion advertised but not implemented", nullptr);
}
Containers::Array<char> AbstractImageConverter::convertToData(const Containers::ArrayView<const CompressedImageView2D> imageLevels) {
CORRADE_ASSERT(features() >= ImageConverterFeature::ConvertCompressedLevels2DToData,
"Trade::AbstractImageConverter::convertToData(): multi-level compressed 2D image conversion not supported", nullptr);
#ifndef CORRADE_NO_ASSERT
/* Explicitly return if checks fail for CORRADE_GRACEFUL_ASSERT builds */
if(!checkImageValidity("Trade::AbstractImageConverter::convertToData():", imageLevels))
return {};
#endif
Containers::Array<char> out = doConvertToData(imageLevels);
CORRADE_ASSERT(!out.deleter(), "Trade::AbstractImageConverter::convertToData(): implementation is not allowed to use a custom Array deleter", {});
return out;
}
Containers::Array<char> AbstractImageConverter::convertToData(const std::initializer_list<CompressedImageView2D> imageLevels) {
return convertToData(Containers::arrayView(imageLevels));
}
Containers::Array<char> AbstractImageConverter::doConvertToData(Containers::ArrayView<const CompressedImageView2D>) {
CORRADE_ASSERT_UNREACHABLE("Trade::AbstractImageConverter::convertToData(): multi-level compressed 2D image conversion advertised but not implemented", nullptr);
}
Containers::Array<char> AbstractImageConverter::convertToData(const Containers::ArrayView<const CompressedImageView3D> imageLevels) {
CORRADE_ASSERT(features() >= ImageConverterFeature::ConvertCompressedLevels3DToData,
"Trade::AbstractImageConverter::convertToData(): multi-level compressed 3D image conversion not supported", nullptr);
#ifndef CORRADE_NO_ASSERT
/* Explicitly return if checks fail for CORRADE_GRACEFUL_ASSERT builds */
if(!checkImageValidity("Trade::AbstractImageConverter::convertToData():", imageLevels))
return {};
#endif
Containers::Array<char> out = doConvertToData(imageLevels);
CORRADE_ASSERT(!out.deleter(), "Trade::AbstractImageConverter::convertToData(): implementation is not allowed to use a custom Array deleter", {});
return out;
}
Containers::Array<char> AbstractImageConverter::convertToData(const std::initializer_list<CompressedImageView3D> imageLevels) {
return convertToData(Containers::arrayView(imageLevels));
}
Containers::Array<char> AbstractImageConverter::doConvertToData(Containers::ArrayView<const CompressedImageView3D>) {
CORRADE_ASSERT_UNREACHABLE("Trade::AbstractImageConverter::convertToData(): multi-level compressed 3D image conversion advertised but not implemented", nullptr);
}
bool AbstractImageConverter::convertToFile(const ImageView1D& image, const Containers::StringView filename) {
CORRADE_ASSERT(features() & ImageConverterFeature::Convert1DToFile,
"Trade::AbstractImageConverter::convertToFile(): 1D image conversion not supported", {});
#ifndef CORRADE_NO_ASSERT
/* Explicitly return if checks fail for CORRADE_GRACEFUL_ASSERT builds */
if(!checkImageValidity("Trade::AbstractImageConverter::convertToFile():", image))
return {};
#endif
return doConvertToFile(image, filename);
}
bool AbstractImageConverter::doConvertToFile(const ImageView1D& image, const Containers::StringView filename) {
/* Prefer to go through the ToFile variant instead of ToData assuming that
it could have a more memory-efficient implementation than having to
materialize the whole output in memory first */
if(features() >= ImageConverterFeature::ConvertLevels1DToFile)
return doConvertToFile(Containers::arrayView({image}), filename);
CORRADE_ASSERT(features() >= ImageConverterFeature::Convert1DToData, "Trade::AbstractImageConverter::convertToFile(): 1D image conversion advertised but not implemented", false);
const auto data = doConvertToData(image);
/* No deleter checks as it doesn't matter here */
if(!data) return false;
if(!Utility::Directory::write(filename, data)) {
Error() << "Trade::AbstractImageConverter::convertToFile(): cannot write to file" << filename;
return false;
}
return true;
}
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
bool AbstractImageConverter::convertToFile(const ImageView2D& image, const Containers::StringView filename) {
CORRADE_ASSERT(features() & ImageConverterFeature::Convert2DToFile,
"Trade::AbstractImageConverter::convertToFile(): 2D image conversion not supported", {});
#ifndef CORRADE_NO_ASSERT
/* Explicitly return if checks fail for CORRADE_GRACEFUL_ASSERT builds */
if(!checkImageValidity("Trade::AbstractImageConverter::convertToFile():", image))
return {};
#endif
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
return doConvertToFile(image, filename);
}
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
bool AbstractImageConverter::doConvertToFile(const ImageView2D& image, const Containers::StringView filename) {
/* Prefer to go through the ToFile variant instead of ToData assuming that
it could have a more memory-efficient implementation than having to
materialize the whole output in memory first */
if(features() >= ImageConverterFeature::ConvertLevels2DToFile)
return doConvertToFile(Containers::arrayView({image}), filename);
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
CORRADE_ASSERT(features() >= ImageConverterFeature::Convert2DToData, "Trade::AbstractImageConverter::convertToFile(): 2D image conversion advertised but not implemented", false);
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
const auto data = doConvertToData(image);
/* No deleter checks as it doesn't matter here */
if(!data) return false;
if(!Utility::Directory::write(filename, data)) {
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
Error() << "Trade::AbstractImageConverter::convertToFile(): cannot write to file" << filename;
return false;
}
return true;
}
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
#ifdef MAGNUM_BUILD_DEPRECATED
bool AbstractImageConverter::exportToFile(const ImageView2D& image, const std::string& filename) {
return convertToFile(image, filename);
}
#endif
bool AbstractImageConverter::convertToFile(const ImageView3D& image, const Containers::StringView filename) {
CORRADE_ASSERT(features() & ImageConverterFeature::Convert3DToFile,
"Trade::AbstractImageConverter::convertToFile(): 3D image conversion not supported", {});
#ifndef CORRADE_NO_ASSERT
/* Explicitly return if checks fail for CORRADE_GRACEFUL_ASSERT builds */
if(!checkImageValidity("Trade::AbstractImageConverter::convertToFile():", image))
return {};
#endif
return doConvertToFile(image, filename);
}
bool AbstractImageConverter::doConvertToFile(const ImageView3D& image, const Containers::StringView filename) {
/* Prefer to go through the ToFile variant instead of ToData assuming that
it could have a more memory-efficient implementation than having to
materialize the whole output in memory first */
if(features() >= ImageConverterFeature::ConvertLevels3DToFile)
return doConvertToFile(Containers::arrayView({image}), filename);
CORRADE_ASSERT(features() >= ImageConverterFeature::Convert3DToData, "Trade::AbstractImageConverter::convertToFile(): 3D image conversion advertised but not implemented", false);
const auto data = doConvertToData(image);
/* No deleter checks as it doesn't matter here */
if(!data) return false;
if(!Utility::Directory::write(filename, data)) {
Error() << "Trade::AbstractImageConverter::convertToFile(): cannot write to file" << filename;
return false;
}
return true;
}
bool AbstractImageConverter::convertToFile(const CompressedImageView1D& image, const Containers::StringView filename) {
CORRADE_ASSERT(features() & ImageConverterFeature::ConvertCompressed1DToFile,
"Trade::AbstractImageConverter::convertToFile(): compressed 1D image conversion not supported", {});
#ifndef CORRADE_NO_ASSERT
/* Explicitly return if checks fail for CORRADE_GRACEFUL_ASSERT builds */
if(!checkImageValidity("Trade::AbstractImageConverter::convertToFile():", image))
return {};
#endif
return doConvertToFile(image, filename);
}
bool AbstractImageConverter::doConvertToFile(const CompressedImageView1D& image, Containers::StringView filename) {
/* Prefer to go through the ToFile variant instead of ToData assuming that
it could have a more memory-efficient implementation than having to
materialize the whole output in memory first */
if(features() >= ImageConverterFeature::ConvertCompressedLevels1DToFile)
return doConvertToFile(Containers::arrayView({image}), filename);
CORRADE_ASSERT(features() >= ImageConverterFeature::ConvertCompressed1DToData, "Trade::AbstractImageConverter::convertToFile(): compressed 1D image conversion advertised but not implemented", false);
const auto data = doConvertToData(image);
/* No deleter checks as it doesn't matter here */
if(!data) return false;
if(!Utility::Directory::write(filename, data)) {
Error() << "Trade::AbstractImageConverter::convertToFile(): cannot write to file" << filename;
return false;
}
return true;
}
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
bool AbstractImageConverter::convertToFile(const CompressedImageView2D& image, const Containers::StringView filename) {
CORRADE_ASSERT(features() & ImageConverterFeature::ConvertCompressed2DToFile,
"Trade::AbstractImageConverter::convertToFile(): compressed 2D image conversion not supported", {});
#ifndef CORRADE_NO_ASSERT
/* Explicitly return if checks fail for CORRADE_GRACEFUL_ASSERT builds */
if(!checkImageValidity("Trade::AbstractImageConverter::convertToFile():", image))
return {};
#endif
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
return doConvertToFile(image, filename);
}
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
bool AbstractImageConverter::doConvertToFile(const CompressedImageView2D& image, Containers::StringView filename) {
/* Prefer to go through the ToFile variant instead of ToData assuming that
it could have a more memory-efficient implementation than having to
materialize the whole output in memory first */
if(features() >= ImageConverterFeature::ConvertCompressedLevels2DToFile)
return doConvertToFile(Containers::arrayView({image}), filename);
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
CORRADE_ASSERT(features() >= ImageConverterFeature::ConvertCompressed2DToData, "Trade::AbstractImageConverter::convertToFile(): compressed 2D image conversion advertised but not implemented", false);
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
const auto data = doConvertToData(image);
/* No deleter checks as it doesn't matter here */
if(!data) return false;
if(!Utility::Directory::write(filename, data)) {
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
Error() << "Trade::AbstractImageConverter::convertToFile(): cannot write to file" << filename;
return false;
}
return true;
}
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
#ifdef MAGNUM_BUILD_DEPRECATED
bool AbstractImageConverter::exportToFile(const CompressedImageView2D& image, const std::string& filename) {
return convertToFile(image, filename);
}
#endif
bool AbstractImageConverter::convertToFile(const CompressedImageView3D& image, const Containers::StringView filename) {
CORRADE_ASSERT(features() & ImageConverterFeature::ConvertCompressed3DToFile,
"Trade::AbstractImageConverter::convertToFile(): compressed 3D image conversion not supported", {});
#ifndef CORRADE_NO_ASSERT
/* Explicitly return if checks fail for CORRADE_GRACEFUL_ASSERT builds */
if(!checkImageValidity("Trade::AbstractImageConverter::convertToFile():", image))
return {};
#endif
return doConvertToFile(image, filename);
}
bool AbstractImageConverter::doConvertToFile(const CompressedImageView3D& image, Containers::StringView filename) {
/* Prefer to go through the ToFile variant instead of ToData assuming that
it could have a more memory-efficient implementation than having to
materialize the whole output in memory first */
if(features() >= ImageConverterFeature::ConvertCompressedLevels3DToFile)
return doConvertToFile(Containers::arrayView({image}), filename);
CORRADE_ASSERT(features() >= ImageConverterFeature::ConvertCompressed3DToData, "Trade::AbstractImageConverter::convertToFile(): compressed 3D image conversion advertised but not implemented", false);
const auto data = doConvertToData(image);
/* No deleter checks as it doesn't matter here */
if(!data) return false;
if(!Utility::Directory::write(filename, data)) {
Error() << "Trade::AbstractImageConverter::convertToFile(): cannot write to file" << filename;
return false;
}
return true;
}
bool AbstractImageConverter::convertToFile(const ImageData1D& image, const Containers::StringView filename) {
return image.isCompressed() ? convertToFile(CompressedImageView1D(image), filename) : convertToFile(ImageView1D(image), filename);
}
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
bool AbstractImageConverter::convertToFile(const ImageData2D& image, const Containers::StringView filename) {
return image.isCompressed() ? convertToFile(CompressedImageView2D(image), filename) : convertToFile(ImageView2D(image), filename);
}
#ifdef MAGNUM_BUILD_DEPRECATED
bool AbstractImageConverter::exportToFile(const ImageData2D& image, const std::string& filename) {
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
return convertToFile(image, filename);
}
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
#endif
bool AbstractImageConverter::convertToFile(const ImageData3D& image, const Containers::StringView filename) {
return image.isCompressed() ? convertToFile(CompressedImageView3D(image), filename) : convertToFile(ImageView3D(image), filename);
}
bool AbstractImageConverter::convertToFile(const Containers::ArrayView<const ImageView1D> imageLevels, const Containers::StringView filename) {
CORRADE_ASSERT(features() & ImageConverterFeature::ConvertLevels1DToFile,
"Trade::AbstractImageConverter::convertToFile(): multi-level 1D image conversion not supported", {});
#ifndef CORRADE_NO_ASSERT
/* Explicitly return if checks fail for CORRADE_GRACEFUL_ASSERT builds */
if(!checkImageValidity("Trade::AbstractImageConverter::convertToFile():", imageLevels))
return {};
#endif
return doConvertToFile(imageLevels, filename);
}
bool AbstractImageConverter::convertToFile(const std::initializer_list<ImageView1D> imageLevels, const Containers::StringView filename) {
return convertToFile(Containers::arrayView(imageLevels), filename);
}
bool AbstractImageConverter::doConvertToFile(const Containers::ArrayView<const ImageView1D> imageLevels, const Containers::StringView filename) {
CORRADE_ASSERT(features() >= ImageConverterFeature::ConvertLevels1DToData, "Trade::AbstractImageConverter::convertToFile(): multi-level 1D image conversion advertised but not implemented", false);
const auto data = doConvertToData(imageLevels);
/* No deleter checks as it doesn't matter here */
if(!data) return false;
if(!Utility::Directory::write(filename, data)) {
Error() << "Trade::AbstractImageConverter::convertToFile(): cannot write to file" << filename;
return false;
}
return true;
}
bool AbstractImageConverter::convertToFile(const Containers::ArrayView<const ImageView2D> imageLevels, const Containers::StringView filename) {
CORRADE_ASSERT(features() & ImageConverterFeature::ConvertLevels2DToFile,
"Trade::AbstractImageConverter::convertToFile(): multi-level 2D image conversion not supported", {});
#ifndef CORRADE_NO_ASSERT
/* Explicitly return if checks fail for CORRADE_GRACEFUL_ASSERT builds */
if(!checkImageValidity("Trade::AbstractImageConverter::convertToFile():", imageLevels))
return {};
#endif
return doConvertToFile(imageLevels, filename);
}
bool AbstractImageConverter::convertToFile(const std::initializer_list<ImageView2D> imageLevels, const Containers::StringView filename) {
return convertToFile(Containers::arrayView(imageLevels), filename);
}
bool AbstractImageConverter::doConvertToFile(const Containers::ArrayView<const ImageView2D> imageLevels, const Containers::StringView filename) {
CORRADE_ASSERT(features() >= ImageConverterFeature::ConvertLevels2DToData, "Trade::AbstractImageConverter::convertToFile(): multi-level 2D image conversion advertised but not implemented", false);
const auto data = doConvertToData(imageLevels);
/* No deleter checks as it doesn't matter here */
if(!data) return false;
if(!Utility::Directory::write(filename, data)) {
Error() << "Trade::AbstractImageConverter::convertToFile(): cannot write to file" << filename;
return false;
}
return true;
}
bool AbstractImageConverter::convertToFile(const Containers::ArrayView<const ImageView3D> imageLevels, const Containers::StringView filename) {
CORRADE_ASSERT(features() & ImageConverterFeature::ConvertLevels3DToFile,
"Trade::AbstractImageConverter::convertToFile(): multi-level 3D image conversion not supported", {});
#ifndef CORRADE_NO_ASSERT
/* Explicitly return if checks fail for CORRADE_GRACEFUL_ASSERT builds */
if(!checkImageValidity("Trade::AbstractImageConverter::convertToFile():", imageLevels))
return {};
#endif
return doConvertToFile(imageLevels, filename);
}
bool AbstractImageConverter::convertToFile(const std::initializer_list<ImageView3D> imageLevels, const Containers::StringView filename) {
return convertToFile(Containers::arrayView(imageLevels), filename);
}
bool AbstractImageConverter::doConvertToFile(const Containers::ArrayView<const ImageView3D> imageLevels, const Containers::StringView filename) {
CORRADE_ASSERT(features() >= ImageConverterFeature::ConvertLevels3DToData, "Trade::AbstractImageConverter::convertToFile(): multi-level 3D image conversion advertised but not implemented", false);
const auto data = doConvertToData(imageLevels);
/* No deleter checks as it doesn't matter here */
if(!data) return false;
if(!Utility::Directory::write(filename, data)) {
Error() << "Trade::AbstractImageConverter::convertToFile(): cannot write to file" << filename;
return false;
}
return true;
}
bool AbstractImageConverter::convertToFile(const Containers::ArrayView<const CompressedImageView1D> imageLevels, const Containers::StringView filename) {
CORRADE_ASSERT(features() & ImageConverterFeature::ConvertCompressedLevels1DToFile,
"Trade::AbstractImageConverter::convertToFile(): multi-level compressed 1D image conversion not supported", {});
#ifndef CORRADE_NO_ASSERT
/* Explicitly return if checks fail for CORRADE_GRACEFUL_ASSERT builds */
if(!checkImageValidity("Trade::AbstractImageConverter::convertToFile():", imageLevels))
return {};
#endif
return doConvertToFile(imageLevels, filename);
}
bool AbstractImageConverter::convertToFile(const std::initializer_list<CompressedImageView1D> imageLevels, const Containers::StringView filename) {
return convertToFile(Containers::arrayView(imageLevels), filename);
}
bool AbstractImageConverter::doConvertToFile(const Containers::ArrayView<const CompressedImageView1D> imageLevels, Containers::StringView filename) {
CORRADE_ASSERT(features() >= ImageConverterFeature::ConvertCompressedLevels1DToData, "Trade::AbstractImageConverter::convertToFile(): multi-level compressed 1D image conversion advertised but not implemented", false);
const auto data = doConvertToData(imageLevels);
/* No deleter checks as it doesn't matter here */
if(!data) return false;
if(!Utility::Directory::write(filename, data)) {
Error() << "Trade::AbstractImageConverter::convertToFile(): cannot write to file" << filename;
return false;
}
return true;
}
bool AbstractImageConverter::convertToFile(const Containers::ArrayView<const CompressedImageView2D> imageLevels, const Containers::StringView filename) {
CORRADE_ASSERT(features() & ImageConverterFeature::ConvertCompressedLevels2DToFile,
"Trade::AbstractImageConverter::convertToFile(): multi-level compressed 2D image conversion not supported", {});
#ifndef CORRADE_NO_ASSERT
/* Explicitly return if checks fail for CORRADE_GRACEFUL_ASSERT builds */
if(!checkImageValidity("Trade::AbstractImageConverter::convertToFile():", imageLevels))
return {};
#endif
return doConvertToFile(imageLevels, filename);
}
bool AbstractImageConverter::convertToFile(const std::initializer_list<CompressedImageView2D> imageLevels, const Containers::StringView filename) {
return convertToFile(Containers::arrayView(imageLevels), filename);
}
bool AbstractImageConverter::doConvertToFile(const Containers::ArrayView<const CompressedImageView2D> imageLevels, Containers::StringView filename) {
CORRADE_ASSERT(features() >= ImageConverterFeature::ConvertCompressedLevels2DToData, "Trade::AbstractImageConverter::convertToFile(): multi-level compressed 2D image conversion advertised but not implemented", false);
const auto data = doConvertToData(imageLevels);
/* No deleter checks as it doesn't matter here */
if(!data) return false;
if(!Utility::Directory::write(filename, data)) {
Error() << "Trade::AbstractImageConverter::convertToFile(): cannot write to file" << filename;
return false;
}
return true;
}
bool AbstractImageConverter::convertToFile(const Containers::ArrayView<const CompressedImageView3D> imageLevels, const Containers::StringView filename) {
CORRADE_ASSERT(features() & ImageConverterFeature::ConvertCompressedLevels3DToFile,
"Trade::AbstractImageConverter::convertToFile(): multi-level compressed 3D image conversion not supported", {});
#ifndef CORRADE_NO_ASSERT
/* Explicitly return if checks fail for CORRADE_GRACEFUL_ASSERT builds */
if(!checkImageValidity("Trade::AbstractImageConverter::convertToFile():", imageLevels))
return {};
#endif
return doConvertToFile(imageLevels, filename);
}
bool AbstractImageConverter::convertToFile(const std::initializer_list<CompressedImageView3D> imageLevels, const Containers::StringView filename) {
return convertToFile(Containers::arrayView(imageLevels), filename);
}
bool AbstractImageConverter::doConvertToFile(const Containers::ArrayView<const CompressedImageView3D> imageLevels, Containers::StringView filename) {
CORRADE_ASSERT(features() >= ImageConverterFeature::ConvertCompressedLevels3DToData, "Trade::AbstractImageConverter::convertToFile(): multi-level compressed 3D image conversion advertised but not implemented", false);
const auto data = doConvertToData(imageLevels);
/* No deleter checks as it doesn't matter here */
if(!data) return false;
if(!Utility::Directory::write(filename, data)) {
Error() << "Trade::AbstractImageConverter::convertToFile(): cannot write to file" << filename;
return false;
}
return true;
}
Debug& operator<<(Debug& debug, const ImageConverterFeature value) {
debug << "Trade::ImageConverterFeature" << Debug::nospace;
switch(value) {
/* LCOV_EXCL_START */
#define _c(v) case ImageConverterFeature::v: return debug << "::" #v;
_c(Convert1D)
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
_c(Convert2D)
_c(Convert3D)
_c(ConvertCompressed1D)
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
_c(ConvertCompressed2D)
_c(ConvertCompressed3D)
_c(Convert1DToFile)
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
_c(Convert2DToFile)
_c(Convert3DToFile)
_c(ConvertCompressed1DToFile)
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
_c(ConvertCompressed2DToFile)
_c(ConvertCompressed3DToFile)
_c(Convert1DToData)
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
_c(Convert2DToData)
_c(Convert3DToData)
_c(ConvertCompressed1DToData)
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
_c(ConvertCompressed2DToData)
_c(ConvertCompressed3DToData)
_c(ConvertLevels1DToFile)
_c(ConvertLevels2DToFile)
_c(ConvertLevels3DToFile)
_c(ConvertCompressedLevels1DToFile)
_c(ConvertCompressedLevels2DToFile)
_c(ConvertCompressedLevels3DToFile)
_c(ConvertLevels1DToData)
_c(ConvertLevels2DToData)
_c(ConvertLevels3DToData)
_c(ConvertCompressedLevels1DToData)
_c(ConvertCompressedLevels2DToData)
_c(ConvertCompressedLevels3DToData)
#undef _c
/* LCOV_EXCL_STOP */
}
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
return debug << "(" << Debug::nospace << reinterpret_cast<void*>(UnsignedInt(value)) << Debug::nospace << ")";
}
Debug& operator<<(Debug& debug, const ImageConverterFeatures value) {
return Containers::enumSetDebugOutput(debug, value, "Trade::ImageConverterFeatures{}", {
ImageConverterFeature::Convert1D,
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
ImageConverterFeature::Convert2D,
ImageConverterFeature::Convert3D,
ImageConverterFeature::ConvertCompressed1D,
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
ImageConverterFeature::ConvertCompressed2D,
ImageConverterFeature::ConvertCompressed3D,
ImageConverterFeature::ConvertLevels1DToData,
ImageConverterFeature::ConvertLevels2DToData,
ImageConverterFeature::ConvertLevels3DToData,
ImageConverterFeature::ConvertCompressedLevels1DToData,
ImageConverterFeature::ConvertCompressedLevels2DToData,
ImageConverterFeature::ConvertCompressedLevels3DToData,
/* These 6 are implied by Convert[Compressed]LevelsToData, so have to
be after */
ImageConverterFeature::ConvertLevels1DToFile,
ImageConverterFeature::ConvertLevels2DToFile,
ImageConverterFeature::ConvertLevels3DToFile,
ImageConverterFeature::ConvertCompressedLevels1DToFile,
ImageConverterFeature::ConvertCompressedLevels2DToFile,
ImageConverterFeature::ConvertCompressedLevels3DToFile,
/* These 12 are implied by Convert[Compressed]LevelsTo{File,Data}, so
have to be after */
ImageConverterFeature::Convert1DToData,
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
ImageConverterFeature::Convert2DToData,
ImageConverterFeature::Convert3DToData,
ImageConverterFeature::ConvertCompressed1DToData,
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
ImageConverterFeature::ConvertCompressed2DToData,
ImageConverterFeature::ConvertCompressed3DToData,
/* These 6 are implied by Convert[Compressed]ToData, so have to be
after */
ImageConverterFeature::Convert1DToFile,
Trade: refresh the AbstractImageConverter API. First and foremost I need to expand the interface to support 3D image conversion. But the interface was not great to begin with, so this takes the opportunity of an API break and does several things: * The `export*()` names were rather strange and I don't even remember why I chose that name (maybe because at first I wanted to have an "exporter" API as a counterpart to importers?) * In addition, there was no way to convert a compressed image to a compressed image (or to an uncompressed image) and adding the two missing variants would be a lot of combinations. So instead the new convert() returns an ImageData, which can be both, and thus also allows the converters to produce compressed or uncompressed output based on some runtime setting, without having to implement two (four?) separate functions for that and requiring users to know beforehand what type of an image will be created. * The ImageConverterFeature enum was named in a really strange way as well, with ConvertCompressedImage meaning "convert to a compressed image" while "ConvertCompressedData" instead meant "convert a compressed image to a data". Utter chaos. It also all implied 2D and on the other hand had a redundant `Image` in the name, so I went and remade the whole thing. As mentioned above, two of the enums now mean the same thing, and are both replaced with Convert2D. * Finally, similarly as changes elsewhere, I took this opportunity to get rid of std::string in the convertToFile() APIs.
5 years ago
ImageConverterFeature::Convert2DToFile,
ImageConverterFeature::Convert3DToFile,
ImageConverterFeature::ConvertCompressed1DToFile,
ImageConverterFeature::ConvertCompressed2DToFile,
ImageConverterFeature::ConvertCompressed3DToFile});
}
Debug& operator<<(Debug& debug, const ImageConverterFlag value) {
debug << "Trade::ImageConverterFlag" << Debug::nospace;
switch(value) {
/* LCOV_EXCL_START */
#define _c(v) case ImageConverterFlag::v: return debug << "::" #v;
_c(Verbose)
#undef _c
/* LCOV_EXCL_STOP */
}
return debug << "(" << Debug::nospace << reinterpret_cast<void*>(UnsignedByte(value)) << Debug::nospace << ")";
}
Debug& operator<<(Debug& debug, const ImageConverterFlags value) {
return Containers::enumSetDebugOutput(debug, value, "Trade::ImageConverterFlags{}", {
ImageConverterFlag::Verbose});
}
}}