terminal/src/inc/til/color.h

220 lines
8 KiB
C
Raw Normal View History

// Copyright (c) Microsoft Corporation.
// Licensed under the MIT license.
#pragma once
namespace til // Terminal Implementation Library. Also: "Today I Learned"
{
// color is a universal integral 8bpp RGBA (0-255) color type implicitly convertible to/from
// a number of other color types.
#pragma warning(push)
// we can't depend on GSL here, so we use static_cast for explicit narrowing
#pragma warning(disable : 26472)
struct color
{
// Clang (10) has no trouble optimizing the COLORREF conversion operator, below, to a
// simple 32-bit load with mask (!) even though it's a series of bit shifts across
// multiple struct members.
// CL (19.24) doesn't make the same optimization decision, and it emits three 8-bit loads
// and some shifting.
// In any case, the optimization only applies at -O2 (clang) and above.
// Here, we leverage the spec-legality of using unions for type conversions and the
// overlap of four uint8_ts and a uint32_t to make the conversion very obvious to
// both compilers.
#pragma warning(push)
// CL will complain about the both nameless and anonymous struct.
#pragma warning(disable : 4201)
union
{
struct
{
#if defined(__BYTE_ORDER__) && __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__ // Clang, GCC
uint8_t a, b, g, r;
#else
uint8_t r, g, b, a;
#endif
};
uint32_t abgr;
};
#pragma warning(pop)
constexpr color() noexcept :
r{ 0 },
g{ 0 },
b{ 0 },
a{ 0 }
{
}
constexpr color(uint8_t _r, uint8_t _g, uint8_t _b) noexcept :
r{ _r },
g{ _g },
b{ _b },
a{ 255 } {}
constexpr color(uint8_t _r, uint8_t _g, uint8_t _b, uint8_t _a) noexcept :
r{ _r },
g{ _g },
b{ _b },
a{ _a } {}
constexpr color(const color&) = default;
constexpr color(color&&) = default;
color& operator=(const color&) = default;
color& operator=(color&&) = default;
~color() = default;
#ifdef _WINDEF_
constexpr color(COLORREF c) :
abgr{ static_cast<uint32_t>(c | 0xFF000000u) }
{
}
Reduce usage of Json::Value throughout Terminal.Settings.Model (#11184) This commit reduces the code surface that interacts with raw JSON data, reducing code complexity and improving maintainability. Files that needed to be changed drastically were additionally cleaned up to remove any code cruft that has accrued over time. In order to facility this the following changes were made: * Move JSON handling from `CascadiaSettings` into `SettingsLoader` This allows us to use STL containers for data model instances. For instance profiles are now added to a hashmap for O(1) lookup. * JSON parsing within `SettingsLoader` doesn't differentiate between user, inbox and fragment JSON data, reducing code complexity and size. It also centralizes common concerns, like profile deduplication and ensuring that all profiles are assigned a GUID. * Direct JSON modification, like the insertion of dynamic profiles into settings.json were removed. This vastly reduces code complexity, but unfortunately removes support for comments in JSON on first start. * `ColorScheme`s cannot be layered. As such its `LayerJson` API was replaced with `FromJson`, allowing us to remove JSON-based color scheme validation. * `Profile`s used to test their wish to layer using `ShouldBeLayered`, which was replaced with a GUID-based hashmap lookup on previously parsed profiles. Further changes were made as improvements upon the previous changes: * Compact the JSON files embedded binary, saving 28kB * Prevent double-initialization of the color table in `ColorScheme` * Making `til::color` getters `constexpr`, allow better optimizations The result is a reduction of: * 48kB binary size for the Settings.Model.dll * 5-10% startup duration * 26% code for the `CascadiaSettings` class * 1% overall code in this project Furthermore this results in the following breaking changes: * The long deprecated "globals" settings object will not be detected and no warning will be created during load. * The initial creation of a new settings.json will not produce helpful comments. Both cases are caused by the removal of manual JSON handling and the move to representing the settings file with model objects instead ## PR Checklist * [x] Closes #5276 * [x] Closes #7421 * [x] I work here * [x] Tests added/passed ## Validation Steps Performed * Out-of-box-experience is identical to before ✔️ (Except for the settings.json file lacking comments.) * Existing user settings load correctly ✔️ * New WSL instances are added to user settings ✔️ * New fragments are added to user settings ✔️ * All profiles are assigned GUIDs ✔️
2021-09-22 18:27:31 +02:00
constexpr operator COLORREF() const noexcept
{
return static_cast<COLORREF>(abgr & 0x00FFFFFFu);
}
#endif
// Method Description:
// - Converting constructor for any other color structure type containing integral R, G, B, A (case sensitive.)
// Notes:
// - This and all below conversions make use of std::enable_if and a default parameter to disambiguate themselves.
// enable_if will result in an <error-type> if the constraint within it is not met, which will make this
// template ill-formed. Because SFINAE, ill-formed templated members "disappear" instead of causing an error.
template<typename TOther>
constexpr color(const TOther& other, std::enable_if_t<std::is_integral_v<decltype(std::declval<TOther>().R)> && std::is_integral_v<decltype(std::declval<TOther>().A)>, int> /*sentinel*/ = 0) :
r{ static_cast<uint8_t>(other.R) },
g{ static_cast<uint8_t>(other.G) },
b{ static_cast<uint8_t>(other.B) },
a{ static_cast<uint8_t>(other.A) }
{
}
// Method Description:
// - Converting constructor for any other color structure type containing integral r, g, b, a (case sensitive.)
template<typename TOther>
constexpr color(const TOther& other, std::enable_if_t<std::is_integral_v<decltype(std::declval<TOther>().r)> && std::is_integral_v<decltype(std::declval<TOther>().a)>, int> /*sentinel*/ = 0) :
r{ static_cast<uint8_t>(other.r) },
g{ static_cast<uint8_t>(other.g) },
b{ static_cast<uint8_t>(other.b) },
a{ static_cast<uint8_t>(other.a) }
{
}
// Method Description:
// - Converting constructor for any other color structure type containing floating-point R, G, B, A (case sensitive.)
template<typename TOther>
constexpr color(const TOther& other, std::enable_if_t<std::is_floating_point_v<decltype(std::declval<TOther>().R)> && std::is_floating_point_v<decltype(std::declval<TOther>().A)>, float> /*sentinel*/ = 1.0f) :
r{ static_cast<uint8_t>(other.R * 255.0f) },
g{ static_cast<uint8_t>(other.G * 255.0f) },
b{ static_cast<uint8_t>(other.B * 255.0f) },
a{ static_cast<uint8_t>(other.A * 255.0f) }
{
}
// Method Description:
// - Converting constructor for any other color structure type containing floating-point r, g, b, a (case sensitive.)
template<typename TOther>
constexpr color(const TOther& other, std::enable_if_t<std::is_floating_point_v<decltype(std::declval<TOther>().r)> && std::is_floating_point_v<decltype(std::declval<TOther>().a)>, float> /*sentinel*/ = 1.0f) :
r{ static_cast<uint8_t>(other.r * 255.0f) },
g{ static_cast<uint8_t>(other.g * 255.0f) },
b{ static_cast<uint8_t>(other.b * 255.0f) },
a{ static_cast<uint8_t>(other.a * 255.0f) }
{
}
constexpr color with_alpha(uint8_t alpha) const
{
return color{
r,
g,
b,
alpha
};
}
#ifdef D3DCOLORVALUE_DEFINED
constexpr operator D3DCOLORVALUE() const
{
return D3DCOLORVALUE{ r / 255.0f, g / 255.0f, b / 255.0f, a / 255.0f };
}
#endif
#ifdef WINRT_Windows_UI_H
constexpr color(const winrt::Windows::UI::Color& winUIColor) :
color(winUIColor.R, winUIColor.G, winUIColor.B, winUIColor.A)
{
}
Reduce usage of Json::Value throughout Terminal.Settings.Model (#11184) This commit reduces the code surface that interacts with raw JSON data, reducing code complexity and improving maintainability. Files that needed to be changed drastically were additionally cleaned up to remove any code cruft that has accrued over time. In order to facility this the following changes were made: * Move JSON handling from `CascadiaSettings` into `SettingsLoader` This allows us to use STL containers for data model instances. For instance profiles are now added to a hashmap for O(1) lookup. * JSON parsing within `SettingsLoader` doesn't differentiate between user, inbox and fragment JSON data, reducing code complexity and size. It also centralizes common concerns, like profile deduplication and ensuring that all profiles are assigned a GUID. * Direct JSON modification, like the insertion of dynamic profiles into settings.json were removed. This vastly reduces code complexity, but unfortunately removes support for comments in JSON on first start. * `ColorScheme`s cannot be layered. As such its `LayerJson` API was replaced with `FromJson`, allowing us to remove JSON-based color scheme validation. * `Profile`s used to test their wish to layer using `ShouldBeLayered`, which was replaced with a GUID-based hashmap lookup on previously parsed profiles. Further changes were made as improvements upon the previous changes: * Compact the JSON files embedded binary, saving 28kB * Prevent double-initialization of the color table in `ColorScheme` * Making `til::color` getters `constexpr`, allow better optimizations The result is a reduction of: * 48kB binary size for the Settings.Model.dll * 5-10% startup duration * 26% code for the `CascadiaSettings` class * 1% overall code in this project Furthermore this results in the following breaking changes: * The long deprecated "globals" settings object will not be detected and no warning will be created during load. * The initial creation of a new settings.json will not produce helpful comments. Both cases are caused by the removal of manual JSON handling and the move to representing the settings file with model objects instead ## PR Checklist * [x] Closes #5276 * [x] Closes #7421 * [x] I work here * [x] Tests added/passed ## Validation Steps Performed * Out-of-box-experience is identical to before ✔️ (Except for the settings.json file lacking comments.) * Existing user settings load correctly ✔️ * New WSL instances are added to user settings ✔️ * New fragments are added to user settings ✔️ * All profiles are assigned GUIDs ✔️
2021-09-22 18:27:31 +02:00
constexpr operator winrt::Windows::UI::Color() const
{
Reduce usage of Json::Value throughout Terminal.Settings.Model (#11184) This commit reduces the code surface that interacts with raw JSON data, reducing code complexity and improving maintainability. Files that needed to be changed drastically were additionally cleaned up to remove any code cruft that has accrued over time. In order to facility this the following changes were made: * Move JSON handling from `CascadiaSettings` into `SettingsLoader` This allows us to use STL containers for data model instances. For instance profiles are now added to a hashmap for O(1) lookup. * JSON parsing within `SettingsLoader` doesn't differentiate between user, inbox and fragment JSON data, reducing code complexity and size. It also centralizes common concerns, like profile deduplication and ensuring that all profiles are assigned a GUID. * Direct JSON modification, like the insertion of dynamic profiles into settings.json were removed. This vastly reduces code complexity, but unfortunately removes support for comments in JSON on first start. * `ColorScheme`s cannot be layered. As such its `LayerJson` API was replaced with `FromJson`, allowing us to remove JSON-based color scheme validation. * `Profile`s used to test their wish to layer using `ShouldBeLayered`, which was replaced with a GUID-based hashmap lookup on previously parsed profiles. Further changes were made as improvements upon the previous changes: * Compact the JSON files embedded binary, saving 28kB * Prevent double-initialization of the color table in `ColorScheme` * Making `til::color` getters `constexpr`, allow better optimizations The result is a reduction of: * 48kB binary size for the Settings.Model.dll * 5-10% startup duration * 26% code for the `CascadiaSettings` class * 1% overall code in this project Furthermore this results in the following breaking changes: * The long deprecated "globals" settings object will not be detected and no warning will be created during load. * The initial creation of a new settings.json will not produce helpful comments. Both cases are caused by the removal of manual JSON handling and the move to representing the settings file with model objects instead ## PR Checklist * [x] Closes #5276 * [x] Closes #7421 * [x] I work here * [x] Tests added/passed ## Validation Steps Performed * Out-of-box-experience is identical to before ✔️ (Except for the settings.json file lacking comments.) * Existing user settings load correctly ✔️ * New WSL instances are added to user settings ✔️ * New fragments are added to user settings ✔️ * All profiles are assigned GUIDs ✔️
2021-09-22 18:27:31 +02:00
return { a, r, g, b };
}
#endif
#ifdef WINRT_Microsoft_Terminal_Core_H
constexpr color(const winrt::Microsoft::Terminal::Core::Color& coreColor) :
color(coreColor.R, coreColor.G, coreColor.B, coreColor.A)
{
}
Reduce usage of Json::Value throughout Terminal.Settings.Model (#11184) This commit reduces the code surface that interacts with raw JSON data, reducing code complexity and improving maintainability. Files that needed to be changed drastically were additionally cleaned up to remove any code cruft that has accrued over time. In order to facility this the following changes were made: * Move JSON handling from `CascadiaSettings` into `SettingsLoader` This allows us to use STL containers for data model instances. For instance profiles are now added to a hashmap for O(1) lookup. * JSON parsing within `SettingsLoader` doesn't differentiate between user, inbox and fragment JSON data, reducing code complexity and size. It also centralizes common concerns, like profile deduplication and ensuring that all profiles are assigned a GUID. * Direct JSON modification, like the insertion of dynamic profiles into settings.json were removed. This vastly reduces code complexity, but unfortunately removes support for comments in JSON on first start. * `ColorScheme`s cannot be layered. As such its `LayerJson` API was replaced with `FromJson`, allowing us to remove JSON-based color scheme validation. * `Profile`s used to test their wish to layer using `ShouldBeLayered`, which was replaced with a GUID-based hashmap lookup on previously parsed profiles. Further changes were made as improvements upon the previous changes: * Compact the JSON files embedded binary, saving 28kB * Prevent double-initialization of the color table in `ColorScheme` * Making `til::color` getters `constexpr`, allow better optimizations The result is a reduction of: * 48kB binary size for the Settings.Model.dll * 5-10% startup duration * 26% code for the `CascadiaSettings` class * 1% overall code in this project Furthermore this results in the following breaking changes: * The long deprecated "globals" settings object will not be detected and no warning will be created during load. * The initial creation of a new settings.json will not produce helpful comments. Both cases are caused by the removal of manual JSON handling and the move to representing the settings file with model objects instead ## PR Checklist * [x] Closes #5276 * [x] Closes #7421 * [x] I work here * [x] Tests added/passed ## Validation Steps Performed * Out-of-box-experience is identical to before ✔️ (Except for the settings.json file lacking comments.) * Existing user settings load correctly ✔️ * New WSL instances are added to user settings ✔️ * New fragments are added to user settings ✔️ * All profiles are assigned GUIDs ✔️
2021-09-22 18:27:31 +02:00
constexpr operator winrt::Microsoft::Terminal::Core::Color() const noexcept
{
Reduce usage of Json::Value throughout Terminal.Settings.Model (#11184) This commit reduces the code surface that interacts with raw JSON data, reducing code complexity and improving maintainability. Files that needed to be changed drastically were additionally cleaned up to remove any code cruft that has accrued over time. In order to facility this the following changes were made: * Move JSON handling from `CascadiaSettings` into `SettingsLoader` This allows us to use STL containers for data model instances. For instance profiles are now added to a hashmap for O(1) lookup. * JSON parsing within `SettingsLoader` doesn't differentiate between user, inbox and fragment JSON data, reducing code complexity and size. It also centralizes common concerns, like profile deduplication and ensuring that all profiles are assigned a GUID. * Direct JSON modification, like the insertion of dynamic profiles into settings.json were removed. This vastly reduces code complexity, but unfortunately removes support for comments in JSON on first start. * `ColorScheme`s cannot be layered. As such its `LayerJson` API was replaced with `FromJson`, allowing us to remove JSON-based color scheme validation. * `Profile`s used to test their wish to layer using `ShouldBeLayered`, which was replaced with a GUID-based hashmap lookup on previously parsed profiles. Further changes were made as improvements upon the previous changes: * Compact the JSON files embedded binary, saving 28kB * Prevent double-initialization of the color table in `ColorScheme` * Making `til::color` getters `constexpr`, allow better optimizations The result is a reduction of: * 48kB binary size for the Settings.Model.dll * 5-10% startup duration * 26% code for the `CascadiaSettings` class * 1% overall code in this project Furthermore this results in the following breaking changes: * The long deprecated "globals" settings object will not be detected and no warning will be created during load. * The initial creation of a new settings.json will not produce helpful comments. Both cases are caused by the removal of manual JSON handling and the move to representing the settings file with model objects instead ## PR Checklist * [x] Closes #5276 * [x] Closes #7421 * [x] I work here * [x] Tests added/passed ## Validation Steps Performed * Out-of-box-experience is identical to before ✔️ (Except for the settings.json file lacking comments.) * Existing user settings load correctly ✔️ * New WSL instances are added to user settings ✔️ * New fragments are added to user settings ✔️ * All profiles are assigned GUIDs ✔️
2021-09-22 18:27:31 +02:00
return { r, g, b, a };
}
#endif
constexpr bool operator==(const til::color& other) const
{
return abgr == other.abgr;
}
constexpr bool operator!=(const til::color& other) const
{
return !(*this == other);
}
std::wstring to_string() const
{
std::wstringstream wss;
wss << L"Color " << ToHexString(false);
return wss.str();
}
std::wstring ToHexString(const bool omitAlpha = false) const
{
std::wstringstream wss;
wss << L"#" << std::uppercase << std::setfill(L'0') << std::hex;
// Force the compiler to promote from byte to int. Without it, the
// stringstream will try to write the components as chars
if (!omitAlpha)
{
wss << std::setw(2) << static_cast<int>(a);
}
wss << std::setw(2) << static_cast<int>(r);
wss << std::setw(2) << static_cast<int>(g);
wss << std::setw(2) << static_cast<int>(b);
return wss.str();
}
};
#pragma warning(pop)
}
static_assert(sizeof(til::color) == sizeof(uint32_t));
#ifdef __WEX_COMMON_H__
namespace WEX::TestExecution
{
template<>
class VerifyOutputTraits<::til::color>
{
public:
static WEX::Common::NoThrowString ToString(const ::til::color& color)
{
return WEX::Common::NoThrowString(color.to_string().c_str());
}
};
};
#endif