LibUncode: Parse and generate emoji code point data

According to TR #51, the "best definition of the full set [of emojis] is
in the emoji-test.txt file". This defines not only the emoji themselves,
but the order in which they should be displayed, and what "group" of
emojis they belong to.

4 files changed, 113 insertions, 0 deletions

diff --git a/Userland/Libraries/LibUnicode/CMakeLists.txt b/Userland/Libraries/LibUnicode/CMakeLists.txt
index 1bacfa8375..c6210ab217 100644
--- a/Userland/Libraries/LibUnicode/CMakeLists.txt
+++ b/Userland/Libraries/LibUnicode/CMakeLists.txt
@@ -3,6 +3,7 @@ include(${SerenityOS_SOURCE_DIR}/Meta/CMake/unicode_data.cmake)
 set(SOURCES
     CharacterTypes.cpp
     CurrencyCode.cpp
+    Emoji.cpp
     ${UNICODE_DATA_SOURCES}
 )
 
diff --git a/Userland/Libraries/LibUnicode/Emoji.cpp b/Userland/Libraries/LibUnicode/Emoji.cpp
new file mode 100644
index 0000000000..ce0974b9a8
--- /dev/null
+++ b/Userland/Libraries/LibUnicode/Emoji.cpp
@@ -0,0 +1,13 @@
+/*
+ * Copyright (c) 2022, Tim Flynn <trflynn89@serenityos.org>
+ *
+ * SPDX-License-Identifier: BSD-2-Clause
+ */
+
+#include <LibUnicode/Emoji.h>
+
+namespace Unicode {
+
+Optional<Emoji> __attribute__((weak)) find_emoji_for_code_points(Span<u32 const>) { return {}; }
+
+}
diff --git a/Userland/Libraries/LibUnicode/Emoji.h b/Userland/Libraries/LibUnicode/Emoji.h
new file mode 100644
index 0000000000..ca5d81b839
--- /dev/null
+++ b/Userland/Libraries/LibUnicode/Emoji.h
@@ -0,0 +1,97 @@
+/*
+ * Copyright (c) 2022, Tim Flynn <trflynn89@serenityos.org>
+ *
+ * SPDX-License-Identifier: BSD-2-Clause
+ */
+
+#pragma once
+
+#include <AK/Optional.h>
+#include <AK/StringView.h>
+#include <AK/Types.h>
+
+namespace Unicode {
+
+enum class EmojiGroup : u8 {
+    SmileysAndEmotion,
+    PeopleAndBody,
+    Component,
+    AnimalsAndNature,
+    FoodAndDrink,
+    TravelAndPlaces,
+    Activities,
+    Objects,
+    Symbols,
+    Flags,
+};
+
+struct Emoji {
+    StringView name;
+    EmojiGroup group;
+    u32 display_order { 0 };
+    Span<u32 const> code_points;
+};
+
+Optional<Emoji> find_emoji_for_code_points(Span<u32 const> code_points);
+
+template<size_t Size>
+Optional<Emoji> find_emoji_for_code_points(u32 const (&code_points)[Size])
+{
+    return find_emoji_for_code_points(Span<u32 const> { code_points });
+}
+
+constexpr StringView emoji_group_to_string(EmojiGroup group)
+{
+    switch (group) {
+    case EmojiGroup::SmileysAndEmotion:
+        return "Smileys & Emotion"sv;
+    case EmojiGroup::PeopleAndBody:
+        return "People & Body"sv;
+    case EmojiGroup::Component:
+        return "Component"sv;
+    case EmojiGroup::AnimalsAndNature:
+        return "Animals & Nature"sv;
+    case EmojiGroup::FoodAndDrink:
+        return "Food & Drink"sv;
+    case EmojiGroup::TravelAndPlaces:
+        return "Travel & Places"sv;
+    case EmojiGroup::Activities:
+        return "Activities"sv;
+    case EmojiGroup::Objects:
+        return "Objects"sv;
+    case EmojiGroup::Symbols:
+        return "Symbols"sv;
+    case EmojiGroup::Flags:
+        return "Flags"sv;
+    }
+
+    VERIFY_NOT_REACHED();
+}
+
+constexpr EmojiGroup emoji_group_from_string(StringView group)
+{
+    if (group == "Smileys & Emotion"sv)
+        return EmojiGroup::SmileysAndEmotion;
+    if (group == "People & Body"sv)
+        return EmojiGroup::PeopleAndBody;
+    if (group == "Component"sv)
+        return EmojiGroup::Component;
+    if (group == "Animals & Nature"sv)
+        return EmojiGroup::AnimalsAndNature;
+    if (group == "Food & Drink"sv)
+        return EmojiGroup::FoodAndDrink;
+    if (group == "Travel & Places"sv)
+        return EmojiGroup::TravelAndPlaces;
+    if (group == "Activities"sv)
+        return EmojiGroup::Activities;
+    if (group == "Objects"sv)
+        return EmojiGroup::Objects;
+    if (group == "Symbols"sv)
+        return EmojiGroup::Symbols;
+    if (group == "Flags"sv)
+        return EmojiGroup::Flags;
+
+    VERIFY_NOT_REACHED();
+}
+
+}
diff --git a/Userland/Libraries/LibUnicode/Forward.h b/Userland/Libraries/LibUnicode/Forward.h
index c3b08d3088..22cf7c698e 100644
--- a/Userland/Libraries/LibUnicode/Forward.h
+++ b/Userland/Libraries/LibUnicode/Forward.h
@@ -11,6 +11,7 @@
 namespace Unicode {
 
 enum class Block : u16;
+enum class EmojiGroup : u8;
 enum class GeneralCategory : u8;
 enum class GraphemeBreakProperty : u8;
 enum class Property : u8;
@@ -19,6 +20,7 @@ enum class SentenceBreakProperty : u8;
 enum class WordBreakProperty : u8;
 
 struct CurrencyCode;
+struct Emoji;
 struct SpecialCasing;
 
 }