From f9e5b2074b187b57f4a935e0e66967cc3020e9622a53e88d411873cb8b9555fd Mon Sep 17 00:00:00 2001
From: ACh Sulfate <dex@libart.so>
Date: Fri, 17 May 2024 22:55:06 +0800
Subject: [PATCH] add mutf8

---
 utf_convert.cc | 60 +++++++++++++++++++++++++++++++++++++++++---------
 utf_convert.h  | 10 ++++++---
 2 files changed, 56 insertions(+), 14 deletions(-)

diff --git a/utf_convert.cc b/utf_convert.cc
index f5bdc42..8919860 100644
--- a/utf_convert.cc
+++ b/utf_convert.cc
@@ -5,6 +5,7 @@
 #include "utf_convert.h"
 
 #include <vector>
+#include <cstdint>
 
 namespace utfcvt {
 
@@ -46,30 +47,30 @@ std::u32string UTF16ToUTF32(std::u16string_view utf16) {
     if (utf16.empty()) {
         return {};
     }
-    std::vector<int> codePoints;
+    std::vector<uint32_t> codePoints;
     for (size_t i = 0; i < utf16.size(); ++i) {
-        if (utf16[i] < 0xD800 || utf16[i] > 0xDFFF) {
+        if (utf16[i] < 0xD800u || utf16[i] > 0xDFFFu) {
             codePoints.push_back(utf16[i]);
         } else if (i + 1 < utf16.size()) {
-            codePoints.push_back(0x10000 + ((utf16[i] - 0xD800) << 10) + (utf16[i + 1] - 0xDC00));
+            codePoints.push_back(0x10000u + ((utf16[i] - 0xD800u) << 10) + (utf16[i + 1] - 0xDC00u));
             ++i;
         } else {
-            codePoints.push_back(0xFFFD);
+            codePoints.push_back(0xFFFDu);
         }
     }
     return {codePoints.begin(), codePoints.end()};
 }
 
-std::u16string UTF32ToUTF16(std::u32string_view utf32){
-    std::vector<char16_t> utf16;
+std::u16string UTF32ToUTF16(std::u32string_view utf32) {
+    std::vector<uint16_t> utf16;
     for (auto c: utf32) {
-        if (c <= 0xFFFF) {
+        if (c <= 0xFFFFu) {
             utf16.push_back(static_cast<char16_t>(c));
-        } else if (c <= 0x10FFFF) {
-            utf16.push_back(static_cast<char16_t>(0xD800 + ((c - 0x10000) >> 10)));
-            utf16.push_back(static_cast<char16_t>(0xDC00 + ((c - 0x10000) & 0x3FF)));
+        } else if (c <= 0x10FFFFu) {
+            utf16.push_back(static_cast<char16_t>(0xD800u + ((c - 0x10000u) >> 10)));
+            utf16.push_back(static_cast<char16_t>(0xDC00u + ((c - 0x10000u) & 0x3FF)));
         } else {
-            utf16.push_back(0xFFFD);
+            utf16.push_back(0xFFFDu);
         }
     }
     return {utf16.begin(), utf16.end()};
@@ -97,6 +98,43 @@ std::string UTF32ToUTF8(std::u32string_view utf32) {
     return utf8;
 }
 
+std::string UTF16ToMUTF8(std::u16string_view utf16) {
+    std::string mutf8;
+    for (size_t i = 0; i < utf16.size(); ++i) {
+        if (utf16[i] < 0x80) {
+            mutf8.push_back(static_cast<char>(utf16[i]));
+        } else if (utf16[i] < 0x800) {
+            mutf8.push_back(static_cast<char>(0xC0 | (utf16[i] >> 6)));
+            mutf8.push_back(static_cast<char>(0x80 | (utf16[i] & 0x3F)));
+        } else {
+            mutf8.push_back(static_cast<char>(0xE0 | (utf16[i] >> 12)));
+            mutf8.push_back(static_cast<char>(0x80 | ((utf16[i] >> 6) & 0x3F)));
+            mutf8.push_back(static_cast<char>(0x80 | (utf16[i] & 0x3F)));
+        }
+    }
+    return mutf8;
+}
+
+std::u16string MUTF8ToUTF16(std::string_view mutf8) {
+    std::vector<uint16_t> utf16;
+    for (size_t i = 0; i < mutf8.size(); ++i) {
+        if ((mutf8[i] & 0b10000000) == 0) {
+            // 1 byte code point, ASCII
+            utf16.push_back(mutf8[i]);
+        } else if ((mutf8[i] & 0b11100000) == 0b11000000) {
+            // 2 byte code point
+            utf16.push_back(uint16_t((mutf8[i] & 0b00011111u) << 6) | uint16_t(mutf8[i + 1] & 0b00111111u));
+            i++;
+        } else if ((mutf8[i] & 0b11110000) == 0b11100000) {
+            // 3 byte code point
+            utf16.push_back(uint16_t((mutf8[i] & 0b00001111u) << 12) |
+                            uint16_t((mutf8[i + 1] & 0b00111111u) << 6) | uint16_t(mutf8[i + 2] & 0b00111111u));
+            i += 2;
+        }
+    }
+    return {utf16.begin(), utf16.end()};
+}
+
 std::string UTF16ToUTF8(std::u16string_view utf16) {
     return UTF32ToUTF8(UTF16ToUTF32(utf16));
 }
diff --git a/utf_convert.h b/utf_convert.h
index dae8bca..164da10 100644
--- a/utf_convert.h
+++ b/utf_convert.h
@@ -10,18 +10,22 @@
 
 namespace utfcvt {
 
-std::string UTF16ToUTF8(std::u16string_view utf16);
-
 std::u16string UTF8ToUTF16(std::string_view utf8);
 
 std::u32string UTF8ToUTF32(std::string_view utf8);
 
-std::string UTF32ToUTF8(std::u32string_view utf32);
+std::string UTF16ToUTF8(std::u16string_view utf16);
 
 std::u32string UTF16ToUTF32(std::u16string_view utf16);
 
+std::string UTF32ToUTF8(std::u32string_view utf32);
+
 std::u16string UTF32ToUTF16(std::u32string_view utf32);
 
+std::string UTF16ToMUTF8(std::u16string_view utf16);
+
+std::u16string MUTF8ToUTF16(std::string_view mutf8);
+
 // for windows, they are not implemented on linux
 
 std::wstring UTF8ToWide(std::string_view utf8);