Fix garbling when copying multibyte text via OSC 52 (#7870)

This commit adds a missing conversion utf8 to utf16 in decoding base64
for handling multibyte text in copying via OSC 52.

## Validation Steps Performed
* automatically
    * Tests w/ multibyte characters
* manually
    * case1
        * Executed `printf "\x1b]52;;%s\x1b\\" "$(printf '👍👍🏻👍🏼👍🏽👍🏾👍🏿' | base64)"`
        * Verified `👍👍🏻👍🏼👍🏽👍🏾👍🏿` in my clipboard
    * case2
        * Copied `👍👍🏻👍🏼👍🏽👍🏾👍🏿` by tmux 2.6 default copy function (OSC 52)
        * Verified `👍👍🏻👍🏼👍🏽👍🏾👍🏿` in my clipboard

Closes #7819

.github/actions/spell-check/dictionary/apis.txt

@@ -49,6 +49,7 @@ GETDESKWALLPAPER
 UPDATEINIFILE
 spsc
 STDCPP
+strchr
 syscall
 tmp
 tx

.github/actions/spell-check/patterns/patterns.txt

@@ -19,5 +19,5 @@ TestUtils::VerifyExpectedString\(tb, L"[^"]+"
 0x[0-9A-Za-z]+
 Base64::s_(?:En|De)code\(L"[^"]+"
 VERIFY_ARE_EQUAL\(L"[^"]+"
-L"ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789\+/"
+"ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789\+/"
 std::memory_order_[\w]+

src/terminal/parser/base64.cpp

@@ -6,8 +6,8 @@
 
 using namespace Microsoft::Console::VirtualTerminal;
 
-static const wchar_t base64Chars[] = L"ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/";
+static const char base64Chars[] = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/";
-static const wchar_t padChar = L'=';
+static const char padChar = '=';
 
 #pragma warning(disable : 26446 26447 26482 26485 26493 26494)
 
@@ -75,15 +75,16 @@ std::wstring Base64::s_Encode(const std::wstring_view src) noexcept
 // - true if decoding successfully, otherwise false.
 bool Base64::s_Decode(const std::wstring_view src, std::wstring& dst) noexcept
 {
+    std::string mbStr;
     int state = 0;
-    wchar_t tmp;
+    char tmp;
 
     const auto len = src.size() / 4 * 3;
     if (len == 0)
     {
         return false;
     }
-    dst.reserve(len);
+    mbStr.reserve(len);
 
     auto iter = src.cbegin();
     while (iter < src.cend())
@@ -99,7 +100,7 @@ bool Base64::s_Decode(const std::wstring_view src, std::wstring& dst) noexcept
             break;
         }
 
-        auto pos = wcschr(base64Chars, *iter);
+        auto pos = strchr(base64Chars, *iter);
         if (!pos) // A non-base64 character found.
         {
             return false;
@@ -108,24 +109,24 @@ bool Base64::s_Decode(const std::wstring_view src, std::wstring& dst) noexcept
         switch (state)
         {
         case 0:
-            tmp = (wchar_t)(pos - base64Chars) << 2;
+            tmp = (char)(pos - base64Chars) << 2;
             state = 1;
             break;
         case 1:
-            tmp |= (pos - base64Chars) >> 4;
+            tmp |= (char)(pos - base64Chars) >> 4;
-            dst.push_back(tmp);
+            mbStr += tmp;
-            tmp = (wchar_t)((pos - base64Chars) & 0x0f) << 4;
+            tmp = (char)((pos - base64Chars) & 0x0f) << 4;
             state = 2;
             break;
         case 2:
-            tmp |= (pos - base64Chars) >> 2;
+            tmp |= (char)(pos - base64Chars) >> 2;
-            dst.push_back(tmp);
+            mbStr += tmp;
-            tmp = (wchar_t)((pos - base64Chars) & 0x03) << 6;
+            tmp = (char)((pos - base64Chars) & 0x03) << 6;
             state = 3;
             break;
         case 3:
             tmp |= pos - base64Chars;
-            dst.push_back(tmp);
+            mbStr += tmp;
             state = 0;
             break;
         default:
@@ -176,7 +177,7 @@ bool Base64::s_Decode(const std::wstring_view src, std::wstring& dst) noexcept
         return false;
     }
 
-    return true;
+    return SUCCEEDED(til::u8u16(mbStr, dst));
 }
 
 // Routine Description:

src/terminal/parser/ut_parser/Base64Test.cpp

@@ -84,5 +84,18 @@ class Microsoft::Console::VirtualTerminal::Base64Test
 
         success = Base64::s_Decode(L"Zm9vYg=", result);
         VERIFY_ARE_EQUAL(false, success);
+
+        // U+306b U+307b U+3093 U+3054 U+6c49 U+8bed U+d55c U+ad6d
+        result = L"";
+        success = Base64::s_Decode(L"44Gr44G744KT44GU5rGJ6K+t7ZWc6rWt", result);
+        VERIFY_ARE_EQUAL(true, success);
+        VERIFY_ARE_EQUAL(L"にほんご汉语한국", result);
+
+        // U+d83d U+dc4d U+d83d U+dc4d U+d83c U+dffb U+d83d U+dc4d U+d83c U+dffc U+d83d
+        // U+dc4d U+d83c U+dffd U+d83d U+dc4d U+d83c U+dffe U+d83d U+dc4d U+d83c U+dfff
+        result = L"";
+        success = Base64::s_Decode(L"8J+RjfCfkY3wn4+78J+RjfCfj7zwn5GN8J+PvfCfkY3wn4++8J+RjfCfj78=", result);
+        VERIFY_ARE_EQUAL(true, success);
+        VERIFY_ARE_EQUAL(L"👍👍🏻👍🏼👍🏽👍🏾👍🏿", result);
     }
 };

src/terminal/parser/ut_parser/OutputEngineTest.cpp

@@ -3139,6 +3139,21 @@ class StateMachineExternalTest final
 
         pDispatch->ClearState();
 
+        // Passing an empty `Pc` param and a base64-encoded multibyte text `Pd` works.
+        // U+306b U+307b U+3093 U+3054 U+6c49 U+8bed U+d55c U+ad6d
+        mach.ProcessString(L"\x1b]52;;44Gr44G744KT44GU5rGJ6K+t7ZWc6rWt\x07");
+        VERIFY_ARE_EQUAL(L"にほんご汉语한국", pDispatch->_copyContent);
+
+        pDispatch->ClearState();
+
+        // Passing an empty `Pc` param and a base64-encoded multibyte text w/ emoji sequences `Pd` works.
+        // U+d83d U+dc4d U+d83d U+dc4d U+d83c U+dffb U+d83d U+dc4d U+d83c U+dffc U+d83d
+        // U+dc4d U+d83c U+dffd U+d83d U+dc4d U+d83c U+dffe U+d83d U+dc4d U+d83c U+dfff
+        mach.ProcessString(L"\x1b]52;;8J+RjfCfkY3wn4+78J+RjfCfj7zwn5GN8J+PvfCfkY3wn4++8J+RjfCfj78=\x07");
+        VERIFY_ARE_EQUAL(L"👍👍🏻👍🏼👍🏽👍🏾👍🏿", pDispatch->_copyContent);
+
+        pDispatch->ClearState();
+
         // Passing a non-empty `Pc` param (`s0` is ignored) and a valid `Pd` param works.
         mach.ProcessString(L"\x1b]52;s0;Zm9v\x07");
         VERIFY_ARE_EQUAL(L"foo", pDispatch->_copyContent);