Android: Use correct encoding when converting strings

The functions with "UTF" in the name use "modified UTF-8" rather
than the standard UTF-8 which Dolphin uses, at least according
to Oracle's documentation, so it is incorrect for us to use them.
This change fixes the problem by converting between UTF-8 and
UTF-16 manually instead of letting JNI do it for us.
This commit is contained in:
JosJuice 2020-06-28 19:16:23 +02:00
parent 38791eec18
commit f5da6e07d7
3 changed files with 27 additions and 10 deletions

View File

@ -5,25 +5,28 @@
#include "jni/AndroidCommon/AndroidCommon.h" #include "jni/AndroidCommon/AndroidCommon.h"
#include <string> #include <string>
#include <string_view>
#include <vector> #include <vector>
#include <jni.h> #include <jni.h>
#include "Common/StringUtil.h"
std::string GetJString(JNIEnv* env, jstring jstr) std::string GetJString(JNIEnv* env, jstring jstr)
{ {
std::string result = ""; const jchar* jchars = env->GetStringChars(jstr, nullptr);
if (!jstr) const jsize length = env->GetStringLength(jstr);
return result; const std::u16string_view string_view(reinterpret_cast<const char16_t*>(jchars), length);
const std::string converted_string = UTF16ToUTF8(string_view);
const char* s = env->GetStringUTFChars(jstr, nullptr); env->ReleaseStringChars(jstr, jchars);
result = s; return converted_string;
env->ReleaseStringUTFChars(jstr, s);
return result;
} }
jstring ToJString(JNIEnv* env, const std::string& str) jstring ToJString(JNIEnv* env, const std::string& str)
{ {
return env->NewStringUTF(str.c_str()); const std::u16string converted_string = UTF8ToUTF16(str);
return env->NewString(reinterpret_cast<const jchar*>(converted_string.data()),
converted_string.size());
} }
std::vector<std::string> JStringArrayToVector(JNIEnv* env, jobjectArray array) std::vector<std::string> JStringArrayToVector(JNIEnv* env, jobjectArray array)

View File

@ -5,6 +5,7 @@
#include "Common/StringUtil.h" #include "Common/StringUtil.h"
#include <algorithm> #include <algorithm>
#include <codecvt>
#include <cstdarg> #include <cstdarg>
#include <cstddef> #include <cstddef>
#include <cstdio> #include <cstdio>
@ -33,7 +34,6 @@
constexpr u32 CODEPAGE_SHIFT_JIS = 932; constexpr u32 CODEPAGE_SHIFT_JIS = 932;
constexpr u32 CODEPAGE_WINDOWS_1252 = 1252; constexpr u32 CODEPAGE_WINDOWS_1252 = 1252;
#else #else
#include <codecvt>
#include <errno.h> #include <errno.h>
#include <iconv.h> #include <iconv.h>
#include <locale.h> #include <locale.h>
@ -590,6 +590,18 @@ std::string UTF16BEToUTF8(const char16_t* str, size_t max_size)
#endif #endif
std::string UTF16ToUTF8(std::u16string_view input)
{
std::wstring_convert<std::codecvt_utf8_utf16<char16_t>, char16_t> converter;
return converter.to_bytes(input.data(), input.data() + input.size());
}
std::u16string UTF8ToUTF16(std::string_view input)
{
std::wstring_convert<std::codecvt_utf8_utf16<char16_t>, char16_t> converter;
return converter.from_bytes(input.data(), input.data() + input.size());
}
#ifdef HAS_STD_FILESYSTEM #ifdef HAS_STD_FILESYSTEM
// This is a replacement for path::u8path, which is deprecated starting with C++20. // This is a replacement for path::u8path, which is deprecated starting with C++20.
std::filesystem::path StringToPath(std::string_view path) std::filesystem::path StringToPath(std::string_view path)

View File

@ -173,6 +173,8 @@ std::string SHIFTJISToUTF8(std::string_view str);
std::string UTF8ToSHIFTJIS(std::string_view str); std::string UTF8ToSHIFTJIS(std::string_view str);
std::string WStringToUTF8(std::wstring_view str); std::string WStringToUTF8(std::wstring_view str);
std::string UTF16BEToUTF8(const char16_t* str, size_t max_size); // Stops at \0 std::string UTF16BEToUTF8(const char16_t* str, size_t max_size); // Stops at \0
std::string UTF16ToUTF8(std::u16string_view str);
std::u16string UTF8ToUTF16(std::string_view str);
#ifdef _WIN32 #ifdef _WIN32