specify null length

GUI/host/textthread.cpp

@@ -58,7 +58,10 @@ void TextThread::Flush()
 	std::vector<std::wstring> sentences;
 	queuedSentences->swap(sentences);
 	for (auto& sentence : sentences)
+	{
+		sentence.erase(std::remove(sentence.begin(), sentence.end(), L'\0'));
 		if (Output(*this, sentence)) storage->append(sentence);
+	}
 
 	std::scoped_lock lock(bufferMutex);
 	if (buffer.empty()) return;

GUI/host/util.cpp

@@ -27,6 +27,13 @@ namespace
 		}
 		RCode.erase(0, 1);
 
+		// [null_length<]
+		if (std::regex_search(RCode, match, std::wregex(L"^([0-9]+)<")))
+		{
+			hp.null_length = std::stoi(match[1]);
+			RCode.erase(0, match[0].length());
+		}
+
 		// [codepage#]
 		if (std::regex_search(RCode, match, std::wregex(L"^([0-9]+)#")))
 		{
@@ -104,6 +111,13 @@ namespace
 		}
 		HCode.erase(0, 1);
 
+		// [null_length<]
+		if ((hp.type & USING_STRING) && std::regex_search(HCode, match, std::wregex(L"^([0-9]+)<")))
+		{
+			hp.null_length = std::stoi(match[1]);
+			HCode.erase(0, match[0].length());
+		}
+
 		// [N]
 		if (HCode[0] == L'N')
 		{
@@ -181,10 +195,12 @@ namespace
 		if (hp.type & USING_UNICODE)
 		{
 			RCode << "Q";
+			if (hp.null_length != 0) RCode << hp.null_length << "<";
 		}
 		else
 		{
 			RCode << "S";
+			if (hp.null_length != 0) RCode << hp.null_length << "<";
 			if (hp.codepage != 0) RCode << hp.codepage << "#";
 		}
 
@@ -213,6 +229,9 @@ namespace
 			else if (hp.type & BIG_ENDIAN) HCode << "A";
 			else HCode << "B";
 		}
+
+		if (hp.null_length != 0) HCode << hp.null_length << "<";
+
 		if (hp.type & NO_CONTEXT) HCode << "N";
 		if (hp.text_fun || hp.filter_fun || hp.hook_fun) HCode << "X"; // no AGTH equivalent
 
@@ -287,7 +306,8 @@ namespace Util
 	std::optional<std::wstring> StringToWideString(const std::string& text, UINT encoding)
 	{
 		std::vector<wchar_t> buffer(text.size() + 1);
-		if (MultiByteToWideChar(encoding, 0, text.c_str(), -1, buffer.data(), buffer.size())) return buffer.data();
+		if (int length = MultiByteToWideChar(encoding, 0, text.c_str(), text.size() + 1, buffer.data(), buffer.size())) 
+			return std::wstring(buffer.data(), length - 1);
 		return {};
 	}
 

include/types.h

@@ -29,7 +29,8 @@ struct HookParam
 	int offset, // offset of the data in the memory
 		index, // deref_offset1
 		split, // offset of the split character
-		split_index; // deref_offset2
+		split_index, // deref_offset2
+		null_length;
 	union
 	{
 		wchar_t module[MAX_MODULE_SIZE];

text.cpp

@@ -21,15 +21,16 @@ const char* CODE_INFODUMP = u8R"(Search for text
 S[codepage#]text
 OR
 Enter read code
-R{S|Q|V}[codepage#][*deref_offset]@addr
+R{S|Q|V}[null_length<][codepage#][*deref_offset]@addr
 OR
 Enter hook code
-H{A|B|W|S|Q|V}[N][codepage#]data_offset[*deref_offset1][:split_offset[*deref_offset2]]@addr[:module[:func]]
+H{A|B|W|S|Q|V}[null_length<][N][codepage#]data_offset[*deref_offset1][:split_offset[*deref_offset2]]@addr[:module[:func]]
 All numbers except codepage in hexadecimal
 Default codepage is 932 (Shift-JIS) but this can be changed in settings
 A/B: codepage char little/big endian
 W: UTF-16 char
 S/Q/V: codepage/UTF-16/UTF-8 string
+null_length: length of null terminator used for string
 Negatives for data/split offset refer to registers
 -4 for EAX, -8 for ECX, -C for EDX, -10 for EBX, -14 for ESP, -18 for EBP, -1C for ESI, -20 for EDI
 * means dereference pointer+deref_offset)";

texthook/texthook.cc

@@ -229,14 +229,14 @@ DWORD WINAPI TextHook::Reader(LPVOID hookPtr)
 {
 	TextHook* This = (TextHook*)hookPtr;
 	BYTE buffer[TEXT_BUFFER_SIZE] = {};
-	int changeCount = 0, dataLen = 0;
+	int changeCount = 0, dataLen = 1;
 	__try
 	{
 		uint64_t currentAddress = This->address;
 		while (WaitForSingleObject(This->readerEvent, 500) == WAIT_TIMEOUT)
 		{
 			if (This->hp.type & DATA_INDIRECT) currentAddress = *(uintptr_t*)This->address + This->hp.index;
-			if (memcmp(buffer, (void*)currentAddress, dataLen + 2) == 0)
+			if (memcmp(buffer, (void*)currentAddress, dataLen) == 0)
 			{
 				changeCount = 0;
 				continue;
@@ -248,10 +248,8 @@ DWORD WINAPI TextHook::Reader(LPVOID hookPtr)
 				break;
 			}
 
-			if (This->hp.type & USING_UNICODE) dataLen = wcslen((wchar_t*)currentAddress) * 2;
-			else dataLen = strlen((char*)currentAddress);
-			if (dataLen > TEXT_BUFFER_SIZE - 2) dataLen = TEXT_BUFFER_SIZE - 2;
-			memcpy(buffer, (void*)currentAddress, dataLen + 2);
+			dataLen = min(This->HookStrlen((BYTE*)currentAddress), TEXT_BUFFER_SIZE);
+			memcpy(buffer, (void*)currentAddress, dataLen);
 			TextOutput({ GetCurrentProcessId(), This->address, 0, 0 }, buffer, dataLen);
 		}
 	}
@@ -308,10 +306,7 @@ int TextHook::GetLength(uintptr_t base, uintptr_t in)
 			break;
 		//len == -1 then continue to case 0.
 	case 0:
-		if (hp.type & USING_UNICODE)
-			len = wcslen((const wchar_t *)in) << 1;
-		else
-			len = strlen((const char *)in);
+		len = HookStrlen((BYTE*)in);
 		break;
 	case 1:
 		if (hp.type & USING_UNICODE)
@@ -328,4 +323,12 @@ int TextHook::GetLength(uintptr_t base, uintptr_t in)
 	return max(0, len);
 }
 
+int TextHook::HookStrlen(BYTE* data)
+{
+	BYTE* orig = data;
+	for (int nullsRemaining = hp.null_length ? hp.null_length : hp.type & USING_UNICODE ? 2 : 1; nullsRemaining >= 0; ++data)
+		if (*data == 0) nullsRemaining -= 1;
+	return data - orig;
+}
+
 // EOF

texthook/texthook.h

@@ -34,6 +34,7 @@ private:
 	bool InsertReadCode();
 	void Send(uintptr_t dwDatabase);
 	int GetLength(uintptr_t base, uintptr_t in); // jichi 12/25/2013: Return 0 if failed
+	int HookStrlen(BYTE* data);
 	void RemoveHookCode();
 	void RemoveReadCode();