Ren'py 2.x and 3.x engine hook

This commit is contained in:
Blu3train 2023-12-17 18:42:15 +01:00
parent f3fbe04409
commit f316813283
2 changed files with 539 additions and 25 deletions

View File

@ -16945,6 +16945,7 @@ bool InsertAdobeFlash10Hook()
*/
bool InsertRenpyHook()
{
bool ok = false;
wchar_t python[] = L"python2X.dll", libpython[] = L"libpython2.X.dll";
for (wchar_t* name : { python, libpython })
{
@ -16957,32 +16958,77 @@ bool InsertRenpyHook()
wcscpy_s(spDefault.exportModule, name);
HookParam hp = {};
hp.address = (DWORD)GetProcAddress(module, "PyUnicodeUCS2_Format");
if (!hp.address)
if (hp.address)
{
ConsoleOutput("Textractor: Ren'py failed: failed to find PyUnicodeUCS2_Format");
return false;
}
hp.offset = 4;
hp.index = 0xc;
hp.length_offset = 0;
//hp.split = pusha_ebx_off - 4;
hp.text_fun = [](auto, auto, auto, DWORD* data, DWORD* split, DWORD* count)
{
*data = *(DWORD*)(*data + 0xc);
*count = wcslen((wchar_t*)*data) * sizeof(wchar_t);
*split = wcschr((wchar_t*)*data, L'%') == nullptr;
};
hp.type = USING_STRING | USING_UNICODE | NO_CONTEXT | DATA_INDIRECT/* | USING_SPLIT*/;
//hp.filter_fun = [](void* str, auto, auto, auto) { return *(wchar_t*)str != L'%'; };
NewHook(hp, "Ren'py");
return true;
}
}
}
ConsoleOutput("Textractor: Ren'py failed: failed to find python2X.dll");
return false;
}
hp.filter_fun = [](LPVOID data, DWORD *size, HookParam *, BYTE)
{
static std::wstring prevText;
auto text = reinterpret_cast<LPWSTR>(data);
auto len = reinterpret_cast<size_t *>(size);
if (cpp_wcsnstr(text, L"%", *len/sizeof(wchar_t)))
return false;
if (cpp_wcsnstr(text, L"{", *len/sizeof(wchar_t))) {
WideStringCharReplacer(text, len, L"{i}", 3, L'\'');
WideStringCharReplacer(text, len, L"{/i}", 4, L'\'');
WideStringFilterBetween(text, len, L"{", 1, L"}", 1);
}
WideStringFilter(text, len, L"^", 2); // remove ^ followed by 1 char
WideCharReplacer(text, len, L'\n', L' ');
if (prevText.length()==*len/sizeof(wchar_t) && prevText.find(text, 0, *len/sizeof(wchar_t))!=std::string::npos) // Check if the string is the same as the previous one
return false;
prevText.assign(text, *len/sizeof(wchar_t));
return true;
};
NewHook(hp, "Ren'py UCS2Format");
ok = true;
}
hp.address = (DWORD)GetProcAddress(module, "PyUnicodeUCS2_Replace");
if (hp.address)
{
hp.offset = 3 * 4; //arg 3;
hp.index = 0xC;
hp.length_offset = 0;
hp.split = 2 * 4; //arg 2;
hp.type = USING_STRING | USING_UNICODE | DATA_INDIRECT | USING_SPLIT;
hp.filter_fun = [](LPVOID data, DWORD *size, HookParam *, BYTE)
{
static std::wstring prevText;
auto text = reinterpret_cast<LPWSTR>(data);
auto len = reinterpret_cast<size_t *>(size);
if (cpp_wcsnstr(text, L"{fast}", *len/sizeof(wchar_t)))
return false;
if (cpp_wcsnstr(text, L"{", *len/sizeof(wchar_t))) {
WideStringCharReplacer(text, len, L"{i}", 3, L'\'');
WideStringCharReplacer(text, len, L"{/i}", 4, L'\'');
WideStringFilterBetween(text, len, L"{", 1, L"}", 1);
}
WideCharReplacer(text, len, L'\n', L' ');
if (prevText.length()==*len/sizeof(wchar_t) && prevText.find(text, 0, *len/sizeof(wchar_t))!=std::string::npos) // Check if the string is the same as the previous one
return false;
prevText.assign(text, *len/sizeof(wchar_t));
return true;
};
NewHook(hp, "Ren'py UCS2Replace");
ok = true;
}
}
}
}
if ( !ok )
ConsoleOutput("Textractor: Ren'py failed: failed to find python2X.dll");
return ok;
}
void InsertMonoHook(HMODULE h)
{
static HMODULE mono = h;

View File

@ -6,9 +6,336 @@
#include "mono/funcinfo.h"
#include "engine.h"
#include "util.h"
#include "cpputil/cppcstring.h"
// Warning: The offset in ITH has -4 offset comparing to pusha and AGTH
enum pusha_off
{
pusha_rax_off = -0xC,
pusha_rbx_off = -0x14,
pusha_rcx_off = -0x1c,
pusha_rdx_off = -0x24,
pusha_rsp_off = -0x2c,
pusha_rbp_off = -0x34,
pusha_rsi_off = -0x3c,
pusha_rdi_off = -0x44,
pusha_r8_off = -0x4c,
pusha_r9_off = -0x54,
pusha_r10_off = -0x5c,
pusha_r11_off = -0x64,
pusha_r12_off = -0x6c,
pusha_r13_off = -0x74,
pusha_r14_off = -0x7c,
pusha_r15_off = -0x84,
pusha_off = -0x8c // pushad offset
};
#define retof(rsp_base) *(uintptr_t *)(rsp_base) // return address
#define regof(name, rsp_base) *(uintptr_t *)((rsp_base) + pusha_##name##_off - 4)
#define argof(count, rsp_base) *(uintptr_t *)((rsp_base) + 4 * (count)) // starts from 1 instead of 0
enum { VNR_TEXT_CAPACITY = 1500 }; // estimated max number of bytes allowed in VNR, slightly larger than VNR's text limit (1000)
namespace { // unnamed helpers
#define XX2 XX,XX // WORD
#define XX4 XX2,XX2 // DWORD
#define XX8 XX4,XX4 // QWORD
// jichi 8/18/2013: Original maximum relative address in ITH
//enum { MAX_REL_ADDR = 0x200000 };
// jichi 10/1/2013: Increase relative address limit. Certain game engine like Artemis has larger code region
enum : DWORD { MAX_REL_ADDR = 0x00300000 };
static union {
char text_buffer[0x1000];
wchar_t wc_buffer[0x800];
struct { // CodeSection
DWORD base;
DWORD size;
} code_section[0x200];
};
DWORD text_buffer_length;
// 7/29/2014 jichi: I should move these functions to different files
// String utilities
// Return the address of the first non-zero address
LPCSTR reverse_search_begin(const char *s, int maxsize = VNR_TEXT_CAPACITY)
{
if (*s)
for (int i = 0; i < maxsize; i++, s--)
if (!*s)
return s + 1;
return nullptr;
}
bool all_ascii(const char *s, int maxsize = VNR_TEXT_CAPACITY)
{
if (s)
for (int i = 0; i < maxsize && *s; i++, s++)
if ((BYTE)*s > 127) // unsigned char
return false;
return true;
}
bool all_ascii(const wchar_t *s, int maxsize = VNR_TEXT_CAPACITY)
{
if (s)
for (int i = 0; i < maxsize && *s; i++, s++)
if (*s > 127) // unsigned char
return false;
return true;
}
// String filters
void CharReplacer(char *str, size_t *size, char fr, char to)
{
size_t len = *size;
for (size_t i = 0; i < len; i++)
if (str[i] == fr)
str[i] = to;
}
void WideCharReplacer(wchar_t *str, size_t *size, wchar_t fr, wchar_t to)
{
size_t len = *size / 2;
for (size_t i = 0; i < len; i++)
if (str[i] == fr)
str[i] = to;
}
void CharFilter(char *str, size_t *size, char ch)
{
size_t len = *size,
curlen;
for (char *cur = (char *)::memchr(str, ch, len);
(cur && --len && (curlen = len - (cur - str)));
cur = (char *)::memchr(cur, ch, curlen))
::memmove(cur, cur + 1, curlen);
*size = len;
}
void WideCharFilter(wchar_t *str, size_t *size, wchar_t ch)
{
size_t len = *size / 2,
curlen;
for (wchar_t *cur = cpp_wcsnchr(str, ch, len);
(cur && --len && (curlen = len - (cur - str)));
cur = cpp_wcsnchr(cur, ch, curlen))
::memmove(cur, cur + 1, 2 * curlen);
*size = len * 2;
}
void CharsFilter(char *str, size_t *size, const char *chars)
{
size_t len = *size,
curlen;
for (char *cur = cpp_strnpbrk(str, chars, len);
(cur && --len && (curlen = len - (cur - str)));
cur = cpp_strnpbrk(cur, chars, curlen))
::memmove(cur, cur + 1, curlen);
*size = len;
}
void WideCharsFilter(wchar_t *str, size_t *size, const wchar_t *chars)
{
size_t len = *size / 2,
curlen;
for (wchar_t *cur = cpp_wcsnpbrk(str, chars, len);
(cur && --len && (curlen = len - (cur - str)));
cur = cpp_wcsnpbrk(cur, chars, curlen))
::memmove(cur, cur + 1, 2 * curlen);
*size = len * 2;
}
void StringFilter(char *str, size_t *size, const char *remove, size_t removelen)
{
size_t len = *size,
curlen;
for (char *cur = cpp_strnstr(str, remove, len);
(cur && (len -= removelen) && (curlen = len - (cur - str)));
cur = cpp_strnstr(cur, remove, curlen))
::memmove(cur, cur + removelen, curlen);
*size = len;
}
void WideStringFilter(wchar_t *str, size_t *size, const wchar_t *remove, size_t removelen)
{
size_t len = *size / 2,
curlen;
for (wchar_t *cur = cpp_wcsnstr(str, remove, len);
(cur && (len -= removelen) && (curlen = len - (cur - str)));
cur = cpp_wcsnstr(cur, remove, curlen))
::memmove(cur, cur + removelen, 2 * curlen);
*size = len * 2;
}
void StringFilterBetween(char *str, size_t *size, const char *fr, size_t frlen, const char *to, size_t tolen)
{
size_t len = *size,
curlen;
for (char *cur = cpp_strnstr(str, fr, len);
cur;
cur = cpp_strnstr(cur, fr, curlen)) {
curlen = (len - frlen) - (cur - str);
auto end = cpp_strnstr(cur + frlen, to, curlen);
if (!end)
break;
curlen = len - (end - str) - tolen;
::memmove(cur, end + tolen, curlen);
len -= tolen + (end - cur);
}
*size = len;
}
void WideStringFilterBetween(wchar_t *str, size_t *size, const wchar_t *fr, size_t frlen, const wchar_t *to, size_t tolen)
{
size_t len = *size / 2,
curlen;
for (wchar_t *cur = cpp_wcsnstr(str, fr, len);
cur;
cur = cpp_wcsnstr(cur, fr, curlen)) {
curlen = (len - frlen) - (cur - str);
auto end = cpp_wcsnstr(cur + frlen, to, curlen);
if (!end)
break;
curlen = len - (end - str) - tolen;
::memmove(cur, end + tolen, 2 * curlen);
len -= tolen + (end - cur);
}
*size = len * 2;
}
void StringCharReplacer(char *str, size_t *size, const char *src, size_t srclen, char ch)
{
size_t len = *size,
curlen;
for (char *cur = cpp_strnstr(str, src, len);
cur && len;
cur = cpp_strnstr(cur, src, curlen)) {
*cur++ = ch;
len -= srclen - 1;
curlen = len - (cur - str);
if (curlen == 0)
break;
::memmove(cur, cur + srclen - 1, curlen);
}
*size = len;
}
void WideStringCharReplacer(wchar_t *str, size_t *size, const wchar_t *src, size_t srclen, wchar_t ch)
{
size_t len = *size / 2,
curlen;
for (wchar_t *cur = cpp_wcsnstr(str, src, len);
cur && len;
cur = cpp_wcsnstr(cur, src, curlen)) {
*cur++ = ch;
len -= srclen - 1;
curlen = len - (cur - str);
if (curlen == 0)
break;
::memmove(cur, cur + srclen -1, 2 * curlen);
}
*size = len * 2;
}
// NOTE: I assume srclen >= dstlen
void StringReplacer(char *str, size_t *size, const char *src, size_t srclen, const char *dst, size_t dstlen)
{
size_t len = *size,
curlen;
for (char *cur = cpp_strnstr(str, src, len);
cur && len;
cur = cpp_strnstr(cur, src, curlen)) {
::memcpy(cur, dst, dstlen);
cur += dstlen;
len -= srclen - dstlen;
curlen = len - (cur - str);
if (curlen == 0)
break;
if (srclen > dstlen)
::memmove(cur, cur + srclen - dstlen, curlen);
}
*size = len;
}
void WideStringReplacer(wchar_t *str, size_t *size, const wchar_t *src, size_t srclen, const wchar_t *dst, size_t dstlen)
{
size_t len = *size / 2,
curlen;
for (wchar_t *cur = cpp_wcsnstr(str, src, len);
cur && len;
cur = cpp_wcsnstr(cur, src, curlen)) {
::memcpy(cur, dst, 2 * dstlen);
cur += dstlen;
len -= srclen - dstlen;
curlen = len - (cur - str);
if (curlen == 0)
break;
if (srclen > dstlen)
::memmove(cur, cur + srclen - dstlen, 2 * curlen);
}
*size = len * 2;
}
bool NewLineCharFilter(LPVOID data, DWORD *size, HookParam *, BYTE)
{
CharFilter(reinterpret_cast<LPSTR>(data), reinterpret_cast<size_t *>(size),
'\n');
return true;
}
bool NewLineWideCharFilter(LPVOID data, DWORD *size, HookParam *, BYTE)
{
CharFilter(reinterpret_cast<LPSTR>(data), reinterpret_cast<size_t *>(size),
L'\n');
return true;
}
bool NewLineStringFilter(LPVOID data, DWORD *size, HookParam *, BYTE)
{
StringFilter(reinterpret_cast<LPSTR>(data), reinterpret_cast<size_t *>(size),
"\\n", 2);
return true;
}
bool NewLineWideStringFilter(LPVOID data, DWORD *size, HookParam *, BYTE)
{
WideStringFilter(reinterpret_cast<LPWSTR>(data), reinterpret_cast<size_t *>(size),
L"\\n", 2);
return true;
}
bool NewLineCharToSpaceFilter(LPVOID data, DWORD *size, HookParam *, BYTE)
{
CharReplacer(reinterpret_cast<LPSTR>(data), reinterpret_cast<size_t *>(size), '\n', ' ');
return true;
}
bool NewLineWideCharToSpaceFilter(LPVOID data, DWORD *size, HookParam *, BYTE)
{
WideCharReplacer(reinterpret_cast<LPWSTR>(data), reinterpret_cast<size_t *>(size), L'\n', L' ');
return true;
}
// Remove every characters <= 0x1f (i.e. before space ' ') except 0xa and 0xd.
bool IllegalCharsFilter(LPVOID data, DWORD *size, HookParam *, BYTE)
{
CharsFilter(reinterpret_cast<LPSTR>(data), reinterpret_cast<size_t *>(size),
"\x01\x02\x03\x04\x05\x06\x07\x08\x09\x0b\x0c\x0e\x0f\x10\x11\x12\x12\x14\x15\x16\x17\x18\x19\x1a\x1b\x1c\x1d\x1e\x1f");
return true;
}
bool IllegalWideCharsFilter(LPVOID data, DWORD *size, HookParam *, BYTE)
{
WideCharsFilter(reinterpret_cast<LPWSTR>(data), reinterpret_cast<size_t *>(size),
L"\x01\x02\x03\x04\x05\x06\x07\x08\x09\x0b\x0c\x0e\x0f\x10\x11\x12\x12\x14\x15\x16\x17\x18\x19\x1a\x1b\x1c\x1d\x1e\x1f");
return true;
}
} // unnamed namespace
namespace Engine
{
enum : DWORD { X64_MAX_REL_ADDR = 0x00300000 };
/** Artikash 6/7/2019
* PPSSPP JIT code has pointers, but they are all added to an offset before being used.
Find that offset so that hook searching works properly.
@ -193,18 +520,40 @@ namespace Engine
{
wcscpy_s(spDefault.exportModule, name);
HookParam hp = {};
hp.address = (DWORD)GetProcAddress(module, "PyUnicodeUCS2_Format");
hp.address = (uintptr_t)GetProcAddress(module, "PyUnicodeUCS2_Format");
if (!hp.address)
{
ConsoleOutput("Textractor: Ren'py failed: failed to find PyUnicodeUCS2_Format");
return false;
}
hp.offset = -0x20; // rcx
hp.offset = pusha_rcx_off -4; // rcx
hp.index = 0x18;
hp.length_offset = 0;
//hp.split = pusha_ebx_off - 4;
//hp.split = pusha_rsp_off -4;
hp.type = USING_STRING | USING_UNICODE | NO_CONTEXT | DATA_INDIRECT /* | USING_SPLIT*/;
//hp.filter_fun = [](void* str, auto, auto, auto) { return *(wchar_t*)str != L'%'; };
hp.filter_fun = [](LPVOID data, DWORD *size, HookParam *, BYTE)
{
static std::wstring prevText;
auto text = reinterpret_cast<LPWSTR>(data);
auto len = reinterpret_cast<size_t *>(size);
if (cpp_wcsnstr(text, L"%", *len/sizeof(wchar_t)))
return false;
if (cpp_wcsnstr(text, L"{", *len/sizeof(wchar_t))) {
WideStringCharReplacer(text, len, L"{i}", 3, L'\'');
WideStringCharReplacer(text, len, L"{/i}", 4, L'\'');
WideStringFilterBetween(text, len, L"{", 1, L"}", 1);
}
WideStringFilter(text, len, L"^", 2); // remove ^ followed by 1 char
WideCharReplacer(text, len, L'\n', L' ');
if (prevText.length()==*len/sizeof(wchar_t) && prevText.find(text, 0, *len/sizeof(wchar_t))!=std::string::npos) // Check if the string is the same as the previous one
return false;
prevText.assign(text, *len/sizeof(wchar_t));
return true;
};
NewHook(hp, "Ren'py");
return true;
}
@ -213,7 +562,126 @@ namespace Engine
ConsoleOutput("Textractor: Ren'py failed: failed to find python2X.dll");
return false;
}
bool InsertRenpy3Hook()
{
//by Blu3train
/*
* Sample games:
* https://vndb.org/v45820
* https://vndb.org/v26318
* https://vndb.org/v39954
* https://vndb.org/r110220
* https://vndb.org/r114981
* https://vndb.org/v33647
* https://vndb.org/r73160
* https://vndb.org/v44518
* https://vndb.org/v31994
* https://vndb.org/r115756
*/
wchar_t python[] = L"python3X.dll", libpython[] = L"libpython3.X.dll";
for (wchar_t* name : { python, libpython })
{
wchar_t* pos = wcschr(name, L'X');
for (int pythonMinorVersion = 0; pythonMinorVersion <= 9; ++pythonMinorVersion)
{
*pos = L'0' + pythonMinorVersion;
if (HMODULE module = GetModuleHandleW(name))
{
wcscpy_s(spDefault.exportModule, name);
HookParam hp = {};
hp.address = (uintptr_t)GetProcAddress(module, "PyUnicode_Format");
if (!hp.address)
{
ConsoleOutput("Textractor: Ren'py 3 failed: failed to find PyUnicode_Format");
return false;
}
hp.offset = pusha_rcx_off -4; // rcx
hp.padding = 0x48;
hp.length_offset = 0;
hp.text_fun = [](DWORD rsp_base, HookParam *pHp, BYTE, DWORD* data, DWORD* split, DWORD* count)
{
uint64_t r10 = regof(r10, rsp_base);
uint64_t r11 = regof(r11, rsp_base);
if (r10==0x03FF || r11==0x03FF) {
uint64_t rcx = regof(rcx, rsp_base);
BYTE unicode = !(*(BYTE*)(rcx + 0x20) & 0x40); // [rcx+0x20) bit 0x40 == 0
if (unicode) {
*data += 0x48; //padding
*count = wcslen((wchar_t*)*data) * sizeof(wchar_t);
return;
}
}
*count = 0;
};
hp.type = USING_STRING | USING_UNICODE | NO_CONTEXT;
hp.filter_fun = [](LPVOID data, DWORD *size, HookParam *, BYTE)
{
auto text = reinterpret_cast<LPWSTR>(data);
auto len = reinterpret_cast<size_t *>(size);
if (cpp_wcsnstr(text, L"%", *len/sizeof(wchar_t)))
return false;
if (cpp_wcsnstr(text, L"{", *len/sizeof(wchar_t))) {
WideStringCharReplacer(text, len, L"{i}", 3, L'\'');
WideStringCharReplacer(text, len, L"{/i}", 4, L'\'');
WideStringFilterBetween(text, len, L"{", 1, L"}", 1);
}
//CP_OEMCP -The current system OEM code page
WideCharToMultiByte(CP_OEMCP, 0, text, -1, text_buffer, 0x1000, NULL, NULL);
text_buffer_length = *len/sizeof(wchar_t); // saved for not unicode hook
return true;
};
NewHook(hp, "Ren'py 3 unicode");
hp.address += 6;
hp.padding = 0x30;
hp.text_fun = [](DWORD rsp_base, HookParam *pHp, BYTE, DWORD* data, DWORD* split, DWORD* count)
{
uint64_t r10 = regof(r10, rsp_base);
uint64_t r11 = regof(r11, rsp_base);
if (r10==0x03FF || r11==0x03FF) {
uint64_t rcx = regof(rcx, rsp_base);
BYTE unicode = !(*(BYTE*)(rcx + 0x20) & 0x40); // [rcx+0x20) bit 0x40 == 0
*data += unicode ? 0x48 : 0x30; //padding
*count = ::strlen((char*)*data);
if (!cpp_strnstr((char*)*data, "%", *count)) // not garbage
return;
}
*count = 0;
};
hp.type = USING_STRING | NO_CONTEXT;
hp.filter_fun = [](LPVOID data, DWORD *size, HookParam *, BYTE)
{
auto text = reinterpret_cast<LPSTR>(data);
auto len = reinterpret_cast<size_t *>(size);
if (text[0]!=0 && text[1]==0) {
// text from unicode hook
*len = text_buffer_length;
::memmove(text, text_buffer, *len);
}
if (cpp_strnstr(text, "%", *len))
return false;
if (cpp_strnstr(text, "{", *len)) {
StringCharReplacer(text, len, "{i}", 3, L'\'');
StringCharReplacer(text, len, "{/i}", 4, L'\'');
StringFilterBetween(text, len, "{", 1, "}", 1);
}
return true;
};
NewHook(hp, "Ren'py 3");
return true;
}
}
}
ConsoleOutput("Textractor: Ren'py 3 failed: failed to find python3X.dll");
return false;
}
bool UnsafeDetermineEngineType()
{
if (Util::CheckFile(L"PPSSPP*.exe") && FindPPSSPP()) return true;
@ -228,7 +696,7 @@ namespace Engine
return true;
}
if (Util::CheckFile(L"*.py") && InsertRenpyHook()) return true;
if (Util::CheckFile(L"*.py") && InsertRenpyHook() || InsertRenpy3Hook()) return true;
for (const wchar_t* monoName : { L"mono.dll", L"mono-2.0-bdwgc.dll" }) if (HMODULE module = GetModuleHandleW(monoName)) if (InsertMonoHooks(module)) return true;