lbrycrd/util.h

443 lines
12 KiB
C++

// Copyright (c) 2009 Satoshi Nakamoto
// Distributed under the MIT/X11 software license, see the accompanying
// file license.txt or http://www.opensource.org/licenses/mit-license.php.
#if defined(_MSC_VER) || defined(__BORLANDC__)
typedef __int64 int64;
typedef unsigned __int64 uint64;
#else
typedef long long int64;
typedef unsigned long long uint64;
#endif
#if defined(_MSC_VER) && _MSC_VER < 1300
#define for if (false) ; else for
#endif
#ifndef _MSC_VER
#define __forceinline inline
#endif
#define foreach BOOST_FOREACH
#define loop for (;;)
#define BEGIN(a) ((char*)&(a))
#define END(a) ((char*)&((&(a))[1]))
#define UBEGIN(a) ((unsigned char*)&(a))
#define UEND(a) ((unsigned char*)&((&(a))[1]))
#define ARRAYLEN(array) (sizeof(array)/sizeof((array)[0]))
#define printf OutputDebugStringF
#ifdef snprintf
#undef snprintf
#endif
#define snprintf my_snprintf
#ifndef PRI64d
#if defined(_MSC_VER) || defined(__BORLANDC__) || defined(__MSVCRT__)
#define PRI64d "I64d"
#define PRI64u "I64u"
#define PRI64x "I64x"
#else
#define PRI64d "lld"
#define PRI64u "llu"
#define PRI64x "llx"
#endif
#endif
// This is needed because the foreach macro can't get over the comma in pair<t1, t2>
#define PAIRTYPE(t1, t2) pair<t1, t2>
// Used to bypass the rule against non-const reference to temporary
// where it makes sense with wrappers such as CFlatData or CTxDB
template<typename T>
inline T& REF(const T& val)
{
return (T&)val;
}
extern bool fDebug;
extern bool fPrintToDebugger;
extern bool fPrintToConsole;
extern map<string, string> mapArgs;
void RandAddSeed();
void RandAddSeedPerfmon();
int my_snprintf(char* buffer, size_t limit, const char* format, ...);
string strprintf(const char* format, ...);
bool error(const char* format, ...);
void PrintException(std::exception* pex, const char* pszThread);
void LogException(std::exception* pex, const char* pszThread);
void ParseString(const string& str, char c, vector<string>& v);
string FormatMoney(int64 n, bool fPlus=false);
bool ParseMoney(const char* pszIn, int64& nRet);
vector<unsigned char> ParseHex(const char* psz);
vector<unsigned char> ParseHex(const std::string& str);
bool FileExists(const char* psz);
int GetFilesize(FILE* file);
uint64 GetRand(uint64 nMax);
int64 GetTime();
int64 GetAdjustedTime();
void AddTimeData(unsigned int ip, int64 nTime);
// Wrapper to automatically initialize critical sections
class CCriticalSection
{
#ifdef __WXMSW__
protected:
CRITICAL_SECTION cs;
public:
explicit CCriticalSection() { InitializeCriticalSection(&cs); }
~CCriticalSection() { DeleteCriticalSection(&cs); }
void Enter() { EnterCriticalSection(&cs); }
void Leave() { LeaveCriticalSection(&cs); }
bool TryEnter() { return TryEnterCriticalSection(&cs); }
#else
protected:
wxMutex mutex;
public:
explicit CCriticalSection() { }
~CCriticalSection() { }
void Enter() { mutex.Lock(); }
void Leave() { mutex.Unlock(); }
bool TryEnter() { return mutex.TryLock() == wxMUTEX_NO_ERROR; }
#endif
public:
char* pszFile;
int nLine;
};
// Automatically leave critical section when leaving block, needed for exception safety
class CCriticalBlock
{
protected:
CCriticalSection* pcs;
public:
CCriticalBlock(CCriticalSection& csIn) { pcs = &csIn; pcs->Enter(); }
~CCriticalBlock() { pcs->Leave(); }
};
// WARNING: This will catch continue and break!
// break is caught with an assertion, but there's no way to detect continue.
// I'd rather be careful than suffer the other more error prone syntax.
// The compiler will optimise away all this loop junk.
#define CRITICAL_BLOCK(cs) \
for (bool fcriticalblockonce=true; fcriticalblockonce; assert(("break caught by CRITICAL_BLOCK!", !fcriticalblockonce)), fcriticalblockonce=false) \
for (CCriticalBlock criticalblock(cs); fcriticalblockonce && (cs.pszFile=__FILE__, cs.nLine=__LINE__, true); fcriticalblockonce=false, cs.pszFile=NULL, cs.nLine=0)
class CTryCriticalBlock
{
protected:
CCriticalSection* pcs;
public:
CTryCriticalBlock(CCriticalSection& csIn) { pcs = (csIn.TryEnter() ? &csIn : NULL); }
~CTryCriticalBlock() { if (pcs) pcs->Leave(); }
bool Entered() { return pcs != NULL; }
};
#define TRY_CRITICAL_BLOCK(cs) \
for (bool fcriticalblockonce=true; fcriticalblockonce; assert(("break caught by TRY_CRITICAL_BLOCK!", !fcriticalblockonce)), fcriticalblockonce=false) \
for (CTryCriticalBlock criticalblock(cs); fcriticalblockonce && (fcriticalblockonce = criticalblock.Entered()) && (cs.pszFile=__FILE__, cs.nLine=__LINE__, true); fcriticalblockonce=false, cs.pszFile=NULL, cs.nLine=0)
inline int OutputDebugStringF(const char* pszFormat, ...)
{
int ret = 0;
#ifdef __WXDEBUG__
if (!fPrintToConsole)
{
// print to debug.log
FILE* fileout = fopen("debug.log", "a");
if (fileout)
{
va_list arg_ptr;
va_start(arg_ptr, pszFormat);
ret = vfprintf(fileout, pszFormat, arg_ptr);
va_end(arg_ptr);
fclose(fileout);
}
}
#ifdef __WXMSW__
if (fPrintToDebugger)
{
// accumulate a line at a time
static CCriticalSection cs_OutputDebugStringF;
CRITICAL_BLOCK(cs_OutputDebugStringF)
{
static char pszBuffer[50000];
static char* pend;
if (pend == NULL)
pend = pszBuffer;
va_list arg_ptr;
va_start(arg_ptr, pszFormat);
int limit = END(pszBuffer) - pend - 2;
int ret = _vsnprintf(pend, limit, pszFormat, arg_ptr);
va_end(arg_ptr);
if (ret < 0 || ret >= limit)
{
pend = END(pszBuffer) - 2;
*pend++ = '\n';
}
else
pend += ret;
*pend = '\0';
char* p1 = pszBuffer;
char* p2;
while (p2 = strchr(p1, '\n'))
{
p2++;
char c = *p2;
*p2 = '\0';
OutputDebugString(p1);
*p2 = c;
p1 = p2;
}
if (p1 != pszBuffer)
memmove(pszBuffer, p1, pend - p1 + 1);
pend -= (p1 - pszBuffer);
}
}
#endif
#endif
if (fPrintToConsole)
{
// print to console
va_list arg_ptr;
va_start(arg_ptr, pszFormat);
ret = vprintf(pszFormat, arg_ptr);
va_end(arg_ptr);
}
return ret;
}
inline string i64tostr(int64 n)
{
return strprintf("%"PRI64d, n);
}
inline string itostr(int n)
{
return strprintf("%d", n);
}
inline int64 atoi64(const char* psz)
{
#ifdef _MSC_VER
return _atoi64(psz);
#else
return strtoll(psz, NULL, 10);
#endif
}
inline int64 atoi64(const string& str)
{
#ifdef _MSC_VER
return _atoi64(str.c_str());
#else
return strtoll(str.c_str(), NULL, 10);
#endif
}
inline int atoi(const string& str)
{
return atoi(str.c_str());
}
inline int roundint(double d)
{
return (int)(d > 0 ? d + 0.5 : d - 0.5);
}
template<typename T>
string HexStr(const T itbegin, const T itend, bool fSpaces=true)
{
const unsigned char* pbegin = (const unsigned char*)&itbegin[0];
const unsigned char* pend = pbegin + (itend - itbegin) * sizeof(itbegin[0]);
string str;
for (const unsigned char* p = pbegin; p != pend; p++)
str += strprintf((fSpaces && p != pend-1 ? "%02x " : "%02x"), *p);
return str;
}
inline string HexStr(vector<unsigned char> vch, bool fSpaces=true)
{
return HexStr(vch.begin(), vch.end(), fSpaces);
}
template<typename T>
string HexNumStr(const T itbegin, const T itend, bool f0x=true)
{
const unsigned char* pbegin = (const unsigned char*)&itbegin[0];
const unsigned char* pend = pbegin + (itend - itbegin) * sizeof(itbegin[0]);
string str = (f0x ? "0x" : "");
for (const unsigned char* p = pend-1; p >= pbegin; p--)
str += strprintf("%02X", *p);
return str;
}
template<typename T>
void PrintHex(const T pbegin, const T pend, const char* pszFormat="%s", bool fSpaces=true)
{
printf(pszFormat, HexStr(pbegin, pend, fSpaces).c_str());
}
inline void PrintHex(vector<unsigned char> vch, const char* pszFormat="%s", bool fSpaces=true)
{
printf(pszFormat, HexStr(vch, fSpaces).c_str());
}
inline int64 PerformanceCounter()
{
int64 nCounter = 0;
QueryPerformanceCounter((LARGE_INTEGER*)&nCounter);
return nCounter;
}
#ifndef __WXMSW__
inline void Sleep(unsigned int nMilliseconds)
{
wxMilliSleep(nMilliseconds);
}
#endif
inline void heapchk()
{
if (_heapchk() != _HEAPOK)
DebugBreak();
}
// Randomize the stack to help protect against buffer overrun exploits
#define IMPLEMENT_RANDOMIZE_STACK(ThreadFn) \
{ \
static char nLoops; \
if (nLoops <= 0) \
nLoops = GetRand(50) + 1; \
if (nLoops-- > 1) \
{ \
ThreadFn; \
return; \
} \
}
#define CATCH_PRINT_EXCEPTION(pszFn) \
catch (std::exception& e) { \
PrintException(&e, (pszFn)); \
} catch (...) { \
PrintException(NULL, (pszFn)); \
}
template<typename T1>
inline uint256 Hash(const T1 pbegin, const T1 pend)
{
uint256 hash1;
SHA256((unsigned char*)&pbegin[0], (pend - pbegin) * sizeof(pbegin[0]), (unsigned char*)&hash1);
uint256 hash2;
SHA256((unsigned char*)&hash1, sizeof(hash1), (unsigned char*)&hash2);
return hash2;
}
template<typename T1, typename T2>
inline uint256 Hash(const T1 p1begin, const T1 p1end,
const T2 p2begin, const T2 p2end)
{
uint256 hash1;
SHA256_CTX ctx;
SHA256_Init(&ctx);
SHA256_Update(&ctx, (unsigned char*)&p1begin[0], (p1end - p1begin) * sizeof(p1begin[0]));
SHA256_Update(&ctx, (unsigned char*)&p2begin[0], (p2end - p2begin) * sizeof(p2begin[0]));
SHA256_Final((unsigned char*)&hash1, &ctx);
uint256 hash2;
SHA256((unsigned char*)&hash1, sizeof(hash1), (unsigned char*)&hash2);
return hash2;
}
template<typename T1, typename T2, typename T3>
inline uint256 Hash(const T1 p1begin, const T1 p1end,
const T2 p2begin, const T2 p2end,
const T3 p3begin, const T3 p3end)
{
uint256 hash1;
SHA256_CTX ctx;
SHA256_Init(&ctx);
SHA256_Update(&ctx, (unsigned char*)&p1begin[0], (p1end - p1begin) * sizeof(p1begin[0]));
SHA256_Update(&ctx, (unsigned char*)&p2begin[0], (p2end - p2begin) * sizeof(p2begin[0]));
SHA256_Update(&ctx, (unsigned char*)&p3begin[0], (p3end - p3begin) * sizeof(p3begin[0]));
SHA256_Final((unsigned char*)&hash1, &ctx);
uint256 hash2;
SHA256((unsigned char*)&hash1, sizeof(hash1), (unsigned char*)&hash2);
return hash2;
}
template<typename T>
uint256 SerializeHash(const T& obj, int nType=SER_GETHASH, int nVersion=VERSION)
{
// Most of the time is spent allocating and deallocating CDataStream's
// buffer. If this ever needs to be optimized further, make a CStaticStream
// class with its buffer on the stack.
CDataStream ss(nType, nVersion);
ss.reserve(10000);
ss << obj;
return Hash(ss.begin(), ss.end());
}
inline uint160 Hash160(const vector<unsigned char>& vch)
{
uint256 hash1;
SHA256(&vch[0], vch.size(), (unsigned char*)&hash1);
uint160 hash2;
RIPEMD160((unsigned char*)&hash1, sizeof(hash1), (unsigned char*)&hash2);
return hash2;
}