version 1.115, 2005/11/27 03:11:36
|
version 1.129.2.1, 2007/11/08 09:15:06
|
|
|
//%2005//////////////////////////////////////////////////////////////////////// |
//%2006//////////////////////////////////////////////////////////////////////// |
// | // |
// Copyright (c) 2000, 2001, 2002 BMC Software; Hewlett-Packard Development | // Copyright (c) 2000, 2001, 2002 BMC Software; Hewlett-Packard Development |
// Company, L.P.; IBM Corp.; The Open Group; Tivoli Systems. | // Company, L.P.; IBM Corp.; The Open Group; Tivoli Systems. |
|
|
// IBM Corp.; EMC Corporation; VERITAS Software Corporation; The Open Group. | // IBM Corp.; EMC Corporation; VERITAS Software Corporation; The Open Group. |
// Copyright (c) 2005 Hewlett-Packard Development Company, L.P.; IBM Corp.; | // Copyright (c) 2005 Hewlett-Packard Development Company, L.P.; IBM Corp.; |
// EMC Corporation; VERITAS Software Corporation; The Open Group. | // EMC Corporation; VERITAS Software Corporation; The Open Group. |
|
// Copyright (c) 2006 Hewlett-Packard Development Company, L.P.; IBM Corp.; |
|
// EMC Corporation; Symantec Corporation; The Open Group. |
// | // |
// Permission is hereby granted, free of charge, to any person obtaining a copy | // Permission is hereby granted, free of charge, to any person obtaining a copy |
// of this software and associated documentation files (the "Software"), to | // of this software and associated documentation files (the "Software"), to |
|
|
// | // |
//============================================================================== | //============================================================================== |
// | // |
// Author: Mike Brasher (mbrasher@austin.rr.com) |
|
// |
|
// Modified By: |
|
// Roger Kumpf, Hewlett-Packard Company (roger_kumpf@hp.com) |
|
// Josephine Eskaline Joyce, IBM (jojustin@in.ibm.com) for Bug#3297 |
|
// David Dillard, VERITAS Software Corp. (david.dillard@veritas.com) |
|
// Mike Brasher (mike-brasher@austin.rr.com) |
|
// |
|
//%///////////////////////////////////////////////////////////////////////////// | //%///////////////////////////////////////////////////////////////////////////// |
| |
#include <Pegasus/Common/PegasusAssert.h> | #include <Pegasus/Common/PegasusAssert.h> |
|
|
// | // |
// Compile-time macros (undefined by default). | // Compile-time macros (undefined by default). |
// | // |
// PEGASUS_STRING_NO_THROW -- suppresses throwing of exceptions |
|
// |
|
// PEGASUS_STRING_NO_UTF8 -- don't generate slower UTF8 code. | // PEGASUS_STRING_NO_UTF8 -- don't generate slower UTF8 code. |
// | // |
//============================================================================== | //============================================================================== |
|
|
// Rounds x up to the nearest power of two (or just returns 8 if x < 8). | // Rounds x up to the nearest power of two (or just returns 8 if x < 8). |
static Uint32 _roundUpToPow2(Uint32 x) | static Uint32 _roundUpToPow2(Uint32 x) |
{ | { |
#ifndef PEGASUS_STRING_NO_THROW |
// Check for potential overflow in x |
|
PEGASUS_CHECK_CAPACITY_OVERFLOW(x); |
if (x > 0x0FFFFFFF) |
|
throw PEGASUS_STD(bad_alloc)(); |
|
|
|
#endif |
|
| |
if (x < 8) | if (x < 8) |
return 8; | return 8; |
|
|
return c1 - c2; | return c1 - c2; |
} | } |
| |
static int _compare(const Uint16* s1, const Uint16* s2, size_t n) |
|
{ |
|
// This should only be called when s1 and s2 have the same length. |
|
|
|
while (n-- && (*s1++ - *s2++) == 0) |
|
; |
|
|
|
return s1[-1] - s2[-1]; |
|
} |
|
|
|
static inline void _copy(Uint16* s1, const Uint16* s2, size_t n) | static inline void _copy(Uint16* s1, const Uint16* s2, size_t n) |
{ | { |
memcpy(s1, s2, n * sizeof(Uint16)); | memcpy(s1, s2, n * sizeof(Uint16)); |
|
|
| |
inline void _checkNullPointer(const void* ptr) | inline void _checkNullPointer(const void* ptr) |
{ | { |
#ifdef PEGASUS_STRING_NO_THROW |
|
|
|
if (!ptr) | if (!ptr) |
throw NullPointer(); | throw NullPointer(); |
|
|
#endif |
|
} | } |
| |
static void _StringThrowBadUTF8(Uint32 index) | static void _StringThrowBadUTF8(Uint32 index) |
|
|
| |
inline StringRep* StringRep::alloc(size_t cap) | inline StringRep* StringRep::alloc(size_t cap) |
{ | { |
#ifndef PEGASUS_STRING_NO_THROW |
// Check for potential overflow in cap |
|
PEGASUS_CHECK_CAPACITY_OVERFLOW(cap); |
// Any string bigger than this is seriously suspect. |
|
if (cap > 0x0FFFFFFF) |
|
throw PEGASUS_STD(bad_alloc)(); |
|
|
|
#endif |
|
| |
StringRep* rep = (StringRep*)::operator new( | StringRep* rep = (StringRep*)::operator new( |
sizeof(StringRep) + cap * sizeof(Uint16)); | sizeof(StringRep) + cap * sizeof(Uint16)); |
|
|
size_t utf8_error_index; | size_t utf8_error_index; |
rep->size = _convert((Uint16*)rep->data, data, size, utf8_error_index); | rep->size = _convert((Uint16*)rep->data, data, size, utf8_error_index); |
| |
#ifndef PEGASUS_STRING_NO_THROW |
|
if (rep->size == size_t(-1)) | if (rep->size == size_t(-1)) |
{ | { |
StringRep::free(rep); | StringRep::free(rep); |
_StringThrowBadUTF8(utf8_error_index); |
_StringThrowBadUTF8((Uint32)utf8_error_index); |
} | } |
#endif |
|
| |
rep->data[rep->size] = '\0'; | rep->data[rep->size] = '\0'; |
| |
|
|
while (*end++) | while (*end++) |
; | ; |
| |
return end - str - 1; |
return (Uint32)(end - str - 1); |
} | } |
| |
//============================================================================== | //============================================================================== |
|
|
size_t utf8_error_index; | size_t utf8_error_index; |
size_t tmp = _convert((Uint16*)_rep->data + n1, s2, n2, utf8_error_index); | size_t tmp = _convert((Uint16*)_rep->data + n1, s2, n2, utf8_error_index); |
| |
#ifndef PEGASUS_STRING_NO_THROW |
|
if (tmp == size_t(-1)) | if (tmp == size_t(-1)) |
{ | { |
StringRep::free(_rep); | StringRep::free(_rep); |
_rep = &StringRep::_emptyRep; | _rep = &StringRep::_emptyRep; |
_StringThrowBadUTF8(utf8_error_index); |
_StringThrowBadUTF8((Uint32)utf8_error_index); |
} | } |
#endif |
|
| |
_rep->size = n1 + tmp; | _rep->size = n1 + tmp; |
_rep->data[_rep->size] = '\0'; | _rep->data[_rep->size] = '\0'; |
|
|
size_t utf8_error_index; | size_t utf8_error_index; |
size_t tmp = _convert((Uint16*)_rep->data, s1, n1, utf8_error_index); | size_t tmp = _convert((Uint16*)_rep->data, s1, n1, utf8_error_index); |
| |
#ifndef PEGASUS_STRING_NO_THROW |
|
if (tmp == size_t(-1)) | if (tmp == size_t(-1)) |
{ | { |
StringRep::free(_rep); | StringRep::free(_rep); |
_rep = &StringRep::_emptyRep; | _rep = &StringRep::_emptyRep; |
_StringThrowBadUTF8(utf8_error_index); |
_StringThrowBadUTF8((Uint32)utf8_error_index); |
} | } |
#endif |
|
| |
_rep->size = n2 + tmp; | _rep->size = n2 + tmp; |
_copy(_rep->data + n1, s2._rep->data, n2); | _copy(_rep->data + n1, s2._rep->data, n2); |
|
|
size_t utf8_error_index; | size_t utf8_error_index; |
_rep->size = _convert(_rep->data, str, n, utf8_error_index); | _rep->size = _convert(_rep->data, str, n, utf8_error_index); |
| |
#ifndef PEGASUS_STRING_NO_THROW |
|
if (_rep->size == size_t(-1)) | if (_rep->size == size_t(-1)) |
{ | { |
StringRep::free(_rep); | StringRep::free(_rep); |
_rep = &StringRep::_emptyRep; | _rep = &StringRep::_emptyRep; |
_StringThrowBadUTF8(utf8_error_index); |
_StringThrowBadUTF8((Uint32)utf8_error_index); |
} | } |
#endif |
|
| |
_rep->data[_rep->size] = 0; | _rep->data[_rep->size] = 0; |
| |
|
|
str[_rep->size] = '\0'; | str[_rep->size] = '\0'; |
return CString(str); | return CString(str); |
#else | #else |
Uint32 n = 3 * _rep->size; |
Uint32 n = (Uint32)(3 * _rep->size); |
char* str = (char*)operator new(n + 1); | char* str = (char*)operator new(n + 1); |
size_t size = _copyToUTF8(str, _rep->data, _rep->size); | size_t size = _copyToUTF8(str, _rep->data, _rep->size); |
str[size] = '\0'; | str[size] = '\0'; |
|
|
| |
size_t oldSize = _rep->size; | size_t oldSize = _rep->size; |
size_t newSize = oldSize + n; | size_t newSize = oldSize + n; |
_reserve(_rep, newSize); |
_reserve(_rep, (Uint32)newSize); |
_copy(_rep->data + oldSize, (Uint16*)str, n); | _copy(_rep->data + oldSize, (Uint16*)str, n); |
_rep->size = newSize; | _rep->size = newSize; |
_rep->data[newSize] = '\0'; | _rep->data[newSize] = '\0'; |
|
|
| |
String& String::append(const String& str) | String& String::append(const String& str) |
{ | { |
return append((Char16*)str._rep->data, str._rep->size); |
return append((Char16*)(&(str._rep->data[0])), (Uint32)str._rep->size); |
} | } |
| |
String& String::append(const char* str, Uint32 size) | String& String::append(const char* str, Uint32 size) |
|
|
size_t oldSize = _rep->size; | size_t oldSize = _rep->size; |
size_t cap = oldSize + size; | size_t cap = oldSize + size; |
| |
_reserve(_rep, cap); |
_reserve(_rep, (Uint32)cap); |
size_t utf8_error_index; | size_t utf8_error_index; |
size_t tmp = _convert( | size_t tmp = _convert( |
(Uint16*)_rep->data + oldSize, str, size, utf8_error_index); | (Uint16*)_rep->data + oldSize, str, size, utf8_error_index); |
| |
#ifndef PEGASUS_STRING_NO_THROW |
|
if (tmp == size_t(-1)) | if (tmp == size_t(-1)) |
{ | { |
StringRep::free(_rep); | StringRep::free(_rep); |
_rep = &StringRep::_emptyRep; | _rep = &StringRep::_emptyRep; |
_StringThrowBadUTF8(utf8_error_index); |
_StringThrowBadUTF8((Uint32)utf8_error_index); |
} | } |
#endif |
|
| |
_rep->size += tmp; | _rep->size += tmp; |
_rep->data[_rep->size] = '\0'; | _rep->data[_rep->size] = '\0'; |
|
|
void String::remove(Uint32 index, Uint32 n) | void String::remove(Uint32 index, Uint32 n) |
{ | { |
if (n == PEG_NOT_FOUND) | if (n == PEG_NOT_FOUND) |
n = _rep->size - index; |
n = (Uint32)(_rep->size - index); |
| |
_checkBounds(index + n, _rep->size); | _checkBounds(index + n, _rep->size); |
| |
|
|
if (index < _rep->size) | if (index < _rep->size) |
{ | { |
if (n == PEG_NOT_FOUND || n > _rep->size - index) | if (n == PEG_NOT_FOUND || n > _rep->size - index) |
n = _rep->size - index; |
n = (Uint32)(_rep->size - index); |
| |
return String((Char16*)_rep->data + index, n); |
return String((Char16*)(_rep->data + index), n); |
} | } |
| |
return String(); | return String(); |
|
|
Uint16* p = (Uint16*)_find(_rep->data, _rep->size, c); | Uint16* p = (Uint16*)_find(_rep->data, _rep->size, c); |
| |
if (p) | if (p) |
return p - _rep->data; |
return static_cast<Uint32>(p - _rep->data); |
| |
return PEG_NOT_FOUND; | return PEG_NOT_FOUND; |
} | } |
|
|
Uint16* p = (Uint16*)_find(_rep->data + index, _rep->size - index, c); | Uint16* p = (Uint16*)_find(_rep->data + index, _rep->size - index, c); |
| |
if (p) | if (p) |
return p - _rep->data; |
return static_cast<Uint32>(p - _rep->data); |
| |
return PEG_NOT_FOUND; | return PEG_NOT_FOUND; |
} | } |
|
|
break; | break; |
| |
if (memcmp(p, s, n * sizeof(Uint16)) == 0) | if (memcmp(p, s, n * sizeof(Uint16)) == 0) |
return p - _rep->data; |
return static_cast<Uint32>(p - _rep->data); |
| |
p++; | p++; |
rem -= p - data; | rem -= p - data; |
|
|
while (q != p) | while (q != p) |
{ | { |
if (*--q == x) | if (*--q == x) |
return q - p; |
return static_cast<Uint32>(q - p); |
} | } |
| |
return PEG_NOT_FOUND; | return PEG_NOT_FOUND; |
|
|
| |
int String::compare(const String& s1, const String& s2, Uint32 n) | int String::compare(const String& s1, const String& s2, Uint32 n) |
{ | { |
PEGASUS_ASSERT(n <= s1._rep->size); |
const Uint16* p1 = s1._rep->data; |
PEGASUS_ASSERT(n <= s2._rep->size); |
const Uint16* p2 = s2._rep->data; |
| |
// Ignoring error in which n is greater than s1.size() or s2.size() |
while (n--) |
return _compare(s1._rep->data, s2._rep->data, n); |
{ |
|
int r = *p1++ - *p2++; |
|
if (r) |
|
{ |
|
return r; |
|
} |
|
else if (!p1[-1]) |
|
{ |
|
// We must have encountered a null terminator in both s1 and s2 |
|
return 0; |
|
} |
|
} |
|
return 0; |
} | } |
| |
int String::compare(const String& s1, const String& s2) | int String::compare(const String& s1, const String& s2) |
|
|
if (InitializeICU::initICUSuccessful()) | if (InitializeICU::initICUSuccessful()) |
{ | { |
return u_strcasecmp( | return u_strcasecmp( |
str1._rep->data, str2._rep->data, U_FOLD_CASE_DEFAULT); |
(const UChar*)str1._rep->data, |
|
(const UChar*)str2._rep->data, |
|
U_FOLD_CASE_DEFAULT |
|
); |
} | } |
| |
#endif /* PEGASUS_HAS_ICU */ | #endif /* PEGASUS_HAS_ICU */ |
|
|
| |
Boolean String::equal(const String& s1, const String& s2) | Boolean String::equal(const String& s1, const String& s2) |
{ | { |
return s1._rep->size == s2._rep->size && memcmp(s1._rep->data, |
return (s1._rep == s2._rep) || |
s2._rep->data, s1._rep->size * sizeof(Uint16)) == 0; |
(s1._rep->size == s2._rep->size) && |
|
memcmp(s1._rep->data, |
|
s2._rep->data, |
|
s1._rep->size * sizeof(Uint16)) == 0; |
} | } |
| |
Boolean String::equal(const String& s1, const char* s2) | Boolean String::equal(const String& s1, const char* s2) |
|
|
| |
PEGASUS_STD(ostream)& operator<<(PEGASUS_STD(ostream)& os, const String& str) | PEGASUS_STD(ostream)& operator<<(PEGASUS_STD(ostream)& os, const String& str) |
{ | { |
#if defined(PEGASUS_OS_OS400) |
|
|
|
CString cstr = str.getCString(); |
|
const char* utf8str = cstr; |
|
os << utf8str; |
|
return os; |
|
#else |
|
|
|
#if defined(PEGASUS_HAS_ICU) | #if defined(PEGASUS_HAS_ICU) |
| |
if (InitializeICU::initICUSuccessful()) | if (InitializeICU::initICUSuccessful()) |
|
|
} | } |
| |
return os; | return os; |
#endif // PEGASUS_OS_OS400 |
|
} | } |
| |
void StringAppendCharAux(StringRep*& _rep) | void StringAppendCharAux(StringRep*& _rep) |
|
|
_rep = tmp; | _rep = tmp; |
} | } |
| |
|
void AssignASCII(String& s, const char* str, Uint32 n) |
|
{ |
|
class StringLayout |
|
{ |
|
public: |
|
StringRep* rep; |
|
}; |
|
|
|
StringLayout* that = (StringLayout*)&s; |
|
|
|
_checkNullPointer(str); |
|
|
|
if (n > that->rep->cap || that->rep->refs.get() != 1) |
|
{ |
|
StringRep::unref(that->rep); |
|
that->rep = StringRep::alloc(n); |
|
} |
|
|
|
_copy(that->rep->data, str, n); |
|
that->rep->size = n; |
|
that->rep->data[that->rep->size] = 0; |
|
} |
|
|
PEGASUS_NAMESPACE_END | PEGASUS_NAMESPACE_END |
| |
/* | /* |
|
|
This avoids slower UTF8 processing when not needed. | This avoids slower UTF8 processing when not needed. |
| |
================================================================================ | ================================================================================ |
|
|
TO-DO: |
|
|
|
(+) [DONE] Use PEGASUS_USE_EXPERIMENTAL_INTERFACES |
|
|
|
(+) [DONE] Submit BUG-2754 (Windows buffer limit). |
|
|
|
(+) [DONE] Eliminate char versions of find() and append(). |
|
|
|
(+) [DONE] Remove PEGASUS_MAX_PRINTABLE_CHARACTER from Config.h |
|
|
|
(+) [DONE] Change _next_pow_2() to _roundUpToPow2(). |
|
|
|
(+) [DONE] Change '99' to '2' in StringRep constructor (comment as well). |
|
|
|
(+) [DONE] Comment StringRep allocation layout. |
|
|
|
(+) [DONE] Conceal private inline functions. |
|
|
|
(+) [DONE] Shorten inclusion of StringInline.h in String.h. |
|
|
|
(+) [DONE] Change USE_INTERNAL_INLINE TO DISABLE_INTERNAL_INLINE or get |
|
rid of altogether. |
|
|
|
(+) [DONE] useCamelNotationOnAllFunctionNames. |
|
|
|
(+) [DONE] Check for overlow condition in StringRep::alloc(). |
|
|
|
(+) [DONE] Remove tabs (used vim ":set expandtab" and ":retab"). |
|
|
|
(+) [DONE] Fix throw-related memory leak. |
|
|
|
(+) [DONE] Look at PEP223 for coding security guidelines. |
|
|
|
(+) [DONE] Use old AtomicInt for now (new AtomicInt part of bug #4250). |
|
|
|
(+) [DONE] Removed appendASCII() and the ASCII form of the constructor. |
|
|
|
(+) DOC++ String.h - will open new bug? |
|
|
|
(+) Added PEGASUS_DISABLE_INTERNAL_INLINES macro (to permit suppression |
|
on certain platforms). |
|
|
|
================================================================================ |
|
*/ | */ |