X-Git-Url: https://git.saurik.com/wxWidgets.git/blobdiff_plain/eaba6eba3a1ca393e6ecbafa601cc3b54e705901..f5f93ea573cb7dbbf8c69bfa69466a66d59b1b64:/src/common/string.cpp diff --git a/src/common/string.cpp b/src/common/string.cpp index 1832e5017d..db4b5fe618 100644 --- a/src/common/string.cpp +++ b/src/common/string.cpp @@ -35,6 +35,11 @@ #endif #include + +#ifndef __WXWINCE__ + #include +#endif + #include #include @@ -47,20 +52,36 @@ // NB: EXTRA_ALLOC must be >= 0! #define EXTRA_ALLOC (19 - nLen % 16) + +// string handling functions used by wxString: +#if wxUSE_UNICODE_UTF8 + #define wxStringMemcpy memcpy + #define wxStringMemcmp memcmp + #define wxStringMemchr memchr + #define wxStringStrlen strlen +#else + #define wxStringMemcpy wxTmemcpy + #define wxStringMemcmp wxTmemcmp + #define wxStringMemchr wxTmemchr + #define wxStringStrlen wxStrlen +#endif + + // --------------------------------------------------------------------------- // static class variables definition // --------------------------------------------------------------------------- -#if !wxUSE_STL +#if !wxUSE_STL_BASED_WXSTRING //According to STL _must_ be a -1 size_t - const size_t wxStringBase::npos = (size_t) -1; + const size_t wxStringImpl::npos = (size_t) -1; #endif +const size_t wxString::npos = (size_t) -1; // ---------------------------------------------------------------------------- // static data // ---------------------------------------------------------------------------- -#if wxUSE_STL +#if wxUSE_STL_BASED_WXSTRING extern const wxChar WXDLLIMPEXP_BASE *wxEmptyString = _T(""); @@ -86,52 +107,20 @@ extern const wxChar WXDLLIMPEXP_BASE *wxEmptyString = &g_strEmpty.dummy; #if wxUSE_STD_IOSTREAM -// MS Visual C++ version 5.0 provides the new STL headers as well as the old -// iostream ones. -// -// ATTN: you can _not_ use both of these in the same program! - #include -wxSTD istream& operator>>(wxSTD istream& is, wxString& WXUNUSED(str)) -{ -#if 0 - int w = is.width(0); - if ( is.ipfx(0) ) { - streambuf *sb = is.rdbuf(); - str.erase(); - while ( true ) { - int ch = sb->sbumpc (); - if ( ch == EOF ) { - is.setstate(ios::eofbit); - break; - } - else if ( isspace(ch) ) { - sb->sungetc(); - break; - } - - str += ch; - if ( --w == 1 ) - break; - } - } - - is.isfx(); - if ( str.length() == 0 ) - is.setstate(ios::failbit); -#endif - return is; +wxSTD ostream& operator<<(wxSTD ostream& os, const wxString& str) +{ + return os << str.c_str(); } -wxSTD ostream& operator<<(wxSTD ostream& os, const wxString& str) +wxSTD ostream& operator<<(wxSTD ostream& os, const wxCStrData& str) { -#ifdef __BORLANDC__ - os << str.mb_str(); +#if wxUSE_UNICODE && !defined(__BORLANDC__) + return os << str.AsWChar(); #else - os << str.c_str(); + return os << str.AsChar(); #endif - return os; } #endif // wxUSE_STD_IOSTREAM @@ -165,7 +154,7 @@ wxSTD ostream& operator<<(wxSTD ostream& os, const wxString& str) #define STATISTICS_ADD(av, val) #endif // WXSTRING_STATISTICS -#if !wxUSE_STL +#if !wxUSE_STL_BASED_WXSTRING // =========================================================================== // wxStringData class deallocation @@ -180,11 +169,11 @@ void wxStringData::Free() #endif // =========================================================================== -// wxStringBase +// wxStringImpl // =========================================================================== // takes nLength elements of psz starting at nPos -void wxStringBase::InitWith(const wxChar *psz, size_t nPos, size_t nLength) +void wxStringImpl::InitWith(const wxChar *psz, size_t nPos, size_t nLength) { Init(); @@ -200,21 +189,29 @@ void wxStringBase::InitWith(const wxChar *psz, size_t nPos, size_t nLength) if ( nLength > 0 ) { // trailing '\0' is written in AllocBuffer() if ( !AllocBuffer(nLength) ) { - wxFAIL_MSG( _T("out of memory in wxStringBase::InitWith") ); + wxFAIL_MSG( _T("out of memory in wxStringImpl::InitWith") ); return; } - wxTmemcpy(m_pchData, psz + nPos, nLength); + wxStringMemcpy(m_pchData, psz + nPos, nLength); } } // poor man's iterators are "void *" pointers -wxStringBase::wxStringBase(const void *pStart, const void *pEnd) +wxStringImpl::wxStringImpl(const void *pStart, const void *pEnd) { - InitWith((const wxChar *)pStart, 0, - (const wxChar *)pEnd - (const wxChar *)pStart); + if ( pEnd >= pStart ) + { + InitWith((const wxChar *)pStart, 0, + (const wxChar *)pEnd - (const wxChar *)pStart); + } + else + { + wxFAIL_MSG( _T("pStart is not before pEnd") ); + Init(); + } } -wxStringBase::wxStringBase(size_type n, wxChar ch) +wxStringImpl::wxStringImpl(size_type n, wxStringCharType ch) { Init(); append(n, ch); @@ -225,7 +222,7 @@ wxStringBase::wxStringBase(size_type n, wxChar ch) // --------------------------------------------------------------------------- // allocates memory needed to store a C string of length nLen -bool wxStringBase::AllocBuffer(size_t nLen) +bool wxStringImpl::AllocBuffer(size_t nLen) { // allocating 0 sized buffer doesn't make sense, all empty strings should // reuse g_strEmpty @@ -257,7 +254,7 @@ bool wxStringBase::AllocBuffer(size_t nLen) } // must be called before changing this string -bool wxStringBase::CopyBeforeWrite() +bool wxStringImpl::CopyBeforeWrite() { wxStringData* pData = GetStringData(); @@ -268,7 +265,7 @@ bool wxStringBase::CopyBeforeWrite() // allocation failures are handled by the caller return false; } - wxTmemcpy(m_pchData, pData->data(), nLen); + wxStringMemcpy(m_pchData, pData->data(), nLen); } wxASSERT( !GetStringData()->IsShared() ); // we must be the only owner @@ -277,7 +274,7 @@ bool wxStringBase::CopyBeforeWrite() } // must be called before replacing contents of this string -bool wxStringBase::AllocBeforeWrite(size_t nLen) +bool wxStringImpl::AllocBeforeWrite(size_t nLen) { wxASSERT( nLen != 0 ); // doesn't make any sense @@ -323,12 +320,12 @@ bool wxStringBase::AllocBeforeWrite(size_t nLen) return true; } -wxStringBase& wxStringBase::append(size_t n, wxChar ch) +wxStringImpl& wxStringImpl::append(size_t n, wxStringCharType ch) { size_type len = length(); if ( !Alloc(len + n) || !CopyBeforeWrite() ) { - wxFAIL_MSG( _T("out of memory in wxStringBase::append") ); + wxFAIL_MSG( _T("out of memory in wxStringImpl::append") ); } GetStringData()->nDataLength = len + n; m_pchData[len + n] = '\0'; @@ -337,7 +334,7 @@ wxStringBase& wxStringBase::append(size_t n, wxChar ch) return *this; } -void wxStringBase::resize(size_t nSize, wxChar ch) +void wxStringImpl::resize(size_t nSize, wxStringCharType ch) { size_t len = length(); @@ -353,7 +350,7 @@ void wxStringBase::resize(size_t nSize, wxChar ch) } // allocate enough memory for nLen characters -bool wxStringBase::Alloc(size_t nLen) +bool wxStringImpl::Alloc(size_t nLen) { wxStringData *pData = GetStringData(); if ( pData->nAllocLength <= nLen ) { @@ -407,41 +404,41 @@ bool wxStringBase::Alloc(size_t nLen) return true; } -wxStringBase::iterator wxStringBase::begin() +wxStringImpl::iterator wxStringImpl::begin() { if (length() > 0) CopyBeforeWrite(); return m_pchData; } -wxStringBase::iterator wxStringBase::end() +wxStringImpl::iterator wxStringImpl::end() { if (length() > 0) CopyBeforeWrite(); return m_pchData + length(); } -wxStringBase::iterator wxStringBase::erase(iterator it) +wxStringImpl::iterator wxStringImpl::erase(iterator it) { size_type idx = it - begin(); erase(idx, 1); return begin() + idx; } -wxStringBase& wxStringBase::erase(size_t nStart, size_t nLen) +wxStringImpl& wxStringImpl::erase(size_t nStart, size_t nLen) { wxASSERT(nStart <= length()); size_t strLen = length() - nStart; // delete nLen or up to the end of the string characters nLen = strLen < nLen ? strLen : nLen; - wxString strTmp(c_str(), nStart); + wxStringImpl strTmp(c_str(), nStart); strTmp.append(c_str() + nStart + nLen, length() - nStart - nLen); swap(strTmp); return *this; } -wxStringBase& wxStringBase::insert(size_t nPos, const wxChar *sz, size_t n) +wxStringImpl& wxStringImpl::insert(size_t nPos, const wxChar *sz, size_t n) { wxASSERT( nPos <= length() ); @@ -449,7 +446,7 @@ wxStringBase& wxStringBase::insert(size_t nPos, const wxChar *sz, size_t n) if ( n == 0 ) return *this; if ( !Alloc(length() + n) || !CopyBeforeWrite() ) { - wxFAIL_MSG( _T("out of memory in wxStringBase::insert") ); + wxFAIL_MSG( _T("out of memory in wxStringImpl::insert") ); } memmove(m_pchData + nPos + n, m_pchData + nPos, @@ -461,62 +458,77 @@ wxStringBase& wxStringBase::insert(size_t nPos, const wxChar *sz, size_t n) return *this; } -void wxStringBase::swap(wxStringBase& str) +void wxStringImpl::swap(wxStringImpl& str) { - wxChar* tmp = str.m_pchData; + wxStringCharType* tmp = str.m_pchData; str.m_pchData = m_pchData; m_pchData = tmp; } -size_t wxStringBase::find(const wxStringBase& str, size_t nStart) const +size_t wxStringImpl::find(const wxStringImpl& str, size_t nStart) const { + // deal with the special case of empty string first + const size_t nLen = length(); + const size_t nLenOther = str.length(); + + if ( !nLenOther ) + { + // empty string is a substring of anything + return 0; + } + + if ( !nLen ) + { + // the other string is non empty so can't be our substring + return npos; + } + wxASSERT( str.GetStringData()->IsValid() ); - wxASSERT( nStart <= length() ); + wxASSERT( nStart <= nLen ); - //anchor - const wxChar* p = (const wxChar*)wxTmemchr(c_str() + nStart, - str.c_str()[0], - length() - nStart); + const wxStringCharType * const other = str.c_str(); - if(!p) + // anchor + const wxStringCharType* p = + (const wxStringCharType*)wxStringMemchr(c_str() + nStart, + *other, + nLen - nStart); + + if ( !p ) return npos; - while(p - c_str() + str.length() <= length() && - wxTmemcmp(p, str.c_str(), str.length()) ) + while ( p - c_str() + nLenOther <= nLen && + wxStringMemcmp(p, other, nLenOther) ) { - //Previosly passed as the first argument to wxTmemchr, - //but C/C++ standard does not specify evaluation order - //of arguments to functions - - //http://embedded.com/showArticle.jhtml?articleID=9900607 - ++p; - - //anchor again - p = (const wxChar*)wxTmemchr(p, - str.c_str()[0], - length() - (p - c_str())); - - if(!p) + p++; + + // anchor again + p = (const wxStringCharType*) + wxStringMemchr(p, *other, nLen - (p - c_str())); + + if ( !p ) return npos; } - return (p - c_str() + str.length() <= length()) ? p - c_str() : npos; + return p - c_str() + nLenOther <= nLen ? p - c_str() : npos; } -size_t wxStringBase::find(const wxChar* sz, size_t nStart, size_t n) const +size_t wxStringImpl::find(const wxChar* sz, size_t nStart, size_t n) const { - return find(wxStringBase(sz, n), nStart); + return find(wxStringImpl(sz, n), nStart); } -size_t wxStringBase::find(wxChar ch, size_t nStart) const +size_t wxStringImpl::find(wxStringCharType ch, size_t nStart) const { wxASSERT( nStart <= length() ); - const wxChar *p = (const wxChar*)wxTmemchr(c_str() + nStart, ch, length() - nStart); + const wxStringCharType *p = (const wxStringCharType*) + wxStringMemchr(c_str() + nStart, ch, length() - nStart); return p == NULL ? npos : p - c_str(); } -size_t wxStringBase::rfind(const wxStringBase& str, size_t nStart) const +size_t wxStringImpl::rfind(const wxStringImpl& str, size_t nStart) const { wxASSERT( str.GetStringData()->IsValid() ); wxASSERT( nStart == npos || nStart <= length() ); @@ -535,11 +547,10 @@ size_t wxStringBase::rfind(const wxStringBase& str, size_t nStart) const if ( nStart < top ) top = nStart; - const wxChar *cursor = c_str() + top; + const wxStringCharType *cursor = c_str() + top; do { - if ( wxTmemcmp(cursor, str.c_str(), - str.length()) == 0 ) + if ( wxStringMemcmp(cursor, str.c_str(), str.length()) == 0 ) { return cursor - c_str(); } @@ -549,12 +560,12 @@ size_t wxStringBase::rfind(const wxStringBase& str, size_t nStart) const return npos; } -size_t wxStringBase::rfind(const wxChar* sz, size_t nStart, size_t n) const +size_t wxStringImpl::rfind(const wxChar* sz, size_t nStart, size_t n) const { - return rfind(wxStringBase(sz, n), nStart); + return rfind(wxStringImpl(sz, n), nStart); } -size_t wxStringBase::rfind(wxChar ch, size_t nStart) const +size_t wxStringImpl::rfind(wxStringCharType ch, size_t nStart) const { if ( nStart == npos ) { @@ -565,7 +576,7 @@ size_t wxStringBase::rfind(wxChar ch, size_t nStart) const wxASSERT( nStart <= length() ); } - const wxChar *actual; + const wxStringCharType *actual; for ( actual = c_str() + ( nStart == npos ? length() : nStart + 1 ); actual > c_str(); --actual ) { @@ -576,162 +587,15 @@ size_t wxStringBase::rfind(wxChar ch, size_t nStart) const return npos; } -size_t wxStringBase::find_first_of(const wxChar* sz, size_t nStart) const -{ - wxASSERT(nStart <= length()); - - size_t len = wxStrlen(sz); - - size_t i; - for(i = nStart; i < this->length(); ++i) - { - if (wxTmemchr(sz, *(c_str() + i), len)) - break; - } - - if(i == this->length()) - return npos; - else - return i; -} - -size_t wxStringBase::find_first_of(const wxChar* sz, size_t nStart, - size_t n) const -{ - return find_first_of(wxStringBase(sz, n), nStart); -} - -size_t wxStringBase::find_last_of(const wxChar* sz, size_t nStart) const -{ - if ( nStart == npos ) - { - nStart = length() - 1; - } - else - { - wxASSERT_MSG( nStart <= length(), - _T("invalid index in find_last_of()") ); - } - - size_t len = wxStrlen(sz); - - for ( const wxChar *p = c_str() + nStart; p >= c_str(); --p ) - { - if ( wxTmemchr(sz, *p, len) ) - return p - c_str(); - } - - return npos; -} - -size_t wxStringBase::find_last_of(const wxChar* sz, size_t nStart, - size_t n) const -{ - return find_last_of(wxStringBase(sz, n), nStart); -} - -size_t wxStringBase::find_first_not_of(const wxChar* sz, size_t nStart) const -{ - if ( nStart == npos ) - { - nStart = length(); - } - else - { - wxASSERT( nStart <= length() ); - } - - size_t len = wxStrlen(sz); - - size_t i; - for(i = nStart; i < this->length(); ++i) - { - if (!wxTmemchr(sz, *(c_str() + i), len)) - break; - } - - if(i == this->length()) - return npos; - else - return i; -} - -size_t wxStringBase::find_first_not_of(const wxChar* sz, size_t nStart, - size_t n) const -{ - return find_first_not_of(wxStringBase(sz, n), nStart); -} - -size_t wxStringBase::find_first_not_of(wxChar ch, size_t nStart) const -{ - wxASSERT( nStart <= length() ); - - for ( const wxChar *p = c_str() + nStart; *p; p++ ) - { - if ( *p != ch ) - return p - c_str(); - } - - return npos; -} - -size_t wxStringBase::find_last_not_of(const wxChar* sz, size_t nStart) const -{ - if ( nStart == npos ) - { - nStart = length() - 1; - } - else - { - wxASSERT( nStart <= length() ); - } - - size_t len = wxStrlen(sz); - - for ( const wxChar *p = c_str() + nStart; p >= c_str(); --p ) - { - if ( !wxTmemchr(sz, *p,len) ) - return p - c_str(); - } - - return npos; -} - -size_t wxStringBase::find_last_not_of(const wxChar* sz, size_t nStart, - size_t n) const -{ - return find_last_not_of(wxStringBase(sz, n), nStart); -} - -size_t wxStringBase::find_last_not_of(wxChar ch, size_t nStart) const -{ - if ( nStart == npos ) - { - nStart = length() - 1; - } - else - { - wxASSERT( nStart <= length() ); - } - - for ( const wxChar *p = c_str() + nStart; p >= c_str(); --p ) - { - if ( *p != ch ) - return p - c_str(); - } - - return npos; -} - -wxStringBase& wxStringBase::replace(size_t nStart, size_t nLen, +wxStringImpl& wxStringImpl::replace(size_t nStart, size_t nLen, const wxChar *sz) { wxASSERT_MSG( nStart <= length(), - _T("index out of bounds in wxStringBase::replace") ); + _T("index out of bounds in wxStringImpl::replace") ); size_t strLen = length() - nStart; nLen = strLen < nLen ? strLen : nLen; - wxStringBase strTmp; + wxStringImpl strTmp; strTmp.reserve(length()); // micro optimisation to avoid multiple mem allocs //This is kind of inefficient, but its pretty good considering... @@ -755,34 +619,34 @@ wxStringBase& wxStringBase::replace(size_t nStart, size_t nLen, return *this; } -wxStringBase& wxStringBase::replace(size_t nStart, size_t nLen, - size_t nCount, wxChar ch) +wxStringImpl& wxStringImpl::replace(size_t nStart, size_t nLen, + size_t nCount, wxStringCharType ch) { - return replace(nStart, nLen, wxStringBase(nCount, ch).c_str()); + return replace(nStart, nLen, wxStringImpl(nCount, ch).c_str()); } -wxStringBase& wxStringBase::replace(size_t nStart, size_t nLen, - const wxStringBase& str, +wxStringImpl& wxStringImpl::replace(size_t nStart, size_t nLen, + const wxStringImpl& str, size_t nStart2, size_t nLen2) { return replace(nStart, nLen, str.substr(nStart2, nLen2)); } -wxStringBase& wxStringBase::replace(size_t nStart, size_t nLen, +wxStringImpl& wxStringImpl::replace(size_t nStart, size_t nLen, const wxChar* sz, size_t nCount) { - return replace(nStart, nLen, wxStringBase(sz, nCount).c_str()); + return replace(nStart, nLen, wxStringImpl(sz, nCount).c_str()); } -wxStringBase wxStringBase::substr(size_t nStart, size_t nLen) const +wxStringImpl wxStringImpl::substr(size_t nStart, size_t nLen) const { if ( nLen == npos ) nLen = length() - nStart; - return wxStringBase(*this, nStart, nLen); + return wxStringImpl(*this, nStart, nLen); } // assigns one string to another -wxStringBase& wxStringBase::operator=(const wxStringBase& stringSrc) +wxStringImpl& wxStringImpl::operator=(const wxStringImpl& stringSrc) { wxASSERT( stringSrc.GetStringData()->IsValid() ); @@ -803,25 +667,26 @@ wxStringBase& wxStringBase::operator=(const wxStringBase& stringSrc) } // assigns a single character -wxStringBase& wxStringBase::operator=(wxChar ch) +wxStringImpl& wxStringImpl::operator=(wxStringCharType ch) { - if ( !AssignCopy(1, &ch) ) { - wxFAIL_MSG( _T("out of memory in wxStringBase::operator=(wxChar)") ); + wxChar c(ch); + if ( !AssignCopy(1, &c) ) { + wxFAIL_MSG( _T("out of memory in wxStringImpl::operator=(wxChar)") ); } return *this; } // assigns C string -wxStringBase& wxStringBase::operator=(const wxChar *psz) +wxStringImpl& wxStringImpl::operator=(const wxChar *psz) { if ( !AssignCopy(wxStrlen(psz), psz) ) { - wxFAIL_MSG( _T("out of memory in wxStringBase::operator=(const wxChar *)") ); + wxFAIL_MSG( _T("out of memory in wxStringImpl::operator=(const wxChar *)") ); } return *this; } // helper function: does real copy -bool wxStringBase::AssignCopy(size_t nSrcLen, const wxChar *pszSrcData) +bool wxStringImpl::AssignCopy(size_t nSrcLen, const wxChar *pszSrcData) { if ( nSrcLen == 0 ) { Reinit(); @@ -843,7 +708,7 @@ bool wxStringBase::AssignCopy(size_t nSrcLen, const wxChar *pszSrcData) // --------------------------------------------------------------------------- // add something to this string -bool wxStringBase::ConcatSelf(size_t nSrcLen, const wxChar *pszSrcData, +bool wxStringImpl::ConcatSelf(size_t nSrcLen, const wxChar *pszSrcData, size_t nMaxLen) { STATISTICS_ADD(SummandLength, nSrcLen); @@ -898,70 +763,115 @@ bool wxStringBase::ConcatSelf(size_t nSrcLen, const wxChar *pszSrcData, return true; } -// --------------------------------------------------------------------------- -// simple sub-string extraction -// --------------------------------------------------------------------------- +#endif // !wxUSE_STL_BASED_WXSTRING + +#ifdef HAVE_STD_STRING_COMPARE -// helper function: clone the data attached to this string -bool wxStringBase::AllocCopy(wxString& dest, int nCopyLen, int nCopyIndex) const +// NB: Comparison code (both if HAVE_STD_STRING_COMPARE and if not) works with +// UTF-8 encoded strings too, thanks to UTF-8's design which allows us to +// sort strings in characters code point order by sorting the byte sequence +// in byte values order (i.e. what strcmp() and memcmp() do). + +int wxString::compare(const wxString& str) const { - if ( nCopyLen == 0 ) { - dest.Init(); - } - else { - if ( !dest.AllocBuffer(nCopyLen) ) { - // allocation failure handled by caller - return false; - } - memcpy(dest.m_pchData, m_pchData + nCopyIndex, nCopyLen*sizeof(wxChar)); - } - return true; + return m_impl.compare(str.m_impl); } -#endif // !wxUSE_STL +int wxString::compare(size_t nStart, size_t nLen, + const wxString& str) const +{ + size_t pos, len; + PosLenToImpl(nStart, nLen, &pos, &len); + return m_impl.compare(pos, len, str.m_impl); +} -#if !wxUSE_STL || !defined(HAVE_STD_STRING_COMPARE) +int wxString::compare(size_t nStart, size_t nLen, + const wxString& str, + size_t nStart2, size_t nLen2) const +{ + size_t pos, len; + PosLenToImpl(nStart, nLen, &pos, &len); -#if !wxUSE_STL - #define STRINGCLASS wxStringBase -#else - #define STRINGCLASS wxString -#endif + size_t pos2, len2; + str.PosLenToImpl(nStart2, nLen2, &pos2, &len2); + + return m_impl.compare(pos, len, str.m_impl, pos2, len2); +} + +int wxString::compare(const char* sz) const +{ + return m_impl.compare(ImplStr(sz)); +} + +int wxString::compare(const wchar_t* sz) const +{ + return m_impl.compare(ImplStr(sz)); +} + +int wxString::compare(size_t nStart, size_t nLen, + const char* sz, size_t nCount) const +{ + size_t pos, len; + PosLenToImpl(nStart, nLen, &pos, &len); -static inline int wxDoCmp(const wxChar* s1, size_t l1, - const wxChar* s2, size_t l2) + SubstrBufFromMB str(sz, nCount); + + return m_impl.compare(pos, len, str.data, str.len); +} + +int wxString::compare(size_t nStart, size_t nLen, + const wchar_t* sz, size_t nCount) const +{ + size_t pos, len; + PosLenToImpl(nStart, nLen, &pos, &len); + + SubstrBufFromWC str(sz, nCount); + + return m_impl.compare(pos, len, str.data, str.len); +} + +#else // !HAVE_STD_STRING_COMPARE + +static inline int wxDoCmp(const wxStringCharType* s1, size_t l1, + const wxStringCharType* s2, size_t l2) { if( l1 == l2 ) - return wxTmemcmp(s1, s2, l1); + return wxStringMemcmp(s1, s2, l1); else if( l1 < l2 ) { - int ret = wxTmemcmp(s1, s2, l1); + int ret = wxStringMemcmp(s1, s2, l1); return ret == 0 ? -1 : ret; } else { - int ret = wxTmemcmp(s1, s2, l2); + int ret = wxStringMemcmp(s1, s2, l2); return ret == 0 ? +1 : ret; } } -int STRINGCLASS::compare(const wxStringBase& str) const +int wxString::compare(const wxString& str) const { - return ::wxDoCmp(data(), length(), str.data(), str.length()); + return ::wxDoCmp(m_impl.data(), m_impl.length(), + str.m_impl.data(), str.m_impl.length()); } -int STRINGCLASS::compare(size_t nStart, size_t nLen, - const wxStringBase& str) const +int wxString::compare(size_t nStart, size_t nLen, + const wxString& str) const { wxASSERT(nStart <= length()); size_type strLen = length() - nStart; nLen = strLen < nLen ? strLen : nLen; - return ::wxDoCmp(data() + nStart, nLen, str.data(), str.length()); + + size_t pos, len; + PosLenToImpl(nStart, nLen, &pos, &len); + + return ::wxDoCmp(m_impl.data() + pos, len, + str.m_impl.data(), str.m_impl.length()); } -int STRINGCLASS::compare(size_t nStart, size_t nLen, - const wxStringBase& str, - size_t nStart2, size_t nLen2) const +int wxString::compare(size_t nStart, size_t nLen, + const wxString& str, + size_t nStart2, size_t nLen2) const { wxASSERT(nStart <= length()); wxASSERT(nStart2 <= str.length()); @@ -969,30 +879,67 @@ int STRINGCLASS::compare(size_t nStart, size_t nLen, strLen2 = str.length() - nStart2; nLen = strLen < nLen ? strLen : nLen; nLen2 = strLen2 < nLen2 ? strLen2 : nLen2; - return ::wxDoCmp(data() + nStart, nLen, str.data() + nStart2, nLen2); + + size_t pos, len; + PosLenToImpl(nStart, nLen, &pos, &len); + size_t pos2, len2; + str.PosLenToImpl(nStart2, nLen2, &pos2, &len2); + + return ::wxDoCmp(m_impl.data() + pos, len, + str.m_impl.data() + pos2, len2); } -int STRINGCLASS::compare(const wxChar* sz) const +int wxString::compare(const char* sz) const { - size_t nLen = wxStrlen(sz); - return ::wxDoCmp(data(), length(), sz, nLen); + SubstrBufFromMB str(ImplStr(sz, npos)); + if ( str.len == npos ) + str.len = wxStringStrlen(str.data); + return ::wxDoCmp(m_impl.data(), m_impl.length(), str.data, str.len); } -int STRINGCLASS::compare(size_t nStart, size_t nLen, - const wxChar* sz, size_t nCount) const +int wxString::compare(const wchar_t* sz) const +{ + SubstrBufFromWC str(ImplStr(sz, npos)); + if ( str.len == npos ) + str.len = wxStringStrlen(str.data); + return ::wxDoCmp(m_impl.data(), m_impl.length(), str.data, str.len); +} + +int wxString::compare(size_t nStart, size_t nLen, + const char* sz, size_t nCount) const { wxASSERT(nStart <= length()); size_type strLen = length() - nStart; nLen = strLen < nLen ? strLen : nLen; - if( nCount == npos ) - nCount = wxStrlen(sz); - return ::wxDoCmp(data() + nStart, nLen, sz, nCount); + size_t pos, len; + PosLenToImpl(nStart, nLen, &pos, &len); + + SubstrBufFromMB str(ImplStr(sz, nCount)); + if ( str.len == npos ) + str.len = wxStringStrlen(str.data); + + return ::wxDoCmp(m_impl.data() + pos, len, str.data, str.len); } -#undef STRINGCLASS +int wxString::compare(size_t nStart, size_t nLen, + const wchar_t* sz, size_t nCount) const +{ + wxASSERT(nStart <= length()); + size_type strLen = length() - nStart; + nLen = strLen < nLen ? strLen : nLen; + + size_t pos, len; + PosLenToImpl(nStart, nLen, &pos, &len); + + SubstrBufFromWC str(ImplStr(sz, nCount)); + if ( str.len == npos ) + str.len = wxStringStrlen(str.data); + + return ::wxDoCmp(m_impl.data() + pos, len, str.data, str.len); +} -#endif // !wxUSE_STL || !defined(HAVE_STD_STRING_COMPARE) +#endif // HAVE_STD_STRING_COMPARE/!HAVE_STD_STRING_COMPARE // =========================================================================== // wxString class core @@ -1002,11 +949,54 @@ int STRINGCLASS::compare(size_t nStart, size_t nLen, // construction and conversion // --------------------------------------------------------------------------- +#if wxUSE_UNICODE +/* static */ +wxString::SubstrBufFromMB wxString::ConvertStr(const char *psz, size_t nLength, + const wxMBConv& conv) +{ + // anything to do? + if ( !psz || nLength == 0 ) + return SubstrBufFromMB(); + + if ( nLength == npos ) + nLength = wxNO_LEN; + + size_t wcLen; + wxWCharBuffer wcBuf(conv.cMB2WC(psz, nLength, &wcLen)); + if ( !wcLen ) + return SubstrBufFromMB(); + else + return SubstrBufFromMB(wcBuf, wcLen); +} +#else +/* static */ +wxString::SubstrBufFromWC wxString::ConvertStr(const wchar_t *pwz, size_t nLength, + const wxMBConv& conv) +{ + // anything to do? + if ( !pwz || nLength == 0 ) + return SubstrBufFromWC(); + + if ( nLength == npos ) + nLength = wxNO_LEN; + + size_t mbLen; + wxCharBuffer mbBuf(conv.cWC2MB(pwz, nLength, &mbLen)); + if ( !mbLen ) + return SubstrBufFromWC(); + else + return SubstrBufFromWC(mbBuf, mbLen); +} +#endif + + #if wxUSE_UNICODE // from multibyte string wxString::wxString(const char *psz, const wxMBConv& conv, size_t nLength) { + // FIXME-UTF8: this will need changes + // anything to do? if ( psz && nLength != 0 ) { @@ -1023,6 +1013,11 @@ wxString::wxString(const char *psz, const wxMBConv& conv, size_t nLength) } } +wxString::wxString(const char *psz, size_t nLength) +{ + assign(psz, nLength); +} + //Convert wxString in Unicode mode to a multi-byte string const wxCharBuffer wxString::mb_str(const wxMBConv& conv) const { @@ -1036,6 +1031,8 @@ const wxCharBuffer wxString::mb_str(const wxMBConv& conv) const // from wide string wxString::wxString(const wchar_t *pwz, const wxMBConv& conv, size_t nLength) { + // FIXME-UTF8: this will need changes + // anything to do? if ( pwz && nLength != 0 ) { @@ -1050,6 +1047,12 @@ wxString::wxString(const wchar_t *pwz, const wxMBConv& conv, size_t nLength) if ( nLenMB ) assign(buf, nLenMB); } + +} + +wxString::wxString(const wchar_t *pwz, size_t nLength) +{ + assign(pwz, nLength); } //Converts this string to a wide character string if unicode @@ -1071,9 +1074,9 @@ bool wxString::Shrink() return tmp.length() == length(); } -#if !wxUSE_STL +#if !wxUSE_STL_BASED_WXSTRING && !wxUSE_UNICODE_UTF8 // get the pointer to writable buffer of (at least) nLen bytes -wxChar *wxString::GetWriteBuf(size_t nLen) +wxChar *wxStringImpl::DoGetWriteBuf(size_t nLen) { if ( !AllocBeforeWrite(nLen) ) { // allocation failure handled by caller @@ -1087,18 +1090,43 @@ wxChar *wxString::GetWriteBuf(size_t nLen) } // put string back in a reasonable state after GetWriteBuf +void wxStringImpl::DoUngetWriteBuf() +{ + DoUngetWriteBuf(wxStrlen(m_pchData)); +} + +void wxStringImpl::DoUngetWriteBuf(size_t nLen) +{ + wxStringData * const pData = GetStringData(); + + wxASSERT_MSG( nLen < pData->nAllocLength, _T("buffer overrun") ); + + // the strings we store are always NUL-terminated + pData->data()[nLen] = _T('\0'); + pData->nDataLength = nLen; + pData->Validate(true); +} + +// deprecated compatibility code: +#if WXWIN_COMPATIBILITY_2_8 +wxChar *wxString::GetWriteBuf(size_t nLen) +{ + return DoGetWriteBuf(nLen); +} + void wxString::UngetWriteBuf() { - GetStringData()->nDataLength = wxStrlen(m_pchData); - GetStringData()->Validate(true); + DoUngetWriteBuf(); } void wxString::UngetWriteBuf(size_t nLen) { - GetStringData()->nDataLength = nLen; - GetStringData()->Validate(true); + DoUngetWriteBuf(nLen); } -#endif +#endif // WXWIN_COMPATIBILITY_2_8 + +#endif // !wxUSE_STL_BASED_WXSTRING && !wxUSE_UNICODE_UTF8 + // --------------------------------------------------------------------------- // data access @@ -1139,131 +1167,319 @@ wxString& wxString::operator=(const wchar_t *pwz) wxString operator+(const wxString& str1, const wxString& str2) { -#if !wxUSE_STL - wxASSERT( str1.GetStringData()->IsValid() ); - wxASSERT( str2.GetStringData()->IsValid() ); +#if !wxUSE_STL_BASED_WXSTRING + wxASSERT( str1.IsValid() ); + wxASSERT( str2.IsValid() ); +#endif + + wxString s = str1; + s += str2; + + return s; +} + +wxString operator+(const wxString& str, wxUniChar ch) +{ +#if !wxUSE_STL_BASED_WXSTRING + wxASSERT( str.IsValid() ); +#endif + + wxString s = str; + s += ch; + + return s; +} + +wxString operator+(wxUniChar ch, const wxString& str) +{ +#if !wxUSE_STL_BASED_WXSTRING + wxASSERT( str.IsValid() ); +#endif + + wxString s = ch; + s += str; + + return s; +} + +wxString operator+(const wxString& str, const char *psz) +{ +#if !wxUSE_STL_BASED_WXSTRING + wxASSERT( str.IsValid() ); +#endif + + wxString s; + if ( !s.Alloc(strlen(psz) + str.length()) ) { + wxFAIL_MSG( _T("out of memory in wxString::operator+") ); + } + s += str; + s += psz; + + return s; +} + +wxString operator+(const wxString& str, const wchar_t *pwz) +{ +#if !wxUSE_STL_BASED_WXSTRING + wxASSERT( str.IsValid() ); #endif - wxString s = str1; - s += str2; + wxString s; + if ( !s.Alloc(wxWcslen(pwz) + str.length()) ) { + wxFAIL_MSG( _T("out of memory in wxString::operator+") ); + } + s += str; + s += pwz; - return s; + return s; } -wxString operator+(const wxString& str, wxChar ch) +wxString operator+(const char *psz, const wxString& str) { -#if !wxUSE_STL - wxASSERT( str.GetStringData()->IsValid() ); +#if !wxUSE_STL_BASED_WXSTRING + wxASSERT( str.IsValid() ); #endif - wxString s = str; - s += ch; + wxString s; + if ( !s.Alloc(strlen(psz) + str.length()) ) { + wxFAIL_MSG( _T("out of memory in wxString::operator+") ); + } + s = psz; + s += str; - return s; + return s; } -wxString operator+(wxChar ch, const wxString& str) +wxString operator+(const wchar_t *pwz, const wxString& str) { -#if !wxUSE_STL - wxASSERT( str.GetStringData()->IsValid() ); +#if !wxUSE_STL_BASED_WXSTRING + wxASSERT( str.IsValid() ); #endif - wxString s = ch; - s += str; + wxString s; + if ( !s.Alloc(wxWcslen(pwz) + str.length()) ) { + wxFAIL_MSG( _T("out of memory in wxString::operator+") ); + } + s = pwz; + s += str; + + return s; +} + +// --------------------------------------------------------------------------- +// find_{first,last}_[not]_of functions +// --------------------------------------------------------------------------- + +#if !wxUSE_STL_BASED_WXSTRING || wxUSE_UNICODE_UTF8 - return s; +// NB: All these functions are implemented with the argument being wxChar*, +// i.e. widechar string in any Unicode build, even though native string +// representation is char* in the UTF-8 build. This is because we couldn't +// use memchr() to determine if a character is in a set encoded as UTF-8. + +size_t wxString::find_first_of(const wxChar* sz, size_t nStart) const +{ + return find_first_of(sz, nStart, wxStrlen(sz)); } -wxString operator+(const wxString& str, const wxChar *psz) +size_t wxString::find_first_not_of(const wxChar* sz, size_t nStart) const { -#if !wxUSE_STL - wxASSERT( str.GetStringData()->IsValid() ); -#endif + return find_first_not_of(sz, nStart, wxStrlen(sz)); +} - wxString s; - if ( !s.Alloc(wxStrlen(psz) + str.Len()) ) { - wxFAIL_MSG( _T("out of memory in wxString::operator+") ); - } - s += str; - s += psz; +size_t wxString::find_first_of(const wxChar* sz, size_t nStart, size_t n) const +{ + wxASSERT_MSG( nStart <= length(), _T("invalid index") ); + + size_t idx = nStart; + for ( const_iterator i = begin() + nStart; i != end(); ++idx, ++i ) + { + if ( wxTmemchr(sz, *i, n) ) + return idx; + } - return s; + return npos; } -wxString operator+(const wxChar *psz, const wxString& str) +size_t wxString::find_first_not_of(const wxChar* sz, size_t nStart, size_t n) const { -#if !wxUSE_STL - wxASSERT( str.GetStringData()->IsValid() ); -#endif + wxASSERT_MSG( nStart <= length(), _T("invalid index") ); - wxString s; - if ( !s.Alloc(wxStrlen(psz) + str.Len()) ) { - wxFAIL_MSG( _T("out of memory in wxString::operator+") ); - } - s = psz; - s += str; + size_t idx = nStart; + for ( const_iterator i = begin() + nStart; i != end(); ++idx, ++i ) + { + if ( !wxTmemchr(sz, *i, n) ) + return idx; + } - return s; + return npos; } -// =========================================================================== -// other common string functions -// =========================================================================== -int wxString::Cmp(const wxString& s) const +size_t wxString::find_last_of(const wxChar* sz, size_t nStart) const { - return compare(s); + return find_last_of(sz, nStart, wxStrlen(sz)); } -int wxString::Cmp(const wxChar* psz) const +size_t wxString::find_last_not_of(const wxChar* sz, size_t nStart) const { - return compare(psz); + return find_last_not_of(sz, nStart, wxStrlen(sz)); } -static inline int wxDoCmpNoCase(const wxChar* s1, size_t l1, - const wxChar* s2, size_t l2) +size_t wxString::find_last_of(const wxChar* sz, size_t nStart, size_t n) const { - size_t i; + size_t len = length(); - if( l1 == l2 ) + if ( nStart == npos ) { - for(i = 0; i < l1; ++i) - { - if(wxTolower(s1[i]) != wxTolower(s2[i])) - break; - } - return i == l1 ? 0 : wxTolower(s1[i]) < wxTolower(s2[i]) ? -1 : 1; + nStart = len - 1; } - else if( l1 < l2 ) + else { - for(i = 0; i < l1; ++i) - { - if(wxTolower(s1[i]) != wxTolower(s2[i])) - break; - } - return i == l1 ? -1 : wxTolower(s1[i]) < wxTolower(s2[i]) ? -1 : 1; + wxASSERT_MSG( nStart <= len, _T("invalid index") ); + } + + size_t idx = nStart; + for ( const_reverse_iterator i = rbegin() + (len - nStart - 1); + i != rend(); --idx, ++i ) + { + if ( wxTmemchr(sz, *i, n) ) + return idx; + } + + return npos; +} + +size_t wxString::find_last_not_of(const wxChar* sz, size_t nStart, size_t n) const +{ + size_t len = length(); + + if ( nStart == npos ) + { + nStart = len - 1; } else { - for(i = 0; i < l2; ++i) - { - if(wxTolower(s1[i]) != wxTolower(s2[i])) - break; - } - return i == l2 ? 1 : wxTolower(s1[i]) < wxTolower(s2[i]) ? -1 : 1; + wxASSERT_MSG( nStart <= len, _T("invalid index") ); + } + + size_t idx = nStart; + for ( const_reverse_iterator i = rbegin() + (len - nStart - 1); + i != rend(); --idx, ++i ) + { + if ( !wxTmemchr(sz, *i, n) ) + return idx; } + + return npos; } -int wxString::CmpNoCase(const wxString& s) const +size_t wxString::find_first_not_of(wxUniChar ch, size_t nStart) const +{ + wxASSERT_MSG( nStart <= length(), _T("invalid index") ); + + size_t idx = nStart; + for ( const_iterator i = begin() + nStart; i != end(); ++idx, ++i ) + { + if ( *i != ch ) + return idx; + } + + return npos; +} + +size_t wxString::find_last_not_of(wxUniChar ch, size_t nStart) const { - return wxDoCmpNoCase(data(), length(), s.data(), s.length()); + size_t len = length(); + + if ( nStart == npos ) + { + nStart = len - 1; + } + else + { + wxASSERT_MSG( nStart <= len, _T("invalid index") ); + } + + size_t idx = nStart; + for ( const_reverse_iterator i = rbegin() + (len - nStart - 1); + i != rend(); --idx, ++i ) + { + if ( *i != ch ) + return idx; + } + + return npos; } -int wxString::CmpNoCase(const wxChar* psz) const +// the functions above were implemented for wchar_t* arguments in Unicode +// build and char* in ANSI build; below are implementations for the other +// version: +#if wxUSE_UNICODE + #define wxOtherCharType char + #define STRCONV (const wxChar*)wxConvLibc.cMB2WC +#else + #define wxOtherCharType wchar_t + #define STRCONV (const wxChar*)wxConvLibc.cWC2MB +#endif + +size_t wxString::find_first_of(const wxOtherCharType* sz, size_t nStart) const + { return find_first_of(STRCONV(sz), nStart); } + +size_t wxString::find_first_of(const wxOtherCharType* sz, size_t nStart, + size_t n) const + { return find_first_of(STRCONV(sz, n, NULL), nStart, n); } +size_t wxString::find_last_of(const wxOtherCharType* sz, size_t nStart) const + { return find_last_of(STRCONV(sz), nStart); } +size_t wxString::find_last_of(const wxOtherCharType* sz, size_t nStart, + size_t n) const + { return find_last_of(STRCONV(sz, n, NULL), nStart, n); } +size_t wxString::find_first_not_of(const wxOtherCharType* sz, size_t nStart) const + { return find_first_not_of(STRCONV(sz), nStart); } +size_t wxString::find_first_not_of(const wxOtherCharType* sz, size_t nStart, + size_t n) const + { return find_first_not_of(STRCONV(sz, n, NULL), nStart, n); } +size_t wxString::find_last_not_of(const wxOtherCharType* sz, size_t nStart) const + { return find_last_not_of(STRCONV(sz), nStart); } +size_t wxString::find_last_not_of(const wxOtherCharType* sz, size_t nStart, + size_t n) const + { return find_last_not_of(STRCONV(sz, n, NULL), nStart, n); } + +#undef wxOtherCharType +#undef STRCONV + +#endif // !wxUSE_STL_BASED_WXSTRING || wxUSE_UNICODE_UTF8 + +// =========================================================================== +// other common string functions +// =========================================================================== + +int wxString::CmpNoCase(const wxString& s) const { - int nLen = wxStrlen(psz); + // FIXME-UTF8: use wxUniChar::ToLower/ToUpper once added + + size_t idx = 0; + const_iterator i1 = begin(); + const_iterator end1 = end(); + const_iterator i2 = s.begin(); + const_iterator end2 = s.end(); + + for ( ; i1 != end1 && i2 != end2; ++idx, ++i1, ++i2 ) + { + wxUniChar lower1 = (wxChar)wxTolower(*i1); + wxUniChar lower2 = (wxChar)wxTolower(*i2); + if ( lower1 != lower2 ) + return lower1 < lower2 ? -1 : 1; + } + + size_t len1 = length(); + size_t len2 = s.length(); - return wxDoCmpNoCase(data(), length(), psz, nLen); + if ( len1 < len2 ) + return -1; + else if ( len1 > len2 ) + return 1; + return 0; } @@ -1402,7 +1618,7 @@ bool wxString::EndsWith(const wxChar *suffix, wxString *rest) const wxASSERT_MSG( suffix, _T("invalid parameter in wxString::EndssWith") ); int start = length() - wxStrlen(suffix); - if ( start < 0 || wxStrcmp(c_str() + start, suffix) != 0 ) + if ( start < 0 || wxStrcmp(wx_str() + start, suffix) != 0 ) return false; if ( rest ) @@ -1430,14 +1646,14 @@ wxString wxString::Right(size_t nCount) const // get all characters after the last occurence of ch // (returns the whole string if ch not found) -wxString wxString::AfterLast(wxChar ch) const +wxString wxString::AfterLast(wxUniChar ch) const { wxString str; int iPos = Find(ch, true); if ( iPos == wxNOT_FOUND ) str = *this; else - str = c_str() + iPos + 1; + str = wx_str() + iPos + 1; return str; } @@ -1457,7 +1673,7 @@ wxString wxString::Left(size_t nCount) const // get all characters before the first occurence of ch // (returns the whole string if ch not found) -wxString wxString::BeforeFirst(wxChar ch) const +wxString wxString::BeforeFirst(wxUniChar ch) const { int iPos = Find(ch); if ( iPos == wxNOT_FOUND ) iPos = length(); @@ -1466,7 +1682,7 @@ wxString wxString::BeforeFirst(wxChar ch) const /// get all characters before the last occurence of ch /// (returns empty string if ch not found) -wxString wxString::BeforeLast(wxChar ch) const +wxString wxString::BeforeLast(wxUniChar ch) const { wxString str; int iPos = Find(ch, true); @@ -1478,12 +1694,12 @@ wxString wxString::BeforeLast(wxChar ch) const /// get all characters after the first occurence of ch /// (returns empty string if ch not found) -wxString wxString::AfterFirst(wxChar ch) const +wxString wxString::AfterFirst(wxUniChar ch) const { wxString str; int iPos = Find(ch); if ( iPos != wxNOT_FOUND ) - str = c_str() + iPos + 1; + str = wx_str() + iPos + 1; return str; } @@ -1606,65 +1822,65 @@ inline int wxSafeIsspace(wxChar ch) { return (ch < 127) && wxIsspace(ch); } // trims spaces (in the sense of isspace) from left or right side wxString& wxString::Trim(bool bFromRight) { - // first check if we're going to modify the string at all - if ( !empty() && - ( - (bFromRight && wxSafeIsspace(GetChar(Len() - 1))) || - (!bFromRight && wxSafeIsspace(GetChar(0u))) + // first check if we're going to modify the string at all + if ( !empty() && + ( + (bFromRight && wxSafeIsspace(GetChar(length() - 1))) || + (!bFromRight && wxSafeIsspace(GetChar(0u))) + ) ) - ) - { - if ( bFromRight ) { - // find last non-space character - iterator psz = begin() + length() - 1; - while ( wxSafeIsspace(*psz) && (psz >= begin()) ) - psz--; + if ( bFromRight ) + { + // find last non-space character + reverse_iterator psz = rbegin(); + while ( (psz != rend()) && wxSafeIsspace(*psz) ) + psz++; - // truncate at trailing space start - *++psz = wxT('\0'); - erase(psz, end()); - } - else - { - // find first non-space character - iterator psz = begin(); - while ( wxSafeIsspace(*psz) ) - psz++; + // truncate at trailing space start + erase(psz.base(), end()); + } + else + { + // find first non-space character + iterator psz = begin(); + while ( (psz != end()) && wxSafeIsspace(*psz) ) + psz++; - // fix up data and length - erase(begin(), psz); + // fix up data and length + erase(begin(), psz); + } } - } - return *this; + return *this; } // adds nCount characters chPad to the string from either side -wxString& wxString::Pad(size_t nCount, wxChar chPad, bool bFromRight) +wxString& wxString::Pad(size_t nCount, wxUniChar chPad, bool bFromRight) { - wxString s(chPad, nCount); + wxString s(chPad, nCount); - if ( bFromRight ) - *this += s; - else - { - s += *this; - swap(s); - } + if ( bFromRight ) + *this += s; + else + { + s += *this; + swap(s); + } - return *this; + return *this; } // truncate the string wxString& wxString::Truncate(size_t uiLen) { - if ( uiLen < Len() ) { - erase(begin() + uiLen, end()); - } - //else: nothing to do, string is already short enough + if ( uiLen < length() ) + { + erase(begin() + uiLen, end()); + } + //else: nothing to do, string is already short enough - return *this; + return *this; } // --------------------------------------------------------------------------- @@ -1672,64 +1888,106 @@ wxString& wxString::Truncate(size_t uiLen) // --------------------------------------------------------------------------- // find a character -int wxString::Find(wxChar ch, bool bFromEnd) const +int wxString::Find(wxUniChar ch, bool bFromEnd) const { - size_type idx = bFromEnd ? find_last_of(ch) : find_first_of(ch); + size_type idx = bFromEnd ? find_last_of(ch) : find_first_of(ch); - return (idx == npos) ? wxNOT_FOUND : (int)idx; + return (idx == npos) ? wxNOT_FOUND : (int)idx; } // find a sub-string (like strstr) int wxString::Find(const wxChar *pszSub) const { - size_type idx = find(pszSub); + size_type idx = find(pszSub); - return (idx == npos) ? wxNOT_FOUND : (int)idx; + return (idx == npos) ? wxNOT_FOUND : (int)idx; } // ---------------------------------------------------------------------------- // conversion to numbers // ---------------------------------------------------------------------------- -bool wxString::ToLong(long *val, int base) const +// the implementation of all the functions below is exactly the same so factor +// it out + +template +bool wxStringToIntType(const wxChar *start, + T *val, + int base, + F func) { - wxCHECK_MSG( val, false, _T("NULL pointer in wxString::ToLong") ); + wxCHECK_MSG( val, false, _T("NULL output pointer") ); wxASSERT_MSG( !base || (base > 1 && base <= 36), _T("invalid base") ); - const wxChar *start = c_str(); +#ifndef __WXWINCE__ + errno = 0; +#endif + wxChar *end; - *val = wxStrtol(start, &end, base); + *val = (*func)(start, &end, base); // return true only if scan was stopped by the terminating NUL and if the - // string was not empty to start with - return !*end && (end != start); + // string was not empty to start with and no under/overflow occurred + return !*end && (end != start) +#ifndef __WXWINCE__ + && (errno != ERANGE) +#endif + ; +} + +bool wxString::ToLong(long *val, int base) const +{ + return wxStringToIntType((const wxChar*)c_str(), val, base, wxStrtol); } bool wxString::ToULong(unsigned long *val, int base) const { - wxCHECK_MSG( val, false, _T("NULL pointer in wxString::ToULong") ); - wxASSERT_MSG( !base || (base > 1 && base <= 36), _T("invalid base") ); + return wxStringToIntType((const wxChar*)c_str(), val, base, wxStrtoul); +} - const wxChar *start = c_str(); - wxChar *end; - *val = wxStrtoul(start, &end, base); +bool wxString::ToLongLong(wxLongLong_t *val, int base) const +{ +#ifdef wxHAS_STRTOLL + return wxStringToIntType((const wxChar*)c_str(), val, base, wxStrtoll); +#else + // TODO: implement this ourselves + wxUnusedVar(val); + wxUnusedVar(base); + return false; +#endif // wxHAS_STRTOLL +} - // return true only if scan was stopped by the terminating NUL and if the - // string was not empty to start with - return !*end && (end != start); +bool wxString::ToULongLong(wxULongLong_t *val, int base) const +{ +#ifdef wxHAS_STRTOLL + return wxStringToIntType((const wxChar*)c_str(), val, base, wxStrtoull); +#else + // TODO: implement this ourselves + wxUnusedVar(val); + wxUnusedVar(base); + return false; +#endif } bool wxString::ToDouble(double *val) const { wxCHECK_MSG( val, false, _T("NULL pointer in wxString::ToDouble") ); +#ifndef __WXWINCE__ + errno = 0; +#endif + const wxChar *start = c_str(); wxChar *end; *val = wxStrtod(start, &end); // return true only if scan was stopped by the terminating NUL and if the - // string was not empty to start with - return !*end && (end != start); + // string was not empty to start with and no under/overflow occurred + return !*end && (end != start) +#ifndef __WXWINCE__ + && (errno != ERANGE) +#endif + ; } // --------------------------------------------------------------------------- @@ -1737,13 +1995,17 @@ bool wxString::ToDouble(double *val) const // --------------------------------------------------------------------------- /* static */ -wxString wxString::Format(const wxChar *pszFormat, ...) +#ifdef wxNEEDS_WXSTRING_PRINTF_MIXIN +wxString wxStringPrintfMixinBase::DoFormat(const wxChar *format, ...) +#else +wxString wxString::DoFormat(const wxChar *format, ...) +#endif { va_list argptr; - va_start(argptr, pszFormat); + va_start(argptr, format); wxString s; - s.PrintfV(pszFormat, argptr); + s.PrintfV(format, argptr); va_end(argptr); @@ -1751,26 +2013,39 @@ wxString wxString::Format(const wxChar *pszFormat, ...) } /* static */ -wxString wxString::FormatV(const wxChar *pszFormat, va_list argptr) +wxString wxString::FormatV(const wxString& format, va_list argptr) { wxString s; - s.PrintfV(pszFormat, argptr); + s.PrintfV(format, argptr); return s; } -int wxString::Printf(const wxChar *pszFormat, ...) +#ifdef wxNEEDS_WXSTRING_PRINTF_MIXIN +int wxStringPrintfMixinBase::DoPrintf(const wxChar *format, ...) +#else +int wxString::DoPrintf(const wxChar *format, ...) +#endif { va_list argptr; - va_start(argptr, pszFormat); + va_start(argptr, format); + +#ifdef wxNEEDS_WXSTRING_PRINTF_MIXIN + // get a pointer to the wxString instance; we have to use dynamic_cast<> + // because it's the only cast that works safely for downcasting when + // multiple inheritance is used: + wxString *str = static_cast(this); +#else + wxString *str = this; +#endif - int iLen = PrintfV(pszFormat, argptr); + int iLen = str->PrintfV(format, argptr); va_end(argptr); return iLen; } -int wxString::PrintfV(const wxChar* pszFormat, va_list argptr) +int wxString::PrintfV(const wxString& format, va_list argptr) { int size = 1024; @@ -1789,7 +2064,7 @@ int wxString::PrintfV(const wxChar* pszFormat, va_list argptr) // only a copy va_list argptrcopy; wxVaCopy(argptrcopy, argptr); - int len = wxVsnprintf(buf, size, pszFormat, argptrcopy); + int len = wxVsnprintf(buf, size, format, argptrcopy); va_end(argptrcopy); // some implementations of vsnprintf() don't NUL terminate @@ -1802,13 +2077,29 @@ int wxString::PrintfV(const wxChar* pszFormat, va_list argptr) // buffer were large enough (newer standards such as Unix98) if ( len < 0 ) { +#if wxUSE_WXVSNPRINTF + // we know that our own implementation of wxVsnprintf() returns -1 + // only for a format error - thus there's something wrong with + // the user's format string + return -1; +#else // assume that system version only returns error if not enough space // still not enough, as we don't know how much we need, double the // current size of the buffer size *= 2; +#endif // wxUSE_WXVSNPRINTF/!wxUSE_WXVSNPRINTF } - else if ( len > size ) + else if ( len >= size ) { - size = len; +#if wxUSE_WXVSNPRINTF + // we know that our own implementation of wxVsnprintf() returns + // size+1 when there's not enough space but that's not the size + // of the required buffer! + size *= 2; // so we just double the current size of the buffer +#else + // some vsnprintf() implementations NUL-terminate the buffer and + // some don't in len == size case, to be safe always add 1 + size = len + 1; +#endif } else // ok, there was enough space { @@ -1819,7 +2110,7 @@ int wxString::PrintfV(const wxChar* pszFormat, va_list argptr) // we could have overshot Shrink(); - return Len(); + return length(); } // ---------------------------------------------------------------------------- @@ -1965,13 +2256,12 @@ match: } // Count the number of chars -int wxString::Freq(wxChar ch) const +int wxString::Freq(wxUniChar ch) const { int count = 0; - int len = Len(); - for (int i = 0; i < len; i++) + for ( const_iterator i = begin(); i != end(); ++i ) { - if (GetChar(i) == ch) + if ( *i == ch ) count ++; } return count; @@ -1984,15 +2274,6 @@ wxString wxString::Upper() const // convert to lower case, return the copy of the string wxString wxString::Lower() const { wxString s(*this); return s.MakeLower(); } -int wxString::sprintf(const wxChar *pszFormat, ...) - { - va_list argptr; - va_start(argptr, pszFormat); - int iLen = PrintfV(pszFormat, argptr); - va_end(argptr); - return iLen; - } - // ============================================================================ // ArrayString // ============================================================================ @@ -2026,14 +2307,12 @@ wxArrayString::wxArrayString(size_t sz, const wxString* a) #define ARRAY_DEFAULT_INITIAL_SIZE (16) #endif -#define STRING(p) ((wxString *)(&(p))) - // ctor void wxArrayString::Init(bool autoSort) { m_nSize = m_nCount = 0; - m_pItems = (wxChar **) NULL; + m_pItems = NULL; m_autoSort = autoSort; } @@ -2083,7 +2362,7 @@ void wxArrayString::Grow(size_t nIncrement) m_nSize = ARRAY_DEFAULT_INITIAL_SIZE; if (m_nSize < nIncrement) m_nSize = nIncrement; - m_pItems = new wxChar *[m_nSize]; + m_pItems = new wxString[m_nSize]; } else { // otherwise when it's called for the first time, nIncrement would be 0 @@ -2096,10 +2375,11 @@ void wxArrayString::Grow(size_t nIncrement) if ( nIncrement < ndefIncrement ) nIncrement = ndefIncrement; m_nSize += nIncrement; - wxChar **pNew = new wxChar *[m_nSize]; + wxString *pNew = new wxString[m_nSize]; // copy data to new location - memcpy(pNew, m_pItems, m_nCount*sizeof(wxChar *)); + for ( size_t j = 0; j < m_nCount; j++ ) + pNew[j] = m_pItems[j]; // delete old memory (but do not release the strings!) wxDELETEA(m_pItems); @@ -2109,26 +2389,15 @@ void wxArrayString::Grow(size_t nIncrement) } } -void wxArrayString::Free() -{ - for ( size_t n = 0; n < m_nCount; n++ ) { - STRING(m_pItems[n])->GetStringData()->Unlock(); - } -} - // deletes all the strings from the list void wxArrayString::Empty() { - Free(); - m_nCount = 0; } // as Empty, but also frees memory void wxArrayString::Clear() { - Free(); - m_nSize = m_nCount = 0; @@ -2138,8 +2407,6 @@ void wxArrayString::Clear() // dtor wxArrayString::~wxArrayString() { - Free(); - wxDELETEA(m_pItems); } @@ -2153,13 +2420,17 @@ void wxArrayString::Alloc(size_t nSize) { // only if old buffer was not big enough if ( nSize > m_nSize ) { - Free(); - wxDELETEA(m_pItems); - m_pItems = new wxChar *[nSize]; + wxString *pNew = new wxString[nSize]; + if ( !pNew ) + return; + + for ( size_t j = 0; j < m_nCount; j++ ) + pNew[j] = m_pItems[j]; + delete [] m_pItems; + + m_pItems = pNew; m_nSize = nSize; } - - m_nCount = 0; } // minimizes the memory usage by freeing unused memory @@ -2168,39 +2439,16 @@ void wxArrayString::Shrink() // only do it if we have some memory to free if( m_nCount < m_nSize ) { // allocates exactly as much memory as we need - wxChar **pNew = new wxChar *[m_nCount]; + wxString *pNew = new wxString[m_nCount]; // copy data to new location - memcpy(pNew, m_pItems, m_nCount*sizeof(wxChar *)); + for ( size_t j = 0; j < m_nCount; j++ ) + pNew[j] = m_pItems[j]; delete [] m_pItems; m_pItems = pNew; } } -#if WXWIN_COMPATIBILITY_2_4 - -// return a wxString[] as required for some control ctors. -wxString* wxArrayString::GetStringArray() const -{ - wxString *array = 0; - - if( m_nCount > 0 ) - { - array = new wxString[m_nCount]; - for( size_t i = 0; i < m_nCount; i++ ) - array[i] = m_pItems[i]; - } - - return array; -} - -void wxArrayString::Remove(size_t nIndex, size_t nRemove) -{ - RemoveAt(nIndex, nRemove); -} - -#endif // WXWIN_COMPATIBILITY_2_4 - // searches the array for an item (forward or backwards) int wxArrayString::Index(const wxChar *sz, bool bCase, bool bFromEnd) const { @@ -2233,7 +2481,7 @@ int wxArrayString::Index(const wxChar *sz, bool bCase, bool bFromEnd) const if ( m_nCount > 0 ) { size_t ui = m_nCount; do { - if ( STRING(m_pItems[--ui])->IsSameAs(sz, bCase) ) + if ( m_pItems[--ui].IsSameAs(sz, bCase) ) return ui; } while ( ui != 0 ); @@ -2241,7 +2489,7 @@ int wxArrayString::Index(const wxChar *sz, bool bCase, bool bFromEnd) const } else { for( size_t ui = 0; ui < m_nCount; ui++ ) { - if( STRING(m_pItems[ui])->IsSameAs(sz, bCase) ) + if( m_pItems[ui].IsSameAs(sz, bCase) ) return ui; } } @@ -2280,17 +2528,12 @@ size_t wxArrayString::Add(const wxString& str, size_t nInsert) return (size_t)lo; } else { - wxASSERT( str.GetStringData()->IsValid() ); - Grow(nInsert); for (size_t i = 0; i < nInsert; i++) { - // the string data must not be deleted! - str.GetStringData()->Lock(); - // just append - m_pItems[m_nCount + i] = (wxChar *)str.c_str(); // const_cast + m_pItems[m_nCount + i] = str; } size_t ret = m_nCount; m_nCount += nInsert; @@ -2301,21 +2544,18 @@ size_t wxArrayString::Add(const wxString& str, size_t nInsert) // add item at the given position void wxArrayString::Insert(const wxString& str, size_t nIndex, size_t nInsert) { - wxASSERT( str.GetStringData()->IsValid() ); - wxCHECK_RET( nIndex <= m_nCount, wxT("bad index in wxArrayString::Insert") ); wxCHECK_RET( m_nCount <= m_nCount + nInsert, wxT("array size overflow in wxArrayString::Insert") ); Grow(nInsert); - memmove(&m_pItems[nIndex + nInsert], &m_pItems[nIndex], - (m_nCount - nIndex)*sizeof(wxChar *)); + for (int j = m_nCount - nIndex - 1; j >= 0; j--) + m_pItems[nIndex + nInsert + j] = m_pItems[nIndex + j]; for (size_t i = 0; i < nInsert; i++) { - str.GetStringData()->Lock(); - m_pItems[nIndex + i] = (wxChar *)str.c_str(); + m_pItems[nIndex + i] = str; } m_nCount += nInsert; } @@ -2351,7 +2591,7 @@ void wxArrayString::SetCount(size_t count) wxString s; while ( m_nCount < count ) - m_pItems[m_nCount++] = (wxChar *)s.c_str(); + m_pItems[m_nCount++] = s; } // removes item from array (by index) @@ -2361,12 +2601,9 @@ void wxArrayString::RemoveAt(size_t nIndex, size_t nRemove) wxCHECK_RET( nIndex + nRemove <= m_nCount, wxT("removing too many elements in wxArrayString::Remove") ); - // release our lock - for (size_t i = 0; i < nRemove; i++) - Item(nIndex + i).GetStringData()->Unlock(); + for ( size_t j = 0; j < m_nCount - nIndex -nRemove; j++) + m_pItems[nIndex + j] = m_pItems[nIndex + nRemove + j]; - memmove(&m_pItems[nIndex], &m_pItems[nIndex + nRemove], - (m_nCount - nIndex - nRemove)*sizeof(wxChar *)); m_nCount -= nRemove; } @@ -2447,7 +2684,7 @@ extern "C" void wxArrayString::Sort(CompareFunction2 compareFunction) { - qsort(m_pItems, m_nCount, sizeof(wxChar *), (wxStringCompareFn)compareFunction); + qsort(m_pItems, m_nCount, sizeof(wxString), (wxStringCompareFn)compareFunction); } void wxArrayString::Sort(bool reverseOrder) @@ -2459,9 +2696,7 @@ void wxArrayString::DoSort() { wxCHECK_RET( !m_autoSort, wxT("can't use this method with sorted arrays") ); - // just sort the pointers using qsort() - of course it only works because - // wxString() *is* a pointer to its data - qsort(m_pItems, m_nCount, sizeof(wxChar *), wxStringCompareFunction); + qsort(m_pItems, m_nCount, sizeof(wxString), wxStringCompareFunction); } bool wxArrayString::operator==(const wxArrayString& a) const @@ -2489,3 +2724,106 @@ int wxCMPFUNC_CONV wxStringSortDescending(wxString* s1, wxString* s2) { return -s1->Cmp(*s2); } + + + +// =========================================================================== +// wxJoin and wxSplit +// =========================================================================== + +#include "wx/tokenzr.h" + +wxString wxJoin(const wxArrayString& arr, const wxChar sep, const wxChar escape) +{ + size_t count = arr.size(); + if ( count == 0 ) + return wxEmptyString; + + wxString str; + + // pre-allocate memory using the estimation of the average length of the + // strings in the given array: this is very imprecise, of course, but + // better than nothing + str.reserve(count*(arr[0].length() + arr[count-1].length()) / 2); + + if ( escape == wxT('\0') ) + { + // escaping is disabled: + for ( size_t i = 0; i < count; i++ ) + { + if ( i ) + str += sep; + str += arr[i]; + } + } + else // use escape character + { + for ( size_t n = 0; n < count; n++ ) + { + if ( n ) + str += sep; + + for ( wxString::const_iterator i = arr[n].begin(), + end = arr[n].end(); + i != end; + ++i ) + { + const wxChar ch = *i; + if ( ch == sep ) + str += escape; // escape this separator + str += ch; + } + } + } + + str.Shrink(); // release extra memory if we allocated too much + return str; +} + +wxArrayString wxSplit(const wxString& str, const wxChar sep, const wxChar escape) +{ + if ( escape == wxT('\0') ) + { + // simple case: we don't need to honour the escape character + return wxStringTokenize(str, sep, wxTOKEN_RET_EMPTY_ALL); + } + + wxArrayString ret; + wxString curr; + wxChar prev = wxT('\0'); + + for ( wxString::const_iterator i = str.begin(), + end = str.end(); + i != end; + ++i ) + { + const wxChar ch = *i; + + if ( ch == sep ) + { + if ( prev == escape ) + { + // remove the escape character and don't consider this + // occurrence of 'sep' as a real separator + *curr.rbegin() = sep; + } + else // real separator + { + ret.push_back(curr); + curr.clear(); + } + } + else // normal character + { + curr += ch; + } + + prev = ch; + } + + // add the last token + if ( !curr.empty() || prev == sep ) + ret.Add(curr); + + return ret; +}