X-Git-Url: https://git.saurik.com/wxWidgets.git/blobdiff_plain/d36c9347ea16171bc2f855076d8b9b11801ec622..68a9527d5185cdd9d3fef3d9421415c50de8794d:/include/wx/strconv.h?ds=inline diff --git a/include/wx/strconv.h b/include/wx/strconv.h index 73620ad125..c25a05017c 100644 --- a/include/wx/strconv.h +++ b/include/wx/strconv.h @@ -1,11 +1,12 @@ /////////////////////////////////////////////////////////////////////////////// // Name: strconv.h // Purpose: conversion routines for char sets any Unicode -// Author: Robert Roebling, Ove Kaaven +// Author: Ove Kaaven, Robert Roebling, Vadim Zeitlin // Modified by: // Created: 29/01/98 // RCS-ID: $Id$ -// Copyright: (c) 1998 Ove Kaaven, Robert Roebling, Vadim Zeitlin +// Copyright: (c) 1998 Ove Kaaven, Robert Roebling +// (c) 1998-2006 Vadim Zeitlin // Licence: wxWindows licence /////////////////////////////////////////////////////////////////////////////// @@ -13,7 +14,7 @@ #define _WX_STRCONV_H_ #include "wx/defs.h" -#include "wx/wxchar.h" +#include "wx/chartype.h" #include "wx/buffer.h" #ifdef __DIGITALMARS__ @@ -31,6 +32,10 @@ // the error value returned by wxMBConv methods #define wxCONV_FAILED ((size_t)-1) +// the default value for some length parameters meaning that the string is +// NUL-terminated +#define wxNO_LEN ((size_t)-1) + // ---------------------------------------------------------------------------- // wxMBConv (abstract base class for conversions) // ---------------------------------------------------------------------------- @@ -54,12 +59,13 @@ public: // characters, not bytes) of the converted string including any trailing // L'\0' or (possibly multiple) '\0'(s). If the conversion fails or if // there is not enough space for everything, including the trailing NUL - // character(s), in the output buffer, (size_t)-1 is returned. + // character(s), in the output buffer, wxCONV_FAILED is returned. // // In the special case when dstLen is 0 (outputBuf may be NULL then) the // return value is the length of the needed buffer but nothing happens - // otherwise. If srcLen is -1, the entire string, up to and including the - // trailing NUL(s), is converted, otherwise exactly srcLen bytes are. + // otherwise. If srcLen is wxNO_LEN, the entire string, up to and + // including the trailing NUL(s), is converted, otherwise exactly srcLen + // bytes are. // // Typical usage: // @@ -70,10 +76,10 @@ public: // conv.ToWChar(wbuf, dstLen, src); // virtual size_t ToWChar(wchar_t *dst, size_t dstLen, - const char *src, size_t srcLen = -1) const; + const char *src, size_t srcLen = wxNO_LEN) const; virtual size_t FromWChar(char *dst, size_t dstLen, - const wchar_t *src, size_t srcLen = -1) const; + const wchar_t *src, size_t srcLen = wxNO_LEN) const; // Convenience functions for translating NUL-terminated strings: returns @@ -185,7 +191,8 @@ class WXDLLIMPEXP_BASE wxConvBrokenFileNames : public wxMBConv public: wxConvBrokenFileNames(const wxChar *charset); wxConvBrokenFileNames(const wxConvBrokenFileNames& conv) - : m_conv(conv.m_conv ? conv.m_conv->Clone() : NULL) + : wxMBConv(), + m_conv(conv.m_conv ? conv.m_conv->Clone() : NULL) { } virtual ~wxConvBrokenFileNames() { delete m_conv; } @@ -261,7 +268,16 @@ private: class WXDLLIMPEXP_BASE wxMBConvUTF16Base : public wxMBConv { public: - virtual size_t GetMBNulLen() const { return 2; } + enum { BYTES_PER_CHAR = 2 }; + + virtual size_t GetMBNulLen() const { return BYTES_PER_CHAR; } + +protected: + // return the length of the buffer using srcLen if it's not wxNO_LEN and + // computing the length ourselves if it is; also checks that the length is + // even if specified as we need an entire number of UTF-16 characters and + // returns wxNO_LEN which indicates error if it is odd + static size_t GetLength(const char *src, size_t srcLen); }; // ---------------------------------------------------------------------------- @@ -271,8 +287,10 @@ public: class WXDLLIMPEXP_BASE wxMBConvUTF16LE : public wxMBConvUTF16Base { public: - virtual size_t MB2WC(wchar_t *outputBuf, const char *psz, size_t outputSize) const; - virtual size_t WC2MB(char *outputBuf, const wchar_t *psz, size_t outputSize) const; + virtual size_t ToWChar(wchar_t *dst, size_t dstLen, + const char *src, size_t srcLen = wxNO_LEN) const; + virtual size_t FromWChar(char *dst, size_t dstLen, + const wchar_t *src, size_t srcLen = wxNO_LEN) const; virtual wxMBConv *Clone() const { return new wxMBConvUTF16LE; } }; @@ -283,8 +301,10 @@ public: class WXDLLIMPEXP_BASE wxMBConvUTF16BE : public wxMBConvUTF16Base { public: - virtual size_t MB2WC(wchar_t *outputBuf, const char *psz, size_t outputSize) const; - virtual size_t WC2MB(char *outputBuf, const wchar_t *psz, size_t outputSize) const; + virtual size_t ToWChar(wchar_t *dst, size_t dstLen, + const char *src, size_t srcLen = wxNO_LEN) const; + virtual size_t FromWChar(char *dst, size_t dstLen, + const wchar_t *src, size_t srcLen = wxNO_LEN) const; virtual wxMBConv *Clone() const { return new wxMBConvUTF16BE; } }; @@ -295,7 +315,15 @@ public: class WXDLLIMPEXP_BASE wxMBConvUTF32Base : public wxMBConv { public: - virtual size_t GetMBNulLen() const { return 4; } + enum { BYTES_PER_CHAR = 4 }; + + virtual size_t GetMBNulLen() const { return BYTES_PER_CHAR; } + +protected: + // this is similar to wxMBConvUTF16Base method with the same name except + // that, of course, it verifies that length is divisible by 4 if given and + // not by 2 + static size_t GetLength(const char *src, size_t srcLen); }; // ---------------------------------------------------------------------------- @@ -305,8 +333,10 @@ public: class WXDLLIMPEXP_BASE wxMBConvUTF32LE : public wxMBConvUTF32Base { public: - virtual size_t MB2WC(wchar_t *outputBuf, const char *psz, size_t outputSize) const; - virtual size_t WC2MB(char *outputBuf, const wchar_t *psz, size_t outputSize) const; + virtual size_t ToWChar(wchar_t *dst, size_t dstLen, + const char *src, size_t srcLen = wxNO_LEN) const; + virtual size_t FromWChar(char *dst, size_t dstLen, + const wchar_t *src, size_t srcLen = wxNO_LEN) const; virtual wxMBConv *Clone() const { return new wxMBConvUTF32LE; } }; @@ -317,8 +347,10 @@ public: class WXDLLIMPEXP_BASE wxMBConvUTF32BE : public wxMBConvUTF32Base { public: - virtual size_t MB2WC(wchar_t *outputBuf, const char *psz, size_t outputSize) const; - virtual size_t WC2MB(char *outputBuf, const wchar_t *psz, size_t outputSize) const; + virtual size_t ToWChar(wchar_t *dst, size_t dstLen, + const char *src, size_t srcLen = wxNO_LEN) const; + virtual size_t FromWChar(char *dst, size_t dstLen, + const wchar_t *src, size_t srcLen = wxNO_LEN) const; virtual wxMBConv *Clone() const { return new wxMBConvUTF32BE; } }; @@ -341,13 +373,21 @@ public: wxCSConv& operator=(const wxCSConv& conv); + virtual size_t ToWChar(wchar_t *dst, size_t dstLen, + const char *src, size_t srcLen = wxNO_LEN) const; + virtual size_t FromWChar(char *dst, size_t dstLen, + const wchar_t *src, size_t srcLen = wxNO_LEN) const; virtual size_t MB2WC(wchar_t *outputBuf, const char *psz, size_t outputSize) const; virtual size_t WC2MB(char *outputBuf, const wchar_t *psz, size_t outputSize) const; virtual size_t GetMBNulLen() const; + virtual wxMBConv *Clone() const { return new wxCSConv(*this); } void Clear(); + // return true if the conversion could be initilized successfully + bool IsOk() const; + private: // common part of all ctors void Init(); @@ -403,9 +443,14 @@ extern WXDLLIMPEXP_DATA_BASE(wxMBConv *) wxConvFileName; // default in a couple of places inside wx (initially same as wxConvLibc) extern WXDLLIMPEXP_DATA_BASE(wxMBConv *) wxConvCurrent; -// ??? +// the conversion corresponding to the current locale extern WXDLLIMPEXP_DATA_BASE(wxCSConv&) wxConvLocal; +// the conversion corresponding to the encoding of the standard UI elements +// +// by default this is the same as wxConvLocal but may be changed if the program +// needs to use a fixed encoding +extern WXDLLIMPEXP_DATA_BASE(wxMBConv *) wxConvUI; // ---------------------------------------------------------------------------- // endianness-dependent conversions @@ -423,8 +468,8 @@ extern WXDLLIMPEXP_DATA_BASE(wxCSConv&) wxConvLocal; // filename conversion macros // ---------------------------------------------------------------------------- -// filenames are multibyte on Unix and probably widechar on Windows? -#if defined(__UNIX__) || defined(__BORLANDC__) || defined(__WXMAC__ ) +// filenames are multibyte on Unix and widechar on Windows +#if defined(__UNIX__) || defined(__WXMAC__) #define wxMBFILES 1 #else #define wxMBFILES 0 @@ -476,10 +521,25 @@ extern WXDLLIMPEXP_DATA_BASE(wxMBConv *) wxConvCurrent; #if wxUSE_UNICODE #define wxConvertWX2MB(s) wxConvCurrent->cWX2MB(s) #define wxConvertMB2WX(s) wxConvCurrent->cMB2WX(s) + + // these functions should be used when the conversions really, really have + // to succeed (usually because we pass their results to a standard C + // function which would crash if we passed NULL to it), so these functions + // always return a valid pointer if their argument is non-NULL + + // this function safety is achieved by trying wxConvLibc first, wxConvUTF8 + // next if it fails and, finally, wxConvISO8859_1 which always succeeds + extern WXDLLIMPEXP_BASE wxWCharBuffer wxSafeConvertMB2WX(const char *s); + + // this function uses wxConvLibc and wxConvUTF8(MAP_INVALID_UTF8_TO_OCTAL) + // if it fails + extern WXDLLIMPEXP_BASE wxCharBuffer wxSafeConvertWX2MB(const wchar_t *ws); #else // ANSI // no conversions to do #define wxConvertWX2MB(s) (s) #define wxConvertMB2WX(s) (s) + #define wxSafeConvertMB2WX(s) (s) + #define wxSafeConvertWX2MB(s) (s) #endif // Unicode/ANSI #endif // _WX_STRCONV_H_