FLTK 1.3.0
fl_utf8.h
Go to the documentation of this file.
00001 /* "$Id: fl_utf8.h 8585 2011-04-13 15:43:22Z ianmacarthur $"
00002  *
00003  * Author: Jean-Marc Lienher ( http://oksid.ch )
00004  * Copyright 2000-2010 by O'ksi'D.
00005  *
00006  * This library is free software; you can redistribute it and/or
00007  * modify it under the terms of the GNU Library General Public
00008  * License as published by the Free Software Foundation; either
00009  * version 2 of the License, or (at your option) any later version.
00010  *
00011  * This library is distributed in the hope that it will be useful,
00012  * but WITHOUT ANY WARRANTY; without even the implied warranty of
00013  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
00014  * Library General Public License for more details.
00015  *
00016  * You should have received a copy of the GNU Library General Public
00017  * License along with this library; if not, write to the Free Software
00018  * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307
00019  * USA.
00020  *
00021  * Please report all bugs and problems on the following page:
00022  *
00023  *     http://www.fltk.org/str.php
00024  */
00025 
00026 /* Merged in some functionality from the fltk-2 version. IMM.
00027  * The following code is an attempt to merge the functions incorporated in FLTK2
00028  * with the functions provided in OksiD's fltk-1.1.6-utf8 port
00029  */
00030 
00036 #ifndef _HAVE_FL_UTF8_HDR_
00037 #define _HAVE_FL_UTF8_HDR_
00038 
00039 #include "Fl_Export.H"
00040 #include "fl_types.h"
00041 
00042 #include <stdio.h>
00043 #include <string.h>
00044 #include <stdlib.h>
00045 
00046 #ifdef WIN32
00047 #  include <sys/types.h>
00048 #  include <sys/stat.h>
00049 #  include <locale.h>
00050 #  include <ctype.h>
00051 #  define xchar wchar_t
00052 #  if !defined(FL_DLL) && !defined(__CYGWIN__)
00053 #    undef strdup
00054 #    define strdup _strdup
00055 #    undef putenv
00056 #    define putenv _putenv
00057 #    undef stricmp
00058 #    define stricmp _stricmp
00059 #    undef strnicmp
00060 #    define strnicmp _strnicmp
00061 #    undef hypot
00062 #    define hypot _hypot
00063 #    undef chdir
00064 #    define chdir _chdir
00065 #  endif
00066 #elif defined(__APPLE__)
00067 #  include <wchar.h>
00068 #  include <sys/stat.h>
00069 #  define xchar wchar_t
00070 #else /* X11 */
00071 #  include <sys/types.h>
00072 #  include <sys/stat.h>
00073 #  include "Xutf8.h"
00074 #  include <X11/Xlocale.h>
00075 #  include <X11/Xlib.h>
00076 #  include <locale.h>
00077 #  define xchar unsigned short
00078 #endif
00079 
00080 #  ifdef __cplusplus
00081 extern "C" {
00082 #  endif
00083 
00088 /* F2: comes from FLTK2 */
00089 /* OD: comes from OksiD */
00090 
00096 FL_EXPORT int fl_utf8bytes(unsigned ucs);
00097 
00098 /* OD: returns the byte length of the first UTF-8 char sequence (returns -1 if not valid) */
00099 FL_EXPORT int fl_utf8len(char c);
00100 
00101 /* OD: returns the byte length of the first UTF-8 char sequence (returns +1 if not valid) */
00102 FL_EXPORT int fl_utf8len1(char c);
00103 
00104 /* OD: returns the number of Unicode chars in the UTF-8 string */
00105 FL_EXPORT int fl_utf_nb_char(const unsigned char *buf, int len);
00106 
00107 /* F2: Convert the next UTF8 char-sequence into a Unicode value (and say how many bytes were used) */
00108 FL_EXPORT unsigned fl_utf8decode(const char* p, const char* end, int* len);
00109 
00110 /* F2: Encode a Unicode value into a UTF8 sequence, return the number of bytes used */
00111 FL_EXPORT int fl_utf8encode(unsigned ucs, char* buf);
00112 
00113 /* F2: Move forward to the next valid UTF8 sequence start betwen start and end */
00114 FL_EXPORT const char* fl_utf8fwd(const char* p, const char* start, const char* end);
00115 
00116 /* F2: Move backward to the previous valid UTF8 sequence start */
00117 FL_EXPORT const char* fl_utf8back(const char* p, const char* start, const char* end);
00118 
00119 /* XX: Convert a single 32-bit Unicode value into UTF16 */
00120 FL_EXPORT unsigned fl_ucs_to_Utf16(const unsigned ucs, unsigned short *dst, const unsigned dstlen);
00121 
00122 /* F2: Convert a UTF8 string into UTF16 */
00123 FL_EXPORT unsigned fl_utf8toUtf16(const char* src, unsigned srclen, unsigned short* dst, unsigned dstlen);
00124 
00125 /* F2: Convert a UTF8 string into a wide character string - makes UTF16 on win32, "UCS4" elsewhere */
00126 FL_EXPORT unsigned fl_utf8towc(const char *src, unsigned srclen, wchar_t *dst, unsigned dstlen);
00127 
00128 /* F2: Convert a wide character string to UTF8 - takes in UTF16 on win32, "UCS4" elsewhere */
00129 FL_EXPORT unsigned fl_utf8fromwc(char *dst, unsigned dstlen, const wchar_t *src, unsigned srclen);
00130 
00131 /* F2: Convert a UTF8 string into ASCII, eliding untranslatable glyphs */
00132 FL_EXPORT unsigned fl_utf8toa (const char *src, unsigned srclen, char *dst, unsigned dstlen);
00133 /* OD: convert UTF-8 string to latin1 */
00134 /* FL_EXPORT int fl_utf2latin1(const unsigned char *src, int srclen, char *dst); */
00135 
00136 /* F2: Convert 8859-1 string to UTF8 */
00137 FL_EXPORT unsigned fl_utf8froma (char *dst, unsigned dstlen, const char *src, unsigned srclen);
00138 /* OD: convert latin1 str to UTF-8 */
00139 /* FL_EXPORT int fl_latin12utf(const unsigned char *src, int srclen, char *dst); */
00140 
00141 /* F2: Returns true if the current O/S locale is UTF8 */
00142 FL_EXPORT int fl_utf8locale();
00143 
00144 /* F2: Examine the first len characters of src, to determine if the input text is UTF8 or not
00145  * NOTE: The value returned is not simply boolean - it contains information about the probable
00146  * type of the src text. */
00147 FL_EXPORT int fl_utf8test(const char *src, unsigned len);
00148 
00149 /* XX: return width of "raw" ucs character in columns.
00150  * for internal use only */
00151 FL_EXPORT int fl_wcwidth_(unsigned int ucs);
00152 
00153 /* XX: return width of utf-8 character string in columns.
00154  * NOTE: this may also do C1 control character (0x80 to 0x9f) to CP1252 mapping,
00155  * depending on original build options */
00156 FL_EXPORT int fl_wcwidth(const char *src);
00157 
00158 /* OD: Return true if the character is non-spacing */
00159 FL_EXPORT unsigned int fl_nonspacing(unsigned int ucs);
00160 
00161 /* F2: Convert UTF8 to a local multi-byte encoding - mainly for win32? */
00162 FL_EXPORT unsigned fl_utf8to_mb(const char *src, unsigned srclen, char *dst, unsigned dstlen);
00163 /* OD: Convert UTF8 to a local multi-byte encoding */
00164 FL_EXPORT char* fl_utf2mbcs(const char *src);
00165 
00166 /* F2: Convert a local multi-byte encoding to UTF8 - mainly for win32? */
00167 FL_EXPORT unsigned fl_utf8from_mb(char *dst, unsigned dstlen, const char *src, unsigned srclen);
00168 /* OD: Convert a local multi-byte encoding to UTF8 */
00169 /* FL_EXPORT char* fl_mbcs2utf(const char *src); */
00170 
00171 /*****************************************************************************/
00172 #ifdef WIN32
00173 /* OD: Attempt to convert the UTF8 string to the current locale */
00174 FL_EXPORT char *fl_utf8_to_locale(const char *s, int len, unsigned int codepage);
00175 
00176 /* OD: Attempt to convert a string in the current locale to UTF8 */
00177 FL_EXPORT char *fl_locale_to_utf8(const char *s, int len, unsigned int codepage);
00178 #endif
00179 
00180 /*****************************************************************************
00181  * The following functions are intended to provide portable, UTF8 aware
00182  * versions of standard functions
00183  */
00184 
00185 /* OD: UTF8 aware strncasecmp - converts to lower case Unicode and tests */
00186 FL_EXPORT int fl_utf_strncasecmp(const char *s1, const char *s2, int n);
00187 
00188 /* OD: UTF8 aware strcasecmp - converts to Unicode and tests */
00189 FL_EXPORT int fl_utf_strcasecmp(const char *s1, const char *s2);
00190 
00191 /* OD: return the Unicode lower case value of ucs */
00192 FL_EXPORT int fl_tolower(unsigned int ucs);
00193 
00194 /* OD: return the Unicode upper case value of ucs */
00195 FL_EXPORT int fl_toupper(unsigned int ucs);
00196 
00197 /* OD: converts the UTF8 string to the lower case equivalent */
00198 FL_EXPORT int fl_utf_tolower(const unsigned char *str, int len, char *buf);
00199 
00200 /* OD: converts the UTF8 string to the upper case equivalent */
00201 FL_EXPORT int fl_utf_toupper(const unsigned char *str, int len, char *buf);
00202 
00203 /* OD: Portable UTF8 aware chmod wrapper */
00204 FL_EXPORT int fl_chmod(const char* f, int mode);
00205 
00206 /* OD: Portable UTF8 aware access wrapper */
00207 FL_EXPORT int fl_access(const char* f, int mode);
00208 
00209 /* OD: Portable UTF8 aware stat wrapper */
00210 FL_EXPORT int fl_stat( const char *path, struct stat *buffer );
00211 
00212 /* OD: Portable UTF8 aware getcwd wrapper */
00213 FL_EXPORT char* fl_getcwd( char *buf, int maxlen);
00214 
00215 /* OD: Portable UTF8 aware fopen wrapper */
00216 FL_EXPORT FILE *fl_fopen(const char *f, const char *mode);
00217 
00218 /* OD: Portable UTF8 aware system wrapper */
00219 FL_EXPORT int fl_system(const char* f);
00220 
00221 /* OD: Portable UTF8 aware execvp wrapper */
00222 FL_EXPORT int fl_execvp(const char *file, char *const *argv);
00223 
00224 /* OD: Portable UTF8 aware open wrapper */
00225 FL_EXPORT int fl_open(const char* f, int o, ...);
00226 
00227 /* OD: Portable UTF8 aware unlink wrapper */
00228 FL_EXPORT int fl_unlink(const char *f);
00229 
00230 /* OD: Portable UTF8 aware rmdir wrapper */
00231 FL_EXPORT int fl_rmdir(const char *f);
00232 
00233 /* OD: Portable UTF8 aware getenv wrapper */
00234 FL_EXPORT char* fl_getenv(const char *name);
00235 
00236 /* OD: Portable UTF8 aware execvp wrapper */
00237 FL_EXPORT int fl_mkdir(const char* f, int mode);
00238 
00239 /* OD: Portable UTF8 aware rename wrapper */
00240 FL_EXPORT int fl_rename(const char* f, const char *t);
00241 
00242 
00243 /* OD: Given a full pathname, this will create the directory path needed to hold the file named */
00244 FL_EXPORT void fl_make_path_for_file( const char *path );
00245 
00246 /* OD: recursively create a path in the file system */
00247 FL_EXPORT char fl_make_path( const char *path );
00248 
00249 
00252 /*****************************************************************************/
00253 
00254 #ifdef __cplusplus
00255 }
00256 #endif /* __cplusplus */
00257 
00258 
00259 #endif /* _HAVE_FL_UTF8_HDR_ */
00260 
00261 /*
00262  * End of "$Id: fl_utf8.h 8585 2011-04-13 15:43:22Z ianmacarthur $".
00263  */