FL/fl_utf8.h

   1 /* "$Id: fl_utf8.h 8585 2011-04-13 15:43:22Z ianmacarthur $"
   2  *
   3  * Author: Jean-Marc Lienher ( http://oksid.ch )
   4  * Copyright 2000-2010 by O'ksi'D.
   5  *
   6  * This library is free software; you can redistribute it and/or
   7  * modify it under the terms of the GNU Library General Public
   8  * License as published by the Free Software Foundation; either
   9  * version 2 of the License, or (at your option) any later version.
  10  *
  11  * This library is distributed in the hope that it will be useful,
  12  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  13  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  14  * Library General Public License for more details.
  15  *
  16  * You should have received a copy of the GNU Library General Public
  17  * License along with this library; if not, write to the Free Software
  18  * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307
  19  * USA.
  20  *
  21  * Please report all bugs and problems on the following page:
  22  *
  23  *     http://www.fltk.org/str.php
  24  */
  25
  26 /* Merged in some functionality from the fltk-2 version. IMM.
  27  * The following code is an attempt to merge the functions incorporated in FLTK2
  28  * with the functions provided in OksiD's fltk-1.1.6-utf8 port
  29  */
  30
  31 /**
  32   \file fl_utf8.h
  33   \brief header for Unicode and UTF8 chracter handling
  34 */
  35
  36 #ifndef _HAVE_FL_UTF8_HDR_
  37 #define _HAVE_FL_UTF8_HDR_
  38
  39 #include "Fl_Export.H"
  40 #include "fl_types.h"
  41
  42 #include <stdio.h>
  43 #include <string.h>
  44 #include <stdlib.h>
  45
  46 #ifdef WIN32
  47 #  include <sys/types.h>
  48 #  include <sys/stat.h>
  49 #  include <locale.h>
  50 #  include <ctype.h>
  51 #  define xchar wchar_t
  52 #  if !defined(FL_DLL) && !defined(__CYGWIN__)
  53 #    undef strdup
  54 #    define strdup _strdup
  55 #    undef putenv
  56 #    define putenv _putenv
  57 #    undef stricmp
  58 #    define stricmp _stricmp
  59 #    undef strnicmp
  60 #    define strnicmp _strnicmp
  61 #    undef hypot
  62 #    define hypot _hypot
  63 #    undef chdir
  64 #    define chdir _chdir
  65 #  endif
  66 #elif defined(__APPLE__)
  67 #  include <wchar.h>
  68 #  include <sys/stat.h>
  69 #  define xchar wchar_t
  70 #else /* X11 */
  71 #  include <sys/types.h>
  72 #  include <sys/stat.h>
  73 #  include "Xutf8.h"
  74 #  include <X11/Xlocale.h>
  75 #  include <X11/Xlib.h>
  76 #  include <locale.h>
  77 #  define xchar unsigned short
  78 #endif
  79
  80 #  ifdef __cplusplus
  81 extern "C" {
  82 #  endif
  83
  84 /** \addtogroup fl_unicode
  85     @{
  86 */
  87
  88 /* F2: comes from FLTK2 */
  89 /* OD: comes from OksiD */
  90
  91 /**
  92   Return the number of bytes needed to encode the given UCS4 character in UTF8.
  93   \param [in] ucs UCS4 encoded character
  94   \return number of bytes required
  95  */
  96 FL_EXPORT int fl_utf8bytes(unsigned ucs);
  97
  98 /* OD: returns the byte length of the first UTF-8 char sequence (returns -1 if not valid) */
  99 FL_EXPORT int fl_utf8len(char c);
 100
 101 /* OD: returns the byte length of the first UTF-8 char sequence (returns +1 if not valid) */
 102 FL_EXPORT int fl_utf8len1(char c);
 103
 104 /* OD: returns the number of Unicode chars in the UTF-8 string */
 105 FL_EXPORT int fl_utf_nb_char(const unsigned char *buf, int len);
 106
 107 /* F2: Convert the next UTF8 char-sequence into a Unicode value (and say how many bytes were used) */
 108 FL_EXPORT unsigned fl_utf8decode(const char* p, const char* end, int* len);
 109
 110 /* F2: Encode a Unicode value into a UTF8 sequence, return the number of bytes used */
 111 FL_EXPORT int fl_utf8encode(unsigned ucs, char* buf);
 112
 113 /* F2: Move forward to the next valid UTF8 sequence start betwen start and end */
 114 FL_EXPORT const char* fl_utf8fwd(const char* p, const char* start, const char* end);
 115
 116 /* F2: Move backward to the previous valid UTF8 sequence start */
 117 FL_EXPORT const char* fl_utf8back(const char* p, const char* start, const char* end);
 118
 119 /* XX: Convert a single 32-bit Unicode value into UTF16 */
 120 FL_EXPORT unsigned fl_ucs_to_Utf16(const unsigned ucs, unsigned short *dst, const unsigned dstlen);
 121
 122 /* F2: Convert a UTF8 string into UTF16 */
 123 FL_EXPORT unsigned fl_utf8toUtf16(const char* src, unsigned srclen, unsigned short* dst, unsigned dstlen);
 124
 125 /* F2: Convert a UTF8 string into a wide character string - makes UTF16 on win32, "UCS4" elsewhere */
 126 FL_EXPORT unsigned fl_utf8towc(const char *src, unsigned srclen, wchar_t *dst, unsigned dstlen);
 127
 128 /* F2: Convert a wide character string to UTF8 - takes in UTF16 on win32, "UCS4" elsewhere */
 129 FL_EXPORT unsigned fl_utf8fromwc(char *dst, unsigned dstlen, const wchar_t *src, unsigned srclen);
 130
 131 /* F2: Convert a UTF8 string into ASCII, eliding untranslatable glyphs */
 132 FL_EXPORT unsigned fl_utf8toa (const char *src, unsigned srclen, char *dst, unsigned dstlen);
 133 /* OD: convert UTF-8 string to latin1 */
 134 /* FL_EXPORT int fl_utf2latin1(const unsigned char *src, int srclen, char *dst); */
 135
 136 /* F2: Convert 8859-1 string to UTF8 */
 137 FL_EXPORT unsigned fl_utf8froma (char *dst, unsigned dstlen, const char *src, unsigned srclen);
 138 /* OD: convert latin1 str to UTF-8 */
 139 /* FL_EXPORT int fl_latin12utf(const unsigned char *src, int srclen, char *dst); */
 140
 141 /* F2: Returns true if the current O/S locale is UTF8 */
 142 FL_EXPORT int fl_utf8locale();
 143
 144 /* F2: Examine the first len characters of src, to determine if the input text is UTF8 or not
 145  * NOTE: The value returned is not simply boolean - it contains information about the probable
 146  * type of the src text. */
 147 FL_EXPORT int fl_utf8test(const char *src, unsigned len);
 148
 149 /* XX: return width of "raw" ucs character in columns.
 150  * for internal use only */
 151 FL_EXPORT int fl_wcwidth_(unsigned int ucs);
 152
 153 /* XX: return width of utf-8 character string in columns.
 154  * NOTE: this may also do C1 control character (0x80 to 0x9f) to CP1252 mapping,
 155  * depending on original build options */
 156 FL_EXPORT int fl_wcwidth(const char *src);
 157
 158 /* OD: Return true if the character is non-spacing */
 159 FL_EXPORT unsigned int fl_nonspacing(unsigned int ucs);
 160
 161 /* F2: Convert UTF8 to a local multi-byte encoding - mainly for win32? */
 162 FL_EXPORT unsigned fl_utf8to_mb(const char *src, unsigned srclen, char *dst, unsigned dstlen);
 163 /* OD: Convert UTF8 to a local multi-byte encoding */
 164 FL_EXPORT char* fl_utf2mbcs(const char *src);
 165
 166 /* F2: Convert a local multi-byte encoding to UTF8 - mainly for win32? */
 167 FL_EXPORT unsigned fl_utf8from_mb(char *dst, unsigned dstlen, const char *src, unsigned srclen);
 168 /* OD: Convert a local multi-byte encoding to UTF8 */
 169 /* FL_EXPORT char* fl_mbcs2utf(const char *src); */
 170
 171 /*****************************************************************************/
 172 #ifdef WIN32
 173 /* OD: Attempt to convert the UTF8 string to the current locale */
 174 FL_EXPORT char *fl_utf8_to_locale(const char *s, int len, unsigned int codepage);
 175
 176 /* OD: Attempt to convert a string in the current locale to UTF8 */
 177 FL_EXPORT char *fl_locale_to_utf8(const char *s, int len, unsigned int codepage);
 178 #endif
 179
 180 /*****************************************************************************
 181  * The following functions are intended to provide portable, UTF8 aware
 182  * versions of standard functions
 183  */
 184
 185 /* OD: UTF8 aware strncasecmp - converts to lower case Unicode and tests */
 186 FL_EXPORT int fl_utf_strncasecmp(const char *s1, const char *s2, int n);
 187
 188 /* OD: UTF8 aware strcasecmp - converts to Unicode and tests */
 189 FL_EXPORT int fl_utf_strcasecmp(const char *s1, const char *s2);
 190
 191 /* OD: return the Unicode lower case value of ucs */
 192 FL_EXPORT int fl_tolower(unsigned int ucs);
 193
 194 /* OD: return the Unicode upper case value of ucs */
 195 FL_EXPORT int fl_toupper(unsigned int ucs);
 196
 197 /* OD: converts the UTF8 string to the lower case equivalent */
 198 FL_EXPORT int fl_utf_tolower(const unsigned char *str, int len, char *buf);
 199
 200 /* OD: converts the UTF8 string to the upper case equivalent */
 201 FL_EXPORT int fl_utf_toupper(const unsigned char *str, int len, char *buf);
 202
 203 /* OD: Portable UTF8 aware chmod wrapper */
 204 FL_EXPORT int fl_chmod(const char* f, int mode);
 205
 206 /* OD: Portable UTF8 aware access wrapper */
 207 FL_EXPORT int fl_access(const char* f, int mode);
 208
 209 /* OD: Portable UTF8 aware stat wrapper */
 210 FL_EXPORT int fl_stat( const char *path, struct stat *buffer );
 211
 212 /* OD: Portable UTF8 aware getcwd wrapper */
 213 FL_EXPORT char* fl_getcwd( char *buf, int maxlen);
 214
 215 /* OD: Portable UTF8 aware fopen wrapper */
 216 FL_EXPORT FILE *fl_fopen(const char *f, const char *mode);
 217
 218 /* OD: Portable UTF8 aware system wrapper */
 219 FL_EXPORT int fl_system(const char* f);
 220
 221 /* OD: Portable UTF8 aware execvp wrapper */
 222 FL_EXPORT int fl_execvp(const char *file, char *const *argv);
 223
 224 /* OD: Portable UTF8 aware open wrapper */
 225 FL_EXPORT int fl_open(const char* f, int o, ...);
 226
 227 /* OD: Portable UTF8 aware unlink wrapper */
 228 FL_EXPORT int fl_unlink(const char *f);
 229
 230 /* OD: Portable UTF8 aware rmdir wrapper */
 231 FL_EXPORT int fl_rmdir(const char *f);
 232
 233 /* OD: Portable UTF8 aware getenv wrapper */
 234 FL_EXPORT char* fl_getenv(const char *name);
 235
 236 /* OD: Portable UTF8 aware execvp wrapper */
 237 FL_EXPORT int fl_mkdir(const char* f, int mode);
 238
 239 /* OD: Portable UTF8 aware rename wrapper */
 240 FL_EXPORT int fl_rename(const char* f, const char *t);
 241
 242
 243 /* OD: Given a full pathname, this will create the directory path needed to hold the file named */
 244 FL_EXPORT void fl_make_path_for_file( const char *path );
 245
 246 /* OD: recursively create a path in the file system */
 247 FL_EXPORT char fl_make_path( const char *path );
 248
 249
 250 /** @} */
 251
 252 /*****************************************************************************/
 253
 254 #ifdef __cplusplus
 255 }
 256 #endif /* __cplusplus */
 257
 258
 259 #endif /* _HAVE_FL_UTF8_HDR_ */
 260
 261 /*
 262  * End of "$Id: fl_utf8.h 8585 2011-04-13 15:43:22Z ianmacarthur $".
 263  */