The Tor Browser: intl/unicharutil/src/ucdata.h@fc2d59ddac77

     1 /*

     2  * Copyright 1996, 1997, 1998 Computing Research Labs,

     3  * New Mexico State University

     4  *

     5  * Permission is hereby granted, free of charge, to any person obtaining a

     6  * copy of this software and associated documentation files (the "Software"),

     7  * to deal in the Software without restriction, including without limitation

     8  * the rights to use, copy, modify, merge, publish, distribute, sublicense,

     9  * and/or sell copies of the Software, and to permit persons to whom the

    10  * Software is furnished to do so, subject to the following conditions:

    11  *

    12  * The above copyright notice and this permission notice shall be included in

    13  * all copies or substantial portions of the Software.

    14  *

    15  * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR

    16  * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,

    17  * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL

    18  * THE COMPUTING RESEARCH LAB OR NEW MEXICO STATE UNIVERSITY BE LIABLE FOR ANY

    19  * CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT

    20  * OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR

    21  * THE USE OR OTHER DEALINGS IN THE SOFTWARE.

    22  */

    23 #ifndef _h_ucdata

    24 #define _h_ucdata

    26 /*

    27  * $Id: ucdata.h,v 1.1 1999/01/08 00:19:12 ftang%netscape.com Exp $

    28  */

    30 #ifdef __cplusplus

    31 extern "C" {

    32 #endif

    34 #undef __

    35 #ifdef __STDC__

    36 #define __(x) x

    37 #else

    38 #define __(x) ()

    39 #endif

    41 #define UCDATA_VERSION "1.9"

    43 /**************************************************************************

    44  *

    45  * Masks and macros for character properties.

    46  *

    47  **************************************************************************/

    49 /*

    50  * Values that can appear in the `mask1' parameter of the ucisprop()

    51  * function.

    52  */

    53 #define UC_MN 0x00000001 /* Mark, Non-Spacing          */

    54 #define UC_MC 0x00000002 /* Mark, Spacing Combining    */

    55 #define UC_ME 0x00000004 /* Mark, Enclosing            */

    56 #define UC_ND 0x00000008 /* Number, Decimal Digit      */

    57 #define UC_NL 0x00000010 /* Number, Letter             */

    58 #define UC_NO 0x00000020 /* Number, Other              */

    59 #define UC_ZS 0x00000040 /* Separator, Space           */

    60 #define UC_ZL 0x00000080 /* Separator, Line            */

    61 #define UC_ZP 0x00000100 /* Separator, Paragraph       */

    62 #define UC_CC 0x00000200 /* Other, Control             */

    63 #define UC_CF 0x00000400 /* Other, Format              */

    64 #define UC_OS 0x00000800 /* Other, Surrogate           */

    65 #define UC_CO 0x00001000 /* Other, Private Use         */

    66 #define UC_CN 0x00002000 /* Other, Not Assigned        */

    67 #define UC_LU 0x00004000 /* Letter, Uppercase          */

    68 #define UC_LL 0x00008000 /* Letter, Lowercase          */

    69 #define UC_LT 0x00010000 /* Letter, Titlecase          */

    70 #define UC_LM 0x00020000 /* Letter, Modifier           */

    71 #define UC_LO 0x00040000 /* Letter, Other              */

    72 #define UC_PC 0x00080000 /* Punctuation, Connector     */

    73 #define UC_PD 0x00100000 /* Punctuation, Dash          */

    74 #define UC_PS 0x00200000 /* Punctuation, Open          */

    75 #define UC_PE 0x00400000 /* Punctuation, Close         */

    76 #define UC_PO 0x00800000 /* Punctuation, Other         */

    77 #define UC_SM 0x01000000 /* Symbol, Math               */

    78 #define UC_SC 0x02000000 /* Symbol, Currency           */

    79 #define UC_SK 0x04000000 /* Symbol, Modifier           */

    80 #define UC_SO 0x08000000 /* Symbol, Other              */

    81 #define UC_L  0x10000000 /* Left-To-Right              */

    82 #define UC_R  0x20000000 /* Right-To-Left              */

    83 #define UC_EN 0x40000000 /* European Number            */

    84 #define UC_ES 0x80000000 /* European Number Separator  */

    86 /*

    87  * Values that can appear in the `mask2' parameter of the ucisprop()

    88  * function.

    89  */

    90 #define UC_ET 0x00000001 /* European Number Terminator */

    91 #define UC_AN 0x00000002 /* Arabic Number              */

    92 #define UC_CS 0x00000004 /* Common Number Separator    */

    93 #define UC_B  0x00000008 /* Block Separator            */

    94 #define UC_S  0x00000010 /* Segment Separator          */

    95 #define UC_WS 0x00000020 /* Whitespace                 */

    96 #define UC_ON 0x00000040 /* Other Neutrals             */

    97 /*

    98  * Implementation specific character properties.

    99  */

   100 #define UC_CM 0x00000080 /* Composite                  */

   101 #define UC_NB 0x00000100 /* Non-Breaking               */

   102 #define UC_SY 0x00000200 /* Symmetric                  */

   103 #define UC_HD 0x00000400 /* Hex Digit                  */

   104 #define UC_QM 0x00000800 /* Quote Mark                 */

   105 #define UC_MR 0x00001000 /* Mirroring                  */

   106 #define UC_SS 0x00002000 /* Space, other               */

   108 #define UC_CP 0x00004000 /* Defined                    */

   110 /*

   111  * Added for UnicodeData-2.1.3.

   112  */

   113 #define UC_PI 0x00008000 /* Punctuation, Initial       */

   114 #define UC_PF 0x00010000 /* Punctuation, Final         */

   116 /*

   117  * This is the primary function for testing to see if a character has some set

   118  * of properties.  The macros that test for various character properties all

   119  * call this function with some set of masks.

   120  */

   121 extern int ucisprop __((unsigned long code, unsigned long mask1,

   122                         unsigned long mask2));

   124 #define ucisalpha(cc) ucisprop(cc, UC_LU|UC_LL|UC_LM|UC_LO|UC_LT, 0)

   125 #define ucisdigit(cc) ucisprop(cc, UC_ND, 0)

   126 #define ucisalnum(cc) ucisprop(cc, UC_LU|UC_LL|UC_LM|UC_LO|UC_LT|UC_ND, 0)

   127 #define uciscntrl(cc) ucisprop(cc, UC_CC|UC_CF, 0)

   128 #define ucisspace(cc) ucisprop(cc, UC_ZS|UC_SS, 0)

   129 #define ucisblank(cc) ucisprop(cc, UC_ZS, 0)

   130 #define ucispunct(cc) ucisprop(cc, UC_PD|UC_PS|UC_PE|UC_PO, UC_PI|UC_PF)

   131 #define ucisgraph(cc) ucisprop(cc, UC_MN|UC_MC|UC_ME|UC_ND|UC_NL|UC_NO|\

   132                                UC_LU|UC_LL|UC_LT|UC_LM|UC_LO|UC_PC|UC_PD|\

   133                                UC_PS|UC_PE|UC_PO|UC_SM|UC_SM|UC_SC|UC_SK|\

   134                                UC_SO, UC_PI|UC_PF)

   135 #define ucisprint(cc) ucisprop(cc, UC_MN|UC_MC|UC_ME|UC_ND|UC_NL|UC_NO|\

   136                                UC_LU|UC_LL|UC_LT|UC_LM|UC_LO|UC_PC|UC_PD|\

   137                                UC_PS|UC_PE|UC_PO|UC_SM|UC_SM|UC_SC|UC_SK|\

   138                                UC_SO|UC_ZS, UC_PI|UC_PF)

   139 #define ucisupper(cc) ucisprop(cc, UC_LU, 0)

   140 #define ucislower(cc) ucisprop(cc, UC_LL, 0)

   141 #define ucistitle(cc) ucisprop(cc, UC_LT, 0)

   142 #define ucisxdigit(cc) ucisprop(cc, 0, UC_HD)

   144 #define ucisisocntrl(cc) ucisprop(cc, UC_CC, 0)

   145 #define ucisfmtcntrl(cc) ucisprop(cc, UC_CF, 0)

   147 #define ucissymbol(cc) ucisprop(cc, UC_SM|UC_SC|UC_SO|UC_SK, 0)

   148 #define ucisnumber(cc) ucisprop(cc, UC_ND|UC_NO|UC_NL, 0)

   149 #define ucisnonspacing(cc) ucisprop(cc, UC_MN, 0)

   150 #define ucisopenpunct(cc) ucisprop(cc, UC_PS, 0)

   151 #define ucisclosepunct(cc) ucisprop(cc, UC_PE, 0)

   152 #define ucisinitialpunct(cc) ucisprop(cc, 0, UC_PI)

   153 #define ucisfinalpunct(cc) ucisprop(cc, 0, UC_PF)

   155 #define uciscomposite(cc) ucisprop(cc, 0, UC_CM)

   156 #define ucishex(cc) ucisprop(cc, 0, UC_HD)

   157 #define ucisquote(cc) ucisprop(cc, 0, UC_QM)

   158 #define ucissymmetric(cc) ucisprop(cc, 0, UC_SY)

   159 #define ucismirroring(cc) ucisprop(cc, 0, UC_MR)

   160 #define ucisnonbreaking(cc) ucisprop(cc, 0, UC_NB)

   162 /*

   163  * Directionality macros.

   164  */

   165 #define ucisrtl(cc) ucisprop(cc, UC_R, 0)

   166 #define ucisltr(cc) ucisprop(cc, UC_L, 0)

   167 #define ucisstrong(cc) ucisprop(cc, UC_L|UC_R, 0)

   168 #define ucisweak(cc) ucisprop(cc, UC_EN|UC_ES, UC_ET|UC_AN|UC_CS)

   169 #define ucisneutral(cc) ucisprop(cc, 0, UC_B|UC_S|UC_WS|UC_ON)

   170 #define ucisseparator(cc) ucisprop(cc, 0, UC_B|UC_S)

   172 /*

   173  * Other macros inspired by John Cowan.

   174  */

   175 #define ucismark(cc) ucisprop(cc, UC_MN|UC_MC|UC_ME, 0)

   176 #define ucismodif(cc) ucisprop(cc, UC_LM, 0)

   177 #define ucisletnum(cc) ucisprop(cc, UC_NL, 0)

   178 #define ucisconnect(cc) ucisprop(cc, UC_PC, 0)

   179 #define ucisdash(cc) ucisprop(cc, UC_PD, 0)

   180 #define ucismath(cc) ucisprop(cc, UC_SM, 0)

   181 #define uciscurrency(cc) ucisprop(cc, UC_SC, 0)

   182 #define ucismodifsymbol(cc) ucisprop(cc, UC_SK, 0)

   183 #define ucisnsmark(cc) ucisprop(cc, UC_MN, 0)

   184 #define ucisspmark(cc) ucisprop(cc, UC_MC, 0)

   185 #define ucisenclosing(cc) ucisprop(cc, UC_ME, 0)

   186 #define ucisprivate(cc) ucisprop(cc, UC_CO, 0)

   187 #define ucissurrogate(cc) ucisprop(cc, UC_OS, 0)

   188 #define ucislsep(cc) ucisprop(cc, UC_ZL, 0)

   189 #define ucispsep(cc) ucisprop(cc, UC_ZP, 0)

   191 #define ucisidentstart(cc) ucisprop(cc, UC_LU|UC_LL|UC_LT|UC_LO|UC_NL, 0)

   192 #define ucisidentpart(cc) ucisprop(cc, UC_LU|UC_LL|UC_LT|UC_LO|UC_NL|\

   193                                    UC_MN|UC_MC|UC_ND|UC_PC|UC_CF, 0)

   195 #define ucisdefined(cc) ucisprop(cc, 0, UC_CP)

   196 #define ucisundefined(cc) !ucisprop(cc, 0, UC_CP)

   198 /*

   199  * Other miscellaneous character property macros.

   200  */

   201 #define ucishan(cc) (((cc) >= 0x4e00 && (cc) <= 0x9fff) ||\

   202                      ((cc) >= 0xf900 && (cc) <= 0xfaff))

   203 #define ucishangul(cc) ((cc) >= 0xac00 && (cc) <= 0xd7ff)

   205 /**************************************************************************

   206  *

   207  * Functions for case conversion.

   208  *

   209  **************************************************************************/

   211 extern unsigned long uctoupper __((unsigned long code));

   212 extern unsigned long uctolower __((unsigned long code));

   213 extern unsigned long uctotitle __((unsigned long code));

   215 /**************************************************************************

   216  *

   217  * Functions for getting decompositions.

   218  *

   219  **************************************************************************/

   221 /*

   222  * This routine determines if the code has a decomposition.  If it returns 0,

   223  * there is no decomposition.  Any other value indicates a decomposition was

   224  * returned.

   225  */

   226 extern int ucdecomp __((unsigned long code, unsigned long *num,

   228                         unsigned long **decomp));

   230 /*

   231  * If the code is a Hangul syllable, this routine decomposes it into the array

   232  * passed.  The array size should be at least 3.

   233  */

   234 extern int ucdecomp_hangul __((unsigned long code, unsigned long *num,

   235                                unsigned long decomp[]));

   237 /**************************************************************************

   238  *

   239  * Functions for getting combining classes.

   240  *

   241  **************************************************************************/

   243 /*

   244  * This will return the combining class for a character to be used with the

   245  * Canonical Ordering algorithm.

   246  */

   247 extern unsigned long uccombining_class __((unsigned long code));

   249 /**************************************************************************

   250  *

   251  * Functions for getting numbers and digits.

   252  *

   253  **************************************************************************/

   255 struct ucnumber {

   256     int numerator;

   257     int denominator;

   258 };

   260 extern int ucnumber_lookup __((unsigned long code, struct ucnumber *num));

   261 extern int ucdigit_lookup __((unsigned long code, int *digit));

   263 /*

   264  * For compatibility with John Cowan's "uctype" package.

   265  */

   266 extern struct ucnumber ucgetnumber __((unsigned long code));

   267 extern int ucgetdigit __((unsigned long code));

   269 /**************************************************************************

   270  *

   271  * Functions library initialization and cleanup.

   272  *

   273  **************************************************************************/

   275 /*

   276  * Macros for specifying the data tables to be loaded for ucdata_load().

   277  */

   278 #define UCDATA_CASE   0x01

   279 #define UCDATA_CTYPE  0x02

   280 #define UCDATA_DECOMP 0x04

   281 #define UCDATA_CMBCL  0x08

   282 #define UCDATA_NUM    0x10

   284 #define UCDATA_ALL (UCDATA_CASE|UCDATA_CTYPE|UCDATA_DECOMP|\

   285                     UCDATA_CMBCL|UCDATA_NUM)

   287 /*

   288  * Functions to load, unload, and reload specific data files.

   289  */

   290 extern void ucdata_load __((char *paths, int mask));

   291 extern void ucdata_unload __((int mask));

   292 extern void ucdata_reload __((char *paths, int mask));

   294 /*

   295  * Deprecated functions, now just compatibility macros.

   296  */

   297 #define ucdata_setup(p) ucdata_load(p, UCDATA_ALL)

   298 #define ucdata_cleanup() ucdata_unload(UCDATA_ALL)

   300 #undef __

   302 #ifdef __cplusplus

   303 }

   304 #endif

   306 #endif /* _h_ucdata */

The Tor Browser / file revision

intl/unicharutil/src/ucdata.h@fc2d59ddac77

intl/unicharutil/src/ucdata.h