TB9.2 Example Applications: examples/PIPS/antiword/src/utf8.c Source File

00001 /*
+00002  * utf8.c
+00003  * Copyright (C) 2001-2004 A.J. van Os; Released under GPL
+00004  *
+00005  *====================================================================
+00006  * This part of the software is based on:
+00007  * An implementation of wcwidth() as defined in
+00008  * "The Single UNIX Specification, Version 2, The Open Group, 1997"
+00009  * <http://www.UNIX-systems.org/online.html>
+00010  * Markus Kuhn -- 2001-01-12 -- public domain
+00011  *====================================================================
+00012  * The credit should go to him, but all the bugs are mine.
+00013  */
+00014 
+00015 #include <stdlib.h>
+00016 #include <string.h>
+00017 #include "antiword.h"
+00018 
+00019 struct interval {
+00020         USHORT  first;
+00021         USHORT  last;
+00022 };
+00023 /* Sorted list of non-overlapping intervals of non-spacing characters */
+00024 static const struct interval combining[] = {
+00025         { 0x0300, 0x034E }, { 0x0360, 0x0362 }, { 0x0483, 0x0486 },
+00026         { 0x0488, 0x0489 }, { 0x0591, 0x05A1 }, { 0x05A3, 0x05B9 },
+00027         { 0x05BB, 0x05BD }, { 0x05BF, 0x05BF }, { 0x05C1, 0x05C2 },
+00028         { 0x05C4, 0x05C4 }, { 0x064B, 0x0655 }, { 0x0670, 0x0670 },
+00029         { 0x06D6, 0x06E4 }, { 0x06E7, 0x06E8 }, { 0x06EA, 0x06ED },
+00030         { 0x070F, 0x070F }, { 0x0711, 0x0711 }, { 0x0730, 0x074A },
+00031         { 0x07A6, 0x07B0 }, { 0x0901, 0x0902 }, { 0x093C, 0x093C },
+00032         { 0x0941, 0x0948 }, { 0x094D, 0x094D }, { 0x0951, 0x0954 },
+00033         { 0x0962, 0x0963 }, { 0x0981, 0x0981 }, { 0x09BC, 0x09BC },
+00034         { 0x09C1, 0x09C4 }, { 0x09CD, 0x09CD }, { 0x09E2, 0x09E3 },
+00035         { 0x0A02, 0x0A02 }, { 0x0A3C, 0x0A3C }, { 0x0A41, 0x0A42 },
+00036         { 0x0A47, 0x0A48 }, { 0x0A4B, 0x0A4D }, { 0x0A70, 0x0A71 },
+00037         { 0x0A81, 0x0A82 }, { 0x0ABC, 0x0ABC }, { 0x0AC1, 0x0AC5 },
+00038         { 0x0AC7, 0x0AC8 }, { 0x0ACD, 0x0ACD }, { 0x0B01, 0x0B01 },
+00039         { 0x0B3C, 0x0B3C }, { 0x0B3F, 0x0B3F }, { 0x0B41, 0x0B43 },
+00040         { 0x0B4D, 0x0B4D }, { 0x0B56, 0x0B56 }, { 0x0B82, 0x0B82 },
+00041         { 0x0BC0, 0x0BC0 }, { 0x0BCD, 0x0BCD }, { 0x0C3E, 0x0C40 },
+00042         { 0x0C46, 0x0C48 }, { 0x0C4A, 0x0C4D }, { 0x0C55, 0x0C56 },
+00043         { 0x0CBF, 0x0CBF }, { 0x0CC6, 0x0CC6 }, { 0x0CCC, 0x0CCD },
+00044         { 0x0D41, 0x0D43 }, { 0x0D4D, 0x0D4D }, { 0x0DCA, 0x0DCA },
+00045         { 0x0DD2, 0x0DD4 }, { 0x0DD6, 0x0DD6 }, { 0x0E31, 0x0E31 },
+00046         { 0x0E34, 0x0E3A }, { 0x0E47, 0x0E4E }, { 0x0EB1, 0x0EB1 },
+00047         { 0x0EB4, 0x0EB9 }, { 0x0EBB, 0x0EBC }, { 0x0EC8, 0x0ECD },
+00048         { 0x0F18, 0x0F19 }, { 0x0F35, 0x0F35 }, { 0x0F37, 0x0F37 },
+00049         { 0x0F39, 0x0F39 }, { 0x0F71, 0x0F7E }, { 0x0F80, 0x0F84 },
+00050         { 0x0F86, 0x0F87 }, { 0x0F90, 0x0F97 }, { 0x0F99, 0x0FBC },
+00051         { 0x0FC6, 0x0FC6 }, { 0x102D, 0x1030 }, { 0x1032, 0x1032 },
+00052         { 0x1036, 0x1037 }, { 0x1039, 0x1039 }, { 0x1058, 0x1059 },
+00053         { 0x1160, 0x11FF }, { 0x17B7, 0x17BD }, { 0x17C6, 0x17C6 },
+00054         { 0x17C9, 0x17D3 }, { 0x180B, 0x180E }, { 0x18A9, 0x18A9 },
+00055         { 0x200B, 0x200F }, { 0x202A, 0x202E }, { 0x206A, 0x206F },
+00056         { 0x20D0, 0x20E3 }, { 0x302A, 0x302F }, { 0x3099, 0x309A },
+00057         { 0xFB1E, 0xFB1E }, { 0xFE20, 0xFE23 }, { 0xFEFF, 0xFEFF },
+00058         { 0xFFF9, 0xFFFB }
+00059 };
+00060 
+00061 /* Auxiliary function for binary search in interval table */
+00062 static BOOL
+00063 bIsZeroWidthChar(ULONG ucs)
+00064 {
+00065         int low = 0;
+00066         int high = elementsof(combining) - 1;
+00067         int mid;
+00068 
+00069         if (ucs < (ULONG)combining[low].first ||
+00070             ucs > (ULONG)combining[high].last) {
+00071                 return FALSE;
+00072         }
+00073 
+00074         while (high >= low) {
+00075                 mid = (low + high) / 2;
+00076                 if (ucs > (ULONG)combining[mid].last) {
+00077                         low = mid + 1;
+00078                 } else if (ucs < (ULONG)combining[mid].first) {
+00079                         high = mid - 1;
+00080                 } else {
+00081                         return TRUE;
+00082                 }
+00083         }
+00084         return FALSE;
+00085 } /* end of bIsZeroWidthChar */
+00086 
+00087 /* The following functions define the column width of an ISO 10646
+00088  * character as follows:
+00089  *
+00090  *    - The null character (U+0000) has a column width of 0.
+00091  *
+00092  *    - Other C0/C1 control characters and DEL will lead to a return
+00093  *      value of -1.
+00094  *
+00095  *    - Non-spacing and enclosing combining characters (general
+00096  *      category code Mn or Me in the Unicode database) have a
+00097  *      column width of 0.
+00098  *
+00099  *    - Other format characters (general category code Cf in the Unicode
+00100  *      database) and ZERO WIDTH SPACE (U+200B) have a column width of 0.
+00101  *
+00102  *    - Hangul Jamo medial vowels and final consonants (U+1160-U+11FF)
+00103  *      have a column width of 0.
+00104  *
+00105  *    - Spacing characters in the East Asian Wide (W) or East Asian
+00106  *      FullWidth (F) category as defined in Unicode Technical
+00107  *      Report #11 have a column width of 2.
+00108  *
+00109  *    - All remaining characters (including all printable
+00110  *      ISO 8859-1 and WGL4 characters, Unicode control characters,
+00111  *      etc.) have a column width of 1.
+00112  *
+00113  * This implementation assumes that all characters are encoded
+00114  * in ISO 10646.
+00115  *
+00116  * This function is not named wcwidth() to prevent name clashes
+00117  */
+00118 static int
+00119 iWcWidth(ULONG ucs)
+00120 {
+00121         /* Test for 8-bit control characters */
+00122         if (ucs == 0) {
+00123                 return 0;
+00124         }
+00125         if (ucs < 0x20 || (ucs >= 0x7f && ucs < 0xa0)) {
+00126                 NO_DBG_HEX(ucs);
+00127                 return -1;
+00128         }
+00129 
+00130         /* Binary search in table of non-spacing characters */
+00131         if (bIsZeroWidthChar(ucs)) {
+00132                 return 0;
+00133         }
+00134 
+00135         /* Ucs is not a combining or C0/C1 control character */
+00136 
+00137         return 1 +
+00138         (ucs >= 0x1100 &&
+00139          (ucs <= 0x115f ||                    /* Hangul Jamo init. consonants */
+00140           (ucs >= 0x2e80 && ucs <= 0xa4cf && (ucs & ~0x0011) != 0x300a &&
+00141            ucs != 0x303f) ||                  /* CJK ... Yi */
+00142           (ucs >= 0xac00 && ucs <= 0xd7a3) || /* Hangul Syllables */
+00143           (ucs >= 0xf900 && ucs <= 0xfaff) || /* CJK Compatibility Ideographs */
+00144           (ucs >= 0xfe30 && ucs <= 0xfe6f) || /* CJK Compatibility Forms */
+00145           (ucs >= 0xff00 && ucs <= 0xff5f) || /* Fullwidth Forms */
+00146           (ucs >= 0xffe0 && ucs <= 0xffe6) ||
+00147           (ucs >= 0x20000 && ucs <= 0x2ffff)));
+00148 } /* end of iWcWidth */
+00149 
+00150 /*
+00151  * utf8_to_ucs - convert from UTF-8 to UCS
+00152  *
+00153  * Returns the UCS character,
+00154  * Fills in the number of bytes in the UTF-8 character
+00155  */
+00156 static ULONG
+00157 utf8_to_ucs(const char *p, int iStrLen, int *piUtfLen)
+00158 {
+00159         ULONG   ulUcs;
+00160         int     iIndex, iCharLen;
+00161 
+00162         fail(p == NULL || piUtfLen == NULL);
+00163         fail(iStrLen < 1);
+00164 
+00165         ulUcs = (ULONG)(UCHAR)p[0];
+00166 
+00167         if (ulUcs < 0x80) {
+00168                 *piUtfLen = 1;
+00169                 return ulUcs;
+00170         }
+00171 
+00172         if (ulUcs < 0xe0){
+00173                 iCharLen = 2;
+00174                 ulUcs &= 0x1f;
+00175         } else if (ulUcs < 0xf0){
+00176                 iCharLen = 3;
+00177                 ulUcs &= 0x0f;
+00178         } else if (ulUcs < 0xf8){
+00179                 iCharLen = 4;
+00180                 ulUcs &= 0x07;
+00181         } else if (ulUcs < 0xfc){
+00182                 iCharLen = 5;
+00183                 ulUcs &= 0x03;
+00184         } else {
+00185                 iCharLen = 6;
+00186                 ulUcs &= 0x01;
+00187         }
+00188         for (iIndex = 1; iIndex < iCharLen; iIndex++) {
+00189                 ulUcs <<= 6;
+00190                 if (iIndex < iStrLen) {
+00191                         ulUcs |= (ULONG)(UCHAR)p[iIndex] & 0x3f;
+00192                 }
+00193         }
+00194         *piUtfLen = iCharLen;
+00195         return ulUcs;
+00196 } /* end of utf8_to_ucs */
+00197 
+00198 /*
+00199  * utf8_strwidth - compute the string width of an UTF-8 string
+00200  *
+00201  * Returns the string width in columns
+00202  */
+00203 long
+00204 utf8_strwidth(const char *pcString, size_t tNumchars)
+00205 {
+00206         ULONG   ulUcs;
+00207         long    lTotal;
+00208         int     iToGo, iWidth, iUtflen;
+00209 
+00210         fail(pcString == NULL || tNumchars > (size_t)INT_MAX);
+00211 
+00212         lTotal = 0;
+00213         iToGo = (int)tNumchars;
+00214 
+00215         while (iToGo > 0 && *pcString != '\0') {
+00216                 ulUcs = utf8_to_ucs(pcString, iToGo, &iUtflen);
+00217                 iWidth = iWcWidth(ulUcs);
+00218                 if (iWidth > 0) {
+00219                         lTotal += iWidth;
+00220                 }
+00221                 pcString += iUtflen;
+00222                 iToGo -= iUtflen;
+00223         }
+00224         NO_DBG_DEC(lTotal);
+00225         return lTotal;
+00226 } /* end of utf8_strwidth */
+00227 
+00228 /*
+00229  * utf8_chrlength - get the number of bytes in an UTF-8 character
+00230  *
+00231  * Returns the number of bytes
+00232  */
+00233 int
+00234 utf8_chrlength(const char *p)
+00235 {
+00236         int     iUtflen;
+00237 
+00238         fail(p == NULL);
+00239 
+00240         iUtflen = -1;           /* Just to make sure */
+00241         (void)utf8_to_ucs(p, INT_MAX, &iUtflen);
+00242         NO_DBG_DEC(iUtflen);
+00243         return iUtflen;
+00244 } /* end of utf8_chrlength */
+00245 
+00246 /*
+00247  * is_locale_utf8 - return TRUE if the locale is UTF-8
+00248  */
+00249 BOOL
+00250 is_locale_utf8(void)
+00251 {
+00252         char    szCodeset[20];
+00253 
+00254         szCodeset[0] = '\0';
+00255         if (!bGetNormalizedCodeset(szCodeset, sizeof(szCodeset), NULL)) {
+00256                 return FALSE;
+00257         }
+00258         DBG_MSG(szCodeset);
+00259         return STREQ(szCodeset, "utf8");
+00260 } /* end of is_locale_utf8 */
+