2000-06-30 00:29:46 +00:00
|
|
|
/*
|
2001-03-21 20:44:20 +00:00
|
|
|
******************************************************************************
|
2000-06-30 00:29:46 +00:00
|
|
|
*
|
2005-01-04 21:58:13 +00:00
|
|
|
* Copyright (C) 2000-2005, International Business Machines
|
2000-06-30 00:29:46 +00:00
|
|
|
* Corporation and others. All Rights Reserved.
|
|
|
|
*
|
2001-03-21 20:44:20 +00:00
|
|
|
******************************************************************************
|
2000-06-30 00:29:46 +00:00
|
|
|
* file name: ushape.c
|
|
|
|
* encoding: US-ASCII
|
|
|
|
* tab size: 8 (not used)
|
|
|
|
* indentation:4
|
|
|
|
*
|
|
|
|
* created on: 2000jun29
|
|
|
|
* created by: Markus W. Scherer
|
2001-03-13 00:34:30 +00:00
|
|
|
*
|
|
|
|
* Arabic letter shaping implemented by Ayman Roshdy
|
2000-06-30 00:29:46 +00:00
|
|
|
*/
|
|
|
|
|
|
|
|
#include "unicode/utypes.h"
|
2000-08-12 01:01:03 +00:00
|
|
|
#include "unicode/uchar.h"
|
|
|
|
#include "unicode/ustring.h"
|
2000-06-30 00:29:46 +00:00
|
|
|
#include "unicode/ushape.h"
|
2004-10-18 02:43:33 +00:00
|
|
|
#include "cmemory.h"
|
|
|
|
#include "putilimp.h"
|
2001-10-08 23:26:58 +00:00
|
|
|
#include "ustr_imp.h"
|
2004-12-31 13:41:56 +00:00
|
|
|
#include "ubidi_props.h"
|
2000-06-30 00:29:46 +00:00
|
|
|
|
2000-08-12 01:01:03 +00:00
|
|
|
#if UTF_SIZE<16
|
|
|
|
/*
|
|
|
|
* This implementation assumes that the internal encoding is UTF-16
|
|
|
|
* or UTF-32, not UTF-8.
|
|
|
|
* The main assumption is that the Arabic characters and their
|
|
|
|
* presentation forms each fit into a single UChar.
|
|
|
|
* With UTF-8, they occupy 2 or 3 bytes, and more than the ASCII
|
|
|
|
* characters.
|
|
|
|
*/
|
|
|
|
# error This implementation assumes UTF-16 or UTF-32 (check UTF_SIZE)
|
|
|
|
#endif
|
|
|
|
|
2001-03-13 00:34:30 +00:00
|
|
|
/*
|
|
|
|
* ### TODO in general for letter shaping:
|
|
|
|
* - the letter shaping code is UTF-16-unaware; needs update
|
|
|
|
* + especially invertBuffer()?!
|
|
|
|
* - needs to handle the "Arabic Tail" that is used in some legacy codepages
|
|
|
|
* as a glyph fragment of wide-glyph letters
|
|
|
|
* + IBM Unicode conversion tables map it to U+200B (ZWSP)
|
|
|
|
* + IBM Egypt has proposed to encode the tail in Unicode among Arabic Presentation Forms
|
|
|
|
*/
|
|
|
|
|
|
|
|
/* definitions for Arabic letter shaping ------------------------------------ */
|
|
|
|
|
|
|
|
#define IRRELEVANT 4
|
|
|
|
#define LAMTYPE 16
|
|
|
|
#define ALEFTYPE 32
|
|
|
|
#define LINKR 1
|
|
|
|
#define LINKL 2
|
|
|
|
|
2001-08-30 16:59:44 +00:00
|
|
|
static const UChar IrrelevantPos[] = {
|
|
|
|
0x0, 0x2, 0x4, 0x6,
|
|
|
|
0x8, 0xA, 0xC, 0xE,
|
2001-03-13 00:34:30 +00:00
|
|
|
};
|
|
|
|
|
2001-08-30 16:59:44 +00:00
|
|
|
static const UChar convertLamAlef[] =
|
2001-03-13 00:34:30 +00:00
|
|
|
{
|
|
|
|
/*FEF5*/ 0x0622,
|
|
|
|
/*FEF6*/ 0x0622,
|
|
|
|
/*FEF7*/ 0x0623,
|
|
|
|
/*FEF8*/ 0x0623,
|
|
|
|
/*FEF9*/ 0x0625,
|
|
|
|
/*FEFA*/ 0x0625,
|
|
|
|
/*FEFB*/ 0x0627,
|
|
|
|
/*FEFC*/ 0x0627
|
|
|
|
};
|
|
|
|
|
2001-08-30 16:59:44 +00:00
|
|
|
static const UChar araLink[178]=
|
2001-03-13 00:34:30 +00:00
|
|
|
{
|
|
|
|
1 + 32 + 256 * 0x11,/*0x0622*/
|
|
|
|
1 + 32 + 256 * 0x13,/*0x0623*/
|
|
|
|
1 + 256 * 0x15,/*0x0624*/
|
|
|
|
1 + 32 + 256 * 0x17,/*0x0625*/
|
|
|
|
1 + 2 + 256 * 0x19,/*0x0626*/
|
|
|
|
1 + 32 + 256 * 0x1D,/*0x0627*/
|
|
|
|
1 + 2 + 256 * 0x1F,/*0x0628*/
|
|
|
|
1 + 256 * 0x23,/*0x0629*/
|
|
|
|
1 + 2 + 256 * 0x25,/*0x062A*/
|
|
|
|
1 + 2 + 256 * 0x29,/*0x062B*/
|
|
|
|
1 + 2 + 256 * 0x2D,/*0x062C*/
|
|
|
|
1 + 2 + 256 * 0x31,/*0x062D*/
|
|
|
|
1 + 2 + 256 * 0x35,/*0x062E*/
|
|
|
|
1 + 256 * 0x39,/*0x062F*/
|
|
|
|
1 + 256 * 0x3B,/*0x0630*/
|
|
|
|
1 + 256 * 0x3D,/*0x0631*/
|
|
|
|
1 + 256 * 0x3F,/*0x0632*/
|
|
|
|
1 + 2 + 256 * 0x41,/*0x0633*/
|
|
|
|
1 + 2 + 256 * 0x45,/*0x0634*/
|
|
|
|
1 + 2 + 256 * 0x49,/*0x0635*/
|
|
|
|
1 + 2 + 256 * 0x4D,/*0x0636*/
|
|
|
|
1 + 2 + 256 * 0x51,/*0x0637*/
|
|
|
|
1 + 2 + 256 * 0x55,/*0x0638*/
|
|
|
|
1 + 2 + 256 * 0x59,/*0x0639*/
|
|
|
|
1 + 2 + 256 * 0x5D,/*0x063A*/
|
|
|
|
0, 0, 0, 0, 0, /*0x063B-0x063F*/
|
|
|
|
1 + 2, /*0x0640*/
|
|
|
|
1 + 2 + 256 * 0x61,/*0x0641*/
|
|
|
|
1 + 2 + 256 * 0x65,/*0x0642*/
|
|
|
|
1 + 2 + 256 * 0x69,/*0x0643*/
|
|
|
|
1 + 2 + 16 + 256 * 0x6D,/*0x0644*/
|
|
|
|
1 + 2 + 256 * 0x71,/*0x0645*/
|
|
|
|
1 + 2 + 256 * 0x75,/*0x0646*/
|
|
|
|
1 + 2 + 256 * 0x79,/*0x0647*/
|
|
|
|
1 + 256 * 0x7D,/*0x0648*/
|
|
|
|
1 + 256 * 0x7F,/*0x0649*/
|
|
|
|
1 + 2 + 256 * 0x81,/*0x064A*/
|
|
|
|
4, 4, 4, 4, /*0x064B-0x064E*/
|
|
|
|
4, 4, 4, 4, /*0x064F-0x0652*/
|
|
|
|
4, 4, 4, 0, 0, /*0x0653-0x0657*/
|
|
|
|
0, 0, 0, 0, /*0x0658-0x065B*/
|
|
|
|
1 + 256 * 0x85,/*0x065C*/
|
|
|
|
1 + 256 * 0x87,/*0x065D*/
|
|
|
|
1 + 256 * 0x89,/*0x065E*/
|
|
|
|
1 + 256 * 0x8B,/*0x065F*/
|
|
|
|
0, 0, 0, 0, 0, /*0x0660-0x0664*/
|
|
|
|
0, 0, 0, 0, 0, /*0x0665-0x0669*/
|
|
|
|
0, 0, 0, 0, 0, 0, /*0x066A-0x066F*/
|
|
|
|
4, /*0x0670*/
|
|
|
|
0, /*0x0671*/
|
|
|
|
1 + 32, /*0x0672*/
|
|
|
|
1 + 32, /*0x0673*/
|
|
|
|
0, /*0x0674*/
|
|
|
|
1 + 32, /*0x0675*/
|
|
|
|
1, 1, /*0x0676-0x0677*/
|
|
|
|
1+2, 1+2, 1+2, 1+2, 1+2, 1+2, /*0x0678-0x067D*/
|
|
|
|
1+2, 1+2, 1+2, 1+2, 1+2, 1+2, /*0x067E-0x0683*/
|
|
|
|
1+2, 1+2, 1+2, 1+2, /*0x0684-0x0687*/
|
|
|
|
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, /*0x0688-0x0691*/
|
|
|
|
1, 1, 1, 1, 1, 1, 1, 1, /*0x0692-0x0699*/
|
|
|
|
1+2, 1+2, 1+2, 1+2, 1+2, 1+2, /*0x069A-0x06A3*/
|
|
|
|
1+2, 1+2, 1+2, 1+2, /*0x069A-0x06A3*/
|
|
|
|
1+2, 1+2, 1+2, 1+2, 1+2, 1+2, /*0x06A4-0x06AD*/
|
|
|
|
1+2, 1+2, 1+2, 1+2, /*0x06A4-0x06AD*/
|
|
|
|
1+2, 1+2, 1+2, 1+2, 1+2, 1+2, /*0x06AE-0x06B7*/
|
|
|
|
1+2, 1+2, 1+2, 1+2, /*0x06AE-0x06B7*/
|
|
|
|
1+2, 1+2, 1+2, 1+2, 1+2, 1+2, /*0x06B8-0x06BF*/
|
|
|
|
1+2, 1+2, /*0x06B8-0x06BF*/
|
|
|
|
1, /*0x06C0*/
|
|
|
|
1+2, /*0x06C1*/
|
|
|
|
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, /*0x06C2-0x06CB*/
|
|
|
|
1+2, /*0x06CC*/
|
|
|
|
1, /*0x06CD*/
|
|
|
|
1+2, 1+2, 1+2, 1+2, /*0x06CE-0x06D1*/
|
|
|
|
1, 1 /*0x06D2-0x06D3*/
|
|
|
|
};
|
|
|
|
|
2001-08-30 16:59:44 +00:00
|
|
|
static const UChar presLink[141]=
|
2001-03-13 00:34:30 +00:00
|
|
|
{
|
|
|
|
1 + 2, /*0xFE70*/
|
|
|
|
1 + 2, /*0xFE71*/
|
|
|
|
1 + 2, 0, 1+ 2, 0, 1+ 2, /*0xFE72-0xFE76*/
|
|
|
|
1 + 2, /*0xFE77*/
|
|
|
|
1+ 2, 1 + 2, 1+2, 1 + 2, /*0xFE78-0xFE81*/
|
|
|
|
1+ 2, 1 + 2, 1+2, 1 + 2, /*0xFE82-0xFE85*/
|
|
|
|
0, 0 + 32, 1 + 32, 0 + 32, /*0xFE86-0xFE89*/
|
|
|
|
1 + 32, 0, 1, 0 + 32, /*0xFE8A-0xFE8D*/
|
|
|
|
1 + 32, 0, 2, 1 + 2, /*0xFE8E-0xFE91*/
|
|
|
|
1, 0 + 32, 1 + 32, 0, /*0xFE92-0xFE95*/
|
|
|
|
2, 1 + 2, 1, 0, /*0xFE96-0xFE99*/
|
|
|
|
1, 0, 2, 1 + 2, /*0xFE9A-0xFE9D*/
|
|
|
|
1, 0, 2, 1 + 2, /*0xFE9E-0xFEA1*/
|
|
|
|
1, 0, 2, 1 + 2, /*0xFEA2-0xFEA5*/
|
|
|
|
1, 0, 2, 1 + 2, /*0xFEA6-0xFEA9*/
|
|
|
|
1, 0, 2, 1 + 2, /*0xFEAA-0xFEAD*/
|
|
|
|
1, 0, 1, 0, /*0xFEAE-0xFEB1*/
|
|
|
|
1, 0, 1, 0, /*0xFEB2-0xFEB5*/
|
|
|
|
1, 0, 2, 1+2, /*0xFEB6-0xFEB9*/
|
|
|
|
1, 0, 2, 1+2, /*0xFEBA-0xFEBD*/
|
|
|
|
1, 0, 2, 1+2, /*0xFEBE-0xFEC1*/
|
|
|
|
1, 0, 2, 1+2, /*0xFEC2-0xFEC5*/
|
|
|
|
1, 0, 2, 1+2, /*0xFEC6-0xFEC9*/
|
|
|
|
1, 0, 2, 1+2, /*0xFECA-0xFECD*/
|
|
|
|
1, 0, 2, 1+2, /*0xFECE-0xFED1*/
|
|
|
|
1, 0, 2, 1+2, /*0xFED2-0xFED5*/
|
|
|
|
1, 0, 2, 1+2, /*0xFED6-0xFED9*/
|
|
|
|
1, 0, 2, 1+2, /*0xFEDA-0xFEDD*/
|
|
|
|
1, 0, 2, 1+2, /*0xFEDE-0xFEE1*/
|
|
|
|
1, 0 + 16, 2 + 16, 1 + 2 +16, /*0xFEE2-0xFEE5*/
|
|
|
|
1 + 16, 0, 2, 1+2, /*0xFEE6-0xFEE9*/
|
|
|
|
1, 0, 2, 1+2, /*0xFEEA-0xFEED*/
|
|
|
|
1, 0, 2, 1+2, /*0xFEEE-0xFEF1*/
|
|
|
|
1, 0, 1, 0, /*0xFEF2-0xFEF5*/
|
|
|
|
1, 0, 2, 1+2, /*0xFEF6-0xFEF9*/
|
|
|
|
1, 0, 1, 0, /*0xFEFA-0xFEFD*/
|
|
|
|
1, 0, 1, 0,
|
|
|
|
1
|
|
|
|
};
|
|
|
|
|
2001-08-30 16:59:44 +00:00
|
|
|
static const UChar convertFEto06[] =
|
2001-03-13 00:34:30 +00:00
|
|
|
{
|
|
|
|
/***********0******1******2******3******4******5******6******7******8******9******A******B******C******D******E******F***/
|
|
|
|
/*FE7*/ 0x64B, 0x64B, 0x64C, 0x64C, 0x64D, 0x64D, 0x64E, 0x64E, 0x64F, 0x64F, 0x650, 0x650, 0x651, 0x651, 0x652, 0x652,
|
|
|
|
/*FE8*/ 0x621, 0x622, 0x622, 0x623, 0x623, 0x624, 0x624, 0x625, 0x625, 0x626, 0x626, 0x626, 0x626, 0x627, 0x627, 0x628,
|
|
|
|
/*FE9*/ 0x628, 0x628, 0x628, 0x629, 0x629, 0x62A, 0x62A, 0x62A, 0x62A, 0x62B, 0x62B, 0x62B, 0x62B, 0x62C, 0x62C, 0x62C,
|
|
|
|
/*FEA*/ 0x62C, 0x62D, 0x62D, 0x62D, 0x62D, 0x62E, 0x62E, 0x62E, 0x62E, 0x62F, 0x62F, 0x630, 0x630, 0x631, 0x631, 0x632,
|
|
|
|
/*FEB*/ 0x632, 0x633, 0x633, 0x633, 0x633, 0x634, 0x634, 0x634, 0x634, 0x635, 0x635, 0x635, 0x635, 0x636, 0x636, 0x636,
|
|
|
|
/*FEC*/ 0x636, 0x637, 0x637, 0x637, 0x637, 0x638, 0x638, 0x638, 0x638, 0x639, 0x639, 0x639, 0x639, 0x63A, 0x63A, 0x63A,
|
|
|
|
/*FED*/ 0x63A, 0x641, 0x641, 0x641, 0x641, 0x642, 0x642, 0x642, 0x642, 0x643, 0x643, 0x643, 0x643, 0x644, 0x644, 0x644,
|
|
|
|
/*FEE*/ 0x644, 0x645, 0x645, 0x645, 0x645, 0x646, 0x646, 0x646, 0x646, 0x647, 0x647, 0x647, 0x647, 0x648, 0x648, 0x649,
|
|
|
|
/*FEF*/ 0x649, 0x64A, 0x64A, 0x64A, 0x64A, 0x65C, 0x65C, 0x65D, 0x65D, 0x65E, 0x65E, 0x65F, 0x65F
|
|
|
|
};
|
|
|
|
|
2001-08-30 16:59:44 +00:00
|
|
|
static const UChar shapeTable[4][4][4]=
|
2001-03-13 00:34:30 +00:00
|
|
|
{
|
|
|
|
{ {0,0,0,0}, {0,0,0,0}, {0,1,0,3}, {0,1,0,1} },
|
|
|
|
{ {0,0,2,2}, {0,0,1,2}, {0,1,1,2}, {0,1,1,3} },
|
|
|
|
{ {0,0,0,0}, {0,0,0,0}, {0,1,0,3}, {0,1,0,3} },
|
|
|
|
{ {0,0,1,2}, {0,0,1,2}, {0,1,1,2}, {0,1,1,3} }
|
|
|
|
};
|
|
|
|
|
2000-08-12 01:01:03 +00:00
|
|
|
/*
|
|
|
|
* This function shapes European digits to Arabic-Indic digits
|
|
|
|
* in-place, writing over the input characters.
|
|
|
|
* Since we know that we are only looking for BMP code points,
|
|
|
|
* we can safely just work with code units (again, at least UTF-16).
|
|
|
|
*/
|
|
|
|
static void
|
|
|
|
_shapeToArabicDigitsWithContext(UChar *s, int32_t length,
|
|
|
|
UChar digitBase,
|
|
|
|
UBool isLogical, UBool lastStrongWasAL) {
|
2004-12-31 13:41:56 +00:00
|
|
|
UBiDiProps *bdp;
|
|
|
|
UErrorCode errorCode;
|
|
|
|
|
2000-08-12 01:01:03 +00:00
|
|
|
int32_t i;
|
|
|
|
UChar c;
|
|
|
|
|
2005-01-04 21:58:13 +00:00
|
|
|
errorCode=U_ZERO_ERROR;
|
2004-12-31 13:41:56 +00:00
|
|
|
bdp=ubidi_getSingleton(&errorCode);
|
|
|
|
if(U_FAILURE(errorCode)) {
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2000-08-12 01:01:03 +00:00
|
|
|
digitBase-=0x30;
|
|
|
|
|
|
|
|
/* the iteration direction depends on the type of input */
|
|
|
|
if(isLogical) {
|
|
|
|
for(i=0; i<length; ++i) {
|
|
|
|
c=s[i];
|
2004-12-31 13:41:56 +00:00
|
|
|
switch(ubidi_getClass(bdp, c)) {
|
2000-08-12 01:01:03 +00:00
|
|
|
case U_LEFT_TO_RIGHT: /* L */
|
|
|
|
case U_RIGHT_TO_LEFT: /* R */
|
|
|
|
lastStrongWasAL=FALSE;
|
|
|
|
break;
|
|
|
|
case U_RIGHT_TO_LEFT_ARABIC: /* AL */
|
|
|
|
lastStrongWasAL=TRUE;
|
|
|
|
break;
|
|
|
|
case U_EUROPEAN_NUMBER: /* EN */
|
|
|
|
if(lastStrongWasAL && (uint32_t)(c-0x30)<10) {
|
2000-08-21 21:40:35 +00:00
|
|
|
s[i]=(UChar)(digitBase+c); /* digitBase+(c-0x30) - digitBase was modified above */
|
2000-08-12 01:01:03 +00:00
|
|
|
}
|
|
|
|
break;
|
|
|
|
default :
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
} else {
|
|
|
|
for(i=length; i>0; /* pre-decrement in the body */) {
|
|
|
|
c=s[--i];
|
2004-12-31 13:41:56 +00:00
|
|
|
switch(ubidi_getClass(bdp, c)) {
|
2000-08-12 01:01:03 +00:00
|
|
|
case U_LEFT_TO_RIGHT: /* L */
|
|
|
|
case U_RIGHT_TO_LEFT: /* R */
|
|
|
|
lastStrongWasAL=FALSE;
|
|
|
|
break;
|
|
|
|
case U_RIGHT_TO_LEFT_ARABIC: /* AL */
|
|
|
|
lastStrongWasAL=TRUE;
|
|
|
|
break;
|
|
|
|
case U_EUROPEAN_NUMBER: /* EN */
|
|
|
|
if(lastStrongWasAL && (uint32_t)(c-0x30)<10) {
|
2000-08-21 21:40:35 +00:00
|
|
|
s[i]=(UChar)(digitBase+c); /* digitBase+(c-0x30) - digitBase was modified above */
|
2000-08-12 01:01:03 +00:00
|
|
|
}
|
|
|
|
break;
|
|
|
|
default :
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2001-03-13 00:34:30 +00:00
|
|
|
/*
|
|
|
|
*Name : invertBuffer
|
|
|
|
*Function : This function inverts the buffer, it's used
|
|
|
|
* in case the user specifies the buffer to be
|
|
|
|
* U_SHAPE_TEXT_DIRECTION_LOGICAL
|
|
|
|
*/
|
|
|
|
static void
|
|
|
|
invertBuffer(UChar *buffer,int32_t size,uint32_t options,int32_t *spacesCountl,int32_t *spacesCountr) {
|
|
|
|
|
|
|
|
UChar temp;
|
|
|
|
int32_t i=0,j=0;
|
|
|
|
int32_t lowlimit = 0, highlimit = 0;
|
|
|
|
|
|
|
|
lowlimit = *spacesCountl;
|
|
|
|
highlimit = *spacesCountr;
|
|
|
|
|
|
|
|
for(i=lowlimit,j=size-highlimit-1;i<j;i++,j--) {
|
|
|
|
temp = buffer[i];
|
|
|
|
buffer[i] = buffer[j];
|
|
|
|
buffer[j] = temp;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
*Name : changeLamAlef
|
|
|
|
*Function : Converts the Alef characters into an equivalent
|
|
|
|
* LamAlef location in the 0x06xx Range, this is an
|
|
|
|
* intermediate stage in the operation of the program
|
|
|
|
* later it'll be converted into the 0xFExx LamAlefs
|
|
|
|
* in the shaping function.
|
|
|
|
*/
|
2001-03-22 01:26:48 +00:00
|
|
|
static UChar
|
2001-03-13 00:34:30 +00:00
|
|
|
changeLamAlef(UChar ch) {
|
|
|
|
|
|
|
|
switch(ch) {
|
|
|
|
case 0x0622 :
|
|
|
|
return(0x065C);
|
|
|
|
break;
|
|
|
|
case 0x0623 :
|
|
|
|
return(0x065D);
|
|
|
|
break;
|
|
|
|
case 0x0625 :
|
|
|
|
return(0x065E);
|
|
|
|
break;
|
|
|
|
case 0x0627 :
|
|
|
|
return(0x065F);
|
|
|
|
break;
|
|
|
|
default :
|
|
|
|
return(0);
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
*Name : specialChar
|
|
|
|
*Function : Special Arabic characters need special handling in the shapeUnicode
|
|
|
|
* function, this function returns 1 or 2 for these special characters
|
|
|
|
*/
|
2001-03-22 01:26:48 +00:00
|
|
|
static int32_t
|
2001-03-13 00:34:30 +00:00
|
|
|
specialChar(UChar ch) {
|
|
|
|
|
|
|
|
if( (ch>0x0621 && ch<0x0626)||(ch==0x0627)||(ch>0x062e && ch<0x0633)||
|
|
|
|
(ch>0x0647 && ch<0x064a)||(ch==0x0629) ) {
|
|
|
|
return (1);
|
|
|
|
}
|
|
|
|
else
|
|
|
|
if( ch>=0x064B && ch<= 0x0652 )
|
|
|
|
return (2);
|
|
|
|
else
|
2001-03-22 01:26:48 +00:00
|
|
|
if( (ch>=0x0653 && ch<= 0x0655) || ch == 0x0670 ||
|
|
|
|
(ch>=0xFE70 && ch<= 0xFE7F) )
|
2001-03-13 00:34:30 +00:00
|
|
|
return (3);
|
|
|
|
else
|
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
*Name : getLink
|
|
|
|
*Function : Resolves the link between the characters as
|
|
|
|
* Arabic characters have four forms :
|
|
|
|
* Isolated, Initial, Middle and Final Form
|
|
|
|
*/
|
2001-03-22 01:26:48 +00:00
|
|
|
static UChar
|
2001-03-13 00:34:30 +00:00
|
|
|
getLink(UChar ch) {
|
|
|
|
|
|
|
|
if(ch >= 0x0622 && ch <= 0x06D3) {
|
|
|
|
return(araLink[ch-0x0622]);
|
|
|
|
} else if(ch == 0x200D) {
|
|
|
|
return(3);
|
|
|
|
} else if(ch >= 0x206D && ch <= 0x206F) {
|
|
|
|
return(4);
|
|
|
|
} else if(ch >= 0xFE70 && ch <= 0xFEFC) {
|
|
|
|
return(presLink[ch-0xFE70]);
|
|
|
|
} else {
|
|
|
|
return(0);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
*Name : countSpaces
|
|
|
|
*Function : Counts the number of spaces
|
|
|
|
* at each end of the logical buffer
|
|
|
|
*/
|
2001-03-22 01:26:48 +00:00
|
|
|
static void
|
2001-03-13 00:34:30 +00:00
|
|
|
countSpaces(UChar *dest,int32_t size,uint32_t options,int32_t *spacesCountl,int32_t *spacesCountr) {
|
|
|
|
|
|
|
|
int32_t i = 0;
|
|
|
|
int32_t countl = 0,countr = 0;
|
|
|
|
|
|
|
|
while(dest[i] == 0x0020) {
|
|
|
|
countl++;
|
|
|
|
i++;
|
|
|
|
}
|
|
|
|
while(dest[size-1] == 0x0020) {
|
|
|
|
countr++;
|
|
|
|
size--;
|
|
|
|
}
|
|
|
|
*spacesCountl = countl;
|
|
|
|
*spacesCountr = countr;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
*Name : isTashkeelChar
|
|
|
|
*Function : Returns 1 for Tashkeel characters else return 0
|
|
|
|
*/
|
2001-03-22 01:26:48 +00:00
|
|
|
static int32_t
|
2001-03-13 00:34:30 +00:00
|
|
|
isTashkeelChar(UChar ch) {
|
|
|
|
|
|
|
|
if( ch>=0x064B && ch<= 0x0652 )
|
|
|
|
return (1);
|
|
|
|
else
|
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
*Name : isAlefChar
|
|
|
|
*Function : Returns 1 for Alef characters else return 0
|
|
|
|
*/
|
2001-03-22 01:26:48 +00:00
|
|
|
static int32_t
|
2001-03-13 00:34:30 +00:00
|
|
|
isAlefChar(UChar ch) {
|
|
|
|
|
|
|
|
if( (ch==0x0622)||(ch==0x0623)||(ch==0x0625)||(ch==0x0627) )
|
|
|
|
return (1);
|
|
|
|
else
|
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
*Name : isLamAlefChar
|
|
|
|
*Function : Returns 1 for LamAlef characters else return 0
|
|
|
|
*/
|
2001-03-22 01:26:48 +00:00
|
|
|
static int32_t
|
2001-03-13 00:34:30 +00:00
|
|
|
isLamAlefChar(UChar ch) {
|
|
|
|
|
|
|
|
if( (ch>=0xFEF5)&&(ch<=0xFEFC) )
|
|
|
|
return (1);
|
|
|
|
else
|
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
*Name : calculateSize
|
|
|
|
*Function : This function calculates the destSize to be used in preflighting
|
|
|
|
* when the destSize is equal to 0
|
|
|
|
*/
|
2001-03-22 01:26:48 +00:00
|
|
|
static int32_t
|
2001-03-13 00:34:30 +00:00
|
|
|
calculateSize(const UChar *source, int32_t sourceLength,
|
|
|
|
int32_t destSize,uint32_t options) {
|
|
|
|
|
|
|
|
int32_t i = 0;
|
|
|
|
destSize = sourceLength;
|
|
|
|
|
|
|
|
switch(options&U_SHAPE_LETTERS_MASK) {
|
|
|
|
|
|
|
|
case U_SHAPE_LETTERS_SHAPE :
|
|
|
|
if((options&U_SHAPE_TEXT_DIRECTION_MASK)==U_SHAPE_TEXT_DIRECTION_VISUAL_LTR) {
|
|
|
|
for(i=0;i<sourceLength;i++) {
|
|
|
|
if( (isAlefChar(source[i]))&&(source[i+1]==0x0644) ) {
|
|
|
|
destSize--;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
else
|
|
|
|
if((options&U_SHAPE_TEXT_DIRECTION_MASK)==U_SHAPE_TEXT_DIRECTION_LOGICAL) {
|
|
|
|
for(i=0;i<sourceLength;i++) {
|
|
|
|
if( (isAlefChar(source[i+1]))&&(source[i]==0x0644) ) {
|
|
|
|
destSize--;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
|
|
|
|
case U_SHAPE_LETTERS_UNSHAPE :
|
|
|
|
for(i=0;i<sourceLength;i++) {
|
|
|
|
if( isLamAlefChar(source[i]) ) {
|
|
|
|
destSize++;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
|
|
|
|
default :
|
|
|
|
/* will never occur because of validity checks at the begin of u_shapeArabic */
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
return destSize;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
*Name : removeLamAlefSpaces
|
|
|
|
*Function : The shapeUnicode function converts Lam + Alef into LamAlef + space,
|
|
|
|
* this function removes the spaces behind the LamAlefs according to
|
|
|
|
* the options the user specifies, the spaces are removed to the end
|
|
|
|
* of the buffer, or shrink the buffer ab=nd remove spaces for good
|
|
|
|
* or leave the buffer as it is LamAlef + space.
|
|
|
|
*/
|
2001-03-22 01:26:48 +00:00
|
|
|
static int32_t
|
2001-03-13 00:34:30 +00:00
|
|
|
removeLamAlefSpaces(UChar *dest, int32_t sourceLength,
|
|
|
|
int32_t destSize,
|
|
|
|
uint32_t options,
|
|
|
|
UErrorCode *pErrorCode) {
|
|
|
|
|
|
|
|
int32_t i = 0, j = 0;
|
|
|
|
int32_t count = 0;
|
|
|
|
UChar *tempbuffer=NULL;
|
|
|
|
|
|
|
|
switch(options&U_SHAPE_LENGTH_MASK) {
|
|
|
|
case U_SHAPE_LENGTH_GROW_SHRINK :
|
|
|
|
tempbuffer = (UChar *)uprv_malloc((sourceLength+1)*U_SIZEOF_UCHAR);
|
2002-07-03 11:59:21 +00:00
|
|
|
/* Test for NULL */
|
|
|
|
if(tempbuffer == NULL) {
|
|
|
|
*pErrorCode = U_MEMORY_ALLOCATION_ERROR;
|
|
|
|
return 0;
|
|
|
|
}
|
2002-06-29 09:31:05 +00:00
|
|
|
|
2001-03-13 00:34:30 +00:00
|
|
|
uprv_memset(tempbuffer, 0, (sourceLength+1)*U_SIZEOF_UCHAR);
|
|
|
|
|
|
|
|
i = j = 0;
|
|
|
|
while(i < sourceLength) {
|
|
|
|
if(dest[i] == 0xFFFF) {
|
|
|
|
j--;
|
|
|
|
count++;
|
|
|
|
}
|
|
|
|
else
|
|
|
|
tempbuffer[j] = dest[i];
|
|
|
|
i++;
|
|
|
|
j++;
|
|
|
|
}
|
|
|
|
|
|
|
|
while(count >= 0) {
|
|
|
|
tempbuffer[i] = 0x0000;
|
|
|
|
i--;
|
|
|
|
count--;
|
|
|
|
}
|
|
|
|
|
|
|
|
uprv_memcpy(dest, tempbuffer, sourceLength*U_SIZEOF_UCHAR);
|
|
|
|
destSize = u_strlen(dest);
|
|
|
|
break;
|
|
|
|
|
|
|
|
case U_SHAPE_LENGTH_FIXED_SPACES_NEAR :
|
|
|
|
/* Lam+Alef is already shaped into LamAlef + FFFF */
|
|
|
|
i = 0;
|
|
|
|
while(i < sourceLength) {
|
|
|
|
if(dest[i] == 0xFFFF)
|
|
|
|
dest[i] = 0x0020;
|
|
|
|
i++;
|
|
|
|
}
|
|
|
|
destSize = sourceLength;
|
|
|
|
break;
|
|
|
|
|
|
|
|
case U_SHAPE_LENGTH_FIXED_SPACES_AT_BEGINNING :
|
|
|
|
tempbuffer = (UChar *)uprv_malloc((sourceLength+1)*U_SIZEOF_UCHAR);
|
2002-07-03 11:59:21 +00:00
|
|
|
|
|
|
|
/* Test for NULL */
|
|
|
|
if(tempbuffer == NULL) {
|
|
|
|
*pErrorCode = U_MEMORY_ALLOCATION_ERROR;
|
|
|
|
return 0;
|
|
|
|
}
|
2002-06-29 09:31:05 +00:00
|
|
|
|
2001-03-13 00:34:30 +00:00
|
|
|
uprv_memset(tempbuffer, 0, (sourceLength+1)*U_SIZEOF_UCHAR);
|
|
|
|
|
|
|
|
i = j = sourceLength;
|
|
|
|
while(i >= 0) {
|
|
|
|
if(dest[i] == 0xFFFF) {
|
|
|
|
j++;
|
|
|
|
count++;
|
|
|
|
}
|
|
|
|
else
|
|
|
|
tempbuffer[j] = dest[i];
|
|
|
|
i--;
|
|
|
|
j--;
|
|
|
|
}
|
|
|
|
for(i=0;i<count;i++)
|
|
|
|
tempbuffer[i] = 0x0020;
|
|
|
|
|
|
|
|
uprv_memcpy(dest, tempbuffer, sourceLength*U_SIZEOF_UCHAR);
|
|
|
|
destSize = sourceLength;
|
|
|
|
break;
|
|
|
|
|
|
|
|
case U_SHAPE_LENGTH_FIXED_SPACES_AT_END :
|
|
|
|
tempbuffer = (UChar *)uprv_malloc((sourceLength+1)*U_SIZEOF_UCHAR);
|
2002-06-29 09:31:05 +00:00
|
|
|
|
2002-07-03 11:59:21 +00:00
|
|
|
/* Test for NULL */
|
|
|
|
if(tempbuffer == NULL) {
|
|
|
|
*pErrorCode = U_MEMORY_ALLOCATION_ERROR;
|
|
|
|
return 0;
|
|
|
|
}
|
2002-06-29 09:31:05 +00:00
|
|
|
|
2001-03-13 00:34:30 +00:00
|
|
|
uprv_memset(tempbuffer, 0, (sourceLength+1)*U_SIZEOF_UCHAR);
|
|
|
|
|
|
|
|
i = j = 0;
|
|
|
|
while(i < sourceLength) {
|
|
|
|
if(dest[i] == 0xFFFF) {
|
|
|
|
j--;
|
|
|
|
count++;
|
|
|
|
}
|
|
|
|
else
|
|
|
|
tempbuffer[j] = dest[i];
|
|
|
|
i++;
|
|
|
|
j++;
|
|
|
|
}
|
|
|
|
|
|
|
|
while(count >= 0) {
|
|
|
|
tempbuffer[i] = 0x0020;
|
|
|
|
i--;
|
|
|
|
count--;
|
|
|
|
}
|
|
|
|
|
|
|
|
uprv_memcpy(dest,tempbuffer, sourceLength*U_SIZEOF_UCHAR);
|
|
|
|
destSize = sourceLength;
|
|
|
|
break;
|
|
|
|
|
|
|
|
default :
|
|
|
|
/* will not occur */
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
if(tempbuffer)
|
|
|
|
uprv_free(tempbuffer);
|
|
|
|
|
|
|
|
return destSize;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
*Name : expandLamAlef
|
|
|
|
*Function : LamAlef needs special handling as the LamAlef is
|
|
|
|
* one character while expanding it will give two
|
|
|
|
* characters Lam + Alef, so we need to expand the LamAlef
|
|
|
|
* in near or far spaces according to the options the user
|
|
|
|
* specifies or increase the buffer size.
|
|
|
|
* If there are no spaces to expand the LamAlef, an error
|
|
|
|
* will be set to U_NO_SPACE_AVAILABLE as defined in utypes.h
|
|
|
|
*/
|
2001-03-22 01:26:48 +00:00
|
|
|
static int32_t
|
2001-03-13 00:34:30 +00:00
|
|
|
expandLamAlef(UChar *dest, int32_t sourceLength,
|
|
|
|
int32_t destSize,uint32_t options,
|
|
|
|
UErrorCode *pErrorCode) {
|
|
|
|
|
|
|
|
int32_t i = 0,j = 0;
|
|
|
|
int32_t countl = 0;
|
|
|
|
int32_t countr = 0;
|
|
|
|
int32_t inpsize = sourceLength;
|
|
|
|
UChar lamalefChar;
|
|
|
|
UChar *tempbuffer=NULL;
|
|
|
|
|
|
|
|
switch(options&U_SHAPE_LENGTH_MASK) {
|
|
|
|
|
|
|
|
case U_SHAPE_LENGTH_GROW_SHRINK :
|
|
|
|
destSize = calculateSize(dest,sourceLength,destSize,options);
|
|
|
|
tempbuffer = (UChar *)uprv_malloc((destSize+1)*U_SIZEOF_UCHAR);
|
2002-06-29 09:31:05 +00:00
|
|
|
|
2002-07-03 11:59:21 +00:00
|
|
|
/* Test for NULL */
|
|
|
|
if(tempbuffer == NULL) {
|
|
|
|
*pErrorCode = U_MEMORY_ALLOCATION_ERROR;
|
|
|
|
return 0;
|
|
|
|
}
|
2002-06-29 09:31:05 +00:00
|
|
|
|
2001-03-13 00:34:30 +00:00
|
|
|
uprv_memset(tempbuffer, 0, (destSize+1)*U_SIZEOF_UCHAR);
|
|
|
|
|
|
|
|
i = j = 0;
|
|
|
|
while(i < destSize && j < destSize) {
|
|
|
|
if( isLamAlefChar(dest[i]) ) {
|
|
|
|
tempbuffer[j] = convertLamAlef[ dest[i] - 0xFEF5 ];
|
|
|
|
tempbuffer[j+1] = 0x0644;
|
|
|
|
j++;
|
|
|
|
}
|
|
|
|
else
|
|
|
|
tempbuffer[j] = dest[i];
|
|
|
|
i++;
|
|
|
|
j++;
|
|
|
|
}
|
|
|
|
|
|
|
|
uprv_memcpy(dest, tempbuffer, destSize*U_SIZEOF_UCHAR);
|
|
|
|
break;
|
|
|
|
|
|
|
|
case U_SHAPE_LENGTH_FIXED_SPACES_NEAR :
|
|
|
|
for(i = 0;i<sourceLength;i++) {
|
|
|
|
if((dest[i] == 0x0020) && isLamAlefChar(dest[i+1])) {
|
|
|
|
lamalefChar = dest[i+1];
|
|
|
|
dest[i+1] = 0x0644;
|
|
|
|
dest[i] = convertLamAlef[ lamalefChar - 0xFEF5 ];
|
|
|
|
}
|
|
|
|
else
|
|
|
|
if((dest[i] != 0x0020) && isLamAlefChar(dest[i+1])) {
|
|
|
|
*pErrorCode=U_NO_SPACE_AVAILABLE;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
destSize = sourceLength;
|
|
|
|
break;
|
|
|
|
|
|
|
|
case U_SHAPE_LENGTH_FIXED_SPACES_AT_BEGINNING :
|
|
|
|
tempbuffer = (UChar *)uprv_malloc((sourceLength+1)*U_SIZEOF_UCHAR);
|
2002-06-29 09:31:05 +00:00
|
|
|
|
2002-07-03 11:59:21 +00:00
|
|
|
/* Test for NULL */
|
|
|
|
if(tempbuffer == NULL) {
|
|
|
|
*pErrorCode = U_MEMORY_ALLOCATION_ERROR;
|
|
|
|
return 0;
|
|
|
|
}
|
2002-06-29 09:31:05 +00:00
|
|
|
|
2001-03-13 00:34:30 +00:00
|
|
|
uprv_memset(tempbuffer, 0, (sourceLength+1)*U_SIZEOF_UCHAR);
|
|
|
|
|
|
|
|
i = 0;
|
|
|
|
while(dest[i] == 0x0020) {
|
|
|
|
countl++;
|
|
|
|
i++;
|
|
|
|
}
|
|
|
|
|
|
|
|
i = j = sourceLength-1;
|
|
|
|
while(i >= 0 && j >= 0) {
|
|
|
|
if( countl>0 && isLamAlefChar(dest[i]) ) {
|
|
|
|
tempbuffer[j] = 0x0644;
|
|
|
|
tempbuffer[j-1] = convertLamAlef[ dest[i] - 0xFEF5 ];
|
|
|
|
j--;
|
|
|
|
countl--;
|
|
|
|
}
|
|
|
|
else {
|
|
|
|
if( countl == 0 && isLamAlefChar(dest[i]) )
|
|
|
|
*pErrorCode=U_NO_SPACE_AVAILABLE;
|
|
|
|
tempbuffer[j] = dest[i];
|
|
|
|
}
|
|
|
|
i--;
|
|
|
|
j--;
|
|
|
|
}
|
|
|
|
|
|
|
|
uprv_memcpy(dest, tempbuffer, sourceLength*U_SIZEOF_UCHAR);
|
|
|
|
destSize = sourceLength;
|
|
|
|
break;
|
|
|
|
|
|
|
|
case U_SHAPE_LENGTH_FIXED_SPACES_AT_END :
|
|
|
|
/* LamAlef expansion below is done from right to left to make sure that we consume
|
|
|
|
* the spaces with the LamAlefs as they appear in the visual buffer from right to left
|
|
|
|
*/
|
|
|
|
tempbuffer = (UChar *)uprv_malloc((sourceLength+1)*U_SIZEOF_UCHAR);
|
2002-06-29 09:31:05 +00:00
|
|
|
|
2002-07-03 11:59:21 +00:00
|
|
|
/* Test for NULL */
|
|
|
|
if(tempbuffer == NULL) {
|
|
|
|
*pErrorCode = U_MEMORY_ALLOCATION_ERROR;
|
|
|
|
return 0;
|
|
|
|
}
|
2002-06-29 09:31:05 +00:00
|
|
|
|
2001-03-13 00:34:30 +00:00
|
|
|
uprv_memset(tempbuffer, 0, (sourceLength+1)*U_SIZEOF_UCHAR);
|
|
|
|
|
|
|
|
while(dest[inpsize-1] == 0x0020) {
|
|
|
|
countr++;
|
|
|
|
inpsize--;
|
|
|
|
}
|
|
|
|
|
|
|
|
i = sourceLength - countr - 1;
|
|
|
|
j = sourceLength - 1;
|
|
|
|
|
|
|
|
while(i >= 0 && j >= 0) {
|
|
|
|
if( countr>0 && isLamAlefChar(dest[i]) ) {
|
|
|
|
tempbuffer[j] = 0x0644;
|
|
|
|
tempbuffer[j-1] = convertLamAlef[ dest[i] - 0xFEF5 ];
|
|
|
|
j--;
|
|
|
|
countr--;
|
|
|
|
}
|
|
|
|
else {
|
|
|
|
if( countr == 0 && isLamAlefChar(dest[i]) )
|
|
|
|
*pErrorCode=U_NO_SPACE_AVAILABLE;
|
|
|
|
tempbuffer[j] = dest[i];
|
|
|
|
}
|
|
|
|
i--;
|
|
|
|
j--;
|
|
|
|
}
|
|
|
|
|
|
|
|
if(countr > 0) {
|
|
|
|
uprv_memcpy(tempbuffer, tempbuffer+countr, sourceLength*U_SIZEOF_UCHAR);
|
|
|
|
if(u_strlen(tempbuffer) < sourceLength) {
|
|
|
|
for(i=sourceLength-1;i>=sourceLength-countr;i--)
|
|
|
|
tempbuffer[i] = 0x0020;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
uprv_memcpy(dest, tempbuffer, sourceLength*U_SIZEOF_UCHAR);
|
|
|
|
|
|
|
|
destSize = sourceLength;
|
|
|
|
break;
|
|
|
|
|
|
|
|
default :
|
|
|
|
/* will never occur because of validity checks */
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
if(tempbuffer)
|
|
|
|
uprv_free(tempbuffer);
|
|
|
|
|
|
|
|
return destSize;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
*Name : shapeUnicode
|
|
|
|
*Function : Converts an Arabic Unicode buffer in 06xx Range into a shaped
|
|
|
|
* arabic Unicode buffer in FExx Range
|
|
|
|
*/
|
2001-03-22 01:26:48 +00:00
|
|
|
static int32_t
|
2001-03-13 00:34:30 +00:00
|
|
|
shapeUnicode(UChar *dest, int32_t sourceLength,
|
|
|
|
int32_t destSize,uint32_t options,
|
|
|
|
UErrorCode *pErrorCode,
|
|
|
|
int tashkeelFlag) {
|
|
|
|
|
|
|
|
int32_t i, iend;
|
|
|
|
int32_t step;
|
|
|
|
int32_t prevPos, lastPos,Nx, Nw;
|
|
|
|
unsigned int Shape;
|
|
|
|
int32_t flag;
|
|
|
|
int32_t lamalef_found = 0;
|
|
|
|
UChar prevLink = 0, lastLink = 0, currLink, nextLink = 0;
|
|
|
|
UChar wLamalef;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Converts the input buffer from FExx Range into 06xx Range
|
|
|
|
* to make sure that all characters are in the 06xx range
|
|
|
|
* even the lamalef is converted to the special region in
|
|
|
|
* the 06xx range
|
|
|
|
*/
|
|
|
|
for (i = 0; i < sourceLength; i++) {
|
|
|
|
UChar inputChar = dest[i];
|
|
|
|
if ( (inputChar >= 0xFE70) && (inputChar <= 0xFEFC)) {
|
|
|
|
dest[i] = convertFEto06 [ (inputChar - 0xFE70) ] ;
|
|
|
|
} else {
|
|
|
|
dest[i] = inputChar ;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* sets the index to the end of the buffer, together with the step point to -1 */
|
|
|
|
i = sourceLength - 1;
|
|
|
|
iend = -1;
|
|
|
|
step = -1;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* This function resolves the link between the characters .
|
|
|
|
* Arabic characters have four forms :
|
|
|
|
* Isolated Form, Initial Form, Middle Form and Final Form
|
|
|
|
*/
|
|
|
|
currLink = getLink(dest[i]);
|
|
|
|
|
|
|
|
prevPos = i;
|
|
|
|
lastPos = i;
|
|
|
|
Nx = -2, Nw = 0;
|
|
|
|
|
|
|
|
while (i != iend) {
|
|
|
|
/* If high byte of currLink > 0 then more than one shape */
|
|
|
|
if ((currLink & 0xFF00) > 0 || isTashkeelChar(dest[i])) {
|
|
|
|
Nw = i + step;
|
|
|
|
while (Nx < 0) { /* we need to know about next char */
|
|
|
|
if(Nw == iend) {
|
|
|
|
nextLink = 0;
|
|
|
|
Nx = 3000;
|
|
|
|
} else {
|
|
|
|
nextLink = getLink(dest[Nw]);
|
|
|
|
if((nextLink & IRRELEVANT) == 0) {
|
|
|
|
Nx = Nw;
|
|
|
|
} else {
|
|
|
|
Nw = Nw + step;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
if ( ((currLink & ALEFTYPE) > 0) && ((lastLink & LAMTYPE) > 0) ) {
|
|
|
|
lamalef_found = 1;
|
|
|
|
wLamalef = changeLamAlef(dest[i]); /*get from 0x065C-0x065f */
|
|
|
|
if ( wLamalef != 0) {
|
|
|
|
dest[i] = 0xFFFF; /* The default case is to drop the Alef and replace */
|
|
|
|
dest[lastPos] =wLamalef; /* it by 0xFFFF which is the last character in the */
|
|
|
|
i=lastPos; /* unicode private use area, this is done to make */
|
|
|
|
} /* sure that removeLamAlefSpaces() handles only the */
|
|
|
|
lastLink = prevLink; /* spaces generated during lamalef generation. */
|
|
|
|
currLink = getLink(wLamalef); /* 0xFFFF is added here and is replaced by spaces */
|
|
|
|
} /* in removeLamAlefSpaces() */
|
|
|
|
/*
|
|
|
|
* get the proper shape according to link ability of neighbors
|
|
|
|
* and of character; depends on the order of the shapes
|
|
|
|
* (isolated, initial, middle, final) in the compatibility area
|
|
|
|
*/
|
|
|
|
flag = specialChar(dest[i]);
|
|
|
|
|
|
|
|
Shape = shapeTable[nextLink & (LINKR + LINKL)]
|
|
|
|
[lastLink & (LINKR + LINKL)]
|
|
|
|
[currLink & (LINKR + LINKL)];
|
|
|
|
|
|
|
|
if (flag == 1) {
|
|
|
|
Shape = (Shape == 1 || Shape == 3) ? 1 : 0;
|
|
|
|
}
|
|
|
|
else
|
|
|
|
if(flag == 2) {
|
|
|
|
if( (lastLink & LINKL) && (nextLink & LINKR) && (tashkeelFlag == 1) &&
|
|
|
|
dest[i] != 0x064C && dest[i] != 0x064D ) {
|
|
|
|
Shape = 1;
|
|
|
|
if( (nextLink&ALEFTYPE) == ALEFTYPE && (lastLink&LAMTYPE) == LAMTYPE )
|
|
|
|
Shape = 0;
|
|
|
|
}
|
|
|
|
else {
|
|
|
|
Shape = 0;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
if(flag == 2) {
|
|
|
|
dest[i] = 0xFE70 + IrrelevantPos[(dest[i] - 0x064B)] + Shape;
|
|
|
|
}
|
|
|
|
else
|
|
|
|
dest[i] = (UChar)(0xFE70 + (currLink >> 8) + Shape);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* move one notch forward */
|
|
|
|
if ((currLink & IRRELEVANT) == 0) {
|
|
|
|
prevLink = lastLink;
|
|
|
|
lastLink = currLink;
|
|
|
|
prevPos = lastPos;
|
|
|
|
lastPos = i;
|
|
|
|
}
|
|
|
|
|
|
|
|
i = i + step;
|
|
|
|
if (i == Nx) {
|
|
|
|
currLink = nextLink;
|
|
|
|
Nx = -2;
|
|
|
|
}
|
|
|
|
else if(i != iend) {
|
|
|
|
currLink = getLink(dest[i]);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* If there is lamalef in the buffer call expandLamAlef */
|
|
|
|
if(lamalef_found != 0)
|
|
|
|
destSize = removeLamAlefSpaces(dest,sourceLength,destSize,options,pErrorCode);
|
|
|
|
else
|
|
|
|
destSize = sourceLength;
|
|
|
|
|
|
|
|
return destSize;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
*Name : deShapeUnicode
|
|
|
|
*Function : Converts an Arabic Unicode buffer in FExx Range into unshaped
|
|
|
|
* arabic Unicode buffer in 06xx Range
|
|
|
|
*/
|
2001-03-22 01:26:48 +00:00
|
|
|
static int32_t
|
2001-03-13 00:34:30 +00:00
|
|
|
deShapeUnicode(UChar *dest, int32_t sourceLength,
|
|
|
|
int32_t destSize,uint32_t options,
|
|
|
|
UErrorCode *pErrorCode) {
|
|
|
|
int32_t i = 0;
|
|
|
|
int32_t lamalef_found = 0;
|
|
|
|
|
|
|
|
/*
|
|
|
|
*This for loop changes the buffer from the Unicode FE range to
|
|
|
|
*the Unicode 06 range
|
|
|
|
*/
|
|
|
|
for(i = 0; i < sourceLength; i++) {
|
|
|
|
UChar inputChar = dest[i];
|
|
|
|
if (( inputChar >= 0xFE70) && (inputChar <= 0xFEF4 )) { /* FExx Arabic range */
|
|
|
|
dest[i] = convertFEto06 [ (inputChar - 0xFE70) ] ;
|
|
|
|
} else {
|
|
|
|
dest[i] = inputChar ;
|
|
|
|
}
|
|
|
|
if( isLamAlefChar(dest[i]) )
|
|
|
|
lamalef_found = 1;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* If there is lamalef in the buffer call expandLamAlef */
|
|
|
|
if(lamalef_found != 0)
|
|
|
|
destSize = expandLamAlef(dest,sourceLength,destSize,options,pErrorCode);
|
|
|
|
else
|
|
|
|
destSize = sourceLength;
|
|
|
|
|
|
|
|
return destSize;
|
|
|
|
}
|
|
|
|
|
2000-06-30 00:29:46 +00:00
|
|
|
U_CAPI int32_t U_EXPORT2
|
|
|
|
u_shapeArabic(const UChar *source, int32_t sourceLength,
|
2002-08-01 20:25:10 +00:00
|
|
|
UChar *dest, int32_t destCapacity,
|
2000-06-30 00:29:46 +00:00
|
|
|
uint32_t options,
|
|
|
|
UErrorCode *pErrorCode) {
|
2001-03-13 00:34:30 +00:00
|
|
|
|
2002-08-01 20:25:10 +00:00
|
|
|
int32_t destLength;
|
2001-03-13 00:34:30 +00:00
|
|
|
|
2000-08-12 01:01:03 +00:00
|
|
|
/* usual error checking */
|
2000-06-30 00:29:46 +00:00
|
|
|
if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2000-08-14 23:05:10 +00:00
|
|
|
/* make sure that no reserved options values are used; allow dest==NULL only for preflighting */
|
|
|
|
if( source==NULL || sourceLength<-1 ||
|
2002-08-01 20:25:10 +00:00
|
|
|
(dest==NULL && destCapacity!=0) || destCapacity<0 ||
|
2000-08-12 01:01:03 +00:00
|
|
|
options>=U_SHAPE_DIGIT_TYPE_RESERVED ||
|
|
|
|
(options&U_SHAPE_DIGITS_MASK)>=U_SHAPE_DIGITS_RESERVED
|
|
|
|
) {
|
|
|
|
*pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* determine the source length */
|
|
|
|
if(sourceLength==-1) {
|
|
|
|
sourceLength=u_strlen(source);
|
|
|
|
}
|
|
|
|
if(sourceLength==0) {
|
2002-08-01 20:25:10 +00:00
|
|
|
return u_terminateUChars(dest, destCapacity, 0, pErrorCode);
|
2000-08-12 01:01:03 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/* check that source and destination do not overlap */
|
2000-08-14 23:05:10 +00:00
|
|
|
if( dest!=NULL &&
|
2000-10-12 19:23:35 +00:00
|
|
|
((source<=dest && dest<source+sourceLength) ||
|
2002-08-01 20:25:10 +00:00
|
|
|
(dest<=source && source<dest+destCapacity))
|
2000-08-12 01:01:03 +00:00
|
|
|
) {
|
|
|
|
*pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
if((options&U_SHAPE_LETTERS_MASK)!=U_SHAPE_LETTERS_NOOP) {
|
2001-10-05 23:25:37 +00:00
|
|
|
UChar buffer[300];
|
|
|
|
UChar *tempbuffer;
|
|
|
|
int32_t outputSize, spacesCountl=0, spacesCountr=0;
|
|
|
|
|
|
|
|
/* calculate destination size */
|
|
|
|
/* TODO: do we ever need to do this pure preflighting? */
|
|
|
|
if((options&U_SHAPE_LENGTH_MASK)==U_SHAPE_LENGTH_GROW_SHRINK) {
|
2002-08-01 20:25:10 +00:00
|
|
|
outputSize=calculateSize(source,sourceLength,destCapacity,options);
|
2001-10-05 23:25:37 +00:00
|
|
|
} else {
|
|
|
|
outputSize=sourceLength;
|
|
|
|
}
|
2002-08-01 20:25:10 +00:00
|
|
|
if(outputSize>destCapacity) {
|
2001-10-05 23:25:37 +00:00
|
|
|
*pErrorCode=U_BUFFER_OVERFLOW_ERROR;
|
|
|
|
return outputSize;
|
2001-03-13 00:34:30 +00:00
|
|
|
}
|
|
|
|
|
2002-08-01 20:25:10 +00:00
|
|
|
/*
|
|
|
|
* need a temporary buffer of size max(outputSize, sourceLength)
|
|
|
|
* because at first we copy source->temp
|
|
|
|
*/
|
|
|
|
if(sourceLength>outputSize) {
|
|
|
|
outputSize=sourceLength;
|
|
|
|
}
|
|
|
|
|
2001-03-13 00:34:30 +00:00
|
|
|
/* Start of Arabic letter shaping part */
|
2001-10-05 23:25:37 +00:00
|
|
|
if(outputSize<=sizeof(buffer)/U_SIZEOF_UCHAR) {
|
|
|
|
outputSize=sizeof(buffer)/U_SIZEOF_UCHAR;
|
|
|
|
tempbuffer=buffer;
|
|
|
|
} else {
|
|
|
|
tempbuffer = (UChar *)uprv_malloc(outputSize*U_SIZEOF_UCHAR);
|
2002-06-29 09:31:05 +00:00
|
|
|
|
2002-07-03 11:59:21 +00:00
|
|
|
/*Test for NULL*/
|
|
|
|
if(tempbuffer == NULL) {
|
|
|
|
*pErrorCode = U_MEMORY_ALLOCATION_ERROR;
|
|
|
|
return 0;
|
|
|
|
}
|
2001-10-05 23:25:37 +00:00
|
|
|
}
|
2001-03-13 00:34:30 +00:00
|
|
|
uprv_memcpy(tempbuffer, source, sourceLength*U_SIZEOF_UCHAR);
|
2001-10-05 23:25:37 +00:00
|
|
|
if(sourceLength<outputSize) {
|
|
|
|
uprv_memset(tempbuffer+sourceLength, 0, (outputSize-sourceLength)*U_SIZEOF_UCHAR);
|
|
|
|
}
|
2001-03-13 00:34:30 +00:00
|
|
|
|
|
|
|
if((options&U_SHAPE_TEXT_DIRECTION_MASK) == U_SHAPE_TEXT_DIRECTION_LOGICAL) {
|
|
|
|
countSpaces(tempbuffer,sourceLength,options,&spacesCountl,&spacesCountr);
|
|
|
|
invertBuffer(tempbuffer,sourceLength,options,&spacesCountl,&spacesCountr);
|
|
|
|
}
|
|
|
|
|
|
|
|
switch(options&U_SHAPE_LETTERS_MASK) {
|
|
|
|
case U_SHAPE_LETTERS_SHAPE :
|
|
|
|
/* Call the shaping function with tashkeel flag == 1 */
|
2002-08-01 20:25:10 +00:00
|
|
|
destLength = shapeUnicode(tempbuffer,sourceLength,destCapacity,options,pErrorCode,1);
|
2001-03-13 00:34:30 +00:00
|
|
|
break;
|
2001-10-13 00:06:29 +00:00
|
|
|
case U_SHAPE_LETTERS_SHAPE_TASHKEEL_ISOLATED :
|
2001-03-13 00:34:30 +00:00
|
|
|
/* Call the shaping function with tashkeel flag == 0 */
|
2002-08-01 20:25:10 +00:00
|
|
|
destLength = shapeUnicode(tempbuffer,sourceLength,destCapacity,options,pErrorCode,0);
|
2001-03-13 00:34:30 +00:00
|
|
|
break;
|
|
|
|
case U_SHAPE_LETTERS_UNSHAPE :
|
|
|
|
/* Call the deshaping function */
|
2002-08-01 20:25:10 +00:00
|
|
|
destLength = deShapeUnicode(tempbuffer,sourceLength,destCapacity,options,pErrorCode);
|
2001-03-13 00:34:30 +00:00
|
|
|
break;
|
|
|
|
default :
|
|
|
|
/* will never occur because of validity checks above */
|
2002-08-01 20:25:10 +00:00
|
|
|
destLength = 0;
|
2001-03-13 00:34:30 +00:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
2002-08-01 20:25:10 +00:00
|
|
|
/*
|
|
|
|
* TODO: (markus 2002aug01)
|
|
|
|
* For as long as we always preflight the outputSize above
|
|
|
|
* we should U_ASSERT(outputSize==destLength)
|
|
|
|
* except for the adjustment above before the tempbuffer allocation
|
|
|
|
*/
|
|
|
|
|
2001-03-13 00:34:30 +00:00
|
|
|
if((options&U_SHAPE_TEXT_DIRECTION_MASK) == U_SHAPE_TEXT_DIRECTION_LOGICAL) {
|
2002-08-01 20:25:10 +00:00
|
|
|
countSpaces(tempbuffer,destLength,options,&spacesCountl,&spacesCountr);
|
|
|
|
invertBuffer(tempbuffer,destLength,options,&spacesCountl,&spacesCountr);
|
2001-03-13 00:34:30 +00:00
|
|
|
}
|
2002-08-01 20:25:10 +00:00
|
|
|
uprv_memcpy(dest, tempbuffer, uprv_min(destLength, destCapacity)*U_SIZEOF_UCHAR);
|
2001-03-13 00:34:30 +00:00
|
|
|
|
2001-10-05 23:25:37 +00:00
|
|
|
if(tempbuffer!=buffer) {
|
2001-03-13 00:34:30 +00:00
|
|
|
uprv_free(tempbuffer);
|
|
|
|
}
|
2001-10-05 23:25:37 +00:00
|
|
|
|
2002-08-01 20:25:10 +00:00
|
|
|
if(destLength>destCapacity) {
|
2001-10-05 23:25:37 +00:00
|
|
|
*pErrorCode=U_BUFFER_OVERFLOW_ERROR;
|
2002-08-01 20:25:10 +00:00
|
|
|
return destLength;
|
2001-10-05 23:25:37 +00:00
|
|
|
}
|
|
|
|
|
2001-03-13 00:34:30 +00:00
|
|
|
/* End of Arabic letter shaping part */
|
2000-08-12 01:01:03 +00:00
|
|
|
} else {
|
|
|
|
/*
|
|
|
|
* No letter shaping:
|
|
|
|
* just make sure the destination is large enough and copy the string.
|
|
|
|
*/
|
2002-08-01 20:25:10 +00:00
|
|
|
if(destCapacity<sourceLength) {
|
2000-08-14 23:05:10 +00:00
|
|
|
/* this catches preflighting, too */
|
2000-08-12 01:01:03 +00:00
|
|
|
*pErrorCode=U_BUFFER_OVERFLOW_ERROR;
|
|
|
|
return sourceLength;
|
|
|
|
}
|
|
|
|
uprv_memcpy(dest, source, sourceLength*U_SIZEOF_UCHAR);
|
2002-08-01 20:25:10 +00:00
|
|
|
destLength=sourceLength;
|
2000-08-12 01:01:03 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Perform number shaping.
|
|
|
|
* With UTF-16 or UTF-32, the length of the string is constant.
|
|
|
|
* The easiest way to do this is to operate on the destination and
|
|
|
|
* "shape" the digits in-place.
|
|
|
|
*/
|
|
|
|
if((options&U_SHAPE_DIGITS_MASK)!=U_SHAPE_DIGITS_NOOP) {
|
|
|
|
UChar digitBase;
|
|
|
|
int32_t i;
|
|
|
|
|
|
|
|
/* select the requested digit group */
|
|
|
|
switch(options&U_SHAPE_DIGIT_TYPE_MASK) {
|
|
|
|
case U_SHAPE_DIGIT_TYPE_AN:
|
|
|
|
digitBase=0x660; /* Unicode: "Arabic-Indic digits" */
|
|
|
|
break;
|
|
|
|
case U_SHAPE_DIGIT_TYPE_AN_EXTENDED:
|
|
|
|
digitBase=0x6f0; /* Unicode: "Eastern Arabic-Indic digits (Persian and Urdu)" */
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
/* will never occur because of validity checks above */
|
2000-08-21 21:40:35 +00:00
|
|
|
digitBase=0;
|
2000-08-12 01:01:03 +00:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* perform the requested operation */
|
|
|
|
switch(options&U_SHAPE_DIGITS_MASK) {
|
|
|
|
case U_SHAPE_DIGITS_EN2AN:
|
|
|
|
/* add (digitBase-'0') to each European (ASCII) digit code point */
|
|
|
|
digitBase-=0x30;
|
2002-08-01 20:25:10 +00:00
|
|
|
for(i=0; i<destLength; ++i) {
|
2000-08-12 01:01:03 +00:00
|
|
|
if(((uint32_t)dest[i]-0x30)<10) {
|
|
|
|
dest[i]+=digitBase;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
case U_SHAPE_DIGITS_AN2EN:
|
|
|
|
/* subtract (digitBase-'0') from each Arabic digit code point */
|
2002-08-01 20:25:10 +00:00
|
|
|
for(i=0; i<destLength; ++i) {
|
2000-08-12 01:01:03 +00:00
|
|
|
if(((uint32_t)dest[i]-(uint32_t)digitBase)<10) {
|
|
|
|
dest[i]-=digitBase-0x30;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
case U_SHAPE_DIGITS_ALEN2AN_INIT_LR:
|
2002-08-01 20:25:10 +00:00
|
|
|
_shapeToArabicDigitsWithContext(dest, destLength,
|
2000-08-12 01:01:03 +00:00
|
|
|
digitBase,
|
|
|
|
(UBool)((options&U_SHAPE_TEXT_DIRECTION_MASK)==U_SHAPE_TEXT_DIRECTION_LOGICAL),
|
|
|
|
FALSE);
|
|
|
|
break;
|
|
|
|
case U_SHAPE_DIGITS_ALEN2AN_INIT_AL:
|
2002-08-01 20:25:10 +00:00
|
|
|
_shapeToArabicDigitsWithContext(dest, destLength,
|
2000-08-12 01:01:03 +00:00
|
|
|
digitBase,
|
|
|
|
(UBool)((options&U_SHAPE_TEXT_DIRECTION_MASK)==U_SHAPE_TEXT_DIRECTION_LOGICAL),
|
|
|
|
TRUE);
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
/* will never occur because of validity checks above */
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2002-08-01 20:25:10 +00:00
|
|
|
return u_terminateUChars(dest, destCapacity, destLength, pErrorCode);
|
2000-06-30 00:29:46 +00:00
|
|
|
}
|