2001-03-17 00:46:46 +00:00
|
|
|
/*
|
|
|
|
*******************************************************************************
|
|
|
|
*
|
2004-05-19 20:42:44 +00:00
|
|
|
* Copyright (C) 2001-2004, International Business Machines
|
2001-03-17 00:46:46 +00:00
|
|
|
* Corporation and others. All Rights Reserved.
|
|
|
|
*
|
|
|
|
*******************************************************************************
|
2001-12-28 20:53:44 +00:00
|
|
|
* file name: ucol_bld.cpp
|
2001-03-17 00:46:46 +00:00
|
|
|
* encoding: US-ASCII
|
|
|
|
* tab size: 8 (not used)
|
|
|
|
* indentation:4
|
|
|
|
*
|
|
|
|
* created 02/22/2001
|
|
|
|
* created by: Vladimir Weinstein
|
|
|
|
*
|
|
|
|
* This module builds a collator based on the rule set.
|
|
|
|
*
|
|
|
|
*/
|
|
|
|
|
2002-09-20 01:54:48 +00:00
|
|
|
#include "unicode/utypes.h"
|
|
|
|
|
|
|
|
#if !UCONFIG_NO_COLLATION
|
|
|
|
|
2002-08-21 19:12:24 +00:00
|
|
|
#include "unicode/ucoleitr.h"
|
|
|
|
#include "unicode/uchar.h"
|
2001-04-19 17:08:07 +00:00
|
|
|
#include "ucol_bld.h"
|
2001-08-30 02:59:19 +00:00
|
|
|
#include "ucln_in.h"
|
2002-01-23 17:17:06 +00:00
|
|
|
#include "umutex.h"
|
2002-10-16 22:34:16 +00:00
|
|
|
#include "unicode/uniset.h"
|
2001-10-20 01:09:31 +00:00
|
|
|
|
2003-09-24 00:17:15 +00:00
|
|
|
static const InverseUCATableHeader* _staticInvUCA = NULL;
|
2001-08-30 02:59:19 +00:00
|
|
|
static UDataMemory* invUCA_DATA_MEM = NULL;
|
2001-03-08 17:40:42 +00:00
|
|
|
|
2001-07-02 22:53:22 +00:00
|
|
|
U_CDECL_BEGIN
|
2001-03-08 17:40:42 +00:00
|
|
|
static UBool U_CALLCONV
|
2001-06-06 20:37:10 +00:00
|
|
|
isAcceptableInvUCA(void * /*context*/,
|
|
|
|
const char * /*type*/, const char * /*name*/,
|
2001-03-08 17:40:42 +00:00
|
|
|
const UDataInfo *pInfo){
|
|
|
|
/* context, type & name are intentionally not used */
|
|
|
|
if( pInfo->size>=20 &&
|
|
|
|
pInfo->isBigEndian==U_IS_BIG_ENDIAN &&
|
|
|
|
pInfo->charsetFamily==U_CHARSET_FAMILY &&
|
2001-10-23 21:48:14 +00:00
|
|
|
pInfo->dataFormat[0]==invUcaDataInfo.dataFormat[0] && /* dataFormat="InvC" */
|
|
|
|
pInfo->dataFormat[1]==invUcaDataInfo.dataFormat[1] &&
|
|
|
|
pInfo->dataFormat[2]==invUcaDataInfo.dataFormat[2] &&
|
|
|
|
pInfo->dataFormat[3]==invUcaDataInfo.dataFormat[3] &&
|
2002-09-17 04:53:35 +00:00
|
|
|
pInfo->formatVersion[0]==invUcaDataInfo.formatVersion[0] &&
|
2003-04-24 07:00:27 +00:00
|
|
|
pInfo->formatVersion[1]>=invUcaDataInfo.formatVersion[1] //&&
|
2002-11-14 21:00:19 +00:00
|
|
|
//pInfo->formatVersion[1]==invUcaDataInfo.formatVersion[1] &&
|
|
|
|
//pInfo->formatVersion[2]==invUcaDataInfo.formatVersion[2] &&
|
|
|
|
//pInfo->formatVersion[3]==invUcaDataInfo.formatVersion[3] &&
|
2003-04-24 07:00:27 +00:00
|
|
|
) {
|
|
|
|
UVersionInfo UCDVersion;
|
|
|
|
u_getUnicodeVersion(UCDVersion);
|
|
|
|
if(pInfo->dataVersion[0]==UCDVersion[0] &&
|
2003-05-28 23:04:30 +00:00
|
|
|
pInfo->dataVersion[1]==UCDVersion[1]) {
|
2002-11-14 21:00:19 +00:00
|
|
|
//pInfo->dataVersion[1]==invUcaDataInfo.dataVersion[1] &&
|
|
|
|
//pInfo->dataVersion[2]==invUcaDataInfo.dataVersion[2] &&
|
|
|
|
//pInfo->dataVersion[3]==invUcaDataInfo.dataVersion[3]) {
|
2003-04-24 07:00:27 +00:00
|
|
|
return TRUE;
|
|
|
|
} else {
|
|
|
|
return FALSE;
|
|
|
|
}
|
2001-03-08 17:40:42 +00:00
|
|
|
} else {
|
|
|
|
return FALSE;
|
|
|
|
}
|
|
|
|
}
|
2001-07-02 22:53:22 +00:00
|
|
|
U_CDECL_END
|
2001-03-08 17:40:42 +00:00
|
|
|
|
2001-10-20 01:09:31 +00:00
|
|
|
static
|
2003-09-24 00:17:15 +00:00
|
|
|
int32_t ucol_inv_findCE(const UColTokenParser *src, uint32_t CE, uint32_t SecondCE) {
|
|
|
|
uint32_t bottom = 0, top = src->invUCA->tableSize;
|
2001-03-08 17:40:42 +00:00
|
|
|
uint32_t i = 0;
|
|
|
|
uint32_t first = 0, second = 0;
|
2003-09-24 00:17:15 +00:00
|
|
|
uint32_t *CETable = (uint32_t *)((uint8_t *)src->invUCA+src->invUCA->table);
|
2001-03-08 17:40:42 +00:00
|
|
|
|
|
|
|
while(bottom < top-1) {
|
|
|
|
i = (top+bottom)/2;
|
|
|
|
first = *(CETable+3*i);
|
|
|
|
second = *(CETable+3*i+1);
|
|
|
|
if(first > CE) {
|
|
|
|
top = i;
|
|
|
|
} else if(first < CE) {
|
|
|
|
bottom = i;
|
|
|
|
} else {
|
|
|
|
if(second > SecondCE) {
|
|
|
|
top = i;
|
|
|
|
} else if(second < SecondCE) {
|
|
|
|
bottom = i;
|
|
|
|
} else {
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2002-03-26 16:46:17 +00:00
|
|
|
/* weiv: */
|
|
|
|
/* in searching for elements, I have removed the failure */
|
|
|
|
/* The reason for this is that the builder does not rely */
|
|
|
|
/* on search mechanism telling it that it didn't find an */
|
|
|
|
/* element. However, indirect positioning relies on being */
|
|
|
|
/* able to find the elements around any CE, even if it is */
|
|
|
|
/* not defined in the UCA. */
|
|
|
|
return i;
|
|
|
|
/*
|
2001-03-08 17:40:42 +00:00
|
|
|
if((first == CE && second == SecondCE)) {
|
|
|
|
return i;
|
|
|
|
} else {
|
|
|
|
return -1;
|
|
|
|
}
|
2002-03-26 16:46:17 +00:00
|
|
|
*/
|
2001-03-08 17:40:42 +00:00
|
|
|
}
|
|
|
|
|
2002-03-29 20:10:05 +00:00
|
|
|
static const uint32_t strengthMask[UCOL_CE_STRENGTH_LIMIT] = {
|
2001-03-08 17:40:42 +00:00
|
|
|
0xFFFF0000,
|
|
|
|
0xFFFFFF00,
|
|
|
|
0xFFFFFFFF
|
|
|
|
};
|
|
|
|
|
2003-09-24 00:17:15 +00:00
|
|
|
U_CAPI int32_t U_EXPORT2 ucol_inv_getNextCE(const UColTokenParser *src,
|
|
|
|
uint32_t CE, uint32_t contCE,
|
2001-03-22 23:48:08 +00:00
|
|
|
uint32_t *nextCE, uint32_t *nextContCE,
|
|
|
|
uint32_t strength) {
|
2003-09-24 00:17:15 +00:00
|
|
|
uint32_t *CETable = (uint32_t *)((uint8_t *)src->invUCA+src->invUCA->table);
|
2001-03-22 23:11:22 +00:00
|
|
|
int32_t iCE;
|
|
|
|
|
2003-09-24 00:17:15 +00:00
|
|
|
iCE = ucol_inv_findCE(src, CE, contCE);
|
2001-03-22 23:11:22 +00:00
|
|
|
|
|
|
|
if(iCE<0) {
|
2001-03-22 23:48:08 +00:00
|
|
|
*nextCE = UCOL_NOT_FOUND;
|
|
|
|
return -1;
|
2001-03-22 23:11:22 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
CE &= strengthMask[strength];
|
2001-03-22 23:48:08 +00:00
|
|
|
contCE &= strengthMask[strength];
|
2001-03-22 23:11:22 +00:00
|
|
|
|
2001-03-22 23:48:08 +00:00
|
|
|
*nextCE = CE;
|
|
|
|
*nextContCE = contCE;
|
2001-03-22 23:11:22 +00:00
|
|
|
|
2001-03-22 23:48:08 +00:00
|
|
|
while((*nextCE & strengthMask[strength]) == CE
|
|
|
|
&& (*nextContCE & strengthMask[strength]) == contCE) {
|
|
|
|
*nextCE = (*(CETable+3*(++iCE)));
|
|
|
|
*nextContCE = (*(CETable+3*(iCE)+1));
|
2001-03-22 23:11:22 +00:00
|
|
|
}
|
|
|
|
|
2001-03-22 23:48:08 +00:00
|
|
|
return iCE;
|
2001-03-22 23:11:22 +00:00
|
|
|
}
|
|
|
|
|
2003-09-24 00:17:15 +00:00
|
|
|
U_CAPI int32_t U_EXPORT2 ucol_inv_getPrevCE(const UColTokenParser *src,
|
|
|
|
uint32_t CE, uint32_t contCE,
|
2001-03-22 23:48:08 +00:00
|
|
|
uint32_t *prevCE, uint32_t *prevContCE,
|
|
|
|
uint32_t strength) {
|
2003-09-24 00:17:15 +00:00
|
|
|
uint32_t *CETable = (uint32_t *)((uint8_t *)src->invUCA+src->invUCA->table);
|
2001-03-22 23:11:22 +00:00
|
|
|
int32_t iCE;
|
|
|
|
|
2003-09-24 00:17:15 +00:00
|
|
|
iCE = ucol_inv_findCE(src, CE, contCE);
|
2001-03-22 23:11:22 +00:00
|
|
|
|
|
|
|
if(iCE<0) {
|
2001-03-22 23:48:08 +00:00
|
|
|
*prevCE = UCOL_NOT_FOUND;
|
|
|
|
return -1;
|
2001-03-22 23:11:22 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
CE &= strengthMask[strength];
|
2001-03-22 23:48:08 +00:00
|
|
|
contCE &= strengthMask[strength];
|
2001-03-22 23:11:22 +00:00
|
|
|
|
2001-03-22 23:48:08 +00:00
|
|
|
*prevCE = CE;
|
|
|
|
*prevContCE = contCE;
|
2001-03-22 23:11:22 +00:00
|
|
|
|
2001-03-22 23:48:08 +00:00
|
|
|
while((*prevCE & strengthMask[strength]) == CE
|
2002-03-26 16:46:17 +00:00
|
|
|
&& (*prevContCE & strengthMask[strength])== contCE
|
|
|
|
&& iCE > 0) { /* this condition should prevent falling off the edge of the world */
|
|
|
|
/* here, we end up in a singularity - zero */
|
2001-03-22 23:48:08 +00:00
|
|
|
*prevCE = (*(CETable+3*(--iCE)));
|
|
|
|
*prevContCE = (*(CETable+3*(iCE)+1));
|
2001-03-22 23:11:22 +00:00
|
|
|
}
|
|
|
|
|
2001-03-22 23:48:08 +00:00
|
|
|
return iCE;
|
2001-03-22 23:11:22 +00:00
|
|
|
}
|
|
|
|
|
2004-05-08 07:59:36 +00:00
|
|
|
U_CAPI uint32_t U_EXPORT2 ucol_getCEStrengthDifference(uint32_t CE, uint32_t contCE,
|
|
|
|
uint32_t prevCE, uint32_t prevContCE) {
|
|
|
|
uint32_t strength = UCOL_TERTIARY;
|
|
|
|
while(((prevCE & strengthMask[strength]) != (CE & strengthMask[strength])
|
|
|
|
|| (prevContCE & strengthMask[strength]) != (contCE & strengthMask[strength]))
|
|
|
|
&& strength) {
|
|
|
|
strength--;
|
|
|
|
}
|
|
|
|
return strength;
|
|
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
2001-10-22 05:30:22 +00:00
|
|
|
static
|
2003-09-24 00:17:15 +00:00
|
|
|
inline int32_t ucol_inv_getPrevious(UColTokenParser *src, UColTokListHeader *lh, uint32_t strength) {
|
2001-03-08 17:40:42 +00:00
|
|
|
|
|
|
|
uint32_t CE = lh->baseCE;
|
|
|
|
uint32_t SecondCE = lh->baseContCE;
|
|
|
|
|
2003-09-24 00:17:15 +00:00
|
|
|
uint32_t *CETable = (uint32_t *)((uint8_t *)src->invUCA+src->invUCA->table);
|
2001-03-08 17:40:42 +00:00
|
|
|
uint32_t previousCE, previousContCE;
|
|
|
|
int32_t iCE;
|
|
|
|
|
2003-09-24 00:17:15 +00:00
|
|
|
iCE = ucol_inv_findCE(src, CE, SecondCE);
|
2001-03-08 17:40:42 +00:00
|
|
|
|
|
|
|
if(iCE<0) {
|
|
|
|
return -1;
|
|
|
|
}
|
|
|
|
|
|
|
|
CE &= strengthMask[strength];
|
|
|
|
SecondCE &= strengthMask[strength];
|
|
|
|
|
|
|
|
previousCE = CE;
|
|
|
|
previousContCE = SecondCE;
|
|
|
|
|
|
|
|
while((previousCE & strengthMask[strength]) == CE && (previousContCE & strengthMask[strength])== SecondCE) {
|
|
|
|
previousCE = (*(CETable+3*(--iCE)));
|
|
|
|
previousContCE = (*(CETable+3*(iCE)+1));
|
|
|
|
}
|
|
|
|
lh->previousCE = previousCE;
|
|
|
|
lh->previousContCE = previousContCE;
|
|
|
|
|
|
|
|
return iCE;
|
|
|
|
}
|
|
|
|
|
2001-10-22 05:30:22 +00:00
|
|
|
static
|
2003-09-24 00:17:15 +00:00
|
|
|
inline int32_t ucol_inv_getNext(UColTokenParser *src, UColTokListHeader *lh, uint32_t strength) {
|
2001-03-08 17:40:42 +00:00
|
|
|
uint32_t CE = lh->baseCE;
|
|
|
|
uint32_t SecondCE = lh->baseContCE;
|
|
|
|
|
2003-09-24 00:17:15 +00:00
|
|
|
uint32_t *CETable = (uint32_t *)((uint8_t *)src->invUCA+src->invUCA->table);
|
2001-03-08 17:40:42 +00:00
|
|
|
uint32_t nextCE, nextContCE;
|
|
|
|
int32_t iCE;
|
|
|
|
|
2003-09-24 00:17:15 +00:00
|
|
|
iCE = ucol_inv_findCE(src, CE, SecondCE);
|
2001-03-08 17:40:42 +00:00
|
|
|
|
|
|
|
if(iCE<0) {
|
|
|
|
return -1;
|
|
|
|
}
|
|
|
|
|
|
|
|
CE &= strengthMask[strength];
|
|
|
|
SecondCE &= strengthMask[strength];
|
|
|
|
|
|
|
|
nextCE = CE;
|
|
|
|
nextContCE = SecondCE;
|
|
|
|
|
|
|
|
while((nextCE & strengthMask[strength]) == CE
|
|
|
|
&& (nextContCE & strengthMask[strength]) == SecondCE) {
|
|
|
|
nextCE = (*(CETable+3*(++iCE)));
|
|
|
|
nextContCE = (*(CETable+3*(iCE)+1));
|
|
|
|
}
|
|
|
|
|
|
|
|
lh->nextCE = nextCE;
|
|
|
|
lh->nextContCE = nextContCE;
|
|
|
|
|
|
|
|
return iCE;
|
|
|
|
}
|
|
|
|
|
2002-06-13 18:29:49 +00:00
|
|
|
U_CFUNC void ucol_inv_getGapPositions(UColTokenParser *src, UColTokListHeader *lh, UErrorCode *status) {
|
2001-03-08 17:40:42 +00:00
|
|
|
/* reset all the gaps */
|
|
|
|
int32_t i = 0;
|
2003-09-24 00:17:15 +00:00
|
|
|
uint32_t *CETable = (uint32_t *)((uint8_t *)src->invUCA+src->invUCA->table);
|
2001-03-08 17:40:42 +00:00
|
|
|
uint32_t st = 0;
|
|
|
|
uint32_t t1, t2;
|
|
|
|
int32_t pos;
|
|
|
|
|
2001-06-06 20:20:43 +00:00
|
|
|
UColToken *tok = lh->first;
|
2001-03-08 17:40:42 +00:00
|
|
|
uint32_t tokStrength = tok->strength;
|
|
|
|
|
|
|
|
for(i = 0; i<3; i++) {
|
|
|
|
lh->gapsHi[3*i] = 0;
|
|
|
|
lh->gapsHi[3*i+1] = 0;
|
|
|
|
lh->gapsHi[3*i+2] = 0;
|
|
|
|
lh->gapsLo[3*i] = 0;
|
|
|
|
lh->gapsLo[3*i+1] = 0;
|
|
|
|
lh->gapsLo[3*i+2] = 0;
|
|
|
|
lh->numStr[i] = 0;
|
|
|
|
lh->fStrToken[i] = NULL;
|
|
|
|
lh->lStrToken[i] = NULL;
|
|
|
|
lh->pos[i] = -1;
|
|
|
|
}
|
|
|
|
|
2002-06-13 18:29:49 +00:00
|
|
|
UCAConstants *consts = (UCAConstants *)((uint8_t *)src->UCA->image + src->UCA->image->UCAConsts);
|
|
|
|
|
2004-01-27 22:15:07 +00:00
|
|
|
if((lh->baseCE & 0xFF000000)>= (consts->UCA_PRIMARY_IMPLICIT_MIN<<24) && (lh->baseCE & 0xFF000000) <= (consts->UCA_PRIMARY_IMPLICIT_MAX<<24) ) { /* implicits - */
|
2002-06-13 18:29:49 +00:00
|
|
|
//if(lh->baseCE >= PRIMARY_IMPLICIT_MIN && lh->baseCE < PRIMARY_IMPLICIT_MAX ) { /* implicits - */
|
2001-04-23 01:52:45 +00:00
|
|
|
lh->pos[0] = 0;
|
|
|
|
t1 = lh->baseCE;
|
|
|
|
t2 = lh->baseContCE;
|
2001-05-15 17:39:41 +00:00
|
|
|
lh->gapsLo[0] = (t1 & UCOL_PRIMARYMASK) | (t2 & UCOL_PRIMARYMASK) >> 16;
|
|
|
|
lh->gapsLo[1] = (t1 & UCOL_SECONDARYMASK) << 16 | (t2 & UCOL_SECONDARYMASK) << 8;
|
|
|
|
lh->gapsLo[2] = (UCOL_TERTIARYORDER(t1)) << 24 | (UCOL_TERTIARYORDER(t2)) << 16;
|
2004-01-16 23:44:58 +00:00
|
|
|
uint32_t primaryCE = t1 & UCOL_PRIMARYMASK | (t2 & UCOL_PRIMARYMASK) >> 16;
|
|
|
|
primaryCE = uprv_uca_getImplicitFromRaw(uprv_uca_getRawFromImplicit(primaryCE)+1);
|
|
|
|
|
|
|
|
t1 = primaryCE & UCOL_PRIMARYMASK | 0x0505;
|
|
|
|
t2 = (primaryCE << 16) & UCOL_PRIMARYMASK | UCOL_CONTINUATION_MARKER;
|
|
|
|
|
2001-05-15 17:39:41 +00:00
|
|
|
lh->gapsHi[0] = (t1 & UCOL_PRIMARYMASK) | (t2 & UCOL_PRIMARYMASK) >> 16;
|
|
|
|
lh->gapsHi[1] = (t1 & UCOL_SECONDARYMASK) << 16 | (t2 & UCOL_SECONDARYMASK) << 8;
|
|
|
|
lh->gapsHi[2] = (UCOL_TERTIARYORDER(t1)) << 24 | (UCOL_TERTIARYORDER(t2)) << 16;
|
2002-03-15 23:41:56 +00:00
|
|
|
} else if(lh->indirect == TRUE && lh->nextCE != 0) {
|
|
|
|
//} else if(lh->baseCE == UCOL_RESET_TOP_VALUE && lh->baseContCE == 0) {
|
2001-03-08 17:40:42 +00:00
|
|
|
lh->pos[0] = 0;
|
2002-03-15 23:41:56 +00:00
|
|
|
t1 = lh->baseCE;
|
|
|
|
t2 = lh->baseContCE;
|
|
|
|
lh->gapsLo[0] = (t1 & UCOL_PRIMARYMASK) | (t2 & UCOL_PRIMARYMASK) >> 16;
|
|
|
|
lh->gapsLo[1] = (t1 & UCOL_SECONDARYMASK) << 16 | (t2 & UCOL_SECONDARYMASK) << 8;
|
|
|
|
lh->gapsLo[2] = (UCOL_TERTIARYORDER(t1)) << 24 | (UCOL_TERTIARYORDER(t2)) << 16;
|
|
|
|
t1 = lh->nextCE;
|
|
|
|
t2 = lh->nextContCE;
|
|
|
|
lh->gapsHi[0] = (t1 & UCOL_PRIMARYMASK) | (t2 & UCOL_PRIMARYMASK) >> 16;
|
|
|
|
lh->gapsHi[1] = (t1 & UCOL_SECONDARYMASK) << 16 | (t2 & UCOL_SECONDARYMASK) << 8;
|
|
|
|
lh->gapsHi[2] = (UCOL_TERTIARYORDER(t1)) << 24 | (UCOL_TERTIARYORDER(t2)) << 16;
|
2001-03-08 17:40:42 +00:00
|
|
|
} else {
|
|
|
|
for(;;) {
|
|
|
|
if(tokStrength < UCOL_CE_STRENGTH_LIMIT) {
|
2003-09-24 00:17:15 +00:00
|
|
|
if((lh->pos[tokStrength] = ucol_inv_getNext(src, lh, tokStrength)) >= 0) {
|
2001-03-08 17:40:42 +00:00
|
|
|
lh->fStrToken[tokStrength] = tok;
|
2001-04-23 01:52:45 +00:00
|
|
|
} else { /* The CE must be implicit, since it's not in the table */
|
2001-03-08 17:40:42 +00:00
|
|
|
/* Error */
|
2001-04-23 01:52:45 +00:00
|
|
|
*status = U_INTERNAL_PROGRAM_ERROR;
|
2001-03-08 17:40:42 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
while(tok != NULL && tok->strength >= tokStrength) {
|
|
|
|
if(tokStrength < UCOL_CE_STRENGTH_LIMIT) {
|
|
|
|
lh->lStrToken[tokStrength] = tok;
|
|
|
|
}
|
|
|
|
tok = tok->next;
|
|
|
|
}
|
|
|
|
if(tokStrength < UCOL_CE_STRENGTH_LIMIT-1) {
|
|
|
|
/* check if previous interval is the same and merge the intervals if it is so */
|
|
|
|
if(lh->pos[tokStrength] == lh->pos[tokStrength+1]) {
|
|
|
|
lh->fStrToken[tokStrength] = lh->fStrToken[tokStrength+1];
|
|
|
|
lh->fStrToken[tokStrength+1] = NULL;
|
|
|
|
lh->lStrToken[tokStrength+1] = NULL;
|
|
|
|
lh->pos[tokStrength+1] = -1;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
if(tok != NULL) {
|
|
|
|
tokStrength = tok->strength;
|
|
|
|
} else {
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
for(st = 0; st < 3; st++) {
|
|
|
|
if((pos = lh->pos[st]) >= 0) {
|
|
|
|
t1 = *(CETable+3*(pos));
|
|
|
|
t2 = *(CETable+3*(pos)+1);
|
|
|
|
lh->gapsHi[3*st] = (t1 & UCOL_PRIMARYMASK) | (t2 & UCOL_PRIMARYMASK) >> 16;
|
|
|
|
lh->gapsHi[3*st+1] = (t1 & UCOL_SECONDARYMASK) << 16 | (t2 & UCOL_SECONDARYMASK) << 8;
|
2001-05-29 22:53:07 +00:00
|
|
|
//lh->gapsHi[3*st+2] = (UCOL_TERTIARYORDER(t1)) << 24 | (UCOL_TERTIARYORDER(t2)) << 16;
|
|
|
|
lh->gapsHi[3*st+2] = (t1&0x3f) << 24 | (t2&0x3f) << 16;
|
2004-05-08 07:59:36 +00:00
|
|
|
//pos--;
|
|
|
|
//t1 = *(CETable+3*(pos));
|
|
|
|
//t2 = *(CETable+3*(pos)+1);
|
|
|
|
t1 = lh->baseCE;
|
|
|
|
t2 = lh->baseContCE;
|
2001-03-08 17:40:42 +00:00
|
|
|
lh->gapsLo[3*st] = (t1 & UCOL_PRIMARYMASK) | (t2 & UCOL_PRIMARYMASK) >> 16;
|
|
|
|
lh->gapsLo[3*st+1] = (t1 & UCOL_SECONDARYMASK) << 16 | (t2 & UCOL_SECONDARYMASK) << 8;
|
2001-05-29 22:53:07 +00:00
|
|
|
lh->gapsLo[3*st+2] = (t1&0x3f) << 24 | (t2&0x3f) << 16;
|
2001-03-08 17:40:42 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
#define ucol_countBytes(value, noOfBytes) \
|
|
|
|
{ \
|
|
|
|
uint32_t mask = 0xFFFFFFFF; \
|
|
|
|
(noOfBytes) = 0; \
|
|
|
|
while(mask != 0) { \
|
|
|
|
if(((value) & mask) != 0) { \
|
|
|
|
(noOfBytes)++; \
|
|
|
|
} \
|
|
|
|
mask >>= 8; \
|
|
|
|
} \
|
|
|
|
}
|
|
|
|
|
2001-03-10 00:52:27 +00:00
|
|
|
U_CFUNC uint32_t ucol_getNextGenerated(ucolCEGenerator *g, UErrorCode *status) {
|
2001-03-22 21:16:20 +00:00
|
|
|
if(U_SUCCESS(*status)) {
|
2001-03-10 00:52:27 +00:00
|
|
|
g->current = ucol_nextWeight(g->ranges, &g->noOfRanges);
|
2001-03-22 21:16:20 +00:00
|
|
|
}
|
2001-03-08 17:40:42 +00:00
|
|
|
return g->current;
|
|
|
|
}
|
|
|
|
|
2001-03-10 00:52:27 +00:00
|
|
|
U_CFUNC uint32_t ucol_getSimpleCEGenerator(ucolCEGenerator *g, UColToken *tok, uint32_t strength, UErrorCode *status) {
|
2001-04-23 23:39:14 +00:00
|
|
|
/* TODO: rename to enum names */
|
2001-03-10 00:52:27 +00:00
|
|
|
uint32_t high, low, count=1;
|
2001-05-26 05:33:48 +00:00
|
|
|
uint32_t maxByte = (strength == UCOL_TERTIARY)?0x3F:0xFF;
|
2001-03-08 17:40:42 +00:00
|
|
|
|
2001-03-10 00:52:27 +00:00
|
|
|
if(strength == UCOL_SECONDARY) {
|
|
|
|
low = UCOL_COMMON_TOP2<<24;
|
|
|
|
high = 0xFFFFFFFF;
|
|
|
|
count = 0xFF - UCOL_COMMON_TOP2;
|
2001-03-09 00:50:37 +00:00
|
|
|
} else {
|
2001-04-23 23:39:14 +00:00
|
|
|
low = UCOL_BYTE_COMMON << 24; //0x05000000;
|
2001-03-10 00:52:27 +00:00
|
|
|
high = 0x40000000;
|
2001-04-23 23:39:14 +00:00
|
|
|
count = 0x40 - UCOL_BYTE_COMMON;
|
2001-03-08 17:40:42 +00:00
|
|
|
}
|
|
|
|
|
2001-03-10 00:52:27 +00:00
|
|
|
if(tok->next != NULL && tok->next->strength == strength) {
|
|
|
|
count = tok->next->toInsert;
|
|
|
|
}
|
|
|
|
|
2001-05-26 05:33:48 +00:00
|
|
|
g->noOfRanges = ucol_allocWeights(low, high, count, maxByte, g->ranges);
|
2001-04-24 05:13:25 +00:00
|
|
|
g->current = UCOL_BYTE_COMMON<<24;
|
2001-03-08 17:40:42 +00:00
|
|
|
|
2001-03-10 00:52:27 +00:00
|
|
|
if(g->noOfRanges == 0) {
|
|
|
|
*status = U_INTERNAL_PROGRAM_ERROR;
|
2001-03-08 17:40:42 +00:00
|
|
|
}
|
|
|
|
return g->current;
|
|
|
|
}
|
|
|
|
|
2001-03-10 00:52:27 +00:00
|
|
|
U_CFUNC uint32_t ucol_getCEGenerator(ucolCEGenerator *g, uint32_t* lows, uint32_t* highs, UColToken *tok, uint32_t fStrength, UErrorCode *status) {
|
|
|
|
uint32_t strength = tok->strength;
|
2001-03-08 17:40:42 +00:00
|
|
|
uint32_t low = lows[fStrength*3+strength];
|
|
|
|
uint32_t high = highs[fStrength*3+strength];
|
2004-01-09 07:31:57 +00:00
|
|
|
uint32_t maxByte = 0;
|
|
|
|
if(strength == UCOL_TERTIARY) {
|
|
|
|
maxByte = 0x3F;
|
|
|
|
} else if(strength == UCOL_PRIMARY) {
|
|
|
|
maxByte = 0xFE;
|
|
|
|
} else {
|
|
|
|
maxByte = 0xFF;
|
|
|
|
}
|
2001-03-08 17:40:42 +00:00
|
|
|
|
2001-05-28 20:57:59 +00:00
|
|
|
uint32_t count = tok->toInsert;
|
2001-03-08 17:40:42 +00:00
|
|
|
|
2001-10-03 01:48:06 +00:00
|
|
|
if(low >= high && strength > UCOL_PRIMARY) {
|
2001-03-22 21:16:20 +00:00
|
|
|
int32_t s = strength;
|
|
|
|
for(;;) {
|
2001-03-10 00:52:27 +00:00
|
|
|
s--;
|
|
|
|
if(lows[fStrength*3+s] != highs[fStrength*3+s]) {
|
|
|
|
if(strength == UCOL_SECONDARY) {
|
|
|
|
low = UCOL_COMMON_TOP2<<24;
|
|
|
|
high = 0xFFFFFFFF;
|
|
|
|
} else {
|
2001-10-03 05:58:10 +00:00
|
|
|
//low = 0x02000000; // This needs to be checked - what if low is
|
|
|
|
// not good...
|
2001-03-10 00:52:27 +00:00
|
|
|
high = 0x40000000;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
}
|
2001-03-22 21:16:20 +00:00
|
|
|
if(s<0) {
|
|
|
|
*status = U_INTERNAL_PROGRAM_ERROR;
|
|
|
|
return 0;
|
|
|
|
}
|
2001-03-10 00:52:27 +00:00
|
|
|
}
|
|
|
|
}
|
2001-03-08 17:40:42 +00:00
|
|
|
|
2001-03-20 00:56:37 +00:00
|
|
|
if(low == 0) {
|
|
|
|
low = 0x01000000;
|
|
|
|
}
|
|
|
|
|
2001-03-10 00:52:27 +00:00
|
|
|
if(strength == UCOL_SECONDARY) { /* similar as simple */
|
2001-05-28 20:57:59 +00:00
|
|
|
if(low >= (UCOL_COMMON_BOT2<<24) && low < (uint32_t)(UCOL_COMMON_TOP2<<24)) {
|
2001-03-10 00:52:27 +00:00
|
|
|
low = UCOL_COMMON_TOP2<<24;
|
2001-03-08 17:40:42 +00:00
|
|
|
}
|
2001-05-28 20:57:59 +00:00
|
|
|
if(high > (UCOL_COMMON_BOT2<<24) && high < (uint32_t)(UCOL_COMMON_TOP2<<24)) {
|
2001-03-10 00:52:27 +00:00
|
|
|
high = UCOL_COMMON_TOP2<<24;
|
|
|
|
}
|
2004-05-08 07:59:36 +00:00
|
|
|
if(low < (UCOL_COMMON_BOT2<<24)) {
|
|
|
|
g->noOfRanges = ucol_allocWeights(UCOL_BYTE_UNSHIFTED_MIN<<24, high, count, maxByte, g->ranges);
|
|
|
|
g->current = ucol_nextWeight(g->ranges, &g->noOfRanges);
|
|
|
|
//g->current = UCOL_COMMON_BOT2<<24;
|
2001-03-10 00:52:27 +00:00
|
|
|
return g->current;
|
|
|
|
}
|
|
|
|
}
|
2001-03-08 17:40:42 +00:00
|
|
|
|
2001-05-26 05:33:48 +00:00
|
|
|
g->noOfRanges = ucol_allocWeights(low, high, count, maxByte, g->ranges);
|
2001-03-10 00:52:27 +00:00
|
|
|
if(g->noOfRanges == 0) {
|
|
|
|
*status = U_INTERNAL_PROGRAM_ERROR;
|
2001-03-08 17:40:42 +00:00
|
|
|
}
|
2001-03-10 00:52:27 +00:00
|
|
|
g->current = ucol_nextWeight(g->ranges, &g->noOfRanges);
|
2001-03-08 17:40:42 +00:00
|
|
|
return g->current;
|
|
|
|
}
|
|
|
|
|
2003-07-08 23:56:15 +00:00
|
|
|
static
|
|
|
|
uint32_t u_toLargeKana(const UChar *source, const uint32_t sourceLen, UChar *resBuf, const uint32_t resLen, UErrorCode *status) {
|
|
|
|
uint32_t i = 0;
|
|
|
|
UChar c;
|
|
|
|
|
|
|
|
if(U_FAILURE(*status)) {
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
if(sourceLen > resLen) {
|
|
|
|
*status = U_MEMORY_ALLOCATION_ERROR;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
for(i = 0; i < sourceLen; i++) {
|
|
|
|
c = source[i];
|
|
|
|
if(0x3042 < c && c < 0x30ef) { /* Kana range */
|
|
|
|
switch(c - 0x3000) {
|
|
|
|
case 0x41: case 0x43: case 0x45: case 0x47: case 0x49: case 0x63: case 0x83: case 0x85: case 0x8E:
|
|
|
|
case 0xA1: case 0xA3: case 0xA5: case 0xA7: case 0xA9: case 0xC3: case 0xE3: case 0xE5: case 0xEE:
|
|
|
|
c++;
|
|
|
|
break;
|
|
|
|
case 0xF5:
|
|
|
|
c = 0x30AB;
|
|
|
|
break;
|
|
|
|
case 0xF6:
|
|
|
|
c = 0x30B1;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
resBuf[i] = c;
|
|
|
|
}
|
|
|
|
return sourceLen;
|
|
|
|
}
|
|
|
|
|
|
|
|
static
|
|
|
|
uint32_t u_toSmallKana(const UChar *source, const uint32_t sourceLen, UChar *resBuf, const uint32_t resLen, UErrorCode *status) {
|
|
|
|
uint32_t i = 0;
|
|
|
|
UChar c;
|
|
|
|
|
|
|
|
if(U_FAILURE(*status)) {
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
if(sourceLen > resLen) {
|
|
|
|
*status = U_MEMORY_ALLOCATION_ERROR;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
for(i = 0; i < sourceLen; i++) {
|
|
|
|
c = source[i];
|
|
|
|
if(0x3042 < c && c < 0x30ef) { /* Kana range */
|
|
|
|
switch(c - 0x3000) {
|
|
|
|
case 0x42: case 0x44: case 0x46: case 0x48: case 0x4A: case 0x64: case 0x84: case 0x86: case 0x8F:
|
|
|
|
case 0xA2: case 0xA4: case 0xA6: case 0xA8: case 0xAA: case 0xC4: case 0xE4: case 0xE6: case 0xEF:
|
|
|
|
c--;
|
|
|
|
break;
|
|
|
|
case 0xAB:
|
|
|
|
c = 0x30F5;
|
|
|
|
break;
|
|
|
|
case 0xB1:
|
|
|
|
c = 0x30F6;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
resBuf[i] = c;
|
|
|
|
}
|
|
|
|
return sourceLen;
|
|
|
|
}
|
|
|
|
|
|
|
|
static
|
|
|
|
uint8_t ucol_uprv_getCaseBits(const UCollator *UCA, const UChar *src, uint32_t len, UErrorCode *status) {
|
|
|
|
uint32_t i = 0;
|
|
|
|
UChar n[128];
|
|
|
|
uint32_t nLen = 0;
|
|
|
|
uint32_t uCount = 0, lCount = 0;
|
|
|
|
|
|
|
|
collIterate s;
|
|
|
|
uint32_t order = 0;
|
|
|
|
|
|
|
|
if(U_FAILURE(*status)) {
|
|
|
|
return UCOL_LOWER_CASE;
|
|
|
|
}
|
|
|
|
|
|
|
|
nLen = unorm_normalize(src, len, UNORM_NFKD, 0, n, 128, status);
|
|
|
|
if(U_SUCCESS(*status)) {
|
|
|
|
for(i = 0; i < nLen; i++) {
|
|
|
|
uprv_init_collIterate(UCA, &n[i], 1, &s);
|
|
|
|
order = ucol_getNextCE(UCA, &s, status);
|
|
|
|
if(isContinuation(order)) {
|
|
|
|
*status = U_INTERNAL_PROGRAM_ERROR;
|
|
|
|
return UCOL_LOWER_CASE;
|
|
|
|
}
|
|
|
|
if((order&UCOL_CASE_BIT_MASK)== UCOL_UPPER_CASE) {
|
|
|
|
uCount++;
|
|
|
|
} else {
|
|
|
|
if(u_islower(n[i])) {
|
|
|
|
lCount++;
|
|
|
|
} else {
|
|
|
|
UChar sk[1], lk[1];
|
|
|
|
u_toSmallKana(&n[i], 1, sk, 1, status);
|
|
|
|
u_toLargeKana(&n[i], 1, lk, 1, status);
|
|
|
|
if(sk[0] == n[i] && lk[0] != n[i]) {
|
|
|
|
lCount++;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
if(uCount != 0 && lCount != 0) {
|
|
|
|
return UCOL_MIXED_CASE;
|
|
|
|
} else if(uCount != 0) {
|
|
|
|
return UCOL_UPPER_CASE;
|
|
|
|
} else {
|
|
|
|
return UCOL_LOWER_CASE;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
U_CFUNC void ucol_doCE(UColTokenParser *src, uint32_t *CEparts, UColToken *tok, UErrorCode *status) {
|
2001-03-08 17:40:42 +00:00
|
|
|
/* this one makes the table and stuff */
|
|
|
|
uint32_t noOfBytes[3];
|
|
|
|
uint32_t i;
|
|
|
|
|
|
|
|
for(i = 0; i<3; i++) {
|
|
|
|
ucol_countBytes(CEparts[i], noOfBytes[i]);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Here we have to pack CEs from parts */
|
|
|
|
|
|
|
|
uint32_t CEi = 0;
|
|
|
|
uint32_t value = 0;
|
|
|
|
|
|
|
|
while(2*CEi<noOfBytes[0] || CEi<noOfBytes[1] || CEi<noOfBytes[2]) {
|
|
|
|
if(CEi > 0) {
|
2001-05-10 22:33:50 +00:00
|
|
|
value = UCOL_CONTINUATION_MARKER; /* Continuation marker */
|
2001-03-08 17:40:42 +00:00
|
|
|
} else {
|
|
|
|
value = 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
if(2*CEi<noOfBytes[0]) {
|
|
|
|
value |= ((CEparts[0]>>(32-16*(CEi+1))) & 0xFFFF) << 16;
|
|
|
|
}
|
|
|
|
if(CEi<noOfBytes[1]) {
|
|
|
|
value |= ((CEparts[1]>>(32-8*(CEi+1))) & 0xFF) << 8;
|
|
|
|
}
|
|
|
|
if(CEi<noOfBytes[2]) {
|
|
|
|
value |= ((CEparts[2]>>(32-8*(CEi+1))) & 0x3F);
|
|
|
|
}
|
|
|
|
tok->CEs[CEi] = value;
|
|
|
|
CEi++;
|
|
|
|
}
|
|
|
|
if(CEi == 0) { /* totally ignorable */
|
|
|
|
tok->noOfCEs = 1;
|
|
|
|
tok->CEs[0] = 0;
|
|
|
|
} else { /* there is at least something */
|
|
|
|
tok->noOfCEs = CEi;
|
|
|
|
}
|
|
|
|
|
2003-07-08 23:56:15 +00:00
|
|
|
|
|
|
|
// we want to set case bits here and now, not later.
|
|
|
|
// Case bits handling
|
|
|
|
tok->CEs[0] &= 0xFFFFFF3F; // Clean the case bits field
|
|
|
|
int32_t cSize = (tok->source & 0xFF000000) >> 24;
|
|
|
|
UChar *cPoints = (tok->source & 0x00FFFFFF) + src->source;
|
|
|
|
|
|
|
|
if(cSize > 1) {
|
|
|
|
// Do it manually
|
|
|
|
tok->CEs[0] |= ucol_uprv_getCaseBits(src->UCA, cPoints, cSize, status);
|
|
|
|
} else {
|
|
|
|
// Copy it from the UCA
|
|
|
|
uint32_t caseCE = ucol_getFirstCE(src->UCA, cPoints[0], status);
|
|
|
|
tok->CEs[0] |= (caseCE & 0xC0);
|
|
|
|
}
|
|
|
|
|
2001-03-08 17:40:42 +00:00
|
|
|
#if UCOL_DEBUG==2
|
|
|
|
fprintf(stderr, "%04X str: %i, [%08X, %08X, %08X]: tok: ", tok->debugSource, tok->strength, CEparts[0] >> (32-8*noOfBytes[0]), CEparts[1] >> (32-8*noOfBytes[1]), CEparts[2]>> (32-8*noOfBytes[2]));
|
|
|
|
for(i = 0; i<tok->noOfCEs; i++) {
|
|
|
|
fprintf(stderr, "%08X ", tok->CEs[i]);
|
|
|
|
}
|
|
|
|
fprintf(stderr, "\n");
|
|
|
|
#endif
|
|
|
|
}
|
|
|
|
|
2002-06-13 18:29:49 +00:00
|
|
|
U_CFUNC void ucol_initBuffers(UColTokenParser *src, UColTokListHeader *lh, UErrorCode *status) {
|
2001-03-08 17:40:42 +00:00
|
|
|
ucolCEGenerator Gens[UCOL_CE_STRENGTH_LIMIT];
|
|
|
|
uint32_t CEparts[UCOL_CE_STRENGTH_LIMIT];
|
|
|
|
|
2001-06-06 20:20:43 +00:00
|
|
|
UColToken *tok = lh->last;
|
2001-03-08 17:40:42 +00:00
|
|
|
uint32_t t[UCOL_STRENGTH_LIMIT];
|
|
|
|
|
2002-06-13 18:29:49 +00:00
|
|
|
uprv_memset(t, 0, UCOL_STRENGTH_LIMIT*sizeof(uint32_t));
|
|
|
|
|
2001-03-08 17:40:42 +00:00
|
|
|
tok->toInsert = 1;
|
|
|
|
t[tok->strength] = 1;
|
|
|
|
|
|
|
|
while(tok->previous != NULL) {
|
|
|
|
if(tok->previous->strength < tok->strength) { /* going up */
|
|
|
|
t[tok->strength] = 0;
|
|
|
|
t[tok->previous->strength]++;
|
|
|
|
} else if(tok->previous->strength > tok->strength) { /* going down */
|
|
|
|
t[tok->previous->strength] = 1;
|
|
|
|
} else {
|
|
|
|
t[tok->strength]++;
|
|
|
|
}
|
|
|
|
tok=tok->previous;
|
|
|
|
tok->toInsert = t[tok->strength];
|
|
|
|
}
|
|
|
|
|
|
|
|
tok->toInsert = t[tok->strength];
|
2002-06-13 18:29:49 +00:00
|
|
|
ucol_inv_getGapPositions(src, lh, status);
|
2001-03-10 00:52:27 +00:00
|
|
|
|
|
|
|
#if UCOL_DEBUG
|
|
|
|
fprintf(stderr, "BaseCE: %08X %08X\n", lh->baseCE, lh->baseContCE);
|
|
|
|
int32_t j = 2;
|
|
|
|
for(j = 2; j >= 0; j--) {
|
|
|
|
fprintf(stderr, "gapsLo[%i] [%08X %08X %08X]\n", j, lh->gapsLo[j*3], lh->gapsLo[j*3+1], lh->gapsLo[j*3+2]);
|
|
|
|
fprintf(stderr, "gapsHi[%i] [%08X %08X %08X]\n", j, lh->gapsHi[j*3], lh->gapsHi[j*3+1], lh->gapsHi[j*3+2]);
|
|
|
|
}
|
2001-03-08 17:40:42 +00:00
|
|
|
tok=lh->first[UCOL_TOK_POLARITY_POSITIVE];
|
|
|
|
|
|
|
|
do {
|
|
|
|
fprintf(stderr,"%i", tok->strength);
|
|
|
|
tok = tok->next;
|
|
|
|
} while(tok != NULL);
|
|
|
|
fprintf(stderr, "\n");
|
|
|
|
|
|
|
|
tok=lh->first[UCOL_TOK_POLARITY_POSITIVE];
|
|
|
|
|
|
|
|
do {
|
|
|
|
fprintf(stderr,"%i", tok->toInsert);
|
|
|
|
tok = tok->next;
|
|
|
|
} while(tok != NULL);
|
|
|
|
#endif
|
|
|
|
|
2001-06-06 20:20:43 +00:00
|
|
|
tok = lh->first;
|
2001-03-10 00:52:27 +00:00
|
|
|
uint32_t fStrength = UCOL_IDENTICAL;
|
|
|
|
uint32_t initStrength = UCOL_IDENTICAL;
|
2001-03-08 17:40:42 +00:00
|
|
|
|
|
|
|
|
2001-03-10 00:52:27 +00:00
|
|
|
CEparts[UCOL_PRIMARY] = (lh->baseCE & UCOL_PRIMARYMASK) | (lh->baseContCE & UCOL_PRIMARYMASK) >> 16;
|
|
|
|
CEparts[UCOL_SECONDARY] = (lh->baseCE & UCOL_SECONDARYMASK) << 16 | (lh->baseContCE & UCOL_SECONDARYMASK) << 8;
|
|
|
|
CEparts[UCOL_TERTIARY] = (UCOL_TERTIARYORDER(lh->baseCE)) << 24 | (UCOL_TERTIARYORDER(lh->baseContCE)) << 16;
|
2001-03-08 17:40:42 +00:00
|
|
|
|
2001-03-10 00:52:27 +00:00
|
|
|
while (tok != NULL && U_SUCCESS(*status)) {
|
2001-03-08 17:40:42 +00:00
|
|
|
fStrength = tok->strength;
|
2001-03-10 00:52:27 +00:00
|
|
|
if(fStrength < initStrength) {
|
|
|
|
initStrength = fStrength;
|
2001-03-08 17:40:42 +00:00
|
|
|
if(lh->pos[fStrength] == -1) {
|
2001-03-10 00:52:27 +00:00
|
|
|
while(lh->pos[fStrength] == -1 && fStrength > 0) {
|
|
|
|
fStrength--;
|
2001-03-08 17:40:42 +00:00
|
|
|
}
|
2001-03-10 00:52:27 +00:00
|
|
|
if(lh->pos[fStrength] == -1) {
|
|
|
|
*status = U_INTERNAL_PROGRAM_ERROR;
|
|
|
|
return;
|
2001-03-08 17:40:42 +00:00
|
|
|
}
|
|
|
|
}
|
2001-03-10 00:52:27 +00:00
|
|
|
if(initStrength == UCOL_TERTIARY) { /* starting with tertiary */
|
|
|
|
CEparts[UCOL_PRIMARY] = lh->gapsLo[fStrength*3];
|
|
|
|
CEparts[UCOL_SECONDARY] = lh->gapsLo[fStrength*3+1];
|
|
|
|
/*CEparts[UCOL_TERTIARY] = ucol_getCEGenerator(&Gens[2], lh->gapsLo[fStrength*3+2], lh->gapsHi[fStrength*3+2], tok, UCOL_TERTIARY); */
|
|
|
|
CEparts[UCOL_TERTIARY] = ucol_getCEGenerator(&Gens[UCOL_TERTIARY], lh->gapsLo, lh->gapsHi, tok, fStrength, status);
|
|
|
|
} else if(initStrength == UCOL_SECONDARY) { /* secondaries */
|
|
|
|
CEparts[UCOL_PRIMARY] = lh->gapsLo[fStrength*3];
|
|
|
|
/*CEparts[1] = ucol_getCEGenerator(&Gens[1], lh->gapsLo[fStrength*3+1], lh->gapsHi[fStrength*3+1], tok, 1);*/
|
|
|
|
CEparts[UCOL_SECONDARY] = ucol_getCEGenerator(&Gens[UCOL_SECONDARY], lh->gapsLo, lh->gapsHi, tok, fStrength, status);
|
|
|
|
CEparts[UCOL_TERTIARY] = ucol_getSimpleCEGenerator(&Gens[UCOL_TERTIARY], tok, UCOL_TERTIARY, status);
|
|
|
|
} else { /* primaries */
|
|
|
|
/*CEparts[UCOL_PRIMARY] = ucol_getCEGenerator(&Gens[0], lh->gapsLo[0], lh->gapsHi[0], tok, UCOL_PRIMARY);*/
|
|
|
|
CEparts[UCOL_PRIMARY] = ucol_getCEGenerator(&Gens[UCOL_PRIMARY], lh->gapsLo, lh->gapsHi, tok, fStrength, status);
|
|
|
|
CEparts[UCOL_SECONDARY] = ucol_getSimpleCEGenerator(&Gens[UCOL_SECONDARY], tok, UCOL_SECONDARY, status);
|
|
|
|
CEparts[UCOL_TERTIARY] = ucol_getSimpleCEGenerator(&Gens[UCOL_TERTIARY], tok, UCOL_TERTIARY, status);
|
2001-03-08 17:40:42 +00:00
|
|
|
}
|
2001-03-10 00:52:27 +00:00
|
|
|
} else {
|
2001-03-08 17:40:42 +00:00
|
|
|
if(tok->strength == UCOL_TERTIARY) {
|
2001-03-10 00:52:27 +00:00
|
|
|
CEparts[UCOL_TERTIARY] = ucol_getNextGenerated(&Gens[UCOL_TERTIARY], status);
|
2001-03-08 17:40:42 +00:00
|
|
|
} else if(tok->strength == UCOL_SECONDARY) {
|
2001-03-10 00:52:27 +00:00
|
|
|
CEparts[UCOL_SECONDARY] = ucol_getNextGenerated(&Gens[UCOL_SECONDARY], status);
|
|
|
|
CEparts[UCOL_TERTIARY] = ucol_getSimpleCEGenerator(&Gens[UCOL_TERTIARY], tok, UCOL_TERTIARY, status);
|
2001-03-08 17:40:42 +00:00
|
|
|
} else if(tok->strength == UCOL_PRIMARY) {
|
2001-03-10 00:52:27 +00:00
|
|
|
CEparts[UCOL_PRIMARY] = ucol_getNextGenerated(&Gens[UCOL_PRIMARY], status);
|
|
|
|
CEparts[UCOL_SECONDARY] = ucol_getSimpleCEGenerator(&Gens[UCOL_SECONDARY], tok, UCOL_SECONDARY, status);
|
|
|
|
CEparts[UCOL_TERTIARY] = ucol_getSimpleCEGenerator(&Gens[UCOL_TERTIARY], tok, UCOL_TERTIARY, status);
|
|
|
|
}
|
2001-03-08 17:40:42 +00:00
|
|
|
}
|
2003-07-08 23:56:15 +00:00
|
|
|
ucol_doCE(src, CEparts, tok, status);
|
2001-03-10 00:52:27 +00:00
|
|
|
tok = tok->next;
|
2001-03-08 17:40:42 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2001-06-06 20:20:43 +00:00
|
|
|
U_CFUNC void ucol_createElements(UColTokenParser *src, tempUCATable *t, UColTokListHeader *lh, UErrorCode *status) {
|
2001-03-08 17:40:42 +00:00
|
|
|
UCAElements el;
|
2001-06-06 20:20:43 +00:00
|
|
|
UColToken *tok = lh->first;
|
2001-03-08 17:40:42 +00:00
|
|
|
UColToken *expt = NULL;
|
2001-04-02 23:12:19 +00:00
|
|
|
uint32_t i = 0, j = 0;
|
2001-03-08 17:40:42 +00:00
|
|
|
|
2001-12-28 20:53:44 +00:00
|
|
|
while(tok != NULL && U_SUCCESS(*status)) {
|
2001-03-08 17:40:42 +00:00
|
|
|
/* first, check if there are any expansions */
|
2001-04-02 23:12:19 +00:00
|
|
|
/* if there are expansions, we need to do a little bit more processing */
|
|
|
|
/* since parts of expansion can be tailored, while others are not */
|
2001-10-05 02:06:28 +00:00
|
|
|
if(tok->expansion != 0) {
|
2001-04-02 23:12:19 +00:00
|
|
|
uint32_t len = tok->expansion >> 24;
|
|
|
|
uint32_t currentSequenceLen = len;
|
|
|
|
uint32_t expOffset = tok->expansion & 0x00FFFFFF;
|
2001-10-17 22:57:15 +00:00
|
|
|
//uint32_t exp = currentSequenceLen | expOffset;
|
|
|
|
UColToken exp;
|
|
|
|
exp.source = currentSequenceLen | expOffset;
|
|
|
|
exp.rulesToParse = src->source;
|
2001-04-02 23:12:19 +00:00
|
|
|
|
|
|
|
while(len > 0) {
|
|
|
|
currentSequenceLen = len;
|
|
|
|
while(currentSequenceLen > 0) {
|
2001-10-17 22:57:15 +00:00
|
|
|
exp.source = (currentSequenceLen << 24) | expOffset;
|
|
|
|
if((expt = (UColToken *)uhash_get(src->tailored, &exp)) != NULL && expt->strength != UCOL_TOK_RESET) { /* expansion is tailored */
|
2001-04-02 23:12:19 +00:00
|
|
|
uint32_t noOfCEsToCopy = expt->noOfCEs;
|
|
|
|
for(j = 0; j<noOfCEsToCopy; j++) {
|
|
|
|
tok->expCEs[tok->noOfExpCEs + j] = expt->CEs[j];
|
|
|
|
}
|
|
|
|
tok->noOfExpCEs += noOfCEsToCopy;
|
2001-05-31 17:05:48 +00:00
|
|
|
// Smart people never try to add codepoints and CEs.
|
|
|
|
// For some odd reason, it won't work.
|
|
|
|
expOffset += currentSequenceLen; //noOfCEsToCopy;
|
|
|
|
len -= currentSequenceLen; //noOfCEsToCopy;
|
2001-04-02 23:12:19 +00:00
|
|
|
break;
|
|
|
|
} else {
|
|
|
|
currentSequenceLen--;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
if(currentSequenceLen == 0) { /* couldn't find any tailored subsequence */
|
|
|
|
/* will have to get one from UCA */
|
|
|
|
/* first, get the UChars from the rules */
|
|
|
|
/* then pick CEs out until there is no more and stuff them into expansion */
|
|
|
|
collIterate s;
|
|
|
|
uint32_t order = 0;
|
2002-07-16 01:46:42 +00:00
|
|
|
uprv_init_collIterate(src->UCA, expOffset + src->source, 1, &s);
|
2001-04-02 23:12:19 +00:00
|
|
|
|
|
|
|
for(;;) {
|
2001-04-23 01:52:45 +00:00
|
|
|
order = ucol_getNextCE(src->UCA, &s, status);
|
2001-04-02 23:12:19 +00:00
|
|
|
if(order == UCOL_NO_MORE_CES) {
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
tok->expCEs[tok->noOfExpCEs++] = order;
|
|
|
|
}
|
|
|
|
expOffset++;
|
|
|
|
len--;
|
|
|
|
}
|
|
|
|
}
|
2001-03-08 17:40:42 +00:00
|
|
|
} else {
|
|
|
|
tok->noOfExpCEs = 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* set the ucaelement with obtained values */
|
|
|
|
el.noOfCEs = tok->noOfCEs + tok->noOfExpCEs;
|
|
|
|
/* copy CEs */
|
|
|
|
for(i = 0; i<tok->noOfCEs; i++) {
|
|
|
|
el.CEs[i] = tok->CEs[i];
|
|
|
|
}
|
|
|
|
for(i = 0; i<tok->noOfExpCEs; i++) {
|
|
|
|
el.CEs[i+tok->noOfCEs] = tok->expCEs[i];
|
|
|
|
}
|
|
|
|
|
|
|
|
/* copy UChars */
|
2001-09-27 23:18:14 +00:00
|
|
|
// We kept prefix and source kind of together, as it is a kind of a contraction.
|
|
|
|
// However, now we have to slice the prefix off the main thing -
|
|
|
|
el.prefix = el.prefixChars;
|
2001-03-08 17:40:42 +00:00
|
|
|
el.cPoints = el.uchars;
|
2001-10-12 20:11:47 +00:00
|
|
|
if(tok->prefix != 0) { // we will just copy the prefix here, and adjust accordingly in the
|
|
|
|
// addPrefix function in ucol_elm. The reason is that we need to add both composed AND
|
|
|
|
// decomposed elements to the unsaf table.
|
|
|
|
el.prefixSize = tok->prefix>>24;
|
|
|
|
uprv_memcpy(el.prefix, src->source + (tok->prefix & 0x00FFFFFF), el.prefixSize*sizeof(UChar));
|
2001-09-27 23:18:14 +00:00
|
|
|
|
|
|
|
el.cSize = (tok->source >> 24)-(tok->prefix>>24);
|
|
|
|
uprv_memcpy(el.uchars, (tok->source & 0x00FFFFFF)+(tok->prefix>>24) + src->source, el.cSize*sizeof(UChar));
|
|
|
|
} else {
|
|
|
|
el.prefixSize = 0;
|
|
|
|
*el.prefix = 0;
|
|
|
|
|
|
|
|
el.cSize = (tok->source >> 24);
|
|
|
|
uprv_memcpy(el.uchars, (tok->source & 0x00FFFFFF) + src->source, el.cSize*sizeof(UChar));
|
|
|
|
}
|
2001-03-08 17:40:42 +00:00
|
|
|
|
|
|
|
if(UCOL_ISTHAIPREVOWEL(el.cPoints[0])) {
|
|
|
|
el.isThai = TRUE;
|
|
|
|
} else {
|
|
|
|
el.isThai = FALSE;
|
|
|
|
}
|
|
|
|
|
2001-03-22 18:45:31 +00:00
|
|
|
if(src->UCA != NULL) {
|
|
|
|
for(i = 0; i<el.cSize; i++) {
|
|
|
|
if(UCOL_ISJAMO(el.cPoints[i])) {
|
|
|
|
t->image->jamoSpecial = TRUE;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2003-07-08 23:56:15 +00:00
|
|
|
#if 0
|
|
|
|
// we do case bits in doCE now, since we will mess up expansions otherwise.
|
2001-05-10 22:33:50 +00:00
|
|
|
// Case bits handling
|
|
|
|
el.CEs[0] &= 0xFFFFFF3F; // Clean the case bits field
|
|
|
|
if(el.cSize > 1) {
|
|
|
|
// Do it manually
|
2001-05-22 22:26:58 +00:00
|
|
|
el.CEs[0] |= ucol_uprv_getCaseBits(src->UCA, el.cPoints, el.cSize, status);
|
2001-05-10 22:33:50 +00:00
|
|
|
} else {
|
|
|
|
// Copy it from the UCA
|
|
|
|
uint32_t caseCE = ucol_getFirstCE(src->UCA, el.cPoints[0], status);
|
|
|
|
el.CEs[0] |= (caseCE & 0xC0);
|
|
|
|
}
|
2003-07-08 23:56:15 +00:00
|
|
|
#endif
|
2001-03-08 17:40:42 +00:00
|
|
|
|
|
|
|
/* and then, add it */
|
|
|
|
#if UCOL_DEBUG==2
|
|
|
|
fprintf(stderr, "Adding: %04X with %08X\n", el.cPoints[0], el.CEs[0]);
|
|
|
|
#endif
|
|
|
|
uprv_uca_addAnElement(t, &el, status);
|
2001-10-25 20:37:02 +00:00
|
|
|
|
|
|
|
#if 0
|
|
|
|
if(el.cSize > 1) { // this is a contraction, we should check whether a composed form should also be included
|
|
|
|
UChar composed[256];
|
|
|
|
uint32_t compLen = unorm_normalize(el.cPoints, el.cSize, UNORM_NFC, 0, composed, 256, status);;
|
|
|
|
|
|
|
|
if(compLen != el.cSize || uprv_memcmp(composed, el.cPoints, el.cSize*sizeof(UChar))) {
|
|
|
|
// composed form of a contraction is different than the decomposed form!
|
|
|
|
// do it!
|
|
|
|
#ifdef UCOL_DEBUG
|
|
|
|
fprintf(stderr, "Adding composed for %04X->%04X\n", *element->cPoints, *composed);
|
|
|
|
#endif
|
|
|
|
el.cSize = compLen;
|
|
|
|
uprv_memcpy(el.cPoints, composed, el.cSize*sizeof(UChar));
|
|
|
|
uprv_uca_addAnElement(t, &el, status);
|
|
|
|
}
|
|
|
|
}
|
2001-11-06 22:55:29 +00:00
|
|
|
#endif
|
2001-10-25 20:37:02 +00:00
|
|
|
|
2001-03-21 00:23:20 +00:00
|
|
|
#if UCOL_DEBUG_DUPLICATES
|
|
|
|
if(*status != U_ZERO_ERROR) {
|
|
|
|
fprintf(stderr, "replaced CE for %04X with CE for %04X\n", el.cPoints[0], tok->debugSource);
|
|
|
|
*status = U_ZERO_ERROR;
|
|
|
|
}
|
|
|
|
#endif
|
2001-03-08 17:40:42 +00:00
|
|
|
|
|
|
|
tok = tok->next;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2002-07-02 22:32:14 +00:00
|
|
|
U_CDECL_BEGIN
|
|
|
|
static UBool U_CALLCONV
|
|
|
|
_processUCACompleteIgnorables(const void *context, UChar32 start, UChar32 limit, uint32_t value) {
|
|
|
|
UErrorCode status = U_ZERO_ERROR;
|
|
|
|
tempUCATable *t = (tempUCATable *)context;
|
|
|
|
if(value == 0) {
|
|
|
|
while(start < limit) {
|
|
|
|
uint32_t CE = utrie_get32(t->mapping, start, NULL);
|
|
|
|
if(CE == UCOL_NOT_FOUND) {
|
|
|
|
UCAElements el;
|
|
|
|
el.isThai = FALSE;
|
|
|
|
el.prefixSize = 0;
|
|
|
|
el.prefixChars[0] = 0;
|
|
|
|
el.prefix = el.prefixChars;
|
|
|
|
el.cPoints = el.uchars;
|
|
|
|
|
|
|
|
el.cSize = 0;
|
|
|
|
UTF_APPEND_CHAR(el.uchars, el.cSize, 1024, start);
|
|
|
|
|
|
|
|
el.noOfCEs = 1;
|
|
|
|
el.CEs[0] = 0;
|
|
|
|
uprv_uca_addAnElement(t, &el, &status);
|
|
|
|
|
|
|
|
}
|
|
|
|
start++;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
if(U_FAILURE(status)) {
|
|
|
|
return FALSE;
|
|
|
|
} else {
|
|
|
|
return TRUE;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
U_CDECL_END
|
|
|
|
|
2002-10-16 22:34:16 +00:00
|
|
|
static void
|
|
|
|
ucol_uprv_bld_copyRangeFromUCA(UColTokenParser *src, tempUCATable *t,
|
|
|
|
UChar32 start, UChar32 end,
|
|
|
|
UErrorCode *status) {
|
|
|
|
//UChar decomp[256];
|
|
|
|
uint32_t CE = UCOL_NOT_FOUND;
|
|
|
|
UChar32 u = 0;
|
|
|
|
UCAElements el;
|
|
|
|
el.isThai = FALSE;
|
|
|
|
el.prefixSize = 0;
|
|
|
|
el.prefixChars[0] = 0;
|
|
|
|
collIterate colIt;
|
|
|
|
|
|
|
|
if(U_SUCCESS(*status)) {
|
|
|
|
for(u = start; u<=end; u++) {
|
|
|
|
if((CE = utrie_get32(t->mapping, u, NULL)) == UCOL_NOT_FOUND
|
|
|
|
/* this test is for contractions that are missing the starting element. */
|
|
|
|
|| ((isCntTableElement(CE)) &&
|
|
|
|
(uprv_cnttab_getCE(t->contractions, CE, 0, status) == UCOL_NOT_FOUND))
|
|
|
|
) {
|
|
|
|
el.cSize = 0;
|
|
|
|
U16_APPEND_UNSAFE(el.uchars, el.cSize, u);
|
|
|
|
//decomp[0] = (UChar)u;
|
|
|
|
//el.uchars[0] = (UChar)u;
|
|
|
|
el.cPoints = el.uchars;
|
|
|
|
//el.cSize = 1;
|
|
|
|
el.noOfCEs = 0;
|
|
|
|
el.prefix = el.prefixChars;
|
|
|
|
el.prefixSize = 0;
|
|
|
|
//uprv_init_collIterate(src->UCA, decomp, 1, &colIt);
|
|
|
|
// We actually want to check whether this element is a special
|
|
|
|
// If it is an implicit element (hangul, CJK - we want to copy the
|
|
|
|
// special, not the resolved CEs) - for hangul, copying resolved
|
|
|
|
// would just make things the same (there is an expansion and it
|
|
|
|
// takes approximately the same amount of time to resolve as
|
|
|
|
// falling back to the UCA).
|
|
|
|
/*
|
|
|
|
UTRIE_GET32(src->UCA->mapping, u, CE);
|
|
|
|
tag = getCETag(CE);
|
|
|
|
if(tag == HANGUL_SYLLABLE_TAG || tag == CJK_IMPLICIT_TAG
|
|
|
|
|| tag == IMPLICIT_TAG || tag == TRAIL_SURROGATE_TAG
|
|
|
|
|| tag == LEAD_SURROGATE_TAG) {
|
|
|
|
el.CEs[el.noOfCEs++] = CE;
|
|
|
|
} else {
|
|
|
|
*/
|
|
|
|
// It turns out that it does not make sense to keep implicits
|
|
|
|
// unresolved. The cost of resolving them is big enough so that
|
|
|
|
// it doesn't make any difference whether we have to go to the UCA
|
|
|
|
// or not.
|
|
|
|
{
|
|
|
|
uprv_init_collIterate(src->UCA, el.uchars, el.cSize, &colIt);
|
|
|
|
while(CE != UCOL_NO_MORE_CES) {
|
|
|
|
CE = ucol_getNextCE(src->UCA, &colIt, status);
|
|
|
|
if(CE != UCOL_NO_MORE_CES) {
|
|
|
|
el.CEs[el.noOfCEs++] = CE;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
uprv_uca_addAnElement(t, &el, status);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
2002-07-02 22:32:14 +00:00
|
|
|
|
2001-03-08 17:40:42 +00:00
|
|
|
UCATableHeader *ucol_assembleTailoringTable(UColTokenParser *src, UErrorCode *status) {
|
|
|
|
uint32_t i = 0;
|
2001-03-22 18:45:31 +00:00
|
|
|
if(U_FAILURE(*status)) {
|
|
|
|
return NULL;
|
|
|
|
}
|
2001-03-08 17:40:42 +00:00
|
|
|
/*
|
|
|
|
2. Eliminate the negative lists by doing the following for each non-null negative list:
|
|
|
|
o if previousCE(baseCE, strongestN) != some ListHeader X's baseCE,
|
|
|
|
create new ListHeader X
|
|
|
|
o reverse the list, add to the end of X's positive list. Reset the strength of the
|
|
|
|
first item you add, based on the stronger strength levels of the two lists.
|
|
|
|
*/
|
|
|
|
/*
|
|
|
|
3. For each ListHeader with a non-null positive list:
|
|
|
|
*/
|
|
|
|
/*
|
|
|
|
o Find all character strings with CEs between the baseCE and the
|
|
|
|
next/previous CE, at the strength of the first token. Add these to the
|
|
|
|
tailoring.
|
|
|
|
? That is, if UCA has ... x <<< X << x' <<< X' < y ..., and the
|
|
|
|
tailoring has & x < z...
|
|
|
|
? Then we change the tailoring to & x <<< X << x' <<< X' < z ...
|
|
|
|
*/
|
|
|
|
/* It is possible that this part should be done even while constructing list */
|
|
|
|
/* The problem is that it is unknown what is going to be the strongest weight */
|
|
|
|
/* So we might as well do it here */
|
|
|
|
|
|
|
|
/*
|
|
|
|
o Allocate CEs for each token in the list, based on the total number N of the
|
|
|
|
largest level difference, and the gap G between baseCE and nextCE at that
|
|
|
|
level. The relation * between the last item and nextCE is the same as the
|
|
|
|
strongest strength.
|
|
|
|
o Example: baseCE < a << b <<< q << c < d < e * nextCE(X,1)
|
|
|
|
? There are 3 primary items: a, d, e. Fit them into the primary gap.
|
|
|
|
Then fit b and c into the secondary gap between a and d, then fit q
|
|
|
|
into the tertiary gap between b and c.
|
|
|
|
|
|
|
|
o Example: baseCE << b <<< q << c * nextCE(X,2)
|
|
|
|
? There are 2 secondary items: b, c. Fit them into the secondary gap.
|
|
|
|
Then fit q into the tertiary gap between b and c.
|
|
|
|
o When incrementing primary values, we will not cross high byte
|
|
|
|
boundaries except where there is only a single-byte primary. That is to
|
|
|
|
ensure that the script reordering will continue to work.
|
|
|
|
*/
|
2001-03-30 00:23:46 +00:00
|
|
|
UCATableHeader *image = (UCATableHeader *)uprv_malloc(sizeof(UCATableHeader));
|
2002-07-02 15:10:30 +00:00
|
|
|
/* test for NULL */
|
2002-06-29 09:31:05 +00:00
|
|
|
if (image == NULL) {
|
|
|
|
*status = U_MEMORY_ALLOCATION_ERROR;
|
|
|
|
return NULL;
|
|
|
|
}
|
2001-03-30 00:23:46 +00:00
|
|
|
uprv_memcpy(image, src->UCA->image, sizeof(UCATableHeader));
|
2001-03-08 17:40:42 +00:00
|
|
|
|
|
|
|
for(i = 0; i<src->resultLen; i++) {
|
|
|
|
/* now we need to generate the CEs */
|
|
|
|
/* We stuff the initial value in the buffers, and increase the appropriate buffer */
|
|
|
|
/* According to strength */
|
2001-03-22 18:45:31 +00:00
|
|
|
if(U_SUCCESS(*status)) {
|
2003-09-11 22:35:13 +00:00
|
|
|
if(src->lh[i].first) { // if there are any elements
|
|
|
|
// due to the way parser works, subsequent tailorings
|
|
|
|
// may remove all the elements from a sequence, therefore
|
|
|
|
// leaving an empty tailoring sequence.
|
|
|
|
ucol_initBuffers(src, &src->lh[i], status);
|
|
|
|
}
|
2001-03-22 18:45:31 +00:00
|
|
|
}
|
2001-10-03 01:48:06 +00:00
|
|
|
if(U_FAILURE(*status)) {
|
|
|
|
return NULL;
|
|
|
|
}
|
2001-08-10 20:30:44 +00:00
|
|
|
|
2001-03-08 17:40:42 +00:00
|
|
|
}
|
|
|
|
|
2001-04-19 17:08:07 +00:00
|
|
|
if(src->varTop != NULL) { /* stuff the variable top value */
|
|
|
|
src->opts->variableTopValue = (*(src->varTop->CEs))>>16;
|
|
|
|
/* remove it from the list */
|
2001-06-06 20:20:43 +00:00
|
|
|
if(src->varTop->listHeader->first == src->varTop) { /* first in list */
|
|
|
|
src->varTop->listHeader->first = src->varTop->next;
|
2001-04-19 17:08:07 +00:00
|
|
|
}
|
2001-06-06 20:20:43 +00:00
|
|
|
if(src->varTop->listHeader->last == src->varTop) { /* first in list */
|
|
|
|
src->varTop->listHeader->last = src->varTop->previous;
|
2001-04-19 17:08:07 +00:00
|
|
|
}
|
|
|
|
if(src->varTop->next != NULL) {
|
|
|
|
src->varTop->next->previous = src->varTop->previous;
|
|
|
|
}
|
|
|
|
if(src->varTop->previous != NULL) {
|
|
|
|
src->varTop->previous->next = src->varTop->next;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
|
2003-07-24 23:23:19 +00:00
|
|
|
tempUCATable *t = uprv_uca_initTempTable(image, src->opts, src->UCA, NOT_FOUND_TAG, NOT_FOUND_TAG, status);
|
2001-03-08 17:40:42 +00:00
|
|
|
|
|
|
|
|
|
|
|
/* After this, we have assigned CE values to all regular CEs */
|
|
|
|
/* now we will go through list once more and resolve expansions, */
|
|
|
|
/* make UCAElements structs and add them to table */
|
|
|
|
for(i = 0; i<src->resultLen; i++) {
|
|
|
|
/* now we need to generate the CEs */
|
|
|
|
/* We stuff the initial value in the buffers, and increase the appropriate buffer */
|
|
|
|
/* According to strength */
|
2001-03-22 18:45:31 +00:00
|
|
|
if(U_SUCCESS(*status)) {
|
2001-06-06 20:20:43 +00:00
|
|
|
ucol_createElements(src, t, &src->lh[i], status);
|
2001-03-22 18:45:31 +00:00
|
|
|
}
|
2001-03-08 17:40:42 +00:00
|
|
|
}
|
|
|
|
|
2002-03-07 19:23:21 +00:00
|
|
|
UCAElements el;
|
|
|
|
el.isThai = FALSE;
|
|
|
|
el.prefixSize = 0;
|
|
|
|
el.prefixChars[0] = 0;
|
|
|
|
|
|
|
|
/* add latin-1 stuff */
|
2002-10-16 22:34:16 +00:00
|
|
|
ucol_uprv_bld_copyRangeFromUCA(src, t, 0, 0xFF, status);
|
|
|
|
|
|
|
|
/* add stuff for copying */
|
|
|
|
if(src->copySet != NULL) {
|
|
|
|
int32_t i = 0;
|
|
|
|
UnicodeSet *set = (UnicodeSet *)src->copySet;
|
|
|
|
for(i = 0; i < set->getRangeCount(); i++) {
|
|
|
|
ucol_uprv_bld_copyRangeFromUCA(src, t, set->getRangeStart(i), set->getRangeEnd(i), status);
|
2001-03-08 17:40:42 +00:00
|
|
|
}
|
2002-03-07 19:23:21 +00:00
|
|
|
}
|
2001-05-15 17:39:41 +00:00
|
|
|
|
2002-03-07 19:23:21 +00:00
|
|
|
if(U_SUCCESS(*status)) {
|
|
|
|
/* copy contractions from the UCA - this is felt mostly for cyrillic*/
|
|
|
|
|
|
|
|
uint32_t tailoredCE = UCOL_NOT_FOUND;
|
2002-07-02 22:32:14 +00:00
|
|
|
//UChar *conts = (UChar *)((uint8_t *)src->UCA->image + src->UCA->image->UCAConsts+sizeof(UCAConstants));
|
|
|
|
UChar *conts = (UChar *)((uint8_t *)src->UCA->image + src->UCA->image->contractionUCACombos);
|
2002-03-07 19:23:21 +00:00
|
|
|
UCollationElements *ucaEl = ucol_openElements(src->UCA, NULL, 0, status);
|
|
|
|
while(*conts != 0) {
|
|
|
|
/*tailoredCE = ucmpe32_get(t->mapping, *conts);*/
|
|
|
|
tailoredCE = utrie_get32(t->mapping, *conts, NULL);
|
|
|
|
if(tailoredCE != UCOL_NOT_FOUND) {
|
|
|
|
UBool needToAdd = TRUE;
|
|
|
|
if(isCntTableElement(tailoredCE)) {
|
|
|
|
if(uprv_cnttab_isTailored(t->contractions, tailoredCE, conts+1, status) == TRUE) {
|
|
|
|
needToAdd = FALSE;
|
2001-06-06 20:20:43 +00:00
|
|
|
}
|
2002-03-07 19:23:21 +00:00
|
|
|
}
|
2002-10-16 22:34:16 +00:00
|
|
|
if(src->removeSet != NULL && uset_contains(src->removeSet, *conts)) {
|
|
|
|
needToAdd = FALSE;
|
|
|
|
}
|
2001-05-15 17:39:41 +00:00
|
|
|
|
2002-03-07 19:23:21 +00:00
|
|
|
if(needToAdd == TRUE) { // we need to add if this contraction is not tailored.
|
|
|
|
el.prefix = el.prefixChars;
|
|
|
|
el.prefixSize = 0;
|
|
|
|
el.cPoints = el.uchars;
|
|
|
|
el.noOfCEs = 0;
|
|
|
|
el.uchars[0] = *conts;
|
|
|
|
el.uchars[1] = *(conts+1);
|
|
|
|
if(*(conts+2)!=0) {
|
|
|
|
el.uchars[2] = *(conts+2);
|
|
|
|
el.cSize = 3;
|
|
|
|
} else {
|
|
|
|
el.cSize = 2;
|
2001-06-05 22:52:56 +00:00
|
|
|
}
|
2002-03-07 19:23:21 +00:00
|
|
|
ucol_setText(ucaEl, el.uchars, el.cSize, status);
|
2004-01-17 00:06:19 +00:00
|
|
|
while ((int32_t)(el.CEs[el.noOfCEs] = ucol_next(ucaEl, status)) != UCOL_NULLORDER) {
|
2002-03-07 19:23:21 +00:00
|
|
|
el.noOfCEs++;
|
|
|
|
}
|
|
|
|
uprv_uca_addAnElement(t, &el, status);
|
2001-06-05 22:52:56 +00:00
|
|
|
}
|
2002-03-07 19:23:21 +00:00
|
|
|
|
2002-10-16 22:34:16 +00:00
|
|
|
} else if(src->removeSet != NULL && uset_contains(src->removeSet, *conts)) {
|
|
|
|
ucol_uprv_bld_copyRangeFromUCA(src, t, *conts, *conts, status);
|
2001-06-05 22:52:56 +00:00
|
|
|
}
|
2002-03-07 19:23:21 +00:00
|
|
|
conts+=3;
|
|
|
|
}
|
|
|
|
ucol_closeElements(ucaEl);
|
2002-06-13 18:29:49 +00:00
|
|
|
}
|
2001-06-06 20:20:43 +00:00
|
|
|
|
2002-07-02 22:32:14 +00:00
|
|
|
// Add completely ignorable elements
|
|
|
|
utrie_enum(t->UCA->mapping, NULL, _processUCACompleteIgnorables, t);
|
|
|
|
|
|
|
|
|
2002-06-13 18:29:49 +00:00
|
|
|
// canonical closure
|
|
|
|
uprv_uca_canonicalClosure(t, status);
|
2001-11-06 22:55:29 +00:00
|
|
|
|
2001-03-08 17:40:42 +00:00
|
|
|
|
2001-05-15 17:39:41 +00:00
|
|
|
/* still need to produce compatibility closure */
|
|
|
|
|
|
|
|
UCATableHeader *myData = uprv_uca_assembleTable(t, status);
|
2001-03-08 17:40:42 +00:00
|
|
|
|
|
|
|
uprv_uca_closeTempTable(t);
|
2001-06-06 22:03:58 +00:00
|
|
|
uprv_free(image);
|
2001-03-08 17:40:42 +00:00
|
|
|
|
|
|
|
return myData;
|
|
|
|
}
|
|
|
|
|
2001-08-30 02:59:19 +00:00
|
|
|
UBool
|
|
|
|
ucol_bld_cleanup(void)
|
|
|
|
{
|
|
|
|
udata_close(invUCA_DATA_MEM);
|
|
|
|
invUCA_DATA_MEM = NULL;
|
2003-09-24 00:17:15 +00:00
|
|
|
_staticInvUCA = NULL;
|
2001-08-30 02:59:19 +00:00
|
|
|
return TRUE;
|
|
|
|
}
|
2001-03-08 17:40:42 +00:00
|
|
|
|
2003-04-30 00:49:01 +00:00
|
|
|
U_CAPI const InverseUCATableHeader * U_EXPORT2
|
2001-08-30 02:59:19 +00:00
|
|
|
ucol_initInverseUCA(UErrorCode *status)
|
|
|
|
{
|
|
|
|
if(U_FAILURE(*status)) return NULL;
|
2003-05-05 21:23:11 +00:00
|
|
|
|
|
|
|
umtx_lock(NULL);
|
2003-09-24 00:17:15 +00:00
|
|
|
UBool f = (_staticInvUCA == NULL);
|
2003-05-05 21:23:11 +00:00
|
|
|
umtx_unlock(NULL);
|
2001-08-30 02:59:19 +00:00
|
|
|
|
2003-05-05 21:23:11 +00:00
|
|
|
if(f) {
|
2003-04-30 00:49:01 +00:00
|
|
|
InverseUCATableHeader *newInvUCA = NULL;
|
2001-08-30 02:59:19 +00:00
|
|
|
UDataMemory *result = udata_openChoice(NULL, INVC_DATA_TYPE, INVC_DATA_NAME, isAcceptableInvUCA, NULL, status);
|
|
|
|
|
|
|
|
if(U_FAILURE(*status)) {
|
|
|
|
if (result) {
|
|
|
|
udata_close(result);
|
|
|
|
}
|
2003-04-30 00:49:01 +00:00
|
|
|
// This is not needed, as we are talking about
|
|
|
|
// memory we got from UData
|
|
|
|
//uprv_free(newInvUCA);
|
2001-08-30 02:59:19 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
if(result != NULL) { /* It looks like sometimes we can fail to find the data file */
|
2003-04-30 00:49:01 +00:00
|
|
|
newInvUCA = (InverseUCATableHeader *)udata_getMemory(result);
|
|
|
|
UCollator *UCA = ucol_initUCA(status);
|
|
|
|
// UCA versions of UCA and inverse UCA should match
|
|
|
|
if(uprv_memcmp(newInvUCA->UCAVersion, UCA->image->UCAVersion, sizeof(UVersionInfo)) != 0) {
|
|
|
|
*status = U_INVALID_FORMAT_ERROR;
|
|
|
|
udata_close(result);
|
|
|
|
return NULL;
|
|
|
|
}
|
2001-08-30 02:59:19 +00:00
|
|
|
|
|
|
|
umtx_lock(NULL);
|
2003-09-24 00:17:15 +00:00
|
|
|
if(_staticInvUCA == NULL) {
|
|
|
|
_staticInvUCA = newInvUCA;
|
2001-08-30 02:59:19 +00:00
|
|
|
invUCA_DATA_MEM = result;
|
|
|
|
result = NULL;
|
|
|
|
newInvUCA = NULL;
|
|
|
|
}
|
|
|
|
umtx_unlock(NULL);
|
|
|
|
|
|
|
|
if(newInvUCA != NULL) {
|
|
|
|
udata_close(result);
|
2003-04-30 00:49:01 +00:00
|
|
|
// This is not needed, as we are talking about
|
|
|
|
// memory we got from UData
|
|
|
|
//uprv_free(newInvUCA);
|
2001-08-30 02:59:19 +00:00
|
|
|
}
|
|
|
|
else {
|
2001-08-31 02:13:00 +00:00
|
|
|
ucln_i18n_registerCleanup();
|
2001-08-30 02:59:19 +00:00
|
|
|
}
|
|
|
|
}
|
2001-03-08 17:40:42 +00:00
|
|
|
}
|
2003-09-24 00:17:15 +00:00
|
|
|
return _staticInvUCA;
|
2001-03-08 17:40:42 +00:00
|
|
|
}
|
|
|
|
|
2002-09-20 01:54:48 +00:00
|
|
|
#endif /* #if !UCONFIG_NO_COLLATION */
|