AuroraMiddleware/scuffed-code
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

ICU-1682 canonical closure for contractions now uses canonical iterator

Browse Source 
X-SVN-Rev: 7953
This commit is contained in:
Vladimir Weinstein 2002-03-12 20:20:08 +00:00
parent ad8cbb01bf
commit 5d8eed9047
1 changed files with 20 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

22
icu4c/source/i18n/ucol_elm.cpp
View File

@ -27,6 +27,8 @@
#include "unicode/uchar.h"
#include "unormimp.h"
#include "cmemory.h"
#include "unicode/unistr.h"
#include "unicode/caniter.h"
U_NAMESPACE_BEGIN
@ -997,9 +999,22 @@ uprv_uca_addAnElement(tempUCATable *t, UCAElements *element, UErrorCode *status)
uprv_free(composed);
}
CE = uprv_uca_finalizeAddition(t, element, status);
// We need to use the canonical iterator here
// the way we do it is to generate the canonically equivalent strings
// for the contraction and then add the sequences that pass FCD check
if(element->cSize > 1 && !(element->cSize==2 && UTF16_IS_LEAD(element->cPoints[0]) && UTF16_IS_TRAIL(element->cPoints[1]))) { // this is a contraction, we should check whether a composed form should also be included
UnicodeString source(element->cPoints, element->cSize);
CanonicalIterator it(source, *status);
source = it.next();
while(source.length() > 0) {
if(Normalizer::quickCheck(source, UNORM_FCD, *status) != UNORM_NO) {
element->cSize = source.extract(element->cPoints, 128, *status);
uprv_uca_finalizeAddition(t, element, status);
}
source = it.next();
}
#if 0
CE = uprv_uca_finalizeAddition(t, element, status);
UChar composed[256];
uint32_t compLen = unorm_normalize(element->cPoints, element->cSize, UNORM_NFC, 0, composed, 256, status);;
@ -1013,6 +1028,9 @@ uprv_uca_addAnElement(tempUCATable *t, UCAElements *element, UErrorCode *status)
uprv_memcpy(element->cPoints, composed, element->cSize*sizeof(UChar));
uprv_uca_finalizeAddition(t, element, status);
}
#endif
} else {
CE = uprv_uca_finalizeAddition(t, element, status);
}
return CE;