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

ICU-1250 backport bug fixes from icu4j re: dangling lead surrogates and cursorpos matching

Browse Source 
X-SVN-Rev: 6012
This commit is contained in:
Alan Liu 2001-10-03 16:34:53 +00:00
parent 5da9d23c58
commit 2c1b306ea4
2 changed files with 13 additions and 14 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
icu4c/source/i18n/rbt_rule.cpp
View File

@ -491,13 +491,13 @@ UMatchDegree TransliterationRule::matchAndReplace(Replaceable& text,
if (segments == NULL) {
text.handleReplaceBetween(pos.start, keyLimit, output);
lenDelta = output.length() - (keyLimit - pos.start);
if (cursorPos >= 0 && cursorPos < keyLength) {
// Within the key, the cursor refers to 16-bit code units
if (cursorPos >= 0 && cursorPos <= output.length()) {
// Within the output string, the cursor refers to 16-bit code units
newStart = pos.start + cursorPos;
} else {
newStart = pos.start;
int32_t n = cursorPos;
// Outside the key, cursorPos counts code points
// Outside the output string, cursorPos counts code points
while (n > 0) {
newStart += UTF_CHAR_LENGTH(text.char32At(newStart));
--n;

21
icu4c/source/i18n/translit.cpp
View File

@ -253,17 +253,7 @@ void Transliterator::transliterate(Replaceable& text,
UChar32 insertion,
UErrorCode& status) const {
UnicodeString str(insertion);
if (UTF_IS_LEAD(insertion)) {
// Oops, the caller passed us a single lead surrogate. In
// general, we don't support this, but we'll do the caller a
// favor in the special case of LEAD followed by TRAIL
// insertion. Anything else won't work.
text.handleReplaceBetween(index.limit, index.limit, str);
++index.limit;
++index.contextLimit;
} else {
_transliterate(text, index, &str, status);
}
_transliterate(text, index, &str, status);
}
/**
@ -329,6 +319,15 @@ void Transliterator::_transliterate(Replaceable& text,
index.contextLimit += insertion->length();
}
if (index.limit > 0 &&
UTF_IS_LEAD(text.charAt(index.limit - 1))) {
// Oops, there is a dangling lead surrogate in the buffer.
// This will break most transliterators, since they will
// assume it is part of a pari. Don't transliterate until
// more text comes in.
return;
}
filteredTransliterate(text, index, TRUE);
#if 0