Magic-Mirror/gecko-dev
1
0
Fork 0
mirror of https://github.com/mozilla/gecko-dev.git synced 2025-04-11 18:32:31 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

bug 672472 - convert hyphenation-point offsets correctly from Unicode characters to UTF16 code units. r=smontagu

Browse Source
This commit is contained in:
Jonathan Kew 2011-07-20 11:15:06 +01:00
parent c62529f8bd
commit 938babc448
1 changed files with 33 additions and 26 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

59
intl/hyphenation/src/nsHyphenator.cpp
View File

@ -85,8 +85,19 @@ nsHyphenator::Hyphenate(const nsAString& aString,
PRBool inWord = PR_FALSE; PRBool inWord = PR_FALSE;
PRUint32 wordStart = 0, wordLimit = 0; PRUint32 wordStart = 0, wordLimit = 0;
for (PRUint32 i = 0; i < aString.Length(); i++) { PRUint32 chLen;
PRUnichar ch = aString[i]; for (PRUint32 i = 0; i < aString.Length(); i += chLen) {
PRUint32 ch = aString[i];
chLen = 1;
if (NS_IS_HIGH_SURROGATE(ch)) {
if (i + 1 < aString.Length() && NS_IS_LOW_SURROGATE(aString[i+1])) {
ch = SURROGATE_TO_UCS4(ch, aString[i+1]);
chLen = 2;
} else {
NS_WARNING("unpaired surrogate found during hyphenation");
}
}
nsIUGenCategory::nsUGenCategory cat = mCategories->Get(ch); nsIUGenCategory::nsUGenCategory cat = mCategories->Get(ch);
if (cat == nsIUGenCategory::kLetter || cat == nsIUGenCategory::kMark) { if (cat == nsIUGenCategory::kLetter || cat == nsIUGenCategory::kMark) {
@ -94,14 +105,15 @@ nsHyphenator::Hyphenate(const nsAString& aString,
inWord = PR_TRUE; inWord = PR_TRUE;
wordStart = i; wordStart = i;
} }
wordLimit = i + 1; wordLimit = i + chLen;
if (i < aString.Length() - 1) { if (i + chLen < aString.Length()) {
continue; continue;
} }
} }
if (inWord) { if (inWord) {
NS_ConvertUTF16toUTF8 utf8(aString.BeginReading() + wordStart, const PRUnichar *begin = aString.BeginReading();
NS_ConvertUTF16toUTF8 utf8(begin + wordStart,
wordLimit - wordStart); wordLimit - wordStart);
nsAutoTArray<char,200> utf8hyphens; nsAutoTArray<char,200> utf8hyphens;
utf8hyphens.SetLength(utf8.Length() + 5); utf8hyphens.SetLength(utf8.Length() + 5);
@ -113,29 +125,24 @@ nsHyphenator::Hyphenate(const nsAString& aString,
utf8hyphens.Elements(), nsnull, utf8hyphens.Elements(), nsnull,
&rep, &pos, &cut); &rep, &pos, &cut);
if (!err) { if (!err) {
PRUint32 utf16offset = wordStart; // Surprisingly, hnj_hyphen_hyphenate2 converts the 'hyphens' buffer
const char *cp = utf8.BeginReading(); // from utf8 code unit indexing (which would match the utf8 input
while (cp < utf8.EndReading()) { // string directly) to Unicode character indexing.
if (UTF8traits::isASCII(*cp)) { // single-byte utf8 char // We then need to convert this to utf16 code unit offsets for Gecko.
cp++; const char *hyphPtr = utf8hyphens.Elements();
utf16offset++; const PRUnichar *cur = begin + wordStart;
} else if (UTF8traits::is2byte(*cp)) { // 2-byte sequence const PRUnichar *end = begin + wordLimit;
cp += 2; while (cur < end) {
utf16offset++; if (*hyphPtr & 0x01) {
} else if (UTF8traits::is3byte(*cp)) { // 3-byte sequence aHyphens[cur - begin] = PR_TRUE;
cp += 3;
utf16offset++;
} else { // must be a 4-byte sequence (no need to check validity,
// as this was just created with NS_ConvertUTF16toUTF8)
NS_ASSERTION(UTF8traits::is4byte(*cp), "unexpected utf8 byte");
cp += 4;
utf16offset += 2;
} }
NS_ASSERTION(cp <= utf8.EndReading(), "incomplete utf8 string?"); cur++;
NS_ASSERTION(utf16offset <= aString.Length(), "length mismatch?"); if (cur < end && NS_IS_LOW_SURROGATE(*cur) &&
if (utf8hyphens[cp - utf8.BeginReading() - 1] & 0x01) { NS_IS_HIGH_SURROGATE(*(cur-1)))
aHyphens[utf16offset - 1] = PR_TRUE; {
cur++;
} }
hyphPtr++;
} }
} }
} }