[SCM] WebKit Debian packaging branch, debian/unstable, updated. debian/1.1.15-1-40151-g37bb677
darin
darin at 268f45cc-cd09-0410-ab3c-d52691b4dbfc
Sat Sep 26 08:16:24 UTC 2009
The following commit has been merged in the debian/unstable branch:
commit 3df10b4867fa7e6df5ac6417517f7c2e74b2fd11
Author: darin <darin at 268f45cc-cd09-0410-ab3c-d52691b4dbfc>
Date: Fri Dec 5 01:27:10 2003 +0000
Reviewed by Richard.
- fixed 3497879: REGRESSION (100-115): all non-BMP characters (including Deseret) are broken
* WebCoreSupport.subproj/WebTextRenderer.m:
(-[WebTextRenderer _CG_pointToOffset:style:position:reversed:]): Bump offset by getting it from
the iterator; don't assume we can just bump it by one each time. It would be even nicer to have
a bit more abstraction.
(initializeCharacterWidthIterator): Remove call to initializeCharacterShapeIterator.
(widthForNextCharacter): Move handling of surrogate pairs (non-BMP) in here and unify it with
the handling of BMP characters; this removes the broken code that was returning the wrong font,
and changes us to use the code that was already doing the right thing for the surrogate pair case.
Also get rid of the use of 0 width to mean "no glyph", which fixes the doubled glyph problem.
Also got rid of remnants of use of the shape iterator.
* Misc.subproj/WebUnicode.h: Remove obsolete shape iterator.
* Misc.subproj/WebUnicode.m: Ditto.
git-svn-id: http://svn.webkit.org/repository/webkit/trunk@5702 268f45cc-cd09-0410-ab3c-d52691b4dbfc
diff --git a/WebKit/ChangeLog b/WebKit/ChangeLog
index dcefeab..0571419 100644
--- a/WebKit/ChangeLog
+++ b/WebKit/ChangeLog
@@ -1,3 +1,23 @@
+2003-12-04 Darin Adler <darin at apple.com>
+
+ Reviewed by Richard.
+
+ - fixed 3497879: REGRESSION (100-115): all non-BMP characters (including Deseret) are broken
+
+ * WebCoreSupport.subproj/WebTextRenderer.m:
+ (-[WebTextRenderer _CG_pointToOffset:style:position:reversed:]): Bump offset by getting it from
+ the iterator; don't assume we can just bump it by one each time. It would be even nicer to have
+ a bit more abstraction.
+ (initializeCharacterWidthIterator): Remove call to initializeCharacterShapeIterator.
+ (widthForNextCharacter): Move handling of surrogate pairs (non-BMP) in here and unify it with
+ the handling of BMP characters; this removes the broken code that was returning the wrong font,
+ and changes us to use the code that was already doing the right thing for the surrogate pair case.
+ Also get rid of the use of 0 width to mean "no glyph", which fixes the doubled glyph problem.
+ Also got rid of remnants of use of the shape iterator.
+
+ * Misc.subproj/WebUnicode.h: Remove obsolete shape iterator.
+ * Misc.subproj/WebUnicode.m: Ditto.
+
=== Safari-116 ===
2003-12-04 Darin Adler <darin at apple.com>
diff --git a/WebKit/Misc.subproj/WebUnicode.h b/WebKit/Misc.subproj/WebUnicode.h
index 7cf7cd7..adc8623 100644
--- a/WebKit/Misc.subproj/WebUnicode.h
+++ b/WebKit/Misc.subproj/WebUnicode.h
@@ -31,13 +31,3 @@ enum {
#define LowSurrogatePair(c) (((c - 0x10000)&0x3ff) + 0xdc00)
#define IsHighSurrogatePair(c) (( c & 0xFC00 ) == HighSurrogateRangeStart )
#define IsLowSurrogatePair(c) (( c & 0xFC00 ) == LowSurrogateRangeStart )
-
-typedef struct CharacterShapeIterator CharacterShapeIterator;
-struct CharacterShapeIterator
-{
- const WebCoreTextRun *run;
- const UniChar *currentCharacter;
-};
-extern bool initializeCharacterShapeIterator (CharacterShapeIterator *iterator, const WebCoreTextRun *run);
-extern UniChar shapeForNextCharacter (CharacterShapeIterator *iterator);
-extern bool hasShapeForNextCharacter (const CharacterShapeIterator *iterator);
diff --git a/WebKit/Misc.subproj/WebUnicode.m b/WebKit/Misc.subproj/WebUnicode.m
index 834fafd..f2ab8f5 100644
--- a/WebKit/Misc.subproj/WebUnicode.m
+++ b/WebKit/Misc.subproj/WebUnicode.m
@@ -46,30 +46,3 @@ void WebKitInitializeUnicode(void)
WebCoreUnicodeLowerFunction = _unicodeLower;
WebCoreUnicodeUpperFunction = _unicodeUpper;
}
-
-bool hasShapeForNextCharacter (const CharacterShapeIterator *iterator)
-{
- if (iterator->currentCharacter - iterator->run->characters >= iterator->run->to)
- return false;
- return true;
-}
-
-UniChar shapeForNextCharacter (CharacterShapeIterator *iterator)
-{
- if (!hasShapeForNextCharacter(iterator))
- return 0;
-
- UniChar currentCharacter = *iterator->currentCharacter;
- iterator->currentCharacter++;
-
- return currentCharacter;
-}
-
-// Assumes input characters are logically ordered.
-bool initializeCharacterShapeIterator (CharacterShapeIterator *iterator, const WebCoreTextRun *run)
-{
- iterator->currentCharacter = run->characters + run->from;
- iterator->run = run;
-
- return true;
-}
diff --git a/WebKit/WebCoreSupport.subproj/WebTextRenderer.m b/WebKit/WebCoreSupport.subproj/WebTextRenderer.m
index 6ab9646..c36d4a1 100644
--- a/WebKit/WebCoreSupport.subproj/WebTextRenderer.m
+++ b/WebKit/WebCoreSupport.subproj/WebTextRenderer.m
@@ -104,8 +104,6 @@ struct CharacterWidthIterator
const WebCoreTextRun *run;
const WebCoreTextStyle *style;
unsigned currentCharacter;
- CharacterShapeIterator shapeIterator;
- unsigned needsShaping;
float runWidthSoFar;
float widthToStart;
int padding;
@@ -1605,7 +1603,7 @@ static const char *joiningNames[] = {
if (reversed) {
width = [self floatWidthForRun:run style:style widths:nil];
delta -= width;
- while(offset < run->length) {
+ while (offset < run->length) {
float w = widthForNextCharacter(&widthIterator, 0, 0);
if (w == INVALID_WIDTH){
// Something very bad happened, like we only have half of a surrogate pair.
@@ -1620,7 +1618,7 @@ static const char *joiningNames[] = {
break;
delta += w;
}
- offset++;
+ offset = widthIterator.currentCharacter;
}
}
} else {
@@ -1639,7 +1637,7 @@ static const char *joiningNames[] = {
break;
delta -= w;
}
- offset++;
+ offset = widthIterator.currentCharacter;
}
}
}
@@ -1731,7 +1729,6 @@ static inline SubstituteFontWidthMap *mapForSubstituteFont(WebTextRenderer *rend
static void initializeCharacterWidthIterator (CharacterWidthIterator *iterator, WebTextRenderer *renderer, const WebCoreTextRun *run , const WebCoreTextStyle *style)
{
- iterator->needsShaping = initializeCharacterShapeIterator (&iterator->shapeIterator, run);
iterator->renderer = renderer;
iterator->run = run;
iterator->style = style;
@@ -1776,38 +1773,6 @@ static void initializeCharacterWidthIterator (CharacterWidthIterator *iterator,
iterator->widthToStart = 0;
}
-static float widthAndGlyphForSurrogate (WebTextRenderer *renderer, UniChar high, UniChar low, ATSGlyphRef *glyphID, NSString **families, NSFont **fontUsed)
-{
- UnicodeChar uc = UnicodeValueForSurrogatePair(high, low);
- float width;
-
- *glyphID = glyphForUnicodeCharacter(renderer->unicodeCharacterToGlyphMap, uc, fontUsed);
- if (*glyphID == nonGlyphID) {
- *glyphID = [renderer _extendUnicodeCharacterToGlyphMapToInclude: uc];
- }
-
- if (*glyphID == 0){
- UniChar surrogates[2];
- unsigned clusterLength;
-
- clusterLength = 2;
- surrogates[0] = high;
- surrogates[1] = low;
- *fontUsed = [renderer _substituteFontForCharacters:&surrogates[0] length: clusterLength families: families];
- if (*fontUsed){
- WebTextRenderer *substituteRenderer = [[WebTextRendererFactory sharedFactory] rendererWithFont:*fontUsed usingPrinterFont:renderer->usingPrinterFont];
- *glyphID = glyphForUnicodeCharacter(substituteRenderer->unicodeCharacterToGlyphMap, uc, fontUsed);
- if (*glyphID == nonGlyphID)
- *glyphID = [substituteRenderer _extendUnicodeCharacterToGlyphMapToInclude: uc];
- return widthForGlyph (substituteRenderer, *glyphID, *fontUsed);
- }
- }
-
- width = widthForGlyph (renderer, *glyphID, *fontUsed);
-
- return width;
-}
-
static inline float ceilCurrentWidth (CharacterWidthIterator *iterator)
{
float delta = CEIL_TO_INT(iterator->widthToStart + iterator->runWidthSoFar) - (iterator->widthToStart + iterator->runWidthSoFar);
@@ -1816,113 +1781,102 @@ static inline float ceilCurrentWidth (CharacterWidthIterator *iterator)
}
// Return INVALID_WIDTH if an error is encountered or we're at the end of the range in the run.
-static float widthForNextCharacter (CharacterWidthIterator *iterator, ATSGlyphRef *glyphUsed, NSFont **fontUsed)
+static float widthForNextCharacter(CharacterWidthIterator *iterator, ATSGlyphRef *glyphUsed, NSFont **fontUsed)
{
WebTextRenderer *renderer = iterator->renderer;
const WebCoreTextRun *run = iterator->run;
- UniChar c;
- unsigned offset = iterator->currentCharacter;
- WebGlyphWidth width;
- NSFont *_fontUsed = 0;
+ unsigned currentCharacter = iterator->currentCharacter;
+
+ NSFont *_fontUsed = nil;
ATSGlyphRef _glyphUsed;
- BOOL useSmallCapsFont = NO;
if (!fontUsed)
fontUsed = &_fontUsed;
if (!glyphUsed)
glyphUsed = &_glyphUsed;
- if (offset >= (unsigned)run->to)
- // Error! Offset specified beyond end of run.
+ if (currentCharacter >= (unsigned)run->to)
+ // Error! Offset specified beyond end of run.
return INVALID_WIDTH;
-
- // Determine if the string requires any shaping, i.e. Arabic.
- if (iterator->needsShaping)
- c = shapeForNextCharacter(&iterator->shapeIterator);
- else
- c = run->characters[offset];
- iterator->currentCharacter++;
-
- // It's legit to sometimes get 0 from the shape iterator, return a zero width.
- if (c == 0)
- return 0;
+
+ const UniChar *cp = &run->characters[currentCharacter];
+ UnicodeChar c = *cp;
+
+ if (IsLowSurrogatePair(c))
+ return INVALID_WIDTH;
+
+ // Do we have a surrogate pair? If so, determine the full Unicode (32 bit)
+ // code point before glyph lookup.
+ unsigned clusterLength = 1;
+ if (IsHighSurrogatePair(c)) {
+ // Make sure we have another character and it's a low surrogate.
+ UniChar low;
+ if (currentCharacter + 1 >= run->length || !IsLowSurrogatePair((low = cp[1]))) {
+ // Error! The second component of the surrogate pair is missing.
+ return INVALID_WIDTH;
+ }
+
+ c = UnicodeValueForSurrogatePair(c, low);
+ clusterLength = 2;
+ }
// If small-caps convert lowercase to upper.
+ BOOL useSmallCapsFont = NO;
if (renderer->isSmallCapsRenderer) {
if (!u_isUUppercase(c)) {
// Only use small cap font if the the uppercase version of the character
// is different than the lowercase.
- UniChar newC = u_toupper(c);
+ UnicodeChar newC = u_toupper(c);
if (newC != c) {
useSmallCapsFont = YES;
c = newC;
}
}
}
-
- // Get a glyph for the next characters. Somewhat complicated by surrogate
- // pairs.
- //
- // Do we have a surrogate pair? If so, determine the full Unicode (32bit)
- // code point before glyph lookup. We only provide a width when the offset
- // specifies the first component of the surrogate pair.
- if (c >= HighSurrogateRangeStart && c <= HighSurrogateRangeEnd) {
- UniChar high = c, low;
-
- // Make sure we have another character and it's a low surrogate.
- if (offset+1 >= run->length || !IsLowSurrogatePair((low = run->characters[offset+1]))) {
- // Error! The second component of the surrogate pair is missing.
- return INVALID_WIDTH;
- }
- width = widthAndGlyphForSurrogate (renderer, high, low, glyphUsed, iterator->style->families, fontUsed);
- }
- else if (c >= LowSurrogateRangeStart && c <= LowSurrogateRangeEnd) {
- // Return 0 width for second component of the surrogate pair.
- return 0;
- }
- else {
+ if (clusterLength == 1) {
*glyphUsed = glyphForCharacter(renderer->characterToGlyphMap, c, fontUsed);
-
if (*glyphUsed == nonGlyphID) {
- *glyphUsed = [renderer _extendCharacterToGlyphMapToInclude: c];
+ *glyphUsed = [renderer _extendCharacterToGlyphMapToInclude:c];
}
-
- // Check to see if we're rendering in 'small-caps' mode.
- // ASSUMPTION: We assume the same font in a smaller size has
- // the same glyphs as the large font.
- if (useSmallCapsFont) {
- if (*fontUsed == 0)
- *fontUsed = [renderer _smallCapsFont];
- else {
- // Potential for optimization. This path should only be taken if we're
- // using a cached substituted font.
- *fontUsed = [[NSFontManager sharedFontManager] convertFont:*fontUsed toSize:[*fontUsed pointSize] * SMALLCAPS_FONTSIZE_MULTIPLIER];
- }
+ } else {
+ *glyphUsed = glyphForUnicodeCharacter(renderer->unicodeCharacterToGlyphMap, c, fontUsed);
+ if (*glyphUsed == nonGlyphID) {
+ *glyphUsed = [renderer _extendUnicodeCharacterToGlyphMapToInclude:c];
}
+ }
- // Now that we have glyph and font get it's width. We special case spaces.
- // They are always an even integer width.
- if (*glyphUsed == renderer->spaceGlyph)
- width = renderer->adjustedSpaceWidth;
- else
- width = widthForGlyph (renderer, *glyphUsed, *fontUsed);
+ // Check to see if we're rendering in 'small-caps' mode.
+ // ASSUMPTION: We assume the same font in a smaller size has
+ // the same glyphs as the large font.
+ if (useSmallCapsFont) {
+ if (*fontUsed == nil)
+ *fontUsed = [renderer _smallCapsFont];
+ else {
+ // Potential for optimization. This path should only be taken if we're
+ // using a cached substituted font.
+ *fontUsed = [[NSFontManager sharedFontManager] convertFont:*fontUsed toSize:[*fontUsed pointSize] * SMALLCAPS_FONTSIZE_MULTIPLIER];
+ }
}
+ // Now that we have glyph and font get it's width. We special case spaces.
+ // They are always an even integer width.
+ WebGlyphWidth width;
+ if (*glyphUsed == renderer->spaceGlyph)
+ width = renderer->adjustedSpaceWidth;
+ else
+ width = widthForGlyph(renderer, *glyphUsed, *fontUsed);
+
// Try to find a substitute font if this font didn't have a glyph for a character in the
// string. If one isn't found we end up drawing and measuring the 0 glyph, usually a box.
if (*glyphUsed == 0 && iterator->style->attemptFontSubstitution) {
- UniChar _characters[1];
- NSFont *substituteFont;
-
- _characters[0] = c;
- substituteFont = [renderer _substituteFontForCharacters:_characters length:1 families:iterator->style->families];
+ NSFont *substituteFont = [renderer _substituteFontForCharacters:cp length:clusterLength families:iterator->style->families];
if (substituteFont) {
int cNumGlyphs = 0;
ATSGlyphRef localGlyphBuffer[4];
WebCoreTextRun clusterRun;
- WebCoreInitializeTextRun(&clusterRun, _characters, 1, 0, 1);
+ WebCoreInitializeTextRun(&clusterRun, cp, clusterLength, 0, clusterLength);
WebCoreTextStyle clusterStyle = *iterator->style;
clusterStyle.padding = 0;
clusterStyle.applyRounding = false;
@@ -1942,14 +1896,12 @@ static float widthForNextCharacter (CharacterWidthIterator *iterator, ATSGlyphRe
*fontUsed = substituteFont;
*glyphUsed = localGlyphBuffer[0];
- if (cNumGlyphs == 1 && localGlyphBuffer[0] != 0){
- [renderer _updateGlyphEntryForCharacter:_characters[0] glyphID:localGlyphBuffer[0] font:substituteFont];
+ if (clusterLength == 1 && cNumGlyphs == 1 && localGlyphBuffer[0] != 0){
+ [renderer _updateGlyphEntryForCharacter:c glyphID:localGlyphBuffer[0] font:substituteFont];
}
- else
- NSLog (@"Unable to find appropriate match for character\n");
}
}
-
+
if (!*fontUsed)
*fontUsed = renderer->font;
@@ -1959,7 +1911,7 @@ static float widthForNextCharacter (CharacterWidthIterator *iterator, ATSGlyphRe
// Account for padding. khtml uses space padding to justify text. We
// distribute the specified padding over the available spaces in the run.
- if (c == SPACE){
+ if (c == SPACE) {
if (iterator->padding > 0){
// Only use left over padding if note evenly divisible by
// number of spaces.
@@ -1975,19 +1927,23 @@ static float widthForNextCharacter (CharacterWidthIterator *iterator, ATSGlyphRe
// Account for word-spacing. We apply additional space between "words" by
// adding width to the space character.
- if (iterator->currentCharacter > 1 && !isSpace(run->characters[iterator->currentCharacter-2]))
+ if (currentCharacter > 0 && !isSpace(cp[-1]))
width += iterator->style->wordSpacing;
}
iterator->runWidthSoFar += width;
-
+
+ // Advance past the character we just dealt with.
+ currentCharacter += clusterLength;
+ iterator->currentCharacter = currentCharacter;
+
// Account for float/integer impedance mismatch between CG and khtml. "Words" (characters
// followed by a character defined by isSpace()) are always an integer width. We adjust the
// width of the last character of a "word" to ensure an integer width. When we move khtml to
// floats we can remove this (and related) hacks.
//
// Check to see if the next character is a space, if so, adjust.
- if (offset+1 < run->length && isSpace(run->characters[offset+1])) {
+ if (currentCharacter < run->length && isSpace(cp[clusterLength])) {
width += ceilCurrentWidth(iterator);
}
@@ -1995,7 +1951,7 @@ static float widthForNextCharacter (CharacterWidthIterator *iterator, ATSGlyphRe
// 1) The string is longer than one character
// 2) or the entire stringLength is one character
int len = run->to - run->from;
- if (iterator->currentCharacter >= (unsigned)run->to && (len > 1 || run->length == 1) && iterator->style->applyRounding){
+ if (currentCharacter >= (unsigned)run->to && (len > 1 || run->length == 1) && iterator->style->applyRounding){
width += ceilCurrentWidth(iterator);
}
--
WebKit Debian packaging
More information about the Pkg-webkit-commits
mailing list