diff options
author | Konstantin Ritt <ritt.ks@gmail.com> | 2014-04-10 13:50:53 +0300 |
---|---|---|
committer | The Qt Project <gerrit-noreply@qt-project.org> | 2014-04-14 06:43:57 +0200 |
commit | 0ec07b68ad34e135451dd5291732bf73d297ba0c (patch) | |
tree | 33a7cafbe977729a2d97beb70c18853a31203154 /src/gui | |
parent | ef77c16cb00ef3788e58da8b35cb91eaf2eaa591 (diff) |
Improve the Unicode script itemization implementation
Make it closer to the Unicode specs (UAX#24):
* Common now inherits the preceding character's script, if any;
* In a combining character sequence, if the base character is
of Common script, the entire sequence is treated like if it were
of the first non-Inherited, non-Common script in the sequence.
See http://www.unicode.org/reports/tr24/tr24-21.html for more details.
[ChangeLog][QtGui] Fixed regression in arabic text rendering.
Task-number: QTBUG-28813
Task-number: QTBUG-29930 (related)
Task-number: QTBUG-35836
Change-Id: Id85761965b08ca94c674d5f3613fe58b82b2ce9c
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@digia.com>
Reviewed-by: Ahmed Saidi <justroftest@gmail.com>
Diffstat (limited to 'src/gui')
-rw-r--r-- | src/gui/text/qtextengine.cpp | 17 |
1 files changed, 1 insertions, 16 deletions
diff --git a/src/gui/text/qtextengine.cpp b/src/gui/text/qtextengine.cpp index 967ba24fcf..34788dc4dc 100644 --- a/src/gui/text/qtextengine.cpp +++ b/src/gui/text/qtextengine.cpp @@ -122,20 +122,9 @@ private: return; const int end = start + length; for (int i = start + 1; i < end; ++i) { - // According to the unicode spec we should be treating characters in the Common script - // (punctuation, spaces, etc) as being the same script as the surrounding text for the - // purpose of splitting up text. This is important because, for example, a fullstop - // (0x2E) can be used to indicate an abbreviation and so must be treated as part of a - // word. Thus it must be passed along with the word in languages that have to calculate - // word breaks. For example the thai word "ครม." has no word breaks but the word "ครม" - // does. - // Unfortuntely because we split up the strings for both wordwrapping and for setting - // the font and because Japanese and Chinese are also aliases of the script "Common", - // doing this would break too many things. So instead we only pass the full stop - // along, and nothing else. if (m_analysis[i].bidiLevel == m_analysis[start].bidiLevel && m_analysis[i].flags == m_analysis[start].flags - && (m_analysis[i].script == m_analysis[start].script || m_string[i] == QLatin1Char('.')) + && m_analysis[i].script == m_analysis[start].script && m_analysis[i].flags < QScriptAnalysis::SpaceTabOrObject && i - start < MaxItemLength) continue; @@ -1515,26 +1504,22 @@ void QTextEngine::itemize() const while (uc < e) { switch (*uc) { case QChar::ObjectReplacementCharacter: - analysis->script = QChar::Script_Common; analysis->flags = QScriptAnalysis::Object; break; case QChar::LineSeparator: if (analysis->bidiLevel % 2) --analysis->bidiLevel; - analysis->script = QChar::Script_Common; analysis->flags = QScriptAnalysis::LineOrParagraphSeparator; if (option.flags() & QTextOption::ShowLineAndParagraphSeparators) *const_cast<ushort*>(uc) = 0x21B5; // visual line separator break; case QChar::Tabulation: - analysis->script = QChar::Script_Common; analysis->flags = QScriptAnalysis::Tab; analysis->bidiLevel = control.baseLevel(); break; case QChar::Space: case QChar::Nbsp: if (option.flags() & QTextOption::ShowTabsAndSpaces) { - analysis->script = QChar::Script_Common; analysis->flags = QScriptAnalysis::Space; analysis->bidiLevel = control.baseLevel(); break; |