RESOLVED FIXED 79336
[Chromium] Add HarfBuzzShaperBase class
https://bugs.webkit.org/show_bug.cgi?id=79336
Summary [Chromium] Add HarfBuzzShaperBase class
Kenichi Ishibashi
Reported 2012-02-22 23:36:40 PST
I'd like add a base class of HarfBuzz text shaper to share the code between old HarfBuzz and HarfBuzz-ng, to address Evan's comment on https://bugs.webkit.org/show_bug.cgi?id=69826#c18.
Attachments
Patch (29.63 KB, patch)
2012-02-22 23:52 PST, Kenichi Ishibashi
no flags
Patch for landing (31.10 KB, patch)
2012-02-23 16:01 PST, Kenichi Ishibashi
no flags
Kenichi Ishibashi
Comment 1 2012-02-22 23:52:45 PST
Kenichi Ishibashi
Comment 2 2012-02-22 23:53:51 PST
Hi Evan, Could you please review the patch?
Evan Martin
Comment 3 2012-02-23 07:52:39 PST
Looks great, thanks for refactoring!
Tony Chang
Comment 4 2012-02-23 10:29:38 PST
Comment on attachment 128414 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=128414&action=review > Source/WebCore/WebCore.gypi:3472 > + 'platform/graphics/harfbuzz/HarfBuzzShaperBase.cpp', > + 'platform/graphics/harfbuzz/HarfBuzzShaperBase.h', Can you also include the cpp file in PlatformBlackBerry.cmake? > Source/WebCore/platform/graphics/harfbuzz/ComplexTextControllerHarfBuzz.cpp:303 > + if (static_cast<unsigned>(logClustersIndex) < m_item.item.length && isWordEnd(m_item.item.pos + logClustersIndex)) Should logClustersIndex be unsigned or size_t? That might remove some of the casts.
Kenichi Ishibashi
Comment 5 2012-02-23 16:01:26 PST
Created attachment 128584 [details] Patch for landing
Kenichi Ishibashi
Comment 6 2012-02-23 16:02:52 PST
Comment on attachment 128414 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=128414&action=review Thank you for review Evan and Tony! >> Source/WebCore/WebCore.gypi:3472 >> + 'platform/graphics/harfbuzz/HarfBuzzShaperBase.h', > > Can you also include the cpp file in PlatformBlackBerry.cmake? Done. >> Source/WebCore/platform/graphics/harfbuzz/ComplexTextControllerHarfBuzz.cpp:303 >> + if (static_cast<unsigned>(logClustersIndex) < m_item.item.length && isWordEnd(m_item.item.pos + logClustersIndex)) > > Should logClustersIndex be unsigned or size_t? That might remove some of the casts. Done.
WebKit Review Bot
Comment 7 2012-02-23 22:59:46 PST
Comment on attachment 128584 [details] Patch for landing Clearing flags on attachment: 128584 Committed r108733: <http://trac.webkit.org/changeset/108733>
WebKit Review Bot
Comment 8 2012-02-23 22:59:51 PST
All reviewed patches have been landed. Closing bug.
Note You need to log in before you can comment on or make changes to this bug.