Given: <table> <caption>My caption</caption> <tr><td>1</td><td>2</td></tr> <tr><td>3</td><td>4</td></tr> </table> Expected results: An accessible object with ATK_ROLE_CAPTION would appear as a descendant of the accessible table. Actual results: There is no accessible object with ATK_ROLE_CAPTION (or any other role) appearing as a descendant of the accessible table.
<rdar://problem/19068639>
Expanding the summary to include rows. Unlike captions, which were always expected for exposure on ATK, rows were deliberately not exposed. But more and more table rows are showing up in other ATK implementations, so we might as well start exposing them. And if we're going to rework ATK table exposure, might as well do it all at once.
Created attachment 242225 [details] Patch
Comment on attachment 242225 [details] Patch Chris and/or Mario: If either of you have time to review this, it would be awesome. (At the moment, the gtk-wk2 bot is hosed, but that's independent of this patch.)
Comment on attachment 242225 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=242225&action=review > Source/WebCore/accessibility/AccessibilityTable.cpp:365 > + Node* tableElement = m_renderer->node(); There might be a tableElement() method on AXTable. If not we should probably add one > Source/WebCore/accessibility/atk/WebKitAccessibleWrapperAtk.cpp:352 > + if (static_cast<unsigned>(index) >= children.size()) Size() returns a size_t so you might want to cast to that
Created attachment 242450 [details] Patch
Comment on attachment 242450 [details] Patch (In reply to comment #5) > Comment on attachment 242225 [details] > Patch > > View in context: > https://bugs.webkit.org/attachment.cgi?id=242225&action=review > > > Source/WebCore/accessibility/AccessibilityTable.cpp:365 > > + Node* tableElement = m_renderer->node(); > > There might be a tableElement() method on AXTable. If not we should probably > add one There is one already. Change made. > > Source/WebCore/accessibility/atk/WebKitAccessibleWrapperAtk.cpp:352 > > + if (static_cast<unsigned>(index) >= children.size()) > > Size() returns a size_t so you might want to cast to that Change made. Thanks!
Comment on attachment 242450 [details] Patch Clearing flags on attachment: 242450 Committed r176706: <http://trac.webkit.org/changeset/176706>
All reviewed patches have been landed. Closing bug.