Context Navigation

rendering/RenderFlow.cpp: (WebCore::RenderFlow::caretRect): Consider border, padding, and the width of the caret properly in the calculation of the caret's X position.

rendering/RenderBox.cpp: (WebCore::RenderBox::caretRect): Fixed some similar issues and rewrote this function for clarity. However, I suspect this function was and remains broken and is almost never called.

3:35 PM Changeset in webkit [13630] by darin

9 edits
3 adds in trunk

LayoutTests:

Reviewed by Justin.

test for http://bugzilla.opendarwin.org/show_bug.cgi?id=8106 REGRESSION (NativeTextField): New text fields don't scroll to the beginning when losing focus

fast/forms/input-text-scroll-left-on-blur-expected.png: Added.
fast/forms/input-text-scroll-left-on-blur-expected.txt: Added.
fast/forms/input-text-scroll-left-on-blur-expected.checksum: Added.
fast/forms/input-text-scroll-left-on-blur.html: Added.

WebCore:

Reviewed by Justin.

fix http://bugzilla.opendarwin.org/show_bug.cgi?id=8106 REGRESSION (NativeTextField): New text fields don't scroll to the beginning when losing focus

Test: fast/forms/input-text-scroll-left-on-blur.html

html/HTMLInputElement.cpp: (WebCore::HTMLInputElement::defaultEventHandler): Pass blur events through to the RenderTextField, as well as mouse, drag, and wheel events.
rendering/RenderTextField.cpp: (WebCore::RenderTextField::forwardEvent): Scroll all the way to the left on a blur.

unrelated small changes

html/HTMLTextFieldInnerElement.cpp: Removed excess includes. (WebCore::HTMLTextFieldInnerElement::defaultEventHandler): Tweaked comments a bit.

dom/Element.cpp: (WebCore::Element::scrollIntoView): Removed unneeded this-> before function call. (WebCore::Element::scrollIntoViewIfNeeded): Ditto.

page/FrameView.cpp: (WebCore::FrameView::dispatchMouseEvent): Removed obsolete comment.

WebKitTools:

Reviewed by Justin.

DumpRenderTree/EventSendingController.m: (-[EventSendingController keyDown:withModifiers:]): Fixed keyDown function so that it will work if you don't pass an array of modifiers or if elements in that array are not strings.

2:27 PM Changeset in webkit [13629] by darin

2 edits in trunk/WebKitLibraries

Reviewed by Beth.

libWebKitSystemInterface.a: Changed alpha in the focus-ring drawing code to use the system default alpha.

12:59 PM Changeset in webkit [13628] by mjs

1 edit in trunk/WebCore/editing/ReplaceSelectionCommand.cpp

Ooops, actually land the fix with this ChangeLog entry:

2006-03-31 Maciej Stachowiak <mjs@apple.com>

Reviewed by Adele.

<rdar://problem/4497684> REGRESSION(NativeTextField): After undoing pasted text in a field, the field changes to only a few pixels in height (8096)

editing/ReplaceSelectionCommand.cpp: (WebCore::ReplaceSelectionCommand::doApply): merge into start block when pasting into an empty editable subtree.

11:39 AM Changeset in webkit [13627] by darin

62 edits in trunk/LayoutTests

re-updated many pixel results to try to create a new baseline

2:14 AM Changeset in webkit [13626] by mjs

2 edits
2 adds in trunk

LayoutTests:

Test cases for:

<rdar://problem/4497684> REGRESSION(NativeTextField): After undoing pasted text in a field, the field changes to only a few pixels in height (8096)

fast/forms/input-paste-undo-expected.txt: Added.
fast/forms/input-paste-undo.html: Added.

WebCore:

Reviewed by Adele.

<rdar://problem/4497684> REGRESSION(NativeTextField): After undoing pasted text in a field, the field changes to only a few pixels in height (8096)

editing/ReplaceSelectionCommand.cpp: (WebCore::ReplaceSelectionCommand::doApply): merge into start block when pasting into an empty editable subtree.

1:40 AM Changeset in webkit [13625] by darin

4 edits
4 adds in trunk

LayoutTests:

test for http://bugzilla.opendarwin.org/show_bug.cgi?id=8063 REGRESSION: double clicking in new text fields won't select whole words

fast/forms/input-text-double-click-expected.checksum: Added.
fast/forms/input-text-double-click-expected.png: Added.
fast/forms/input-text-double-click-expected.txt: Added.
fast/forms/input-text-double-click.html: Added.

WebCore:

Reviewed by Maciej.

fix http://bugzilla.opendarwin.org/show_bug.cgi?id=8063 REGRESSION: double clicking in new text fields won't select whole words

Test: fast/forms/input-text-double-click.html

editing/visible_units.cpp: (WebCore::nextBoundary): Set the end of the range by calling selectNodeContents rather than by calling setEndAfter. The problem with setEndAfter is that it doesn't do anything when the parent of the node is 0, and also it's not really what we want, since the boundary node is one with editable contents -- we want to stay inside the boundary node.

editing/Selection.cpp: (WebCore::Selection::validate): Fix a tiny formatting glitch I noticed at the same time.

1:32 AM Changeset in webkit [13624] by darin

3 edits
4 adds in trunk

LayoutTests:

test for http://bugzilla.opendarwin.org/show_bug.cgi?id=8105 REGRESSION (NativeTextField): Option-delete deletes one space before the deleted word

fast/forms/input-text-option-delete-expected.checksum: Added.
fast/forms/input-text-option-delete-expected.png: Added.
fast/forms/input-text-option-delete-expected.txt: Added.
fast/forms/input-text-option-delete.html: Added.

WebKit:

Reviewed by Maciej.

fix http://bugzilla.opendarwin.org/show_bug.cgi?id=8105 REGRESSION (NativeTextField): Option-delete deletes one space before the deleted word

Test: fast/forms/input-text-option-delete.html

WebView/WebHTMLView.m: (-[WebHTMLView _deleteWithDirection:granularity:killRing:isTypingAction:]): Pass NO for smartDeleteOK. Smart deletion only applies to deleting a word at a time, and none of the callers of this function that delete a word at a time want smart deletion. (-[WebHTMLView deleteToEndOfLine:]): Fixed tiny formatting glitch.

Mar 31, 2006:

4:40 PM Changeset in webkit [13623] by thatcher

1 edit in trunk/WebKitTools/BuildSlaveSupport/run-performance-tests

Knocking the number of PLTs we run down to 3 to see if we come in under the BuildBot timeout.

4:12 PM Changeset in webkit [13622] by thatcher

1 edit in trunk/WebKitTools/BuildSlaveSupport/run-performance-tests

Knocking the number of PLTs we run down to 4 to see if we come in under the BuildBot timeout.

3:14 PM Changeset in webkit [13621] by sullivan

2 edits in trunk/WebCore

Reviewed by Tim Hatcher.

fixed <rdar://problem/4372842> 10.4.4 Regression: control-clicking on a misspelled word doesn't select it or offer corrections (first click only)

bridge/mac/FrameMac.mm: (WebCore::FrameMac::sendContextMenuEvent): Rolled in this one-line change that Hyatt wrote ages ago.

2:39 PM Changeset in webkit [13620] by thatcher

2 edits in trunk/WebKit

Reviewed by Darin.

We need to set reachedTerminalState to YES before we release
the resources to prevent a double dealloc of WebView

Fixes <rdar://problem/4372628> crash deallocating a WebView in -[WebFrame stopLoading]

WebView/WebLoader.m: (-[NSURLProtocol releaseResources]): set reachedTerminalState earlier

1:50 PM Changeset in webkit [13619] by bdakin

3 edits
4 adds in trunk

Reviewed by John.

Fix for http://bugzilla.opendarwin.org/show_bug.cgi?id=8108
REGRESSION (r13590-r13593): Floating table's cells don't paint
their background

This is a regression from my painting patch yesterday. Just a silly
error I didn't catch.

rendering/RenderTable.cpp: (WebCore::RenderTable::paint): Change the phase of our new PaintInfo, not our old one.

1:36 PM Changeset in webkit [13618] by sullivan

2 edits in trunk/WebKit

Reviewed by Tim Omernick.

Some cleanup in the WebIconDatabase code in a fruitless search for the cause
of the mysterious -[WebFileDatabase performSetObject:forKey] crash

Misc/WebIconDatabase.m: (-[WebIconDatabase _createFileDatabase]): removed obsolete comment (-[WebIconDatabase _loadIconDictionaries]): Added ERRORs for unexpected early returns. Made dictionaries be autoreleased until the end of the method where they are retained, so that the early returns don't leak; added ASSERTs that the _private->dictionary values aren't being leaked. (-[WebIconDatabase _updateFileDatabase]): Added an ERROR for an unexpected early return, and made a trivial style fix.

12:58 PM Changeset in webkit [13617] by tomernic

3 edits
2 adds in trunk

LayoutTests:

Reviewed by Adele.

Layout test for:
<http://bugzilla.opendarwin.org/show_bug.cgi?id=7858>
<rdar://problem/4483359> REGRESSION: New text field doesn't recognize the read only attribute

fast/forms/input-appearance-readonly-expected.txt: Added.
fast/forms/input-appearance-readonly.html: Added.

WebCore:

Reviewed by Adele.

<http://bugzilla.opendarwin.org/show_bug.cgi?id=7858>
<rdar://problem/4483359> REGRESSION: New text field doesn't recognize the read only attribute

rendering/RenderTextField.cpp: (WebCore::RenderTextField::createDivStyle): Set user modify based on the form element's readOnly(). (WebCore::RenderTextField::updateFromElement): ditto

12:44 PM Changeset in webkit [13616] by adele

3 edits in trunk/WebCore

Reviewed by Tim Omernick.

Updating shadowAncestorNode so it doesn't check for rootEditableElement. Now we just walk
up the tree to look for a shadowNode, and then we find the shadowParent.

dom/Node.cpp: (WebCore::Node::shadowAncestorNode):
rendering/RenderTextField.cpp: (WebCore::RenderTextField::setSelectionRange): Updated assertion to check for shadowAncestorNode instead of rootEditableElement.

12:30 PM Changeset in webkit [13615] by ap

3 edits
2 copies
7 adds in trunk

Fix by Trey Matteson, reviewed by Maciej.

http://bugzilla.opendarwin.org/show_bug.cgi?id=7739
REGRESSION: Assertion failure loading acid2 test in -[WebCoreFrameBridge installInFrame:]

Tests: http/tests/misc/acid2.html, http/tests/misc/acid2-pixel.html

The gist of this change is that we must cancel the load from the Webkit side when we
realize we're switching to the fallback content. This is somewhat a temp workaround
since control of loading will be moving to WebCore.

WebView/WebMainResourceLoader.m: (-[WebMainResourceLoader continueAfterContentPolicy:response:]):

12:12 PM Changeset in webkit [13614] by adele

3 edits in trunk/WebCore

Reviewed by John Sullivan.

fix http://bugzilla.opendarwin.org/show_bug.cgi?id=8101 REGSRESSION: Fix for bug 7031 causes 30 layout tests to fail

rendering/render_line.cpp: (WebCore::InlineFlowBox::placeBoxesVertically): Change top and bottom positions only if childAffectsTopBottomPos is true.
rendering/RenderFlow.cpp: (WebCore::RenderFlow::paintLines): Redo a part of the patch for bug 7031 that wasn't committed with the rest of the patch.

11:16 AM Changeset in webkit [13613] by bdakin

3 edits
4 adds in trunk

Reviewed by Darin, landed by Beth.

Fix for http://bugzilla.opendarwin.org/show_bug.cgi?id=8081
REGRESSION: Drop-down menu has gap at top

rendering/RenderBox.cpp: (WebCore::RenderBox::calcAbsoluteVerticalValues): When calculating the hypothetical vertical position in normal flow, skip table rows in the ancestor chain, since a table cell's Y position is relative to the table section, not the row.

11:03 AM Changeset in webkit [13612] by hyatt

4 edits in trunk/WebCore

FIx theme borders on win32.

10:51 AM Changeset in webkit [13611] by darin

14 edits
2 adds in trunk

LayoutTests:

updated results for tests that involve scrolling

editing/selection/focus_editable_html-expected.txt: Added expected scroll position.
fast/dom/focus-contenteditable-expected.txt: Ditto.
fast/overflow/scrollRevealButton.html: Ditto.

updated results for tests that involve new exception names

fast/dom/Element/attr-param-typechecking-expected.txt: Updated.

fast/dynamic/insertAdjacentElement-expected.txt: Updated.
fast/dynamic/insertAdjacentElement-expected.png: Added.
fast/dynamic/insertAdjacentElement-expected.checksum: Added.

WebCore:

Reviewed by John Sullivan.

khtml/ecma/kjs_binding.cpp: Added names for VALIDATION_ERR and TYPE_MISMATCH_ERR; new DOM Level 3 errors that need to be listed in the mapping from error code to error name.

dom/Element.idl: Removed comment from bad old days where we had to touch these files to make them rebuild.

WebKitTools:

Reviewed by John Sullivan.

added a "--reset-results" option to run-webkit-tests so you can reset the results without first deleting expected results
<rdar://problem/4185878> add scroll position to dumpRenderTree

Scripts/run-webkit-tests: Rearranged the code a bit to make the sequence more logical. Moved all the subroutines to the end of the file. Added a "--force" option.

DumpRenderTree/DumpRenderTree.m: (main): Added a new --dump-all-pixels option, used when forcing run-webkit-tests to generate new output for all tests it runs. (dump): Dump the scroll position if it's non-zero. Always dump the image when the --dump-all-pixels option is passed. Also tightened up the image dumping code and replaced the incorrect use of +[NSGraphicsContext saveGraphicsState] with code to save and restore the context.

DumpRenderTree/DumpRenderTree.h: Tweaked includes a bit and added (void).
DumpRenderTree/EventSendingController.m: Added now-needed include.

10:02 AM Changeset in webkit [13610] by darin

2 edits in trunk/JavaScriptCore

Reviewed by Geoff.

<rdar://problem/4395622> API: WebScriptObject.h incorrectly reports that -isSelectorExcludedFromWebScript returns NO by default

bindings/objc/WebScriptObject.h: Fixed comment.

9:40 AM Changeset in webkit [13609] by bdakin

1 edit
4 adds in trunk/LayoutTests

Layout test for <rdar://problem/4472371> REGRESSION(417.9-TOT):
Focus ring around link in overflow:auto div isn't clipped to div

8:39 AM Changeset in webkit [13608] by darin

1 edit in trunk/JavaScriptCore/ChangeLog

Removed all the tabs and cleared the "allow tabs" bit.

3:45 AM Changeset in webkit [13607] by hyatt

1 edit in trunk/WebCore/rendering/RenderTheme.cpp

Fix my build bustage with themes.

3:32 AM Changeset in webkit [13606] by mjs

4 edits in trunk/WebCore

fixed Windows build breakage from previous change

platform/ScrollView.h:
platform/win/FontWin.cpp: (WebCore::Font::drawLineForText):
platform/win/ScrollViewWin.cpp: (WebCore::ScrollView::scrollOffset): (WebCore::ScrollView::scrollBy):

3:24 AM Changeset in webkit [13605] by mjs

22 edits in trunk/WebCore

Reviewed by Eric.

various Point / Size related cleanup

First, I clarified the semantics of some operations to keep a
better distinction between IntPoint and IntSize:

platform/IntPoint.h: (WebCore::IntPoint::move): new convenience to move a point by separate x and y deltas. (WebCore::operator+=): You can't add a point to a point, you can only add a size to a point. (WebCore::operator+): ditto (WebCore::operator-): point - point = size; point - size = point (WebCore::operator-=): only allow subtracting a size for the mutating version
platform/IntRect.h: (WebCore::IntRect::move): tweaked to use IntPoint::move, also, move by an IntSize, not an IntPoint.
platform/IntSize.h: (WebCore::IntSize::shrunkTo): analog to expandedTo (WebCore::IntSize::clampNegativeToZero): a handy helper (WebCore::operator-): Added unary minus operator

Made the same changes for FloatPoint:

platform/FloatPoint.h: (WebCore::FloatPoint::move): (WebCore::operator+=): (WebCore::operator-=): (WebCore::operator+): (WebCore::operator-):
platform/FloatRect.h: (WebCore::FloatRect::move):
platform/FloatSize.h: (WebCore::operator-):

Then I changed a bunch of stuff to pass around IntPoint instead of separate x and y
coordinates. The main one was:

platform/ScrollView.h:
platform/mac/ScrollViewMac.mm: (WebCore::ScrollView::scrollOffset): new method, return an IntSize (WebCore::ScrollView::contentsToViewport): take and return an IntPoint (WebCore::ScrollView::viewportToContents): take and return an IntPoint
platform/win/ScrollViewWin.cpp: (WebCore::ScrollView::updateContents): handle things in terms of scrollOffset, not scrollPoint (WebCore::ScrollView::visibleContentRect): (WebCore::ScrollView::contentsX): (WebCore::ScrollView::contentsY): (WebCore::ScrollView::viewportToContents): (WebCore::ScrollView::contentsToViewport): (WebCore::scrollOffset): (WebCore::ScrollView::maximumScroll): (WebCore::ScrollView::scrollBy): (WebCore::ScrollView::updateScrollBars):

The rest is mainly updates for these changes.

bridge/mac/FrameMac.h:
bridge/mac/FrameMac.mm: (WebCore::FrameMac::eventMayStartDrag): (WebCore::FrameMac::dragHysteresisExceeded): (WebCore::FrameMac::handleMouseMoveEvent): (WebCore::FrameMac::mouseDown): (WebCore::FrameMac::shouldDragAutoNode): (WebCore::FrameMac::sendContextMenuEvent):
bridge/mac/WebCoreFrameBridge.mm: (-[WebCoreFrameBridge isPointInsideSelection:]):
dom/EventTargetNode.cpp: (WebCore::EventTargetNode::dispatchMouseEvent): (WebCore::EventTargetNode::dispatchWheelEvent):
khtml/ecma/kjs_window.cpp: (KJS::WindowFunc::callAsFunction):
page/Frame.cpp: (WebCore::Frame::shouldDragAutoNode): (WebCore::Frame::isPointInsideSelection): (WebCore::Frame::selectClosestWordFromMouseEvent): (WebCore::Frame::handleMousePressEventDoubleClick): (WebCore::Frame::handleMousePressEventTripleClick): (WebCore::Frame::handleMousePressEventSingleClick): (WebCore::Frame::handleMouseMoveEventPart2): (WebCore::Frame::handleMouseReleaseEvent):
page/Frame.h:
page/FrameView.cpp: (WebCore::FrameView::dispatchDragEvent): (WebCore::FrameView::prepareMouseEvent): (WebCore::FrameView::handleWheelEvent):
rendering/RenderLayer.cpp: (WebCore::RenderLayer::scrollRectToVisible):
rendering/RenderObject.cpp: (WebCore::RenderObject::draggableNode):
rendering/RenderObject.h: (WebCore::RenderObject::positionForPoint):
rendering/render_list.cpp: (WebCore::RenderListMarker::paint):

3:18 AM Changeset in webkit [13604] by mjs

2 edits in trunk/WebCore

fixed windows build

platform/win/TemporaryLinkStubs.cpp: (Widget::unlockDrawingFocus):

1:15 AM Changeset in webkit [13603] by eseidel

8 edits in trunk/JavaScriptCore

2006-03-31 Eric Seidel <eseidel@apple.com>

Reviewed by mjs.

A bit more code cleanup.

bindings/c/c_utility.cpp: (KJS::Bindings::convertValueToNPVariant):
bindings/objc/objc_runtime.mm: (convertValueToObjcObject):
bindings/objc/objc_utility.mm: (KJS::Bindings::convertValueToObjcValue):
kjs/function.cpp: (KJS::GlobalFuncImp::callAsFunction):
kjs/interpreter.cpp: (KJS::ExecState::lexicalInterpreter):
kjs/interpreter.h:
kjs/operations.cpp: (KJS::equal):

1:14 AM Changeset in webkit [13602] by eseidel

9 edits in trunk/WebCore

2006-03-31 Eric Seidel <eseidel@apple.com>

Reviewed by mjs.

A bit more code cleanup.

bridge/mac/WebCoreScriptDebugger.mm: (-[WebCoreScriptCallFrame evaluateWebScript:]):
html/HTMLFormElement.cpp: (WebCore::HTMLFormElement::submit):
html/HTMLInputElement.cpp: (WebCore::HTMLInputElement::setValue):
kwq/WebCoreTextField.mm: (-[KWQTextFieldController textView:shouldHandleEvent:]): (-[KWQSecureTextField selectText:]):
page/Frame.cpp: (WebCore::Frame::submitForm):
platform/Widget.h:
platform/mac/WidgetMac.mm:
rendering/render_frames.cpp: (WebCore::RenderFrameSet::userResize):

Note: See TracTimeline for information about the timeline view.

Download in other formats:

RSS Feed