Skip to main content
aboutsummaryrefslogtreecommitdiffstats
AgeCommit message (Collapse)AuthorFilesLines
2018-04-13- Resized Button with Image on DPI change.DynamicDPI_v4860fNiraj Modi3-14/+35
Change-Id: Iea25be350d071f9e3e8e5fdfa46908886f2fee3f Signed-off-by: Niraj Modi <niraj.modi@in.ibm.com>
2018-04-12- Updated Label to compute size dynamically.Niraj Modi3-7/+19
- Added 2 new API in Image class: Rectangle getBounds(int zoom) boolean refreshImageForZoom (Control control) Change-Id: I70defc2a4cafd7fa57d89580e420226065794773 Signed-off-by: Niraj Modi <niraj.modi@in.ibm.com>
2018-03-29- Updated javaw.exe.manifest to be backward compatibleNiraj Modi2-2/+39
- Updated javaw.exe.manifest to specifically target Win8.1 and Win10 via manifest - Updated Control.java to read below DPI change event info: 1. New DPI for monitor 2. New dimensions for the shell Change-Id: Ic8c20f89288abb5456b8cb99206e6c3c2629aa18 Signed-off-by: Niraj Modi <niraj.modi@in.ibm.com>
2018-03-12Snippet371 to display/refresh Monitor DPI dynamically.Niraj Modi1-0/+135
Change-Id: Icc74c2140872d28d59b7655c4bb2407314407490 Signed-off-by: Niraj Modi <niraj.modi@in.ibm.com>
2018-03-09- Made Control#refreshControlForDPIChange() as non-publicNiraj Modi3-9/+6
- Monitor handle public(for test purpose) Change-Id: Ib6c96ab76391b6819e55c5bbc0d4cd05d2f50922 Signed-off-by: Niraj Modi <niraj.modi@in.ibm.com>
2018-03-09Bug 515502 - [HiDPI][Win10] SWT to receive/handle DPI changeNiraj Modi16-12/+181
notification from OS Change-Id: I042189547f85b197cfe77cd7b61fc00f30d5fd27 Signed-off-by: Niraj Modi <niraj.modi@in.ibm.com>
2018-03-08v4860v4860fS4_8_0_M6I20180308-0630genie.releng1-1/+1
2018-03-08Bug 519015 - Tab after 4 spaces is not visibleSravan Kumar Lakkimsetti1-1/+4
fixed a regression where the text segments using wrap around are getting over written The fix is to ignore text segments when we are calculating default tabwidth Change-Id: Ie7b013357204577a5a2d35ca305a13003800fdf7 Signed-off-by: Sravan Kumar Lakkimsetti <sravankumarl@in.ibm.com>
2018-03-07v4860v4860eI20180307-2000I20180307-1025I20180307-0900I20180307-0825genie.releng1-1/+1
2018-03-07Bug 532063: Replace calls to getAverageCharWidth() withI20180307-0800Lakshmi Shanmugam6-8/+9
getAverageCharacterWidth() in SWT Update snippets and examples with the new API Change-Id: I68a61091f07c60978f60e99f1fd76da2b51537e9
2018-03-07Revert "Bug 528498 - Combo menu is narrower than combo"Alexander Kurtakov3-84/+3
This reverts commit 2feeddb0172f839c43ea8087619a9ea912eb88fa. Change-Id: I32d4f75cda1fbd9c7e88d212124f06830291fcaa
2018-03-07v4860v4860dI20180306-2000genie.releng1-1/+1
2018-03-06Bug 532063: Replace calls to getAverageCharWidth() withLakshmi Shanmugam3-3/+3
getAverageCharacterWidth() in SWT Replace calls to getAverageCharWidth() with getAverageCharacterWidth() in SWT code with int cast to fix compiler warnings. Change-Id: I9533bdd17af8569f1ea40ccbc54cd4fb0b91ebf4
2018-03-06v4860v4860cI20180306-0800I20180305-2000genie.releng1-1/+1
2018-03-05Bug 530317: [GTK3.20+] Java formatter editor has cut-off buttonsEric Williams1-4/+11
This bug is caused by a corner case where the minimum size is larger than the natural size. In these cases the widget will end up too large, as only the minimum size is taken into account. When performing sizing calculations for GTK3.20+, use the smallest of the minimum and natural sizes to prevent oversized widgets. Tested on GTK3.20+ using Eclipse and the AllNonBrowser JUnit tests. No additional allocation warnings are printed in the console. Change-Id: I30a7ef75abb135b1a23402a834824241b6eb055f Signed-off-by: Eric Williams <ericwill@redhat.com>
2018-03-05Bug 508600: The average character width of a font is not integerLakshmi Shanmugam1-0/+6
Added Junit test for the new API Change-Id: I1f78e5fe55746a57a38ed7a579c39547666b6d05
2018-03-05Bug 531667 - [GTK3] Cannot draw Canvas with Control.print(GC)Simeon Andreev4-2/+320
Calling Canvas.pring(GC) results in no painted canvas on the GC. This is caused by a regression from the fix for Bug 421127. Control.print propagates the GC.handle as the last argument to Control.gtk_draw. With changes for Bug 421127 in Control.gtk_draw, the handle is no longer propagated to the actual onPaint event. The canvas (and canvas derived widgets) therefore no longer paint themselves on the GC. The problem which was fixed with the change in Control.gtk_draw, for Bug 421127, was that CTabFolder tabs were not painted. This is the case, since GC.getClipping returned broken values with GTK 3.10 and above. CTabFolder checks for intersection with the clipping region and doesn't paint itself. The clipping region was not computed properly, due to transformations done with the device space transformation matrix provided by Cairo. From GTK 3.10 onward, this matrix contains absolute coordinates of the widget, with respect to the shell. This causes the method to malfunction. With the fix for Bug 421127, the Cairo matrix is no longer propagated. As a result GC.getClipping only works with the identity matrix, resulting in correct clipping. This change reverts the change for Bug 421127 in Control.gtk_draw, for GTK versions 3.14 and above. The GC.getClipping method is adapted to first transform the local GC bounds to device space, before computing the intersection with the clipping region. This is done for GTK 3.14 and above. We validated the change with GTK 3.14 and GTK 3.22. A bug snippet is provided, show-casing clipping region and Cairo device space transformation matrix. Change-Id: Ib1d6bb866bcffa010a9e8f2a5e4296eba4e1d852 Signed-off-by: Simeon Andreev <simeon.danailov.andreev@gmail.com>
2018-03-05v4860v4860bI20180305-0800genie.releng1-1/+1
2018-03-05Bug 527937 - [HiDPI][Win32] Checkbox/Radiobutton: focus-rect cut offConrad Groth1-1/+1
Change-Id: I6a7ce52c9064c899e5329316145f94c38a6c5970 Signed-off-by: Conrad Groth <info@conrad-groth.de>
2018-03-03v4860v4860aI20180305-0300I20180304-2000I20180304-0800I20180303-1500I20180303-1200I20180303-0800I20180303-0335I20180303-0155I20180303-0055I20180302-2000genie.releng1-1/+1
2018-03-02Bug 531048 [GTK3] Large problems drawing ownerdraw tables (tests).Leo Ufimtsev4-0/+1501
Adding manual JUnit tests to verify changes to table don't break table. Tests are aimed at verifying visual glitches (like cheese). Going through these is quite quick (F1=pass, F2=Fail, F3=Skip, ESC=Quit). A lot of the tests originate from snippets. Bug: https://bugs.eclipse.org/bugs/show_bug.cgi?id=5310483 Change-Id: I3ad13abd45110e528987b04f1061435d38127fa7 Signed-off-by: Leo Ufimtsev <lufimtse@redhat.com>
2018-03-02Bug 479646 - GLib-CRITICAL output when adding MouseHover listenerAlexander Kurtakov1-1/+6
GSource created with g_timeout_add is auto destroyed when the callback returns FALSE and our implementation always does that in order for the hover to be hidden. Change-Id: I0b3fa89de5c7ba2a813c54477166e0f5eb919bb8 Signed-off-by: Alexander Kurtakov <akurtako@redhat.com>
2018-03-02Bug 508600: [API] The average character width of a font is not integerLakshmi Shanmugam5-10/+70
Added new API getAverageCharacterWidth() which returns the width in floating point. Change-Id: I4767aa07f0be6ead821d1db18cd8508864225056
2018-03-02Bug 378202 - Two-finger tap to simulate right-click results inTill Brychcy1-5/+26
multiple click events Control.menuForEvent needs to return the menu to be shown for the right click instead of 0, or the right click will be propagated up the chain. If the menu is result of getMenu(), it is returned. If it is created in the SWT.MenuDetect event listener, it is not shown yet, but registered in display.popups. When this is detected it is removed from that array and returned. Change-Id: Ia2bf133f61b0ce74d036c9188fbf1dd36ca76508 Signed-off-by: Till Brychcy <register.eclipse@brychcy.de>
2018-03-02v4860v4860I20180301-2000genie.releng1-1/+1
2018-03-02v4860genie.releng3-4/+4
2018-03-01Bug 531639 - [StyledText] Improve performance on load whenangelozerr2-3/+27
setLineSpacingProvider is called Change-Id: Ie26e1718b9af2cc5e008c53a4db2012684713777 Signed-off-by: angelozerr <angelo.zerr@gmail.com>
2018-03-01Bug 431423: [GTK3] Context menu should appear by pointer when invokedEric Williams5-0/+85
via mouse Some context menus are positioned incorrectly when the size (number of items) of the menu changes between clicks. The cause of this bug is due to the way SWT populates its menus: often it is done asynchronously outside of SWT based on an SWT.Show listener. The result is that the menu items are added/removed just before showing, and GTK sometimes doesn't have enough time to properly adjust the size of the menu internally. Specifically, the GdkWindow of the toplevel widget (GtkWindow) for the menu lags in height. Since GTK thinks the menu is taller/shorter than it is, the menu is positioned at the wrong y-coordinate. The fix is to check if the number of menu items has changed from one pop-up to the next. If so, calculate the preferred height of the menu and resize the GdkWindow of the menu's toplevel GtkWindow to reflect this change. This way the internal sizing calculations in GTK will be accurate and the menu will be positioned correctly. The bug is fixed on GTK3.22 with X11 only, as this uses the new GTK3.22 GtkMenu API. Older GTK3 versions use the outdated GtkMenu API and fixing these cases would not be worth the effort. Wayland is not affected by this bug. Reliable steps to reproduce the issue can be found here: https://bugs.eclipse.org/bugs/show_bug.cgi?id=431423#c19 Tested JFace/SWT menu snippets, AllNonBrowser JUnit tests, and ControlExample on GTK3.22 with X11. Change-Id: I392983438cc32b9455c53cc464626294f3ec72f9 Signed-off-by: Eric Williams <ericwill@redhat.com>
2018-02-27v4859v4859eI20180228-2000I20180227-2000I20180226-2000genie.releng1-1/+1
2018-02-26Bug 431423: [GTK3] Context menu should appear by pointer when invokedEric Williams1-14/+4
via mouse Some context Menus on GTK3.22+ had an issue where the content of the menu was incorrect: often the content of the previous menu would be shown, or some menu items were missing entirely. This fix ensures the proper event process is followed so that dynamically populated menus have enough time to be created before being shown. This is done by reverting a fix which removed display.[add/remove]Popup logic in Menu.setVisible(). Change-Id: Ice2fe3a1b8ecb58035dbe69e5de1bf9749fc22c2 Signed-off-by: Eric Williams <ericwill@redhat.com>
2018-02-26Bug 531593 – [Webkit2] Crash in webkit on hoverLeo Ufimtsev2-26/+55
I can't reproduce the reported issue, but I can see a rare corner race condition where such a crash could occur. I.e, the race condition could be that webkit has internally disposed itself but we might be referencing it accidentally still. The fix is to guard against this. The fixes don't break anything as the jvm crash is occurring on webkit's disposal anyway. Bug: https://bugs.eclipse.org/bugs/show_bug.cgi?id=531593 Change-Id: Ia7394ce301b237dcf5ac43edb6590f0f5003e49f Signed-off-by: Leo Ufimtsev <lufimtse@redhat.com>
2018-02-24v4859v4859dI20180225-2000I20180224-1500genie.releng1-1/+1
2018-02-24Bug 531633 - Replace usage of System.getProperty("file.separator")Karsten Thoms7-9/+15
Change-Id: I609d4b7e6415a113e2afcf790cc72d9a7c6199c0 Signed-off-by: Karsten Thoms <karsten.thoms@itemis.de>
2018-02-24Bug 248075 update popupcalendar on text changelaeubi1-0/+20
Change-Id: Id7c42c4ac48e67d155d4ab09b85fdc319b1dbbf1 Signed-off-by: laeubi <laeubi@laeubi-soft.de>
2018-02-23Bug 531575 - Control#setBackground Javadoc outdatedLars Vogel3-3/+0
Change-Id: Ied09c450a026b081539e3f305e39e120dd8fcdb0 Signed-off-by: Lars Vogel <Lars.Vogel@vogella.com>
2018-02-21v4859v4859cI20180223-2000I20180222-2000I20180221-2000I20180220-2000genie.releng1-1/+1
2018-02-20Bug 531032 - [StyledText] StyledTextRenderer#isVariableHeight bug &angelozerr3-21/+32
improve Change-Id: I3766e7e55969a29f3ae645b452b0e369dd5e767f Signed-off-by: angelozerr <angelo.zerr@gmail.com>
2018-02-20Bug 528498 - Combo menu is narrower than comboSimeon Andreev3-3/+84
In order to fix Bug 438992, wrap width of 1 was set for read-only combos. This results in drop-down menus which are narrower than the combo itself. Seen e.g. in the Eclipse Help preference page, where the combos are as long as the page, but their entries are relatively short. The drop-down menu is then only as short as the longest entry. This change ensures that the GTK style option -GtkComboBox-appears-as-list is used, for GTK versions above 3.22.5. This changes the drop-down menu style, preventing both Bug 438992 and Bug 528498. The option cannot be used in earlier versions, since due to a bug, the drop-down menu cannot be scrolled. Change-Id: I96aecae46c6edef65e69a9ae590e713bf607dd4b Signed-off-by: Simeon Andreev <simeon.danailov.andreev@gmail.com>
2018-02-17v4859v4859bI20180219-2000I20180218-2000I20180217-1500genie.releng1-1/+1
2018-02-17Bug 530387 - [GTK3] Replace deprecatedAlexander Kurtakov6-27/+23
gdk_device_manager_get_client_pointer() Retrieve pointer from GdkSeat on Gtk 3.20+. Change-Id: I39c148ebd00ccde9e3b0823d2dd92b387bee322b Signed-off-by: Alexander Kurtakov <akurtako@redhat.com>
2018-02-17v4859v4859aI20180216-2000genie.releng1-1/+1
2018-02-16Bug 530950 - [Gtk3][DateTime] Visual glitch, SWT.DROP_DOWN button isAlexander Kurtakov1-1/+1
outside the editbox. Really use the padding when calculating button Change-Id: Ia1749032df0807f78cbb84b15937bb002df18981 Signed-off-by: Alexander Kurtakov <akurtako@redhat.com>
2018-02-16v4859v4859I20180216-0415genie.releng1-1/+1
2018-02-16v4859genie.releng3-4/+4
2018-02-15Bug 510803 [GTK3] Regression in table editing capabilities in tabfolder.I20180215-2000Leo Ufimtsev11-135/+203
Breakage occurred because tabItem reparented an item that had a non-standard gDk parent window. Solution: fix gdk window parent for such special controls. (I.e ControlEditors in tables). Verification: - Open child eclipse, change method signature. Now table cells can be edited via click. - Open attached snippet, with patch cells can be edited properly. - AllTests Gtk3.22 - X11/Wayland. Patchset 3: - Minor rename of methods to clarify that it's a *gdk* window, Bug: https://bugs.eclipse.org/bugs/show_bug.cgi?id=5108033 Change-Id: I8dcaac950eb0847dd97016b2140c607012550d2f Signed-off-by: Leo Ufimtsev <lufimtse@redhat.com>
2018-02-15v4858v4858dI20180214-2000genie.releng1-1/+1
2018-02-14Bug 521387 – [Gtk3] Alt+Enter performs two actions in Package ExplorerLeo Ufimtsev4-13/+32
Fix to make alt+enter not send selection event, to be consitent with win32. Bug: https://bugs.eclipse.org/bugs/show_bug.cgi?id=521387 Change-Id: I79a1208057b598b61d07dcef2457c371b5030cea Signed-off-by: Leo Ufimtsev <lufimtse@redhat.com>
2018-02-13Bug 511133: [GTK3.10+] Some critical icons missing on Ubuntu 16.04Eric Williams5-32/+49
This patch fixes Table/Tree drawing in EGit. It patch reverts the changes introduced for bug 438505, which causes this regression. This patch also contains fixes to prevent bug 438505 from being re-introduced. EGit drawing was broken because TableItem.getBounds() was taking the position of the header into account. This is because we were drawing on the fixedHandle instead of the GtkTreeView. The positioning of the Table header broke simple drawing cases like Snippet051TableCenteredImage (JFace snippet). Reverting the fix back to drawing on the handle has fixed that issue. Unfortunately this means bug 438505 is re-introduced. To fix that issue I have tweaked the re-parenting logic found in Table/Tree.setParentWindow(). GTK3.10+ onward only draws on toplevel or native GdkWindows. This means the Text widgets used for Table/Tree editing continue to get events, but they are not drawn. The solution to this problem is call gdk_window_raise/lower on these Text widgets when setVisible() is called. This raises and lowers the Text's GdkWindow when the widget's visibility is changed, ensuring that it actually is drawn when visible, and not when hidden. This isn't a huge change as the GdkWindow re-parenting already existed in Table/Tree.setParentWindow(), this patch just builds on that. I have tested against the following cases/snippets: Snippet88 Snippet019 (JFace snippets) Snippet025 (JFace snippets) Snippet026 (JFace snippets) Snippet051 (JFace snippets) use case from bug 438505 use case from bug 436324 use case from bug 460581 use case from bug 458630 Additionally, using a child Eclipse I have verified that the missing EGit icons have returned. No AllNonBrowser JUnit tests fail. Change-Id: I02b3fb30037a8f0f74de79144f8e563f5284d571 Signed-off-by: Eric Williams <ericwill@redhat.com>
2018-02-13v4858v4858cI20180213-2000I20180213-0125I20180212-2000genie.releng1-1/+1
2018-02-12Bug 530969: [GTK3] Control.print() doesn't seem to workEric Williams2-0/+126
In GTK3, gtk_widget_draw() relies on the priv->alloc_needed flag to be set to TRUE. This is an internal flag in a private struct so we cannot access it. The fix is to the get the topHandle's allocation, and simply re-set it. This triggers an allocation change without actually changing the size of the widget. Tested on GTK3.8+. No JUnit test failures. Change-Id: I81df53bfceaa988a2cce574ef7629c64c0ec6fb9 Signed-off-by: Eric Williams <ericwill@redhat.com>

Back to the top