Skip to main content
aboutsummaryrefslogtreecommitdiffstats
AgeCommit message (Collapse)AuthorFilesLines
2018-03-23v4866v4866I20180327-0805I20180327-0630I20180326-2000I20180325-2000I20180324-1500I20180323-2000genie.releng1-1/+1
2018-03-23v4866genie.releng3-4/+4
2018-03-23Bug 532825 - Remove hpux/solaris clutter in the codebaseAlexander Kurtakov7-379/+6
Change-Id: I735be5853d393097c7d11c9abb43e9e5cbf93c8d Signed-off-by: Alexander Kurtakov <akurtako@redhat.com>
2018-03-23Bug 488431 - Provide SWT/GTK3 port on WindowsRolf Theunissen5-21/+31
Make GTK build on win32 again Change-Id: Iea9e8fbd53af430b998597449a1445ffd716c5ef Signed-off-by: Rolf Theunissen <rolf.theunissen@altran.com>
2018-03-23Bug 488431 - Provide SWT/GTK3 port on WindowsRolf Theunissen2-8/+10
Test for windows style newlines (os=windows) instead of windows platform (platform=win32) Change-Id: I97d96c4b24c6b7ff9152c715ada973acfeb01075 Signed-off-by: Rolf Theunissen <rolf.theunissen@altran.com>
2018-03-23v4865v4865I20180323-0355genie.releng1-1/+1
2018-03-23v4865genie.releng3-4/+4
2018-03-22Bug 530204: [GTK] Menu appears at mouse regardless of setLocation()I20180322-2000Eric Williams9-21/+238
This patch contains two main parts. Part 1 - functionality implementation: This implements functionality for Menu.setLocation() to actually work as intended. Currently menus are popped up wherever the mouse pointer is, completely ignoring any coordinates given to setLocation(). The functionality provided by this patch only works on X11, as Wayland has no global coordinate system (this has been documented in the javadocs in an earlier patch). Part 2 - debugging support: Menu positioning issues are very hard to pin down, since they cannot be inspected with GtkInspector. Thankfully, GTK3.22+ has API support to remedy this. It's now possible to connect a GtkMenu to the "popped-up" signal. After a GtkMenu is popped up/shown, a callback is triggered. The parameters to the callback function contain information about the menu's position/size, straight from GTK's internal machinery. This patch adds an environment variable, SWT_MENU_LOCATION_DEBUGGING, which (when set) will enable SWT to make use of GTK's "popped-up" signal. The callback function in SWT will print certain information about the menu as provided from GTK. For more information on the "popped-up" signal, see: https://developer.gnome.org/gtk3/stable/GtkMenu.html#GtkMenu-popped-up Tested on Wayland and X11 with GTK3.22. GTK3.20- behaviour is unchanged, as it still uses the older GtkMenu API. No AllNonBrowser JUnit tests fail. IDE testing shows that menus now behave the same as on Win32 or Cocoa (best way to test this: give focus to the "Show In" button in the Open Resource Dialog, and trigger a button press with the keyboard while having the mouse elsewhere -> menu should open up by the button, not the mouse). Change-Id: Ia221734a774d6a18afc4ef879ff883ded7ed81ae Signed-off-by: Eric Williams <ericwill@redhat.com>
2018-03-19v4864v4864bI20180322-0645I20180322-0225I20180321-2000I20180320-2000I20180319-2000genie.releng1-1/+1
2018-03-19Bug 531048 [GTK3] Large problems drawing ownerdraw tables (fix)Leo Ufimtsev8-49/+164
TL;DR: Gtk3 delays some Tree/Table initialization logic. We wait till it's done first to finish some initialization after. It's not a clean fix, but given SWT's custom drawing and requirement for API to be win/cocoa consistent, I couldn't find a better fix. Details: Gtk3's delayed initialization breaks custom drawn Tree/Tables that have invisible columns because internally GTK skips a lot of initialization for invisible items (for optimization). Fix is to wait till this is done and in the mean time 'cache' the 'desired' state of a column in a hash table. Once initialization is done, we flush state (i.e hide/show columns as requested). Technical notes: - Fix is same for Tree and Table. - I put things into Control, because I have another fix in mind that will use the 'delayed' logic. (invisible Rowlayout had such issues). - I piggy back onto 'draw' signal, which is only sent once Gtk3/Caching initialization has occured. - I had to put columns into a hash table because some columns don't have a TableColumn java object associated with it. On the side, this also fixes accessibility warning reported in 399522. Verified against bug Snippet, jUnits, manual tests, child eclipse. @ Reviewer, you can verify against snippet found in bug. Before you get cheese, after thinngs work. Patchset 4: - Minor update, moved some code higher up into table/tree so that check is only done for these two widgets and not for all widgets. Bug: https://bugs.eclipse.org/bugs/show_bug.cgi?id=5310483 Change-Id: I3152f64ddb296879d9bad08bdaeb5c38e0672ef8 Signed-off-by: Leo Ufimtsev <lufimtse@redhat.com>
2018-03-19Bug 532139 – Add Copyright to a java few files.Leo Ufimtsev12-288/+138
Adding missing copyright notices and removing old Command line interface that is no longer being used as such testing can be accomplished via launch groups. Bug: https://bugs.eclipse.org/bugs/show_bug.cgi?id=532139 Change-Id: I9cc3e9a48e70adbd49c993e39100be1301ea5551 Signed-off-by: Leo Ufimtsev <lufimtse@redhat.com>
2018-03-19Bug 531599: [Generic TP Editor] typo in Edit Software Site view whenEric Williams1-4/+14
launched from TP editor Set the font only when/after a Button has its text set. Internal size calculations for fonts are CSS based with GTK3.22+, and setting the font before text is actually set can result in caching issues. The size then reported is too small, resulting in clipped buttons. Tested on GTK3.22 only, 3.20 is unaffected. No AllNonBrowser JUnit test failures, and child Eclipse looks fine. Change-Id: I5e8f45a9023ed88409cc1474c1b1c920a5ed72d9 Signed-off-by: Eric Williams <ericwill@redhat.com>
2018-03-18v4864v4864aI20180318-2000genie.releng1-1/+1
2018-03-18Bug 530019 - StyledText.setStyleRange (and some other) reset more cacheangelozerr1-1/+3
than necessary Change-Id: I3fb57f782bd3167796718dcc16812c0a33926dec Signed-off-by: angelozerr <angelo.zerr@gmail.com>
2018-03-16v4864v4864I20180317-1500I20180316-2000genie.releng1-1/+1
2018-03-16v4864genie.releng3-4/+4
2018-03-16Bug 530019 - StyledText.setStyleRange resets less cacheMickael Istria2-6/+143
This introduces methods for the renderer to reset discontinuous set of lines and usage of these methods in StyledText. This would allow methods to reset less cache of TextLayout and should result in better performances. Change-Id: Ia2896381d0fa5bb00c736958a607c35faf314922 Signed-off-by: Mickael Istria <mistria@redhat.com>
2018-03-16Bug 518723 - [GTK3] Replace deprecated gtk_widget_send_expose()Alexander Kurtakov2-0/+9
It's used only on GTK2 so safe to make it dynamic for now. Change-Id: I3c483d45f777a42259c7f9fdf85a87e9926dea46 Signed-off-by: Alexander Kurtakov <akurtako@redhat.com>
2018-03-16Bug 531048 [GTK3] Problems drawing ownerdraw tables (Tests)Leo Ufimtsev4-219/+741
Adding/updating some tests I used for the fix. Useful for quickly running through a large set of snippets and visually insect that nothing broke. I often found the Waterbed effect, a small change in one place would break thing far far away... Bug: https://bugs.eclipse.org/bugs/show_bug.cgi?id=5310483 Change-Id: I284e01fb077361cc1a21ffff3c77528c9e25b491 Signed-off-by: Leo Ufimtsev <lufimtse@redhat.com>
2018-03-16Bug 532548 - Update surefire to 2.21.0Alexander Kurtakov1-1/+1
Change-Id: I27ee0acc233dd44791f94a00537786882272ca60 Signed-off-by: Alexander Kurtakov <akurtako@redhat.com>
2018-03-16v4863v4863bI20180316-0740genie.releng1-1/+1
2018-03-16Bug 532522 - fixMenus: add disposal checksI20180316-0730Karsten Thoms2-2/+2
Change-Id: I99c1113ed373dd125af75d8e84023c875f68cec2 Signed-off-by: Karsten Thoms <karsten.thoms@itemis.de>
2018-03-16v4863v4863aI20180316-0525I20180316-0510genie.releng1-1/+1
2018-03-16[Tests] Use dedicated SWT native lib cache folderI20180316-0455Mickael Istria1-0/+3
SWT tests often also involve testing a new snapshot of the native libraries. As SWT usually puts the libraries in a cache under user.home, it makes SWT tests more likely to used the cached version then the one that inside platform-specific fragments. SWT Tests are now configured to use a different user.home to not interact with the default cache directory. Signed-off-by: Mickael Istria <mistria@redhat.com>
2018-03-15v4863v4863I20180315-2000genie.releng1-1/+1
2018-03-15v4863genie.releng3-4/+4
2018-03-15Bug 530019 - StyledText.setStyleRange resets less cacheangelozerr2-2/+261
Change-Id: Iaa4e0177de4e9d962fcdd4ea1aeeae6d6e8f2e54 Signed-off-by: angelozerr <angelo.zerr@gmail.com>
2018-03-15Bug 532483 - Remove XFlush usageAlexander Kurtakov8-47/+4
gdk_flush under x11 just calls XFlush so there is no need for calling gdk_flush on non x11 and XFlush on x11. Change-Id: I8939328f053e9fd45e10c0b310ff48d6da78ef32 Signed-off-by: Alexander Kurtakov <akurtako@redhat.com>
2018-03-15Bug 488431 - Provide SWT/GTK3 port on WindowsRolf Theunissen1-0/+3
Prevent warnings for unimplemented method 'gdk_property_get' in GDK-win32 Change-Id: Ia5aec2a2cd1bc9f247dfb719e63e8fa32b14e5eb Signed-off-by: Rolf Theunissen <rolf.theunissen@altran.com>
2018-03-15Bug 488431 - Provide SWT/GTK3 port on WindowsRolf Theunissen1-1/+4
Map the realpath method to _fullpath in windows. This function provides the same functionality, only the windows variant will not return an error if the file does not exist. Change-Id: Iec3b4451bd73053ee91fe571fccefdd47be1c2d9 Signed-off-by: Rolf Theunissen <rolf.theunissen@altran.com>
2018-03-15Bug 532475 - Move away of deprecated gdk_error_trap_push/popAlexander Kurtakov7-9/+122
Deprecated in GTK 3.x as it's X11 only and there is even more efficient pop function which doesn't block. Change-Id: I856a18e340a02f8faf84a8a580030a903dc1c92f Signed-off-by: Alexander Kurtakov <akurtako@redhat.com>
2018-03-14v4862v4862I20180314-2000genie.releng1-1/+1
2018-03-14v4862genie.releng3-4/+4
2018-03-14Bug 532173 - [StyledText] Wrong caret bounds with variable heightMickael Istria2-11/+84
Fix bug in getBoundsAtOffset to make it more reliable. Place caret to better location when it doesn't fit the whole line. May as well fix bug 118612 Change-Id: I381dcc9039dae62dfddbd069487f7ecec85057c4 Signed-off-by: Mickael Istria <mistria@redhat.com> Also-By: Angelo Zerr <angelozerr@gmail.com>
2018-03-14Revert "Bug 532173 - [StyledText] Wrong caret bounds with variable height"Mickael Istria2-84/+11
This reverts commit 2ef119e594da0ed9e93c88f3b32b03e82863870c. Change-Id: Ida4a0312d3a46498cff147982463ee933a508e80
2018-03-14Bug 532173 - [StyledText] Wrong caret bounds with variable heightMickael Istria2-11/+84
Fix bug in getBoundsAtOffset to make it more reliable. Place caret to better location when it doesn't fit the whole line. May as well fix bug 118612 Change-Id: I942787c801090ff504c8ed37183a97b312f78e18 Signed-off-by: Mickael Istria <mistria@redhat.com> Also-By: Angelo Zerr <angelozerr@gmail.com>
2018-03-14Bug 532433 - Don't use deprecated gdk_beepAlexander Kurtakov5-16/+19
Must be replaced with gtk_display_beep. Change-Id: Ic276387881a5e992139b7768a19ac4d4a9090d58 Signed-off-by: Alexander Kurtakov <akurtako@redhat.com>
2018-03-13v4861v4861I20180313-2000genie.releng1-1/+1
2018-03-13v4861genie.releng3-4/+4
2018-03-13Bug 366471: [Cocoa] Slow scrolling in editor on Mac OS XLakshmi Shanmugam7-141/+34
Removes the call to runPaint() in display.readAndDispatch(), the setNeedsDisplay and setNeedsDisplayInRect callbacks and the NSMutableArray instances (Display.isPainting, needsDisplay and needsDisplayInRect). Change-Id: I63d460b6155819c2e90a8ed049b702f0f7bee654
2018-03-12v4860v4860iI20180312-2000genie.releng1-1/+1
2018-03-12Bug 248075 close popup on click at daylaeubi1-18/+14
Change-Id: I5c5039e7680b0cb4e2436f5b9e187fca527dfec1 Signed-off-by: laeubi <laeubi@laeubi-soft.de>
2018-03-10v4860v4860hI20180311-2000I20180310-1500genie.releng1-1/+1
2018-03-10Bug 248075 fix problem with date and +/- buttonslaeubi2-3/+3
Change-Id: I5e9e32caaf5b3a176f2fec3b210e6882f1ff2f3f Signed-off-by: laeubi <laeubi@laeubi-soft.de>
2018-03-10v4860v4860gI20180309-2000genie.releng1-1/+1
2018-03-09Bug 529431: [GTK3.10+] Snippet294 fails to draw RegionEric Williams1-3/+36
Draw the region on the Control using cairo_clip(). This doesn't use the GdkWindow API which no longer works on GTK3.10+. The only caveat is that input handling is not blocked. For example: setting the region on a button will not prevent click events within the region from registering. This needs to be investigated and fixed. Tested on GTK3.8 - GTK3.22. IDE looks fine, no AllNonBrowser JUnit tests fail. Change-Id: If6f6bdc1d349cfb3ab42be3977e5c608e29df43e Signed-off-by: Eric Williams <ericwill@redhat.com>
2018-03-09Bug 532074: [Wayland] Arrow menu from fast view pops up in wrongEric Williams1-33/+37
location Adjust the parent GdkWindow of the menu event on Wayland, so that it is always the window of the mouse pointer. Since Wayland has no global coordinates, there is no reason to use the toplevel window as this just pops up the menu in the wrong place. Change-Id: I66c8a3b4b690d921485045c672a474fe5cf99fa2 Signed-off-by: Eric Williams <ericwill@redhat.com>
2018-03-09Bug 530204: [GTK] Menu appears at mouse regardless of setLocation()Eric Williams3-6/+12
Update Javadocs for Menu.setLocation() to reflect platform specific behavior. Change-Id: I43d5bf906f2ddfad7e8fa8945cb7df1d5135ee2f Signed-off-by: Eric Williams <ericwill@redhat.com>
2018-03-09Bug 532161 - Gtk-CRITICAL **: assertion 'GTK_IS_CONTAINER (container)'Alexander Kurtakov1-2/+2
failed from TabFolder This happens due to iterating over children of disposed TabItem. Change-Id: Ic950cea32fda3c164c9d59c41cebfc38e255d5a3 Signed-off-by: Alexander Kurtakov <akurtako@redhat.com>
2018-03-08v4860v4860fS4_8_0_M6I20180308-0630genie.releng1-1/+1

Back to the top