When SolveSpace tries to display an Unicode code point (e.g. U+EA00 )
that does not exist in the embedded vector font (unifont.hex.gz) it hangs
in an endless loop in
`const BitmapFont::Glyph &BitmapFont::GetGlyph(char32_t codepoint)`.
The reason is that the binary search through the text file unifont-8.0.01.hex
that does the "lazy loading" of glyphs does not end.
Here is a short excerpt from the file:
```
D7FE:00007FFE63866DF66DEE6DDE63DE7FFE7FFE61866FBE638E6FBE6F867FFE0000
D7FF:00007FFE63866DF66DEE6DDE63DE7FFE7FFE61866FBE638E6FBE6FBE7FFE0000
F900:0080108810881FF8100800007FFE00000FF008100FF00810042002447FFE0000
F901:00047FFE008010841FFE10841FFC10841FFC10840880050003000CC0703E0000
```
When searching for `0xEA00` after some iterations of the while loop
2450010bbf/src/resource.cpp (L567)
both `first` and `last` end up pointing to F900. After that on each
consecutive iteration of the loop `last` ends up pointing to the LF (0x0A)
just before F900
2450010bbf/src/resource.cpp (L585)
and then `mid` ends up back on F900 here
2450010bbf/src/resource.cpp (L570)
and this will repeat enlessly.
The solution is to do
```
first++;
```
here
2450010bbf/src/resource.cpp (L591),
which will make `first==last` and change whe while loop contition to `while(first < last) {` thiw will
allow the while loop to exit.
Tested with
- 0xEA00 - non existent, not found in 16 iterations.
- 0xF900 - exists but takes exactly 16 iterations of the binary search to finish
- 0x0000 - found in 16 iterations
- 0xFFFD - the replacement Unicode code point for non-existing glyphs. Also the end of the font.
- 0xFFFFF - a codepoint beyond the end, not found and does not cause an exception
The lazy parsing of the vector front was introduced here.
645c2d90acFixes#1150
If the main window is not high enough allow the default 32 pixel padding
between the menu bar and the toolbar to shrink down to zero.
This allows the main window height to be a minimum of 688 pixels (on
Windows 10) so it is possible to capture 720P video tutorials.
Fixes#1130
* mac: Support for pan, zoom and rotate trackpad gestures
Currently SolveSpace is nearly unusable on a mac if you only have a
buttonless trackpad and not a mouse, because there's no way to pan
(ie right-click-drag) or rotate (ie middle-click-drag). You can zoom,
but only by using two-finger-drag up and down, which ends up getting
interpreted as a scrollwheel event.
This change makes the app behave much more like any other mac app, by
adding 2-finger-drag pan gesture support and pinch-gesture zooming, and
3D rotate using shift-2-finger-drag.
I've also added support for the rotate two-finger trackpad gesture,
which rotates directly around the screen Z axis (rather than in all 3
dimensions) which is actually something I've found myself wanting to do
with the mouse but afaik there's no equivalent way of achieving that.
While I was there, I fixed a bugette in convertMouseEvent which was
incorrectly translating the NSEvent coordinates, and then fixing up the
fact that the sign of the y-coordinate was wrong as a result. Using the
convertPoint API correctly means that fixup is not required because
convertPoint handles it for you.
* Don't do trackpad gestures on anything except the toplevel window
* mac: Fix non-functional scrollbar on text window
Which has not worked quite right since the last major refactor.
* Don't pass right-button drags to the toolbar
This improves the behaviour of trackpad pan/rotate on mac which uses
simulated right-button events.
* Don't pass cmd/ctrl modifier through on trackpad pan/rotate MouseEvents
By no longer always returning NSTerminateCancel in
applicationShouldTerminate.
And implement applicationWillTerminate to ensure the cleanup code in
SolveSpaceUI::Exit() is always called.
By making IsForcedToMesh() always return true for STL link groups,
rather than trying to set forceToMesh=true during the import phase.
STL link groups are now always shown as "model already forced to
triangle mesh" in the details screen, but also (unlike when the model
is forced to mesh by a parent group) show the '∆' icon in the group
list.
Where each entity in the active workplane sketch is projected a
different amount normal to the workplane, to allow inspection and
easier selection of entities that entirely overlap each other and are
thus otherwise difficult to see or select.
The distance between the exploded "layers" can be controlled in the
configuration page. Negative distances mean the layers are projected in
the opposite direction, relative to the workplane normal.
As per Xcode 12.4 you can at least do a warning-free incremental build
with these changes. There are still plenty of warnings in a full build
(mostly from thirdparty components) but with these changes you can at
least develop on mac and see if/when you've added any new warnings when
doing incremental builds.
- Use `std::vector<T> elemstore` to store elements. Avoids manual memory management.
- Add and index (`std::vector<int> elemidx`) that avoids moving large objects
when adding elements.
- Add a free element list (`std::vector<int> freelist`) that speeds up element removal by
avoiding rearranging the element storage. It also avoids reallocations when adding
elements later.
- Add a proper iterator. It will be used to remove NextAfter - which is a performance bottleneck.
If a single point is selected when a "Constrain | Comment" (`;`) is added
then the comment is associated with the point and its position becomes
relative to the point. In this way the comment will move with the point.
If nothing is selected or more than a single point is selected then
the behaviour is as before and the comment is "floating".
Closes#1032
Keep items in selected state while dragging so that property browser
will track them until the left mouse button is released. Also, trigger
a property browser update on mouse move while dragging to make the
display "live"
Fix a bug with the lookup of stipple pattern for default styles when the
config settings are not yet available (install/upgrade/first run). This
caused hidden lines to display as continuous rather than dashed.
Also, this fixes the Printf format for the ambient lighting entry on the
configuration screen. This was causing string behavior when attempting
to edit the ambient lighting value on Windows. Oddly this didn't seem
to affect Linux.
* minor fix open/save dialogue on windows
On windows 10 the open/save dialogue box has an minor error, and I believe I fixed it.
When "Open" is selected from the menu, the title of the dialogue box says "SolveSpace - Save File" and the entered file name is "united". My fix correctly titles the dialoged box, and leaves the address bar blank when a file is being opened because "united" is only needed as a default name when a file being saved.
I found that class FileDialogImplWin32 from guiwin.cpp contains two if statements for "isSaveDialog". This is redundant. I removed the first where the title was originally set, but not working. I then set the title in the second if statement and moved the 'if isEmpty'' to this section.
* Update guiwin.cpp
replaced tabs with spaces
E.g. for when you have more than one user logged in, in which case
spacenavd works only for the first of those X11 displays.
We try to do it the old way first, so there should be no regressions.
Also fixes storing non-booleans in bool variables.
Primarily, this enables the user to export of construction entities in 2D
views such as SVG. Previously constuction entities were always skipped.
The "export these objects" is now available for all default styles.
One may turn off export of constraints or the inactive groups for example.
This also adds the exportable flag to the factory settings and
support for saving the exportable option for default styles in the
configuration. Construction entities with custom styles respect this
option as well. NOTE: Running this version will add new entries to the
configuration (Registry, .config etc.) on exit when testing this code.
Gdk::Cursor::create(Gdk::HAND1) yields a hand cursor more appropriate
for "grabbing" vs. pointing. Use the recommended create by name API
to get a "pointing hand" cursor.
When dragging points we must always update position and mark them dirty
through all code paths.
Ensure the ctrl and shift modifier rotation quarternion is always set to
something reasonable.
separate extraLine tracking code from drag tracking code
...by 1/40th of the scrreen width so that the user can notice it.
This can happen if the window was on a second monitor, which is
disconnected, or if the user knows about `Alt-Space | Move` and then
moving the window with the arrow keys.
Before this, if for example the left edge was off-screen it was moved
to 1920, which is just off-screen, so the window remained invisible.
Fixes: #938
On Windows SolveSpace 2.3 uses a plain dialog for `SolveSpace::Error` and
`SolveSpace::Message` with no icon and no system beep. After the GUI
abstraction was reworked this changed to the default system message boxes
(using MessageBoxIndirectW) that play the "Default Beep" sound and show
red "X" and blue "i" icons respectively.
The beep is annoying since the error and message dialogs are used often to
show required conditions for constraints, new groups and other behaviors.
This disables the beep and uses the SolveSpace icon.
Fixes: 719
On macOS actual scroll delta is used for the zoom amount.
On Windows WHEEL_DELTA is used to allow smooth scrolling if supported.
Shift+Scroll is added for 10x finer zooming.
Went through first the diff of the referenced commit,
then all instances of "Create(Open|Save)FileDialog";
added SuggestFilename() calls where a default exists
This has been previously removed in
6b5db58971Closes#538
The return value of these functions is not an error code and according to
the Win32 API documentation they can not affect `GetLastError`.
Calling sscheck on them normally does not fail since it does
SetLastError(0) before running the checked expression and only then
GetLastError(). However in issue 817 a user discovered that when running
"DisplayFusion" software GetLastError does return an error and SolveSpace
closes.
So while not a strict bug-fix this is a "correctness improvement" for
SolveSpace and works around a possible bug in DisplayFusion.
Similarly the return value of Reg*** functions is now compared to
ERROR_SUCCESS which is zero. Before the sschecks were strictly wrong but
did not cause problems for the same reason as above.
... is enabled.
"err" was first introduced in c2c26e95ad to indicate sketches that may cause
problems in the subsequent 3D groups. But is makes sense not to display the
error if the "check sketch for closed contour" option is turned off. The user
obviously does not want to be warned.
Based on a suggestion in https://github.com/solvespace/solvespace/issues/819
On scroll wheel events convert the mouse coordinates from screen to client
area so that scroll wheel zooming remains centered irrespective of the
window position.
Fixes https://github.com/solvespace/solvespace/issues/806
It was added in 3a3a2755b as a potential way to export colorful meshes
to Horizon EDA but ended up being supported only by SolveSpace. Since no
software can consume the exported q3do files the feature is superfluous.
See https://github.com/solvespace/solvespace/issues/795 for details.
This is another small profiling driven optimization.
Moving the initialization of `const double *c` as part of the definition
also helps with the generated assembler.
This reverts commit 68b1abf77f.
The warnings are valuable and shouldn't be cast aside.
As of 8f509f1, we special case macOS and don't set -fno-sanitize-recover
to allow CI to succeed.
In the future, this could be made stricter again by only suppressing
known bugs, which ideally should also be fixed or reported upstream.
Profiling with MSVC 2019 showed that many of the Vector methods are on
a critical path (not surprising). They are changed to be inline and
unnecessary temporaries are removed.
On the example below generate times decreased from 102s. to 64s.
At the same time the executable size shrank from 5569536 to 5150208 bytes
in release mode (with global optimizations).
This should not stop us from working on optimizing inner loops e.g. https://github.com/solvespace/solvespace/issues/759 .
[Test model](https://github.com/solvespace/solvespace/files/5414683/PrismConeNURBSNormalsTangents300.zip)
Previously sanitizer flags were set unconditionally for
all code, including that of external libraries.
Set them only for targets in src/, tests/ and exposed/.
Unfortunately, the linker equivalent to add_compile_options,
add_link_options, is only available for CMake version >= 3.13.
So add the sanitizer flags manually to each target's linker options.
The implementation may be sub-optimal, since the colour and alpha is
defined for each NURBS surface instead of on group level, but the STEP
export currently does not represent group structure at all and I am not
familiar with the format in order to change this.
Fixes: https://github.com/solvespace/solvespace/issues/452
- Add OpenMP to macOS build
- Use as many cores as possible in CI
- Update travis osx image to xcode12.2
- Ignore .vscode folder
- In `.travis/sign-macos.sh`, only create keychain when `CI` variable is present
- Only run macOS deploy stage when a tag is pushed
`GetSaveFileNameA` `OPENFILENAMEA` does not like UNC ( "\\\\?\\C:\\..." ) file prefixes in `lpstrFile`.
Work around it by not `Expand`-ing parameters passed on the command line too early.
The only user visible change is that "File|Open Recent" will show items as they
were passed instead of expanded to full path for example:
"..\..\NURBSTests\Intersection2.slvs"
Fixes: https://github.com/solvespace/solvespace/issues/622
When a plane coinsides with a seam we need to copy that trim curve. The existing curve belongs to the original shell surfaces and an intersection is otherwise not found. Fixes#540.
Extrusion direction (normal) of arcs and circles were not taken into
account when importing.
- Add method for calculating a quaternion from extrusion direction
according to DXF arbitrary axis algorithm
- Add required workplanes for arcs not on XY origin plane
- Adjust addDimRadial and addDimDiametric to include normal when
creating associated circle request
This fixes issues #499. The --view option changes projUp and
projRight. For --view to affect the camera, the camera needs to be
initialized using these values.
This set of changes covers all triangulation errors in issue 693. A point coincident with a triangle vertex is OK and needed for bridges, but sometimes the middle point has edges cutting through the triangle that make it a non-ear. We fix that and a couple of off-by-one error (that fixes one of the test cases).
