When font family of a private font is retrieved with call to
Gdiplus::Font::GetFamily() then later on there is thrown access violation
exception when array of cached private font families (filled in by
Gdiplus::PrivateFontCollection::GetFamilies()) is deleted in
wxGDIPlusRenderer::Unload(). Call to Font::GetFamily() is done in
wxGDIPlusContext::GetTextExtent() so this issue can be seen once text
extent is retrieved for a private font.
Because it looks that calling to Font::GetFamily() for a private font is
messing up something in the array of cached private font families so
we should avoid calling this method and directly fetch corresponding font
family from the cache instead.
Reference to the cached font family would be stored in
the wxGDIPlusDataFont and it could be fetched from there directly instead
of making a call to Font::GetFamily().
Closes#18704.
Line height calculation based on the font size is not right for wxGTK3
because native buttons (used as in-line editor buttons) don't scale down
well and they are not displayed properly within the line.
Minimal button size when the label is displayed properly is 35 so this
has to be minimal line height under wxGTK3.
Closes#18715.
For the sake of consistency with editor buttons created with
wxPropertyGrid::GenerateEditorButton(), the borders around the buttons
should be taken into account while positioning the buttons
in wxPGMultiButton.
Because the size of the text (usually) shown in the grid changes, the
size of its columns/rows should change as well when DPI changes. Scale
them by the DPI ratio if necessary, i.e. if any of them were set to
non-default values to achieve this.
Note that this is not ideal because it can result in rounding errors and
hence in slight change of row/column size when moving the grid window to
another display with a different DPI and then back, but to prevent this
from happening we'd need to store the column/rows sizes in DIPs inside
wxGrid and convert them to actual physical pixels everywhere, which
would require much more changes. Still, if the round trip problems turn
out to be a real problem in practice, we probably will need to do this.
Also note that this doesn't take care of the grid with a native header,
but this will be addressed by upcoming commits.
This handler redoes wxGrid layout and refreshes it to at least avoid
ugly display artifacts when moving wxGrid window between displays with
different DPI.
Since the changes in a40acbb28e (Add CanOverflow function to
wxGridCellAttr, 2020-02-06), cells with non-default vertical alignment
didn't overflow any more, even if their horizontal alignment was
unchanged and still defaulted to left-aligned.
This was due to assuming that if the alignment of wxGridCellAttr itself
was different from wxALIGN_LEFT, it meant that it wasn't left-aligned,
which seems logical but is in fact false, as the alignment can also be
wxALIGN_INVALID, in which case the real alignment is taken from the
default grid attribute.
Fix this by using GetNonDefaultAlignment() to get the alignment value
effectively used and add a unit test, as well as an example in the
sample, showing that this now works correctly.
In particular, this makes default column width better suited for high
DPI displays, as it was too narrow before.
Also mark a couple of obsolete constants as such with a comment.
In many case SetTable() is called with its takeOwnership parameter set
to true, as shown by the grid sample in which all 3 of the calls to
SetTable() set it to true, but calling it in this case is awkward, as
bare "true" in the caller is unreadable and almost invariably requires
an explanatory comment.
Improve the API by adding AssignTable(), which is the same to SetTable()
as the existing AssignImageList() to SetImageLabel(), which always takes
ownership of the table pointer.
This file uses wxScopedPtr<> and so needs to include this header
explicitly instead of relying on it being included indirectly from some
other header, as happens in the default build, but not always.
Closes#18711.
Don't include wx/private/spinctrl.h header (nor the other ones) at all
in this case, as it doesn't compile in this build configuration.
This fixes one of build problems under iOS.
See https://github.com/wxWidgets/wxWidgets/pull/1773
::ExtCreatePen() doesn't allow the pen width to be 0, unlike
::CreatePen() and failed to create the brush in this case.
Use width of 1 to draw e.g. hatched lines of width 0 correctly.
See #7097.
Use ::ExtCreatePen() for creating pens for such lines, as it results in
much better appearance for them than ::CreatePen(), which draws dashed,
and not dotted, lines in this case.
See #7097.
Explicitly include <commctrl.h> instead of relying on it being done
implicitly by wx/msw/uxtheme.h, as this doesn't happen when
wxUSE_UXTHEME is 0.
Closes#18702.
The change in the value can be greater than 1 when the arrows are held
pressed, contrary to what the logic of determining the wraparound added
in 086793ceef supposed.
Replace this with a check of whether we switch from the min value
directly to the max one or vice versa, which mostly works and avoids
asserts, even if it can still fail and produces wrong events when the
arrows are held pressed for long enough to increase the delta to the
range of the control, in which case we just can't distinguish between
wraparound and passing from min to max (or vice versa) in a single step,
which means that we have no way to determine the right event to send.
But producing a wrong event is better than asserting, so this still
counts as an improvement.
Closes https://github.com/wxWidgets/wxWidgets/pull/1764
See #17957.
Closes#18695.
This variable was written to and read from different threads without any
synchronization, which resulted in tons of TSAN warnings -- in addition
to being incorrect, of course.
Fix this by using C++11 atomic type for it if available and just not
doing anything otherwise, as there doesn't seem to be any simple
workaround for this problem without an atomic 64 bit type that our own
wx/atomic.h doesn't currently provide.
Make the left arrow collapse the item if it's currently expanded before
falling back to its previous behaviour of going to the item parent and
the right arrow to go to the first child if the item is already
expanded.
This is compatible with the native MSW control behaviour, but mostly
just more useful and convenient.
Closes#18684.
Don't deselect all the other items when an item is collapsed or expanded
in wxGenericTreeCtrl with wxTR_MULTIPLE style, this was completely
unexpected and seems to have been accidentally introduced back in
35cf1ec63cCloses#18680.
Don't call GetSelections() unnecessarily if we're not going to use its
result in any case because either Shift or Command/Control key was
pressed when the mouse button was released.
See #18680.
This test remained since 902725eefe but
should have been simply removed back then, as we return if the item is
null due to a test above.
No real changes.
Set LVCFMT_IMAGE, meaning that the column just uses an image from the
image list, instead of LVCFMT_COL_HAS_IMAGES, meaning that the column
contains the image from the image list and do not set LVCF_IMAGE when
resetting the image, as this still reserved space for the (invalid)
image, resulting in an extra blank area in the column header.
Closes https://github.com/wxWidgets/wxWidgets/pull/1739Closes#18617.
GetTextObject() might return null, e.g. it does it for read-only
wxBitmapComboBox, so EnableTextChangedEvents() must account for this
possibility, as it's not really possible to avoid calling it in this
case, as it's called indirectly from e.g. SetSelection().
Check that the entry is valid before enabling or disabling events for it
to avoid several GLib assertion failures every time when e.g.
wxBitmapComboBox::SetValue() is called.
Closes https://github.com/wxWidgets/wxWidgets/pull/1756
Update the column width immediately, as it's being dragged, instead of
drawing a temporary line showing the new column boundary using wxINVERT.
This results in better user experience, as it the effect of changing the
column width can be immediately seen (especially important for non-left
aligned columns or columns using ellipsizition) and, equally if not more
importantly, fixes wxGrid drag-resize not showing any visible UI at all
with wxGTK3 and wxOSX where wxINVERT is not implemented.
Rename the functions used from wxGridHeaderCtrl event handlers to start
with DoHeader prefix to make it clear that they're (only) used by it in
an attempt to make things more clear and more uniform.
No real changes.
Don't reuse the same m_dragRowOrCol variable for both the index of the
row or column being drag-resized and for the index of the column being
drag-moved. Reusing it was confusing and made it more difficult what the
code was doing and what invariants were preserved in it, and just wasn't
worth saving a few bytes per wxGrid object.
No real changes.
Use new ShouldRefresh() helper instead of testing for !GetBatchCount()
before calling Refresh().
Also check for GetBatchCount() inside CalcDimensions() itself, which
means that it can now be called unconditionally.
No real changes.
