Since the 'Enforce palette colors' is a feature of the internal color
selector, this commit moves the functionality from KisPaletteView to
KisDlgInternalColorSelector. Also, that way the color change is forced
only once upon foreground color change, preventing cycling signals and confusing color changes
(KisPaletteView can be used more than once with different color palettes).

Removes KisPaletteView::slotFGColorResourceChanged, the function is
obsolete: it was meant to solve bug 402072 that is now fixed in a better way.

BACKPORT:krita/4.2
BUG:408256
CCBUG:402072

Now switching languages will not automatically select the last
template type because the strings we use to compare the last
used pane in the open dialog isn't the same in, e.g., French
and English...

BUG:408638

defaultProfileForColorSpace needs a colorSpaceId including
the channelDepthId, not just the colorModelId.

BUG:408485
BACKPORT:krita/4.2

... so we shouldn't assert on it.

BUG:408618

Duplication happens in a node juggler that tries to keep the number of
actions/updates as low as possible, therefor it keeps a stroke open for a
couple of seconds after the user did the action. We should properly
close this stroke by emitting KisImage::sigStrokeEndRequested().

BUG:408593

Just set default document style to the lastly used font

BUG:408441

We cannot just skip rendering of layers, which has any clones attached,
becasue clones may fetch data from their projections.

BUG:408167
BUG:405536

Apparently this is only needed if some of the entries are not the
default.

Support all documented features of the format.

Filter to match colours to selected palette and
optionally apply some simple dithering.

Differential Revision: https://phabricator.kde.org/D21682

Implements basic version of high pass filter for Krita. More
versions may or may not to come to Krita from GIMP versions.
However, I decided not to do grayscale version as Krita already
have de-saturation filter which takes care of that part nicely.

Differential Revision: https://phabricator.kde.org/D20475

BUG:408171
BACKPORT:krita/4.2

This fixes loading the scripter when not everything has been
translated yet.

BUG:408400

CCBUG:408441

BUG:408481
BACKPORT:krita/4.2

If we know what the value should be, we can use that and not crap
out.

CCBUG:408481

Before there was no way of checking if an image saved in one color space
is opened in the same color space. For example tiff import can assume
things about color space of the image that isn't necessarily true.
This commit adds a possible test to use to check it.

Previous version of tiff export assumed that if visitor returns
false it means internal error, a bug in the code. Truth is, it can
happen also if the image (or layer) has an unsupported color space.
Hence deletion of the assert and replacement InternalError error code
with just Failure.

BUG:408177

I think there actually was something reshuffled in a dependency to lose
them, no problem though

BACKPORT:krita/4.2

BUG:408442

This also reformats the rather complex if statement in a
couple of lines.

Attempting to set the number of columns will immediately
start allocating QMap instances for every column.

For maliciously crafted files, this can take a long time
and waste a whole lot of memory, forcing the user to kill
krita or waiting for the operating system to OOM-kill it.

On invalid column count, the parser will behave as if no
column count had been present at all.

BUG:408447
BACKPORT:krita/4.2