There were a lot of incertainty of what should happen when we copy layers being
descendant of each other (i.e. when you select a group layer and some of its
children), then when you paste such data. So we sat down with Aryeom and tried
to come up with some consistent behavior which is somewhat expectable, but also
which would allow the most use-case.
Otherwise it was making very weird result when pasting the data, duplicating
some layers and whatnot, which was obviously a buggy behavior and never the
expected result.
We decided that if you select one leaf item, then even if you also selected a
parent item, it would be as though the parent was not selected. This is very
often what you expect anyway when you select a whole bunch of layers and would
work well if, say, you shift-click over many layers in sub-groups. Then you
wouldn't have to manually ctrl-click to unselect every group.
Then what if you were instead expecting to copy many groups? Then you could
shift-click the group arrow, closing all same-level groups. Once they are all
closed, you can shift-click the groups to only select group layers, not their
contents.
This way, both use cases are still quite doable easily with this default choice.
I created a new function gimp_channel_combine_items() which combines a
list of items with a channel. The list of items is first combined
together as an union set, then only combined with the channel with the
desired operation (this is important for operations such as intersect
which was broken in my previous commit because all items would be
intersected with each other and the selection, whereas we actually want
the union of all items to be intersected with the selection). This new
function is now used for "Alpha to Selection".
Also similarly to copy or color-pick on multi-layers, alpha to selection
will now use the composited result of the multi-layers as visible. In
particular it means that opacity, modes and visible properties on the
layers are taken into account. Alpha to selection on a single layer
though still works as previously, only taking the non-composited layer
data into account.
I am actually struggling if alpha to selection on uncomposited layers
(just an union on alpha to selection result for each layer) would not
make sense to on some workflows. To be experimented.
Finally it is to be noted that this function should also work on
channels and vectors (both single or multiple; and of course in such
cases, compositing does not matter) though I haven't tested yet. It
could even work with a source mix of layers, channels and vectors,
though our GUI does not allow such action currently.
In gimpchannel-select, move some of the common functionality of the
various gimp_channel_select_foo() functions to gimpchannel-combine.
Furthermore, don't special-case CHANNEL_OP_INTERSECT, but rather
pass it over to gimpchannel-combine, which is now prepared to
handle it in all functions, as per the previous commits.
In gimpchannel-combine, factor out the common functionality of the
various gimp_channel_combine_foo() functions into a pair of
gimp_channel_combine_{start,end}() functions, which are called
before/after the actual gimp_gegl_mask_combine_foo() function,
respectively. In particular, these functions deal with calculating
the new channel bounds. Previously, the various
gimp_gegl_mask_combine_foo() functions would implicitly invalidate
the channel bounds (since commit
d0ae244fe8), rendering the bounds-
recalculation code ineffective. This avoids manually recalculating
the bounds in many cases, speeding up selection operations.
Improve gimp_gegl_mask_combine_ellipse_rect() -- the funciton
responsible for rendering ellipse/rounded-rectangle selections.
Most notably, this commit significantly improves the function's
performance, by identifying whole tiles, whole rows, or parts of a
row, that are fully inside, or fully outside, the ellipse, and
filling them in bulk, instead of calculating the anti-aliasing
value at each pixel, which is now only done along the
circumference.
This commit also improves anti-aliasing, by more accurately
approximating the distance from a pixel to the ellipse, and by
normalizing the distance according to the pixel's cross-section
length in the direction of the said point. In particular, we
guarantee that pixels that are fully inside/outside the ellipse
have a value of 1/0, respectively, facilitating the aforementioned
optimization.
Additionally, this commit fixes various edge cases where several
primitives coincide at a single pixel (in the rounded-rectangle
case), adds support for CHANNEL_OP_INTERSECT, and parallelizes
processing.
- don't include <gdk-pixbuf/gdk-pixbuf.h> in headers in app/
- instead, include it in many .c files instead of <glib-object.h>,
finally acknowledging the fact that app/ depends on gdk-pixbuf almost
globally
- fix up includes as if libgimpbase depended in GIO, which it soon will
The actual algorithm is still the same sick algorithm that was used
before. But instead of iterating the mask row-by-row and filling
it in small spans, we now use one pixel_regions_process() loop to
process the whole mask. Makes a significant difference for large
elliptical selections.
Remove gimp_channel_add_segment() and gimp_channel_sub_segment()
as they are not needed any longer and were responsible for the
bad performance.
2009-01-29 Sven Neumann <sven@gimp.org>
* app/core/gimpchannel-combine.c
(gimp_channel_combine_ellipse_rect):
fixed incorrect optimization that caused glitches in the rounded
corners on the left side of rectangular selections.
svn path=/trunk/; revision=27968
2009-01-17 Michael Natterer <mitch@gimp.org>
* all files with a GPL header and all COPYING files:
Change licence to GPLv3 (and to LGPLv3 for libgimp).
Cleaned up some copyright headers and regenerated the parsers in
the ImageMap plugin.
svn path=/trunk/; revision=27913
2008-10-09 Michael Natterer <mitch@gimp.org>
Add GEGL_CFLAGS and #includes as if gimpdrawable.h and gimpimage.h
had a GEGL dependency (they will have in the next commit, but I
wanted to keep the commit separate).
* app/dialogs/Makefile.am
* app/file/Makefile.am
* app/gui/Makefile.am
* app/menus/Makefile.am
* app/paint/Makefile.am
* app/plug-in/Makefile.am
* app/text/Makefile.am
* app/vectors/Makefile.am
* app/widgets/Makefile.am
* app/xcf/Makefile.am: add GEGL_CFLAGS.
* app/actions/*.c
* app/core/*.c
* app/dialogs/*.c
* app/display/*.c
* app/file/*.c
* app/gui/*.c
* app/menus/*.c
* app/paint/*.c
* app/pdb/gimppdb-utils.c
* app/pdb/gimpprocedure.c
* app/plug-in/*.c
* app/text/*.c
* app/tools/*.c
* app/vectors/*.c
* app/widgets/*.c
* app/xcf/*.c: add <gegl.h> or replace <glib-object.h> by <gegl.h>
to all files which include a drawable subclass or gimpimage.h
* tools/pdbgen/app.pl: include <gegl.h> instead of <glib-object.h>
in all generated files.
* app/pdb/*-cmds.c: regenerated.
* data/images/gimp-splash.png: the goat is still sleeping.
By Aurore Derriennic.
svn path=/trunk/; revision=27202
2007-03-09 Michael Natterer <mitch@gimp.org>
* app/core/core-types.h: include "libgimpmath/gimpmathtypes.h"
instead of "libgimpmath/gimpmath.h".
* app/core/gimpbrush.h
* app/paint/gimppaintcore.h
* app/paint/gimpperspectiveclone.h
* app/text/gimptext.h
* app/tools/gimptransformtool.h: include gimpvector.h and
gimpmatrix.h explicitely where they are needed in public structs.
* app/*/*.c
* tools/pdbgen/pdb/paths.pdb: include "libgimpmath/gimpmath.h"
where needed.
* app/pdb/paths_cmds.c: regenerated.
svn path=/trunk/; revision=22084
2006-10-18 Michael Natterer <mitch@gimp.org>
Applied modified patch from Martin Nordholts which adds a "Rounded
Corners" option to the rectangle select tool. Fixes bug #86279.
* app/core/gimpchannel-combine.[ch]: added
gimp_channel_combine_ellipse_rect(). Use it from
gimp_channel_combine_ellipse().
* app/core/gimpchannel-select.[ch]: added
gimp_channel_select_round_rect()
* app/tools/gimprectangleselectoptions.[ch]: added properties
"round-corners" and "corner-radius" and GUI for the new propeties.
* app/tools/gimprectangleselecttool.h: added macro
GIMP_RECT_SELECT_TOOL_GET_OPTIONS().
* app/tools/gimprectangleselecttool.c (gimp_rect_select_tool_draw):
draw round corners if enabled.
(gimp_rect_select_tool_real_select): use
gimp_channel_select_round_rect() if enabled.
* app/tools/gimpselectionoptions.[ch]: added "antialias_toggle"
to the GimpSelectionOptions struct so the rect select options
can set its sensitivity.
Unrelated:
* app/tools/gimpellipseselecttool.c (gimp_ellipse_select_tool_draw):
use 360 * 64 instead of 23040.
2006-08-29 Sven Neumann <sven@gimp.org>
* libgimpbase/Makefile.am
* libgimpbase/gimpbase.h
* libgimpbase/gimprectangle.[ch]: added new files that hold
gimp_rectangle_intersect(), factored out of the core.
* libgimpbase/gimpbase.def: updated.
* app/core/gimp-edit.c
* app/core/gimp-utils.c
* app/core/gimp-utils.h
* app/core/gimpchannel-combine.c
* app/core/gimpdrawable-foreground-extract.c
* app/core/gimpdrawable-transform.c
* app/core/gimpdrawable.c
* app/core/gimpimage-preview.c
* app/core/gimplayer.c
* app/core/gimpscanconvert.c
* app/display/gimpdisplayshell-draw.c: changed includes accordingly.
* libgimp/gimpdrawablepreview.c: don't duplicate
gimp_rectangle_intersect() here, use the function in libgimpbase.
* app/base/siox.c: use gimp_rectangle_intersect() to reduce the
working area to the region of interest. Fixes bug #340422.
2005-09-07 Michael Natterer <mitch@gimp.org>
* app/core/gimpchannel-combine.c (gimp_channel_combine_ellipse):
use gimp_rectangle_intersect() instead of just looking at the
passed in w,h to figure whether nothing needs to be done. While
the algorithm does nicely with an entirely out-of-image ellipse,
the channel's bounds were set to a rectangle of zero width or
height, which caused subsequent calls to gimp_channel_bounds() to
return broken bounds. Fixes bug #315417.
(gimp_channel_combine_rect)
(gimp_channel_combine_mask): use gimp_rectangle_intersect() here
too instead of the usual unclear CLAMPing on x1,y1,x2,y2.
2004-07-19 Sven Neumann <sven@gimp.org>
* app/core/gimpchannel-combine.c (gimp_channel_combine_ellipse):
moved variable declarations to the scope they are being used in,
removed trailing whitespace, minor cleanups.
2004-07-19 Michael Natterer <mitch@gimp.org>
* app/core/gimpchannel-combine.c (gimp_channel_combine_ellipse):
comments not intended for gtk-doc must not start with '/**'.
2003-10-06 Michael Natterer <mitch@gimp.org>
Treat changes to the selection like changes to any other drawable:
* app/core/gimpchannel.c
* app/core/gimpchannel-combine.c: call gimp_drawable_update() after
changing the channel.
* app/core/gimpimage.[ch]: added struct GimpImageFlushAccumulator
with one member "gboolean mask_changed". Connect to "update" of
the selection and set accum.mask_changed to TRUE in the callback.
Added default implementation for GimpImage::flush() and emit
"mask_changed" there.
Unrelated:
* app/core/gimpimage.h: removed GimpGuide struct...
* app/core/gimpimage-guides.h: ...and added it here.
* app/core/gimpimage-undo-push.c (undo_pop_mask)
(undo_pop_channel_mod): don't distinguish between selection and
non-selection channels and just call gimp_drawable_update().
* app/core/gimpundo.h
* app/core/gimpimage-undo.c: removed "gboolean mask_changed" from
the GimpUndoAccumulator struct since we don't have to care about
that signal explicitly any more.
* app/display/gimpdisplay-foreach.[ch]: removed gimp_displays_flush().
* tools/pdbgen/pdb/display.pdb (displays_flush_invoker): call
gimp_image_flush() on all images so the flush accumulator is
honored.
This generalization enables the removal of more special purpose
code which was needed to treat the selection different:
* app/core/gimpimage-mask-select.[ch]: removed...
* app/core/gimpchannel-select.[ch]: ...and added under a new name
because it's not selection specific any more.
* app/core/gimpimage-mask.[ch]: removed...
* app/core/gimpselection.[ch]: ...added the two remaining
functions here. Removed all calls to gimp_image_mask_changed().
* app/core/Makefile.am
* app/core/gimp-edit.c
* app/core/gimpdrawable-transform.c
* app/core/gimpimage-scale.c
* app/core/gimpimage-snap.c
* app/display/gimpdisplayshell.c
* app/gui/channels-commands.c
* app/gui/layers-commands.c
* app/gui/select-commands.c
* app/gui/vectors-commands.c
* app/tools/gimpbycolorselecttool.c
* app/tools/gimpeditselectiontool.c
* app/tools/gimpellipseselecttool.c
* app/tools/gimpfreeselecttool.c
* app/tools/gimpfuzzyselecttool.c
* app/tools/gimpiscissorstool.c
* app/tools/gimprectselecttool.c
* app/tools/gimptransformtool.c
* app/widgets/gimpchanneltreeview.c
* app/widgets/gimpselectioneditor.c
* app/widgets/gimpvectorstreeview.c
* app/xcf/xcf-save.c
* tools/pdbgen/pdb/paths.pdb
* tools/pdbgen/pdb/selection.pdb
* tools/pdbgen/pdb/selection_tools.pdb: changed accordingly.
* app/core/gimpdrawable-bucket-fill.c
* app/core/gimpimage-colormap.c
* app/core/gimplayer-floating-sel.c
* app/core/gimplayer.c
* app/gui/image-menu.c
* app/paint/gimppaintcore.c
* app/tools/gimpcroptool.c
* app/tools/gimpinkoptions.c
* app/tools/gimpvectortool.c: removed useless and/or obsolete
#includes.
* app/pdb/display_cmds.c
* app/pdb/paths_cmds.c
* app/pdb/selection_cmds.c
* app/pdb/selection_tools_cmds.c: regenerated.
2003-09-04 Michael Natterer <mitch@gimp.org>
* app/core/gimpdrawable.[ch]: added new pure virtual function
GimpDrawable::invalidate_boundary().
* app/core/gimplayer.[ch]: implement it and removed public
function gimp_layer_invalidate_boundary().
* app/core/gimpchannel.[ch]: implement it.
* app/core/gimpselection.[ch]: implement it and removed public
function gimp_selection_invalidate().
* app/core/gimpimage-mask.c (gimp_image_mask_invalidate)
* app/core/gimpimage-undo-push.c
* app/core/gimpimage.c
* app/core/gimplayer-floating-sel.c
* app/text/gimptextlayer.c: changed accordingly.
* app/core/gimpchannel.[ch]: made gimp_channel_push_undo() a
public function and made it call
gimp_drawable_invalidate_boundary(). Added undo_desc strings for
all undo pushing functions to GimpChannelClass.
* app/core/gimpselection.[ch]: removed gimp_selection_push_undo()
since after the change above it was identical to
gimp_channel_push_undo(). Don't push any undo here since
upchaining does the right thing now. Override GimpChannelClass'
undo_desc strings to say "Selection".
* app/core/gimpimage-mask.c (gimp_image_mask_push_undo): changed
accordingly.
2003-09-03 Michael Natterer <mitch@gimp.org>
* app/core/gimpchannel.[ch]: made all functions which push an
undo step virtual and added them all as default implementations.
* app/core/Makefile.am
* app/core/core-types.h
* app/core/gimpselection.[ch]: new object which is a GimpChannel
subclass and implements all of its virtual functions, pushes
an image_mask undo and chains up with "push_undo = FALSE".
* app/core/gimpimage-mask.[ch]: made most functions simple
wrappers like gimp_channel_invert(gimp_image_get_mask(gimage));
so the API stays the same for now.
* app/core/gimpimage.[ch]: create a GimpSelection object
as gimage->selection_mask. Removed "gboolean mask_stroking"
since it is in GimpSelection now.
* app/xcf/xcf-load.c (xcf_load_channel_props): added an evil hack
which turns a GimpChannel into a GimpSelection once we figured the
loaded channel is the selection.
* app/core/gimplayer.c (gimp_layer_create_mask):
gimp_channel_clear() takes an additional "const gchar *undo_desc"
parameter now.
* app/core/gimpscanconvert.c (gimp_scan_convert_to_channel): set
mask->bounds_known to FALSE before returning the new channel
* app/tools/gimpiscissorstool.c (iscissors_convert): no need to
call gimp_channel_invalidate_boundary() on the channel returned by
the above function.
* app/core/gimpchannel.[ch]: removed
gimp_channel_invalidate_boundary() since it is no longer needed.
2003-09-01 Michael Natterer <mitch@gimp.org>
* app/core/gimpitem.[ch]: added new virtual function
GimpItem::stroke().
* app/core/gimpchannel.c
* app/vectors/gimpvectors.c: implement GimpItem::stroke().
* app/core/gimpimage-mask.[ch] (gimp_image_mask_stroke): changed
signature to match gimp_item_stroke() (the selection mask *really*
should be a GimpChannel subclass).
Removed global variable "gboolean gimp_image_mask_stroking"...
* app/core/gimpimage.[ch]: ...and added "gboolean mask_stroking"
to the GimpImage struct.
* app/gui/vectors-commands.[ch]: removed vectors_stroke_vectors().
* app/widgets/widgets-types.h: removed GimpStrokeItemFunc typedef.
* app/widgets/gimpvectorstreeview.[ch]: removed "stroke_item_func"
member and use gimp_item_stroke() instead.
* app/gui/dialogs-constructors.c (dialogs_vectors_list_view_new)
* app/gui/edit-commands.c (edit_stroke_cmd_callback)
* app/gui/vectors-commands. (vectors_stroke_cmd_callback)
* app/widgets/gimpselectioneditor.c
(gimp_selection_editor_stroke_clicked)
* tools/pdbgen/pdb/edit.pdb (gimp_edit_stroke): changed accordingly.
* app/pdb/edit_cmds.c: regenerated.
Note that there is no GUI for "stroke channel", although it would
be utterly cool to have one, since currently slelection stroking
cannot be masked by a selection (because we stroke the selection).
Anyway, if anyone has an idea how to trigger "stroke channel" with
another drawable active (the one to stroke to), please let me
know...
