* app/pdb/gimppdb-utils.[ch]: add "gboolean writable" to
gimp_pdb_layer_is_text_layer() because that's called on
all text layers anyway.
* tools/pdbgen/pdb/text_layer.pdb
* tools/pdbgen/pdb/vectors.pdb: pass TRUE if we want to modify the layer.
* app/pdb/text-layer-cmds.c
* app/pdb/vectors-cmds.c: regenerated.
Remove GimpTool::initialize() implementation again and check for the
drawable being locked in GimpTool::button_press() so we can't paint if
the tool has been used before on the same drawable when it was
unlocked.
* app/pdb/gimppdb-utils.[ch]: add "gboolean writable" parameter to
gimp_pdb_item_is_attached() because this function is called on
almost all items where a check for locked is needed.
* tools/pdbgen/pdb/*.pdb: pass writable = TRUE in all checks for items
that are modified. Add explicit calls to gimp_pdb_item_is_writable()
in some rare cases.
* app/pdb/*-cmds.c: regenerated.
Check whether the drawable to edit is locked in GimpTool::initialize()
and bail out with an appropriate error if it is. This currently
prevents cloning from locked drawables, will fix that later.
* app/widgets/gimpitemtreeview.[ch]: add class members for the lock
content button's icon, tooltip and help_id and use them when
creating the button. Create the button in constructor() instead of
init() so we have access to our real class structure without the
need for a custom get_type() function.
* app/widgets/gimpdrawabletreeview.c: configure the button as "Lock pixels".
* app/widgets/gimpvectorstreeview.c: configure it as "Lock path strokes".
Took patch from Martin Nordholts and ported it to add a "lock-content"
property, signal and API in the same way as it's done for "visible"
and "linked".
- new API to add widgets to a box of options, for stuff like the paint
mode menu and opacity scale. Set it sensitive automatically and
update its spacings in GtkWidget::style_set().
- new API to get a hbox for "lock" toggles, for stuff like lock
pixels and lock alpha.
This reverts commit d2e1f2ac74. If we
keep the layer group button in 2.7.0 people will expect layer groups
to fully work and get mad when that is not the case. We can enable it
again after the release.
Add support for having obsolete data resources. An obsolete resource
is not shown in the UI or managed in any way, but it will be
considered when plug-ins requests resources. This in order to maintain
backwards compatibility for plug-ins.
Since Ctrl+E previously meant something harmless, don't use that
keyboard shortcut for the destructive command 'File->Overwrite'. We
still keep Ctrl+E for 'File->Export to' though, and we do this by
having 'File->Overwrite' as a separate GtkAction with its own keyboard
shortcut slot.
Don't reuse the GeglProcessor as it can't handle dirty rects when
painting in GIMP, and this makes it harder to make GEGL-wise progress
in other parts of GIMP.
* app/core/gimpimage.[ch]: rename all gimp_image_reposition_foo() to
gimp_image_reorder_foo() and added "new_parent" parameters. Factor
out calculating of the item's new container and index to a utility
function.
* app/core/core-enums.[ch]: rename the REPOSITION undos to REORDER.
* app/core/gimpimage-undo-push.[ch]
* app/core/gimpchannelpropundo.[ch]
* app/core/gimplayerpropundo.[ch]
* app/vectors/gimpvectorspropundo.[ch]: change accordingly. Remember
the old parent item in all item reorder undos.
* app/widgets/gimpitemtreeview.h: change GimpReorderItemFunc prototype
accordingly.
* app/widgets/gimpchanneltreeview.c
* app/widgets/gimplayertreeview.c
* app/widgets/gimpvectorstreeview.c (class_init): follow image API
name changes.
* app/widgets/gimpitemtreeview.c (gimp_item_tree_view_drop_viewable):
implement reordering of items between groups.
* app/widgets/gimpcontainertreeview.c
(gimp_container_tree_view_reorder_item): fix to reorder the item
within its level of the tree (unrelated to above changes, but needed
to make them work).
* app/widgets/gimpcontainertreeview-dnd.c
(gimp_container_tree_view_real_drop_possible): support DND within a
tree and make sure a group item is not dropped into itself.
* app/widgets/gimpitemtreeview.c (gimp_item_tree_view_drop_viewable):
implement intra-container reordering within all branches; print a
message for inter-container reordering because that's not yet
implemented in the core.
The new functions figures if one viewable is another's parent or
grandparent or grandgrand... Note that unlike gtk_widget_is_ancestor(),
this function has its parameters in the right order.
A tree-enabled replacement for gimp_container_add,remove_handler(),
slightly more sane to a degree that makes me think it could replace
GimpContainer's own handler stuff. Public API consists of two
function:
- gimp_tree_handler_connect() which connects recursively to a specific
signal of all items in a container tree.
- gimp_tree_handler_disconnect() is just a wrapper around unrefing the
handler, but makes sure that dispose() is really executed (which
actually disconnects things).
* app/widgets/gimpcontainerview.h: add "gboolean model_is_tree"
to GimpContainerViewInterface.
* app/widgets/gimpcontainerview.c: default to FALSE and enable the
commented-out optimization in remove_container() for list-only
container views.
Use the actual parent item when adding to the image, not
GIMP_IMAGE_ACTIVE_PARENT because the latter would add a duplicated
group inside itself instead of above it
* app/core/gimpimage.h: add define GIMP_IMAGE_ACTIVE_PARENT for magic
the value ((gpointer) 1) which can be passed as parent item to
add_layer(), add_channel() and add_vectors().
* app/core/gimpimage.c: add utility function
gimp_image_get_insert_pos() which honors above new special value and
figures all of parent container, parent item and insert position.
Remove lots of duplicated code from add_layer(), add_channel() and
add_vectors().
* app/core/gimpimage.[ch]: make the parent parameter public in
add_layer(), add_layers(), add_channel() and add_vectors().
* app/vectors/gimpvectors-import.[ch]: add parent parameters to
the vectors import functions.
* app/core/gimpchannelundo.[ch]
* app/core/gimplayerundo.[ch]
* app/vectors/gimpvectorsundo.[ch]
* app/core/gimpimage-undo-push.[ch]: remember the parent item when
removing layers, channels and vectors.
* app/actions/channels-commands.c
* app/actions/debug-commands.c
* app/actions/edit-commands.c
* app/actions/layers-commands.c
* app/actions/vectors-commands.c
* app/core/gimp-edit.c
* app/core/gimpimage-duplicate.c
* app/core/gimpimage-merge.c
* app/core/gimpimage-quick-mask.c
* app/core/gimplayer-floating-sel.c
* app/core/gimpselection.c
* app/core/gimptemplate.c
* app/dialogs/file-open-dialog.c
* app/display/gimpdisplayshell-dnd.c
* app/text/gimptext-compat.c
* app/tools/gimptexttool.c
* app/tools/gimpvectortool.c
* app/widgets/gimptoolbox-dnd.c
* app/xcf/xcf-load.c
* tools/pdbgen/pdb/image.pdb
* tools/pdbgen/pdb/paths.pdb
* tools/pdbgen/pdb/vectors.pdb: pass NULL as parent item to above
functions and add FIXMEs all over the place because there is some
more hacking needed to make adding with index = -1 (on top of the
current item) work again.
* app/pdb/image-cmds.c
* app/pdb/paths-cmds.c
* app/pdb/vectors-cmds.c: regenerated.
* app/core/gimpimage-duplicate.c: duplicate the original image's
tree structure in the copy.
* app/widgets/gimpitemtreeview.[ch]: add parent to GimpAddItemFunc,
add utility function gimp_item_tree_view_get_drop_index() which
figures where to add something dropped to an item tree.
* app/widgets/gimpchanneltreeview.c
* app/widgets/gimplayertreeview.c
* app/widgets/gimpvectorstreeview.c: changed accordingly, using above
new GimpItemTreeView API.
When using GEGL for the projection, use CIE LCH for the Hue,
Saturation and Value layer modes, just as we already do for the Color
layer mode. The Value layer mode will later be renamed to Lightness
since that is the semantics it has now.
Use gimp_image_get_layer,channel,vectors_list() instead of
gimp_image_get_layer,channel,vectors_iter(). As a side-effect,
simplified some code that was making sure we don't modify the list
returned by the old functions (the new functions return newly
allocated lists so we can safely remove items from the image while
iterating it). Some places will need adjusting once we really have
item trees.
Change gimp_image_add,remove_layer,channel,vectors() to deal with
parent items and insert/remove things to/from the right containers.
The "parent" parameter to the add_foo() functions is currently a local
variable that defaults to NULL, so nothing has chaged yet.
Adjust code which auto-generates the gimprc man page to:
- fix spelling mistakes.
- remove unintentional indentation.
- remove out of place punctuation.
When GEGL is used for the projection, use CIE LCH instead of HSL for
the "Color" layer mode. This give much more accurate and intuitive
results. Requires at least 12d5cc4c1bcfb of babl.
Using the new function makes sure the sensitivity of the raise and
lower actions corresponds to what is possible (raising and lowering
within the group).
This is just a utility function like gimp_image_get_layer_iter()
in order to avoid including "core/gimplist.h" whenever a file needs
access to the actual GList inside the container.
* app/core/gimpitemstack.[ch]: add gimp_item_stack_get_item_list()
which returns a GList of all the stack's items, in depth-first
order if the stack is in fact a tree.
* app/core/gimpimage.[ch]: add gimp_image_get_layer_list(),
channel_list(), vectors_list() which use above new function.
Don't use image->layers, ->channels and ->vectors in the raise, lower
and position functions; use gimp_item_get_container() and
gimp_item_get_index() instead.
* app/core/gimpitem.[ch]: add virtual function ::get_container() plus
pubic API wrapper which returns the children of the item's parent
viewable, or the right toplevel container of its image.
Add utility function get_index() which returns the item's index
within its container.
* app/core/gimpchannel.c
* app/vectors/gimpvectors.c: implement get_container() and return
the right image container.
* app/core/gimplayer.[ch]: same here, and remove previously added
public get_container() API.
* app/core/gimplayermask.c
* app/core/gimpselection.c: implement get_container() for these GimpChannel
subclasses and return NULL.
The new function returns a layer's container, which is either its
parent viewable's children, or image->layers. Scheduled for being
moved to a virtual function on GimpItem.
* app/widgets/gimpcontainerview.[ch]: add and remove container trees
recursively. Change virtual function ::add_item() to pass the
insert_data of the parent viewable.
* app/widgets/gimpcontainercombobox.c
* app/widgets/gimpcontainerentry.c
* app/widgets/gimpcontainergridview.c: changed accordingly.
* app/widgets/gimpcontainertreeview.c
* app/widgets/gimpitemtreeview.c
* app/widgets/gimplayertreeview.c: dito, but actually use the passed
parent_insert_data to insert the item at the right place in the
GtkTreeView.
- add member "GimpViewable *parent" and accessors to get/set it
(no property or signals yet)
- add virtual function ::get_children() which is supposed to return
a GimpContainer of the viewable's children
Factor out large portions of identical code into new utility functions
gimp_container_view_connect_context() and
gimp_container_view_disconnect_context().
Keep around the handler IDs for the "name-changed" signals of the
container's children in a hash table that maps containers to handler
IDs and move adding/removing of the handler to
add_container()/remove_container(). This way the name-changed code is
prepared for handling multiple containers.
In preparation of having a tree of containers, added
gimp_container_view_add_container(),
gimp_container_view_remove_container() and
gimp_container_view_remove_foreach()
which do all the job of inserting/removing items and
connecting/disconnecting the "add", "remove" and "reorder" signals.
Also refactored things so when the toplevel container freezes/thaws,
it simply gets removed from the view instead of ignoring its signals.
gimp_container_view_real_set_container()
gimp_container_view_freeze()
gimp_container_view_thaw(): use the new add_container() and
remove_container() APIs and fix the code for the unlikely case
that a frozen container gets added/removed.
GHashTable has g_hash_table_remove_all() since GLib 2.12, so there is
no need any longer to clear the hash table by destroyung it. Instead,
keep the hash around during the view's entire lifetime and remove all
re-creation code and all checks for its existence.
Don't waste the expander space in all GimpContainerTreeViews. We can
later set it to TRUE in individual subclasses which actually display a
tree and not only a list.
Use gtk_tree_view_get_cell_area() instead of
gtk_tree_view_get_background_area() because the latter includes things
like expanders, indentation and padding and messes up the x coordinate
of our naive click detection.
Rename gimptoolview.[ch] to gimptooleditor.[ch]. The contents of the
file has already gone through this change, we do the file name change
separately for better diffs. Part of fix for bug #500930.
Convert the GimpToolView dockable to a non-dockable GimpToolEditor,
but wait with renaming the file so that we get better diffs. Part of
fix for bug #500930.
Remove the transient-docks setting for gimprc. What GIMP tried to
accomplish with this enabled is much better accomplished by the window
manager with the docks set to the 'Utility window' window hint. See
discussion in bug #322577.
Get rid of the last calls to gdk_window_set_title() and instead use
gdk_window_set_title(). The gdk_window_set_title() calls in
gimpdisplayshell-title.c is a legacy from the initial import of the
project.
Get rid of artificial compiler warnings generated with the #warning
directive. They pollute the build output and don't work as incentives
for fixing stuff.
There is GtkSettings:gtk-enable-mnemonics: now, so there is no
reason to do the same in GIMP:
* app/config/gimpguiconfig.[ch]: turn "menu-mnemonics" into a dummy.
* app/dialogs/preferences-dialog.c: remove its GUI.
* app/widgets/gimpactionfactory.[ch]
* app/widgets/gimpactiongroup.[ch]: remove infrastructure for disabling
menu mnemonics.
* app/actions/actions.c: bye bye glue code.
* app/widgets/gimplayertreeview.c (gimp_layer_tree_view_layer_clicked):
when ALT is pressed, select the layer's alpha. SHIFT and CONTROL work
as usual to add, subtract and intersect.
* app/widgets/gimpchanneltreeview.c
* app/widgets/gimpvectorstreeview.c: add "clicked" handlers here too
and do the same select-on-alt-click thing.
Add a 'Use default comment' button to the Comment tab in Image
Properties that if clicked sets the image comment to the default
comment set in Preferences.
(gimp_image_map_tool_control): call gimp_image_map_tool_response() directly
instead of gtk_dialog_response().
(gimp_image_map_tool_key_press): must not call response() if there is
no dialog (if the tool wasn't used yet).
- artificially limit the popup's height again so scrolling gets some testing.
- make sure the scroll buttons' sensitivity is always correct.
- remove obsolete utility function and other cleanups.
- reorder instance struct and add some spacing
- rename member "timeout_id" to "scroll_timeout_id"
- clean up constructor() even more (still the wrong place to
do all these things)
Closing whatever popup by click on dead space within it is a no no,
otherwise one-pixel mis-clicks inside the widget make it go away,
which is totally unexpected.
Always initialize the 'status' member of a newly allocated cairo_path_t
to the value CAIRO_STATUS_SUCCESS as required to successively call
cairo_append_path().
When an import source has been overwritten with File -> Overwrite
<file.png>, or exported by whatever means, reflect this status in the
title bar of the image window. Do this by adding some string building
logic to gimp_display_shell_format_filename().
Since save and export are separate activities we need to keep track of
image dirtiness for both of them. For this purpose, add a
'export_dirty' member to GimpImage which has the same semantics as
'dirty', but for export. Set it to clean whenever a document is
exported by whatever means. Do this with a new function
gimp_image_export_clean_all(). Also add gimp_image_is_export_dirty().
Forget the import source after save. We interpret a save as that the
user is not interested in being able to quickly export back to the
original any longer. The effect of this is that if we have an
'Overwrite' menu item, it becomes insensitive and set to 'Export to'.
With no export target but with an import source, show
"Overwrite <filename>" in the File menu. This also means we can't set
GIMP_FILE_EXPORT_TO_URI_KEY along with import source any longer.
Not all actions have procedures associated with them, for example
unused "plug-in-recent-[N]" actions, so check for NULL before we
invoke the plug-in action
the image and the up or down arrow is pressed.
Applied patch from Massimo as found in bug #567393. This changes ROUND()
to RINT() to correct rounding for negative numbers. Also did this change
in gimp_rectangle_tool_adjust_coord().
- keep around more local "buffer" variables to improve readability
- fix all comparisons between coordinates and bounding boxes,
they were off-by-one on one or both sides.
(gimp_text_tool_move_cursor): use PangoLayoutLines and byte indices to
calculate x positions when moving the cursor up or down. text_tool->x_pos
is not in terms of device units within a PangoLayoutLine.
Reuse the GeglProcessor instead of creating a new one in each call to
gimp_projection_construct_gegl(). Requires at least 0ed57a05da of
GEGL. Does not seem to give any performance improvement at all
unfortunately :(
* app/widgets/gimptextproxy.c: also swallow the "toggle-overwrite"
signal.
* app/tools/gimpdrawtool. [ch] (gimp_draw_tool_draw_text_cursor): add
"gboolean overwrite" which causes the cursor to be drawn as block.
* app/tools/gimptexttool.[ch]: implement overwriting, toggling it,
and changing the text cursor accordingly.
- add some local "buffer" variables to avoid writing text_tool->text_buffer
all the time.
- text_tool->text_buffer always exists, no need to check for it.
Add GimpTextProxy, a GtkTextView subclass which does nothing but
overriding the binding methods with empty implementations, so the text
view has no chance of letting e.g. the display beep or doing anything
else.
Change public text tool API to gimp_text_tool_delete_selection() and
move delete and backspace code to their own handlers in preparation of
handling all text deletion types.
* app/widgets/gimpwindow.c: treat GimpCanvas as a text widget and
dispatch all key events to it before invoking menu shortcuts.
* app/display/gimpdisplayshell-callbacks.c: treat all events on the
empty display as unhandled, not handled.
* app/tools/gimptexttool.c: use the right API for invoking the proxy
text view's bindings. Handle some more cursor navigation request and
swallow text deletion requests we don't handle instead of always
doing what the delete key does.
(gimp_text_tool_move_cursor): for GTK_MOVEMENT_DISPLAY_LINE_ENDS,
check if we are an the end of a line, or the cursor will end up on the
end of the next line.
* app/core/gimpmarshal.list: add marshallers needed for the binding
singnals.
* app/tools/gimptexttool.[ch]: add binding signals "move-cursor",
"delete-from-cursor" and "backspace" and hijack GtkTextView's
binding set to invoke them. Move code from the key_press() handler
to the signals' default handlers. This is how it should work. In
fact that code is #ifdef'ed away and we need an evil proxy
GtkTextView to invoke the bindings on because of reasons stated in
comments in the code. Ugly but works just fine.
(gimp_projection_validate_tile): fix stupid braino that made the
recently added optimization loop infinitely on images with a width
that is a multiple of TILE_WIDTH.
* app/tools/tools-enums.[ch]: add enum GimpClipboardAction which can be
{ CUT, COPY, PASTE }
* app/tools/gimptool.[ch]
* app/tools/tool_manager.[ch]: add GimpTool::clipboard_action() which
returns a boolean indicating whether the tool handled the action.
* app/tools/gimptexttool.c: implement clipboard_action().
* app/actions/edit-commands.c: try the active tool first in the cut,
copy and paste callbacks.
(gimp_text_tool_motion): always keep the "insert" mark at the mouse
location and the "selection_bound" one at the other end of the
selection.
(gimp_text_tool_key_press): make cursor navigation always move the
"insert" mark, and move "selection_bound" along with it unless shift
is pressed. It was exactly the other way around.
* app/tools/tools-enums.[ch]: add enum GimpButtonPressType which can
be { NORMAL, DOUBLE, TRIPLE }
* app/tools/gimptool.[ch]: add press_type paramater to GimpTool::button_press()
* app/tools/gimp*tool.c
* app/tools/tool_manager.[ch]: changed accordingly.
* app/tools/gimptoolcontrol.[ch]: add members and API so tools can choose
to receive double and triple clicks.
* app/display/gimpdisplayshell-callbacks.c (gimp_display_shell_tool_events):
dispatch double and triple clicks to tools if they want them, and if they
became active by the preceding normal button press.
* app/text/gimptextlayout-render.c: remove private function which
creates the transform matrix.
* app/text/gimptextlayout.[ch]: add it here as public API. Also add
function which transform and untransform PangoRectangles, points and
distances using the transform matrix.
* app/tools/gimptexttool.c: convert coordinates using above new
transform functions when drawing selection and cursor, and when
processing mouse events.
(gimp_drawable_transform_affine)
(gimp_drawable_transform_flip)
(gimp_drawable_transform_rotate): also transform the layer mask if we
are transforming a layer and there is no selection (which causes the
entire layer to be transformed).
(gimp_paint_core_get_orig_image)
(gimp_paint_core_get_orig_proj): changed parameters to x,y,width,height.
Update callers accordingly and use x,y,width,height there too except
in gimpperspectiveclone.c which does better with a bounding box.
Use 0666 as permissions instead of 0644 and let the user's umask care
about restricting, so creating a file with open() behaves the same way
as with fopen().
Add gimp_display_shell_sync_config() which copies the prefs settings
to the display shell and call it from gimp_display_shell_new() and
gimp_display_shell_fill().
Check that the cache itself is idle. Previously it would start during
transforms and long pyramid rendering ops and toss writes and large
seeks into the tile cache while it was potentially under heavy pressure.
Also increase its flush rate more to be more in line with modern system
and editing needs.
Although the clean-first strategy gives fast light-load performance,
it also degrades catastrophically under moderate cache pressure. LRU is
not as efficient under light load, but degrades more gracefully under
moderate and heavy load.
Add more profiling In order to analyze efficiency and behavior of the
tile cache. Profiling includes run-time indication of idle swapper
activity.
Also, fixes to original tile profiling code that appears to have
bitrotted some; it was no longer tracking changes made to tile memory
outside of tile.c.
Minor change to TILE_DATA_POINTER that restricts TILE_WIDTH and
TILE_HEIGHT to powers of two, but eliminates two integer divisions
(or, in reality, eliminates the over-complicated assembly resulting
from optimizing out two integer divisions in a C compliant fashion).
The questionable while loops to get GimpCoords.direction into a proper
range were used in two places. This commit fixes the actual problematic
place where, under some rare circumstances, the code used to go into a
loop for a very long time.
Pass the current coords to the function instead of setting them
on the paint core before calling it. Doesn't exactly make the code
in the paint tool simpler, but that needs further refactoring anyway.
Check for SHIFT, CONTROL and MOD1 explicitely so the code doesn't prevent
item activation for esoteric modifiers that are set by whatever X
component (like XKB).
Only activate container tree view items when no modifier keys are
pressed so that for example the layer properties dialog is not shown
when quickly toggling a layer mask with Ctrl + Click.
The version checks for babl and GEGL did not cope correctly with the
minor version upgrade. This change fixes this by introducing a generic
version check.
Also use 2-5 for the zoom-in actions and Shift+2-5 for the zoom-out
ones (the latter don't work, which is probably a regression in
GtkAccelMap, will investigate this...)
* app/actions/actions.h: added prototype for action_message()
* app/actions/actions.c: removed static keyword from action_message()
removed private prototype for action_message()
made action_message() use icons for objects derived from GimpViewable
* app/actions/context-commands.c:
context_brush_radius_cmd_callback()
context_brush_angle_cmd_callback()
context_paint_mode_cmd_callback()
context_brush_shape_cmd_callback()
call acton_message(), these callbacks had use cases where they could not
use generic code
* app/tools/gimpcoloroptions.c: added "average-radius" blurb
and mark for translarion