gimp/libgimpbase/gimpunit.c

1081 lines
28 KiB
C
Raw Normal View History

Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
/* LIBGIMP - The GIMP Library
* Copyright (C) 1995-1997 Peter Mattis and Spencer Kimball
*
* gimpunit.c
* Copyright (C) 2003 Michael Natterer <mitch@gimp.org>
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
*
* This library is free software: you can redistribute it and/or
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 3 of the License, or (at your option) any later version.
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
*
* This library is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* Lesser General Public License for more details.
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
*
* You should have received a copy of the GNU Lesser General Public
* License along with this library. If not, see
* <https://www.gnu.org/licenses/>.
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
*/
#include "config.h"
#include <math.h>
#include <string.h>
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
#include <gegl.h>
#include <glib-object.h>
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
#include "gimpbasetypes.h"
Makefile.am configure.in added new directory libgimpbase/ 2001-05-21 Michael Natterer <mitch@gimp.org> * Makefile.am * configure.in * gimptool-1.4.in: added new directory libgimpbase/ * app/Makefile.am: link against the new lib. * app/appenums.h: removed the PDB enums which are in libgimpbase/gimpbasetypes.h now. They are all "Gimp" prefixed. * app/apptypes.h: #include "libgimpbase/gimpbasetypes.h" * app/[lots] * app/core/[of] * app/gui/[files] * app/tools/: changed includes and all PDB types. * app/pdb/*: regenerated. * libgimp/Makefile.am: don't build libgimpi.a uglyness any more. * libgimp/gimpenv.[ch] * libgimp/gimplimits.[hh] * libgimp/gimpparasite.[ch] * libgimp/gimpparasiteio.[ch] * libgimp/gimpprotocol.[ch] * libgimp/gimpsignal.[ch] * libgimp/gimpunit.h * libgimp/gimputils.[ch] * libgimp/gimpwire.[ch]: removed... * libgimpbase/*: ...and added here as new library. * libgimp/gimp.[ch] * libgimp/gimpdrawable.[ch] * libgimp/gimpenums.h * libgimp/gimpimage.[ch] * libgimp/gimptile.c * libgimp/gimptypes.h * libgimp/gimpunit.c: changed accordingly. Added the gimp_*_add_new_parasite to gimp.[ch], gimpdrawable.[ch] and gimpimage.[ch]. * libgimpwidgets/gimppatheditor.c * libgimpwidgets/gimpquerybox.c * libgimpwidgets/gimpsizeentry.c * libgimpwidgets/gimpunitmenu.c * libgimpwidgets/gimpwidgets.c * libgimpwidgets/gimpwidgetstypes.h: changed includes accordingly. * plug-ins/*/Makefile.am * plug-ins/common/mkgen.pl: link against libgimpbase. * tools/pdbgen/Makefile.am: scan libgimpbase/gimpbasetypes.h, so the enums are known to pdbgen... * tools/pdbgen/enumcode.pl: ...but don't write them out to libgimp/gimpenums.h * tools/pdbgen/app.pl: include libgimp/gimpbase.h in all *_cmds.c files. Added GIMP_ to the type names ganerated in app/. * tools/pdbgen/enums.pl: regenerated. * tools/pdbgen/pdb.pl * tools/pdbgen/pdb/fileops.pdb * tools/pdbgen/pdb/procedural_db.pdb * tools/pdbgen/pdb/unit.pdb: changed includes.
2001-05-21 21:58:46 +08:00
#include "gimpbase-private.h"
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
#include "gimpparamspecs.h"
#include "gimpunit.h"
Makefile.am configure.in added new directory libgimpbase/ 2001-05-21 Michael Natterer <mitch@gimp.org> * Makefile.am * configure.in * gimptool-1.4.in: added new directory libgimpbase/ * app/Makefile.am: link against the new lib. * app/appenums.h: removed the PDB enums which are in libgimpbase/gimpbasetypes.h now. They are all "Gimp" prefixed. * app/apptypes.h: #include "libgimpbase/gimpbasetypes.h" * app/[lots] * app/core/[of] * app/gui/[files] * app/tools/: changed includes and all PDB types. * app/pdb/*: regenerated. * libgimp/Makefile.am: don't build libgimpi.a uglyness any more. * libgimp/gimpenv.[ch] * libgimp/gimplimits.[hh] * libgimp/gimpparasite.[ch] * libgimp/gimpparasiteio.[ch] * libgimp/gimpprotocol.[ch] * libgimp/gimpsignal.[ch] * libgimp/gimpunit.h * libgimp/gimputils.[ch] * libgimp/gimpwire.[ch]: removed... * libgimpbase/*: ...and added here as new library. * libgimp/gimp.[ch] * libgimp/gimpdrawable.[ch] * libgimp/gimpenums.h * libgimp/gimpimage.[ch] * libgimp/gimptile.c * libgimp/gimptypes.h * libgimp/gimpunit.c: changed accordingly. Added the gimp_*_add_new_parasite to gimp.[ch], gimpdrawable.[ch] and gimpimage.[ch]. * libgimpwidgets/gimppatheditor.c * libgimpwidgets/gimpquerybox.c * libgimpwidgets/gimpsizeentry.c * libgimpwidgets/gimpunitmenu.c * libgimpwidgets/gimpwidgets.c * libgimpwidgets/gimpwidgetstypes.h: changed includes accordingly. * plug-ins/*/Makefile.am * plug-ins/common/mkgen.pl: link against libgimpbase. * tools/pdbgen/Makefile.am: scan libgimpbase/gimpbasetypes.h, so the enums are known to pdbgen... * tools/pdbgen/enumcode.pl: ...but don't write them out to libgimp/gimpenums.h * tools/pdbgen/app.pl: include libgimp/gimpbase.h in all *_cmds.c files. Added GIMP_ to the type names ganerated in app/. * tools/pdbgen/enums.pl: regenerated. * tools/pdbgen/pdb.pl * tools/pdbgen/pdb/fileops.pdb * tools/pdbgen/pdb/procedural_db.pdb * tools/pdbgen/pdb/unit.pdb: changed includes.
2001-05-21 21:58:46 +08:00
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
#include "libgimp/libgimp-intl.h"
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
enum
{
PROP_0,
PROP_ID,
PROP_NAME,
PROP_FACTOR,
PROP_DIGITS,
PROP_SYMBOL,
PROP_ABBREVIATION,
};
struct _GimpUnit
{
GObject parent_instance;
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
gint id;
gchar *name;
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
gboolean delete_on_exit;
gdouble factor;
gint digits;
gchar *symbol;
gchar *abbreviation;
};
typedef struct
{
gdouble factor;
gint digits;
gchar *identifier;
gchar *symbol;
gchar *abbreviation;
} GimpUnitDef;
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
/* these are the built-in units
*/
static const GimpUnitDef _gimp_unit_defs[GIMP_UNIT_END] =
{
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
/* pseudo unit */
{
0.0, 0,
NC_("unit-plural", "pixels"),
"px", "px",
},
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
/* standard units */
{
1.0, 2,
NC_("unit-plural", "inches"),
"''", "in",
},
{
25.4, 1,
NC_("unit-plural", "millimeters"),
"mm", "mm",
},
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
/* professional units */
{
72.0, 0,
NC_("unit-plural", "points"),
"pt", "pt",
},
{
6.0, 1,
NC_("unit-plural", "picas"),
"pc", "pc",
}
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
};
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
/* not a unit at all but kept here to have the strings in one place
*/
static const GimpUnitDef _gimp_unit_percent_def =
{
0.0, 0,
NC_("unit-plural", "percent"),
"%", "%",
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
};
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
static void gimp_unit_constructed (GObject *object);
static void gimp_unit_finalize (GObject *object);
static void gimp_unit_set_property (GObject *object,
guint property_id,
const GValue *value,
GParamSpec *pspec);
static void gimp_unit_get_property (GObject *object,
guint property_id,
GValue *value,
GParamSpec *pspec);
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
static gint print (gchar *buf,
gint len,
gint start,
const gchar *fmt,
...) G_GNUC_PRINTF (4, 5);
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
G_DEFINE_TYPE (GimpUnit, gimp_unit, G_TYPE_OBJECT)
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
#define parent_class gimp_unit_parent_class
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
static void
gimp_unit_class_init (GimpUnitClass *klass)
{
GObjectClass *object_class = G_OBJECT_CLASS (klass);
parent_class = g_type_class_peek_parent (klass);
object_class->constructed = gimp_unit_constructed;
object_class->finalize = gimp_unit_finalize;
object_class->set_property = gimp_unit_set_property;
object_class->get_property = gimp_unit_get_property;
g_object_class_install_property (object_class, PROP_ID,
g_param_spec_int ("id", NULL, NULL,
0, G_MAXINT, 0,
GIMP_PARAM_READWRITE |
G_PARAM_CONSTRUCT_ONLY));
g_object_class_install_property (object_class, PROP_NAME,
g_param_spec_string ("name", NULL, NULL,
NULL,
GIMP_PARAM_READWRITE |
G_PARAM_CONSTRUCT_ONLY));
g_object_class_install_property (object_class, PROP_FACTOR,
g_param_spec_double ("factor", NULL, NULL,
0.0, G_MAXDOUBLE, 1.0,
GIMP_PARAM_READWRITE |
G_PARAM_CONSTRUCT_ONLY));
g_object_class_install_property (object_class, PROP_DIGITS,
g_param_spec_int ("digits", NULL, NULL,
0, G_MAXINT, 0,
GIMP_PARAM_READWRITE |
G_PARAM_CONSTRUCT_ONLY));
g_object_class_install_property (object_class, PROP_SYMBOL,
g_param_spec_string ("symbol", NULL, NULL,
NULL,
GIMP_PARAM_READWRITE |
G_PARAM_CONSTRUCT_ONLY));
g_object_class_install_property (object_class, PROP_ABBREVIATION,
g_param_spec_string ("abbreviation", NULL, NULL,
NULL,
GIMP_PARAM_READWRITE |
G_PARAM_CONSTRUCT_ONLY));
/*klass->id_table = gimp_id_table_new ();*/
}
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
static void
gimp_unit_init (GimpUnit *unit)
{
unit->name = NULL;
unit->symbol = NULL;
unit->abbreviation = NULL;
}
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
static void
gimp_unit_constructed (GObject *object)
{
G_OBJECT_CLASS (parent_class)->constructed (object);
}
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
static void
gimp_unit_finalize (GObject *object)
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
{
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
GimpUnit *unit = GIMP_UNIT (object);
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
g_free (unit->name);
g_free (unit->symbol);
g_free (unit->abbreviation);
G_OBJECT_CLASS (parent_class)->finalize (object);
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
}
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
static void
gimp_unit_set_property (GObject *object,
guint property_id,
const GValue *value,
GParamSpec *pspec)
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
{
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
GimpUnit *unit = GIMP_UNIT (object);
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
switch (property_id)
{
case PROP_ID:
unit->id = g_value_get_int (value);
break;
case PROP_NAME:
unit->name = g_value_dup_string (value);
break;
case PROP_FACTOR:
unit->factor = g_value_get_double (value);
break;
case PROP_DIGITS:
unit->digits = g_value_get_int (value);
break;
case PROP_SYMBOL:
unit->symbol = g_value_dup_string (value);
break;
case PROP_ABBREVIATION:
unit->abbreviation = g_value_dup_string (value);
break;
default:
G_OBJECT_WARN_INVALID_PROPERTY_ID (object, property_id, pspec);
break;
}
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
}
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
static void
gimp_unit_get_property (GObject *object,
guint property_id,
GValue *value,
GParamSpec *pspec)
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
{
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
GimpUnit *unit = GIMP_UNIT (object);
switch (property_id)
{
case PROP_ID:
g_value_set_int (value, unit->id);
break;
case PROP_NAME:
g_value_set_string (value, unit->name);
break;
case PROP_FACTOR:
g_value_set_double (value, unit->factor);
break;
case PROP_DIGITS:
g_value_set_int (value, unit->digits);
break;
case PROP_SYMBOL:
g_value_set_string (value, unit->symbol);
break;
case PROP_ABBREVIATION:
g_value_set_string (value, unit->abbreviation);
break;
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
default:
G_OBJECT_WARN_INVALID_PROPERTY_ID (object, property_id, pspec);
break;
}
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
}
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
/* public functions */
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
/**
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
* gimp_unit_get_id:
* @unit: The unit you want to know the integer ID of.
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
*
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
* The ID can be used to retrieve the unit with [func@Unit.get_by_id].
*
* Note that this ID will be stable within a single session of GIMP, but
* you should not expect this ID to stay the same across multiple runs.
*
* Returns: The unit's ID.
**/
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
gint
gimp_unit_get_id (GimpUnit *unit)
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
{
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
g_return_val_if_fail (GIMP_IS_UNIT (unit), -1);
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
return unit->id;
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
}
/**
* gimp_unit_get_name:
* @unit: The unit you want to know the name of.
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
*
* This function returns the usual name of the unit (e.g. "inches").
* It can be used as the long label for the unit in the interface.
* For short labels, use [method@Unit.get_abbreviation].
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
*
* NOTE: This string must not be changed or freed.
*
* Returns: The unit's name.
**/
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
const gchar *
gimp_unit_get_name (GimpUnit *unit)
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
{
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
g_return_val_if_fail (GIMP_IS_UNIT (unit), NULL);
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
return unit->name;
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
}
/**
* gimp_unit_get_factor:
* @unit: The unit you want to know the factor of.
*
* A #GimpUnit's @factor is defined to be:
*
* distance_in_units == (@factor * distance_in_inches)
*
* Returns 0 for @unit == GIMP_UNIT_PIXEL.
*
* Returns: The unit's factor.
**/
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
gdouble
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
gimp_unit_get_factor (GimpUnit *unit)
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
{
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
g_return_val_if_fail (GIMP_IS_UNIT (unit), 1.0);
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
return unit->factor;
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
}
/**
* gimp_unit_get_digits:
* @unit: The unit you want to know the digits.
*
* Returns the number of digits set for @unit.
* Built-in units' accuracy is approximately the same as an inch with
* two digits. User-defined units can suggest a different accuracy.
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
*
* Note: the value is as-set by defaults or by the user and does not
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
* necessary provide enough precision on high-resolution units.
* When the information is needed for a specific unit, the use of
* gimp_unit_get_scaled_digits() may be more appropriate.
*
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
* Returns 0 for @unit == GIMP_UNIT_PIXEL.
*
* Returns: The suggested number of digits.
**/
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
gint
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
gimp_unit_get_digits (GimpUnit *unit)
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
{
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
g_return_val_if_fail (GIMP_IS_UNIT (unit), 0);
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
return unit->digits;
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
}
/**
* gimp_unit_get_scaled_digits:
* @unit: The unit you want to know the digits.
* @resolution: the resolution in PPI.
*
* Returns the number of digits a @unit field should provide to get
* enough accuracy so that every pixel position shows a different
* value from neighboring pixels.
*
* Note: when needing digit accuracy to display a diagonal distance,
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
* the @resolution may not correspond to the unit's horizontal or
* vertical resolution, but instead to the result of:
* `distance_in_pixel / distance_in_inch`.
*
* Returns: The suggested number of digits.
**/
gint
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
gimp_unit_get_scaled_digits (GimpUnit *unit,
gdouble resolution)
{
gint digits;
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
g_return_val_if_fail (GIMP_IS_UNIT (unit), 0);
digits = ceil (log10 (1.0 /
gimp_pixels_to_units (1.0, unit, resolution)));
return MAX (digits, gimp_unit_get_digits (unit));
}
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
/**
* gimp_unit_get_symbol:
* @unit: The unit you want to know the symbol of.
*
* This is e.g. "''" for UNIT_INCH.
*
* NOTE: This string must not be changed or freed.
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
*
* Returns: The unit's symbol.
**/
const gchar *
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
gimp_unit_get_symbol (GimpUnit *unit)
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
{
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
g_return_val_if_fail (GIMP_IS_UNIT (unit), NULL);
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
return unit->symbol;
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
}
/**
* gimp_unit_get_abbreviation:
* @unit: The unit you want to know the abbreviation of.
*
* This function returns the abbreviation of the unit (e.g. "in" for
* inches).
* It can be used as a short label for the unit in the interface.
* For long labels, use [method@Unit.get_name].
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
*
* NOTE: This string must not be changed or freed.
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
*
* Returns: The unit's abbreviation.
**/
const gchar *
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
gimp_unit_get_abbreviation (GimpUnit *unit)
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
{
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
g_return_val_if_fail (GIMP_IS_UNIT (unit), NULL);
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
return unit->abbreviation;
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
}
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
/**
* gimp_unit_get_deletion_flag:
* @unit: The unit you want to know the @deletion_flag of.
*
* Returns: The unit's @deletion_flag.
**/
gboolean
gimp_unit_get_deletion_flag (GimpUnit *unit)
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
{
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
g_return_val_if_fail (GIMP_IS_UNIT (unit), TRUE);
if ((unit->id >= 0 && unit->id < GIMP_UNIT_END) ||
unit->id == GIMP_UNIT_PERCENT)
return FALSE;
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
if (_gimp_unit_vtable.get_deletion_flag != NULL)
/* This code path will only happen in libgimp. */
return _gimp_unit_vtable.get_deletion_flag (unit);
else
return unit->delete_on_exit;
Sven Neumann <sven@gimp.org> 2000-05-31 Michael Natterer <mitch@gimp.org> Sven Neumann <sven@gimp.org> Episode IV... * app/Makefile.am * app/message_cmds.c: new name * app/interface_cmds.c: removed * app/help_cmds.c: new name * app/gimphelp_cmds.c: removed * app/internal_procs.c * app/gimphelp_cmds.c * app/plug_in_cmds.c: moved gimp_plugin_help_register to plug_in_cmds.c, s/gimp_plugin_domain_add/gimp_plugin_domain_register/ * libgimp/Makefile.am * libgimp/gimp_pdb.h * libgimp/gimpfileops_pdb.[ch] * libgimp/gimpguides_pdb.[ch] * libgimp/gimpmessage_pdb.[ch] * libgimp/gimpprocedural_db_pdb.[ch] * libgimp/gimpplugin_pdb.[ch]: new files containing PDB wrappers moved out of libgimp/gimp.[ch] * libgimp/gimpchannel_pdb.c * libgimp/gimpdisplay_pdb.[ch] * libgimp/gimpdrawable_pdb.[ch] * libgimp/gimphelp_pdb.[ch] * libgimp/gimpimage_pdb.[ch] * libgimp/gimplayer_pdb.c * libgimp/gimpparasite_pdb.c * libgimp/gimpselection_pdb.c: various changes. * libgimp/gimpunit.c: new file * libgimp/gimpunit_pdb.[ch]: contains only PDB wrappers prefixed with "_" which are called from libgimp/gimpunit.c only. * libgimp/gimpunit.h: minor stuff * libgimp/gimpcompat.h: declared the following functions obsolete: gimp_query_database gimp_query_images * libgimp/gimp.[ch]: removed all PDB wrappers. * tools/pdbgen/Makefile.am * tools/pdbgen/groups.pl: changed * tools/pdbgen/lib.pl: tried to fix libgimp autogeneration and fixed just a few comments. * tools/pdbgen/pdb/interface.pdb * tools/pdbgen/pdb/gimphelp.pdb: removed * tools/pdbgen/pdb/help.pdb * tools/pdbgen/pdb/message.pdb: new names * plug-ins/script-fu/script-fu.c * plug-ins/perl/Gimp/Lib.xs: s/gimp_plugin_domain_add/gimp_plugin_domain_register/ 2000-05-31 Michael Natterer <mitch@gimp.org> * POTFILES.in: s/gimpunit_pdb/gimpunit/
2000-05-31 21:24:14 +08:00
}
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
/**
* gimp_unit_set_deletion_flag:
* @unit: The unit you want to set the @deletion_flag for.
* @deletion_flag: The new deletion_flag.
*
* Sets a #GimpUnit's @deletion_flag. If the @deletion_flag of a unit is
* %TRUE when GIMP exits, this unit will not be saved in the users's
* "unitrc" file.
*
* Trying to change the @deletion_flag of a built-in unit will be silently
* ignored.
**/
void
gimp_unit_set_deletion_flag (GimpUnit *unit,
gboolean deletion_flag)
{
g_return_if_fail (GIMP_IS_UNIT (unit));
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
if ((unit->id >= 0 && unit->id < GIMP_UNIT_END) ||
unit->id == GIMP_UNIT_PERCENT)
return;
unit->delete_on_exit = deletion_flag;
if (_gimp_unit_vtable.set_deletion_flag != NULL)
/* This code path will only happen in libgimp. */
_gimp_unit_vtable.set_deletion_flag (unit, deletion_flag);
}
/**
* gimp_unit_get_by_id:
* @unit_id: The unit id.
*
* Returns the unique [class@Unit] object corresponding to @unit_id,
* which is the integer identifier as returned by [method@Unit.get_id].
*
* Returns: (transfer none): the #GimpUnit object with ID @unit_id.
*
* Since: 3.0
**/
GimpUnit *
gimp_unit_get_by_id (gint unit_id)
{
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
GimpUnit *unit = NULL;
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
if (unit_id < 0)
return NULL;
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
if (G_UNLIKELY (! _gimp_units))
_gimp_units = g_hash_table_new_full (g_direct_hash,
g_direct_equal,
NULL,
(GDestroyNotify) g_object_unref);
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
unit = g_hash_table_lookup (_gimp_units, GINT_TO_POINTER (unit_id));
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
if (! unit)
{
if (unit_id < GIMP_UNIT_END)
{
GimpUnitDef def = _gimp_unit_defs[unit_id];
unit = g_object_new (GIMP_TYPE_UNIT,
"id", unit_id,
"name", def.identifier,
"factor", def.factor,
"digits", def.digits,
"symbol", def.symbol,
"abbreviation", def.abbreviation,
NULL);
unit->delete_on_exit = FALSE;
}
else if (unit_id == GIMP_UNIT_PERCENT)
{
unit = g_object_new (GIMP_TYPE_UNIT,
"id", unit_id,
"name", _gimp_unit_percent_def.identifier,
"factor", _gimp_unit_percent_def.factor,
"digits", _gimp_unit_percent_def.digits,
"symbol", _gimp_unit_percent_def.symbol,
"abbreviation", _gimp_unit_percent_def.abbreviation,
NULL);
unit->delete_on_exit = FALSE;
}
else if (_gimp_unit_vtable.get_data != NULL)
{
/* This code path should never happen in app/ where get_data()
* is NULL, because non built-in units are created in app/
* whereas they are only queried in libgimp.
*/
gchar *identifier = NULL;
gdouble factor;
gint digits;
gchar *symbol = NULL;
gchar *abbreviation = NULL;
identifier = _gimp_unit_vtable.get_data (unit_id,
&factor,
&digits,
&symbol,
&abbreviation);
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
if (identifier != NULL)
unit = g_object_new (GIMP_TYPE_UNIT,
"id", unit_id,
"name", identifier,
"factor", factor,
"digits", digits,
"symbol", symbol,
"abbreviation", abbreviation,
NULL);
g_free (identifier);
g_free (symbol);
g_free (abbreviation);
}
else if (_gimp_unit_vtable.get_user_unit != NULL)
{
/* This code path should never happen in libgimp, only in app/. */
unit = _gimp_unit_vtable.get_user_unit (unit_id);
if (unit != NULL)
g_object_ref (unit);
}
if (unit != NULL)
g_hash_table_insert (_gimp_units, GINT_TO_POINTER (unit_id), unit);
}
return unit;
}
/**
* gimp_unit_pixel:
*
* Returns the unique object representing pixel unit.
*
* This procedure returns the unit representing pixel. The returned
* object is unique across the whole run.
*
* Returns: (transfer none): The unique pixel unit.
*
* Since: 3.0
**/
GimpUnit *
gimp_unit_pixel (void)
{
return gimp_unit_get_by_id (GIMP_UNIT_PIXEL);
}
/**
* gimp_unit_inch:
*
* Returns the unique object representing inch unit.
*
* This procedure returns the unit representing inch. The returned
* object is unique across the whole run.
*
* Returns: (transfer none): The unique inch unit.
*
* Since: 3.0
**/
GimpUnit *
gimp_unit_inch (void)
{
return gimp_unit_get_by_id (GIMP_UNIT_INCH);
}
/**
* gimp_unit_mm:
*
* Returns the unique object representing millimeter unit.
*
* This procedure returns the unit representing millimeter. The
* returned object is unique across the whole run.
*
* Returns: (transfer none): The unique millimeter unit.
*
* Since: 3.0
**/
GimpUnit *
gimp_unit_mm (void)
{
return gimp_unit_get_by_id (GIMP_UNIT_MM);
}
/**
* gimp_unit_point:
*
* Returns the unique object representing typographical point unit.
*
* This procedure returns the unit representing typographical points.
* The returned object is unique across the whole run.
*
* Returns: (transfer none): The unique point unit.
*
* Since: 3.0
**/
GimpUnit *
gimp_unit_point (void)
{
return gimp_unit_get_by_id (GIMP_UNIT_POINT);
}
/**
* gimp_unit_pica:
*
* Returns the unique object representing Pica unit.
*
* This procedure returns the unit representing Picas.
* The returned object is unique across the whole run.
*
* Returns: (transfer none): The unique pica unit.
*
* Since: 3.0
**/
GimpUnit *
gimp_unit_pica (void)
{
return gimp_unit_get_by_id (GIMP_UNIT_PICA);
}
/**
* gimp_unit_percent:
*
* Returns the unique object representing percent dimensions relatively
* to an image.
*
* This procedure returns the unit representing typographical points.
* The returned object is unique across the whole run.
*
* Returns: (transfer none): The unique percent unit.
*
* Since: 3.0
**/
GimpUnit *
gimp_unit_percent (void)
{
return gimp_unit_get_by_id (GIMP_UNIT_PERCENT);
}
/**
* gimp_unit_is_built_in:
* @unit: the unit.
*
* Returns whether the unit is built-in.
*
* This procedure returns @unit is a built-in unit. In particular the
* deletion flag cannot be set on built-in units.
*
* Returns: Whether @unit is built-in.
*
* Since: 3.0
**/
gboolean
gimp_unit_is_built_in (GimpUnit *unit)
{
g_return_val_if_fail (GIMP_IS_UNIT (unit), FALSE);
return (unit->id >= 0 && unit->id < GIMP_UNIT_END) || unit->id == GIMP_UNIT_PERCENT;
}
/**
* gimp_unit_is_metric:
* @unit: The unit
*
* Checks if the given @unit is metric. A simplistic test is used
* that looks at the unit's factor and checks if it is 2.54 multiplied
* by some common powers of 10. Currently it checks for mm, cm, dm, m.
*
* See also: gimp_unit_get_factor()
*
* Returns: %TRUE if the @unit is metric.
*
* Since: 2.10
**/
gboolean
gimp_unit_is_metric (GimpUnit *unit)
{
gdouble factor;
if (unit == gimp_unit_mm ())
return TRUE;
factor = gimp_unit_get_factor (unit);
if (factor == 0.0)
return FALSE;
return ((ABS (factor - 0.0254) < 1e-7) || /* m */
(ABS (factor - 0.254) < 1e-6) || /* dm */
(ABS (factor - 2.54) < 1e-5) || /* cm */
(ABS (factor - 25.4) < 1e-4)); /* mm */
}
/**
* gimp_unit_format_string:
* @format: A printf-like format string which is used to create the unit
* string.
* @unit: A unit.
*
* The @format string supports the following percent expansions:
*
* * `%n`: Name (long label)
* * `%a`: Abbreviation (short label)
* * `%%`: Literal percent
* * `%f`: Factor (how many units make up an inch)
* * `%y`: Symbol (e.g. `''` for `GIMP_UNIT_INCH`)
*
* Returns: (transfer full): A newly allocated string with above percent
* expressions replaced with the resp. strings for @unit.
*
* Since: 2.8
**/
gchar *
gimp_unit_format_string (const gchar *format,
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
GimpUnit *unit)
{
gchar buffer[1024];
gint i = 0;
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
g_return_val_if_fail (GIMP_IS_UNIT (unit), NULL);
g_return_val_if_fail (format != NULL, NULL);
while (i < (sizeof (buffer) - 1) && *format)
{
switch (*format)
{
case '%':
format++;
switch (*format)
{
case 0:
g_warning ("%s: unit-menu-format string ended within %%-sequence",
G_STRFUNC);
break;
case '%':
buffer[i++] = '%';
break;
case 'f': /* factor (how many units make up an inch) */
i += print (buffer, sizeof (buffer), i, "%f",
gimp_unit_get_factor (unit));
break;
case 'y': /* symbol ("''" for inch) */
i += print (buffer, sizeof (buffer), i, "%s",
gimp_unit_get_symbol (unit));
break;
case 'a': /* abbreviation */
i += print (buffer, sizeof (buffer), i, "%s",
gimp_unit_get_abbreviation (unit));
break;
case 'n': /* full name */
i += print (buffer, sizeof (buffer), i, "%s",
gimp_unit_get_name (unit));
break;
default:
g_warning ("%s: unit-menu-format contains unknown format "
"sequence '%%%c'", G_STRFUNC, *format);
break;
}
break;
default:
buffer[i++] = *format;
break;
}
format++;
}
buffer[MIN (i, sizeof (buffer) - 1)] = 0;
return g_strdup (buffer);
}
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
/**
* gimp_pixels_to_units:
* @pixels: value in pixels
* @unit: unit to convert to
* @resolution: resolution in DPI
*
* Converts a @value specified in pixels to @unit.
*
* Returns: @pixels converted to units.
*
* Since: 2.8
**/
gdouble
gimp_pixels_to_units (gdouble pixels,
GimpUnit *unit,
gdouble resolution)
{
g_return_val_if_fail (gimp_unit_pixel != NULL, 0.0);
if (unit == gimp_unit_pixel ())
return pixels;
return pixels * gimp_unit_get_factor (unit) / resolution;
}
/**
* gimp_units_to_pixels:
* @value: value in units
* @unit: unit of @value
* @resolution: resloution in DPI
*
* Converts a @value specified in @unit to pixels.
*
* Returns: @value converted to pixels.
*
* Since: 2.8
**/
gdouble
gimp_units_to_pixels (gdouble value,
GimpUnit *unit,
gdouble resolution)
{
g_return_val_if_fail (gimp_unit_pixel != NULL, 0.0);
if (unit == gimp_unit_pixel ())
return value;
return value * resolution / gimp_unit_get_factor (unit);
}
/**
* gimp_units_to_points:
* @value: value in units
* @unit: unit of @value
* @resolution: resloution in DPI
*
* Converts a @value specified in @unit to points.
*
* Returns: @value converted to points.
*
* Since: 2.8
**/
gdouble
gimp_units_to_points (gdouble value,
GimpUnit *unit,
gdouble resolution)
{
g_return_val_if_fail (gimp_unit_pixel != NULL, 0.0);
g_return_val_if_fail (gimp_unit_point != NULL, 0.0);
if (unit == gimp_unit_point ())
return value;
if (unit == gimp_unit_pixel ())
return (value * gimp_unit_get_factor (gimp_unit_point ()) / resolution);
return (value *
gimp_unit_get_factor (gimp_unit_point ()) / gimp_unit_get_factor (unit));
}
/*
* GIMP_TYPE_PARAM_UNIT
*/
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
static void gimp_param_unit_class_init (GParamSpecClass *klass);
static void gimp_param_unit_init (GParamSpec *pspec);
static void gimp_param_unit_set_default (GParamSpec *pspec,
GValue *value);
static gboolean gimp_param_unit_validate (GParamSpec *pspec,
GValue *value);
/**
* gimp_param_unit_get_type:
*
* Reveals the object type
*
* Returns: the #GType for a unit param object
*
* Since: 2.4
**/
GType
gimp_param_unit_get_type (void)
{
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
static GType type = 0;
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
if (! type)
{
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
const GTypeInfo info =
{
sizeof (GParamSpecClass),
NULL, NULL,
(GClassInitFunc) gimp_param_unit_class_init,
NULL, NULL,
sizeof (GimpParamSpecUnit),
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
0,
(GInstanceInitFunc) gimp_param_unit_init
};
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
type = g_type_register_static (G_TYPE_PARAM_OBJECT,
"GimpParamUnit", &info, 0);
}
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
return type;
}
static void
gimp_param_unit_class_init (GParamSpecClass *klass)
{
klass->value_type = GIMP_TYPE_UNIT;
klass->value_set_default = gimp_param_unit_set_default;
klass->value_validate = gimp_param_unit_validate;
}
static void
gimp_param_unit_init (GParamSpec *pspec)
{
GimpParamSpecUnit *uspec = GIMP_PARAM_SPEC_UNIT (pspec);
uspec->allow_pixel = TRUE;
uspec->allow_percent = TRUE;
uspec->default_value = gimp_unit_inch ();
}
static void
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
gimp_param_unit_set_default (GParamSpec *pspec,
GValue *value)
{
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
GimpParamSpecUnit *uspec = GIMP_PARAM_SPEC_UNIT (pspec);
g_value_set_object (value, uspec->default_value);
}
static gboolean
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
gimp_param_unit_validate (GParamSpec *pspec,
GValue *value)
{
GimpParamSpecUnit *uspec = GIMP_PARAM_SPEC_UNIT (pspec);
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
GObject *unit = value->data[0].v_pointer;
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
if (unit == NULL ||
(! uspec->allow_percent && value->data[0].v_pointer == gimp_unit_percent ()) ||
(! uspec->allow_pixel && value->data[0].v_pointer == gimp_unit_pixel ()))
{
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
g_clear_object (&unit);
value->data[0].v_pointer = g_object_ref (uspec->default_value);
return TRUE;
}
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
return FALSE;
}
/**
* gimp_param_spec_unit:
* @name: Canonical name of the param
* @nick: Nickname of the param
* @blurb: Brief description of param.
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
* @allow_pixel: Whether "pixels" is an allowed unit.
* @allow_percent: Whether "percent" is an allowed unit.
* @default_value: Unit to use if none is assigned.
* @flags: a combination of #GParamFlags
*
* Creates a param spec to hold a units param.
* See g_param_spec_internal() for more information.
*
* Returns: (transfer full): a newly allocated #GParamSpec instance
*
* Since: 2.4
**/
GParamSpec *
gimp_param_spec_unit (const gchar *name,
const gchar *nick,
const gchar *blurb,
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
gboolean allow_pixel,
gboolean allow_percent,
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
GimpUnit *default_value,
GParamFlags flags)
{
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
GimpParamSpecUnit *uspec;
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
g_return_val_if_fail (GIMP_IS_UNIT (default_value), NULL);
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
uspec = g_param_spec_internal (GIMP_TYPE_PARAM_UNIT,
name, nick, blurb, flags);
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
g_return_val_if_fail (uspec, NULL);
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
uspec->allow_pixel = allow_pixel;
uspec->allow_percent = allow_percent;
uspec->default_value = default_value;
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
return G_PARAM_SPEC (uspec);
}
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
static gint
print (gchar *buf,
gint len,
gint start,
const gchar *fmt,
...)
{
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
va_list args;
gint printed;
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
va_start (args, fmt);
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
printed = g_vsnprintf (buf + start, len - start, fmt, args);
if (printed < 0)
printed = len - start;
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
va_end (args);
Issue #8900 and #9923: reimplementing GimpUnit as a proper class. This fixes all our GObject Introspection issues with GimpUnit which was both an enum and an int-derived type of user-defined units *completing* the enum values. GIR clearly didn't like this! Now GimpUnit is a proper class and units are unique objects, allowing to compare them with an identity test (i.e. `unit == gimp_unit_pixel ()` tells us if unit is the pixel unit or not), which makes it easy to use, just like with int, yet adding also methods, making for nicer introspected API. As an aside, this also fixes #10738, by having all the built-in units retrievable even if libgimpbase had not been properly initialized with gimp_base_init(). I haven't checked in details how GIR works to introspect, but it looks like it loads the library to inspect and runs functions, hence triggering some CRITICALS because virtual methods (supposed to be initialized with gimp_base_init() run by libgimp) are not set. This new code won't trigger any critical because the vtable method are now not necessary, at least for all built-in units. Note that GimpUnit is still in libgimpbase. It could have been moved to libgimp in order to avoid any virtual method table (since we need to keep core and libgimp side's units in sync, PDB is required), but too many libgimpwidgets widgets were already using GimpUnit. And technically most of GimpUnit logic doesn't require PDB (only the creation/sync part). This is one of the reasons why user-created GimpUnit list is handled and stored differently from other types of objects. Globally this simplifies the code a lot too and we don't need separate implementations of various utils for core and libgimp, which means less prone to errors.
2024-07-26 02:55:21 +08:00
return printed;
}