The CLI options now know which procedures are batch procedures or not.
First it means that it won't just randomly try any procedure name one
may pass and will properly output an error if you pass a non-existing
interpreter procedure.
Secondly, there is no default interpreter anymore (unless only one
interpreter exists). If you don't set an interpreter procedure with
--batch-interpreter or if you pass a wrong one, it will output the list
of available batch procedure, thus helping you understanding how to use
the --batch option.
Fixes:
> [33/48] Generating Gimp-3.0.gir with a custom command
> ../../../../../../../dev/src/gimp/libgimp/gimpimageprocedure.c:228: Warning: Gimp: missing ":" at column 41:
> * @run_func: (closure run_data) the run function for the new procedure.
> ^
On Windows our Lua goat extension crashed when run. After some digging
it appeared that incorrect annotations were the cause.
The destroy annotation was clearly an error when loking at the gir API
documentation. However, it also crashed on `(closure run_func)`, which
seemed to be as it should.
After comparing what other libraries do and testing, it seems that the
documentation for (closure CLOSURE) is incorrect. See the discussion
in the above mentioned issue.
The correct use here seems to be to use it as annotation of the run_func
where CLOSURE points to the user data. So, that's what we implement here.
Fix the dependency by making the stamp an actual (yet empty/no-op)
header file which is included by all generated source file. This way, we
ensure that meson rebuild .o files when the .pdb sources are changed.
This is the second solution proposed by eli-schwartz here:
https://github.com/mesonbuild/meson/issues/10196#issuecomment-1080053413
The build now successfully build the PDB files into the source folder
itself. Unfortunately it seems I can't get meson dependencies to work
properly, once more! I added a "sources" argument to the relevant
library() or static_library() but it still uses old versions to build
these. E.g. if I add an error on purpose to a pdb file, the next build
still passes, yet the second-next fails (as it should have before).
Note that I even tested a declare_dependency() with just the "sources"
arguments, because it says "sources to add to targets (or generated
header files that should be built before sources including them are
built)" (so I assume it means that it should be trigger a rebuild,
otherwise it's useless) but it's just not working. I'll investigate
more.
Still going with this for now, because at least generating the PDB
source was a big miss until now. But we should
The initial attempt of this commit was to remove the `GtkAction` usage,
but grew a bit wider than that. The following happened:
* The dialog became a proper GObject, rather than being a big chunk of
static variables that were hopefully initialized before you used them.
* The dialog now uses `GAction`s to implement actions, and converted
some signal handlers to actions as well.
* The plug-in run procedure now uses `GtkApplication`. This is one hand
necessary to be able to set accelerators for actions, but on the other
hand is more future-proof, as GTK4 removes `gtk_main()`
This new function is an alternative to existing
gimp_image_metadata_save_finish, when you want to save metadata
yourself and you need only filtering processing.
It returns filtered metadata allowing the caller
to save the finalized metadata via other means
(via native format’s library for example)
This API can be used to support metadata saving of image formats
not directly supported by gexiv2/exiv2.
… %GIMP_TYPE_SPIN_SCALE in gimp_procedure_dialog_get_widget().
The dedicated function is for when a plug-in wants to use a scale range
multiplied by a factor. Otherwise using the generic function is fine.
Now the warning is:
WARNING: Invalid fragment for 'Gimp.Config': it should be struct
It implements the [iface@Config] interface and therefore has all its
^~~~~~~~~~~~~~
This warning feels wrong as we use GimpConfig for the name of the
interface, yet in the .gir file, I see `<record name="Config" …>` yet
`<interface name="ConfigInterface" …>`.
I suppose gi-docgen would want us to use [iface@ConfigInterface] if we
want to link the interface. But looking at the gir file, all interesting
interface methods are associated to the Config record. So let's just go
with the [struct@Config] proposed by the warning.
In any case, something feels wrong or broken here, but we need to fix
this for the CI. I hope Niels can look at it at some point.
Fixing these 2 warnings in the CI which end up fatal:
WARNING: Invalid fragment for 'Gimp.Parasite': it should be struct
Serializes the object properties of @config to a [class@Parasite].
^~~~~~~~~~~~~~~~
WARNING: Invalid fragment for 'GLib.MainLoop': it should be struct
it has a GUI and is hanging around in a [class@GLib.MainLoop], it must call
^~~~~~~~~~~~~~~~~~~~~
Now using the new GimpLabelColor as new default for RGB properties. It
makes more sense that the default is editable widgets. Also it has a
label, which is better default widget.
Also gimp_procedure_dialog_get_color_widget() now only returns
GimpLabelColor widgets.
As Massimo notes, the issue is not about the callback being broken in
bindings, but simply that bindings fail to handle random data without an
associated size. So let's just add the size. I confirmed testing API in
the Python binding that it now works fine.
GLib has a specific type of NULL-terminated string arrays:
`G_TYPE_STRV`, which is the `GType` of `char**` aka `GStrv`.
By using this type, we can avoid having a `GimpStringArray` which is a
bit cumbersome to use for both the C API, as well as bindings. By using
`GStrv`, we allow other languages to pass on string lists as they are
used to, while the bindings will make sure to do the right thing.
In the end, it makes the API a little bit simpler for everyone, and
reduces confusion for people who are used to working with string arrays
in other C/GLib based code (and not having 2 different types to denote
the same thing).
Related: https://gitlab.gnome.org/GNOME/gimp/-/issues/5919
While we do have quite a few gimp_pdb_run_procedure*() functions now, I
always felt that one based on a config file was missing, even more as we
are getting further and further into using config objects in plug-ins.
In C, the gimp_pdb_run_procedure() function is without a doubt the
easiest one. But such variable arg functions are not available on
bindings, and having to deal with GValue and GimpValueArray is a real
pain.
Also using a config file has the very great advantage that we don't need
to care about order. For instance, if I need to set the 10th argument of
a PDB call (and leave the rest to default values), I don't have to set
all 9 previous arguments. I can set only this one if I want. This
advantage is useful also for C code by the way.
For the record, here is how you could load then export an image with the
"file-png-*" PDB procedures in Python:
> c = Gimp.get_pdb().lookup_procedure('file-png-load').create_config()
> c.set_property('file', Gio.file_new_for_path('/path/sample.png'))
> r = Gimp.get_pdb().run_procedure_config('file-png-load', c)
> d = Gimp.Display.new(r.index(1)) # Give it a display to work on it.
Now exporting:
> img = r.index(1)
> c = Gimp.get_pdb().lookup_procedure('file-png-save').create_config()
> c.set_property('image', img)
> c.set_property('file', Gio.file_new_for_path('/path/exported.png'))
> layers = img.get_layers()
> c.set_property('drawables', Gimp.ObjectArray.new(Gimp.Drawable, layers, False))
> c.set_property('num-drawables', len(layers))
> r = Gimp.get_pdb().run_procedure_config('file-png-save', c)
… assert the existence of GError.
This is even worse as deserialize() method does not even take a GError
parameter anyway so this assert will always go off when a
deserialization failed (which happened in my case as I was working on a
plug-in API, hence gimp_procedure_config_load_last() actually failed to
load a previous version of the plug-in-settings when I changed a
procedure arg's type).
Just fail the deserialization normally and let the calling code handling
this case. Nevertheless it is kind of useful to bubble-up the error to
calling code, so I add a TODO in the interface header (hopefully to see
and improve this before we release GIMP 3.0).
Last deprecated usages in this file. Actually there are a few other
calls but deprecated on GExiv2 0.14.0, hence over our current
requirement. So we'll have to handle these later.
Replace functions gexiv2_metadata_set_xmp_tag_struct() and
gexiv2_metadata_get_tag_type() with their _try_ variants.
Note that I print to stderr rather than raising a warning here as I am
quite unsure where the list of XMP metadata we are gathering comes from.
Is it fully validated by GExiv2, and therefore no errors are ever
supposed to happen? (in such case, we should raise a warning if it does)
Or is it user-provided data (e.g. from a file loaded in GIMP which can
contain broken metadata)? In such a case, we should probably handle the
error slightly differently to warn for non-processable data (hence
possibly metadata loss at export time).
For the time being, then go with this weak handling and take care of
this better when we can look further into this.
… gexiv2_metadata_try_set_tag_string().
These are usage where we have full control over tag names and values so
no error is supposed to happen. This is why we use them as code warnings
and not returned error (because if an error did happen, this would be a
bug rather than a user error or a system error).
Here are the changes:
- Separate the check for comment contents and the one of whether or not
it starts with "charset=Ascii ". Indeed in my tests, we still want to
handle the empty string case or the "binary comment" case even without
the charset prefix (currently it was both or none, so I encountered
cases with a broken "binary comment" comment because the charset
prefix was absent).
- Add some #warning in order not to forget to remove the bogus "binary
comment" test. Indeed after some digging in Exiv2 code, I could
confirm this return value got removed in commit 9b510858 of Exiv2
repository, i.e. since Exiv2 0.27.4. Now in my test case where I had a
tag containing only 0s, Exiv2 returns an empty string, which is
perfectly fine (and it's up to us to keep or ignore it).
- Use gexiv2_metadata_try_get_tag_interpreted_string() instead of their
deprecate non-try counterparses. Right now, I am just outputting any
error message to stderr, as I'm not sure if this is the kind of errors
we want to warn people about. I guess it would depend on which type of
errors exactly are returned, so let's see if we encounter some case in
the future. For now stderr is fine to detect these.
yocto/oe is capable of building gobject introspection despite cross-compiling.
add an option to enable gir build even if cross-compiling
Signed-off-by: Markus Volk <f_l_k@t-online.de>
Reviewer note (Jehan): this whole stuff is a mess. Actually I'd like to
simply get rid of the whole no-gir-when-cross-compiling logics but I
still can't figure out how to cross-build generically with GIR.
Yet since some manage it with yocto, let's unblock them.
See #7208.
gtk-doc has been slowly dying for the past few years; with gi-docgen we
have a nice successor.
This also makes sure the C documentation also uses the GIR file, which
in turn means faster build times (since all the C code doesn't have to
be parsed and recompiled again), and has a clear dependency graph.
See the [gi-docgen tutorial] for more info on how the system works.
[gi-docgen tutorial]: https://gnome.pages.gitlab.gnome.org/gi-docgen/tutorial.html
I cleaned many remaining places where the concept of linked item still
survived.
On loading an XCF file with linked items, we are now going to create a
named sets for the linked items, allowing people to easily select these
back if the relation was still needed.
We don't remove gimp_item_get_linked() yet and in particular, we don't
save stored items into XCF files. This will come in an upcoming change.
When running tests, the data are not meant to be necessarily installed.
Therefore icons won't be found when calling gimp_widgets_init().
Add some special-casing to find them relatively to the install
directory.
Fixes the patch from !470 which is mostly right, except that
g_param_spec_sink() may possibly lead to finalizing the GParamSpec
(typically if it was a just-created floating spec). We don't want to
return pointer to freed data. Let's return NULL instead.
Also looking closer at the memory handling here, it looks the right
annotation for @pspec is (transfer floating). Basically we are sinking a
floating object into a full object and taking ownership of this sunk
object. But if the object was already sunk, we are reffing it and
keeping this additional reference, not the passed argument's. Hopefully
it's right since the annotation and handling of floating object with
GObject Introspection seems very unclear to me (even in core GObject
code, I see what looks like contradictory annotations).
In the normal flow, pspec is persisted in the arguments array, and is
g_param_spec_ref_sink()'d in order to sink a possible floating ref. To
avoid a leak in the error case, we need to add some g_param_spec_sink().
Saving a thumbnail is closely related to the other metadata preferences,
but so far this was the only one that didn't have a preference for a
default user value.
This commit adds a preference in the metadata section where a user can
select whether thumbnail saving is enabled by default or not.
Since it appeared with GLib 2.68.0, we could not change this until we
bumped the dependency which has only become possible a few days ago
(since Debian testing is our baseline for dependency bumps). Cf.
previous commit.
As this is a drop-in replacement (just a guint parameter changed to
gsize to avoid integer overflow), search-and-replace with:
> sed -i 's/g_memdup\>/g_memdup2/g' `grep -rIl 'g_memdup\>' *`
… followed by a few manual alignment tweaks when necessary.
This gets rid of the many deprecation warnings which we had lately when
building with a recent GLib version.
Jehan