AuroraMiddleware/gtk2
1
0
Fork 0
forked from AuroraMiddleware/gtk
Code Issues Pull Requests Projects Releases Wiki Activity
a7a2dbc3ea
gtk2/gsk/gsktransform.c

2222 lines
63 KiB
C
Raw Normal View History

gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
/*
* Copyright © 2019 Benjamin Otte
*
* This library is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2.1 of the License, or (at your option) any later version.
*
* 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.
*
* You should have received a copy of the GNU Lesser General Public
* License along with this library. If not, see <http://www.gnu.org/licenses/>.
*
* Authors: Benjamin Otte <otte@gnome.org>
*/
/**
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
* GskTransform: (ref-func gsk_transform_ref) (unref-func gsk_transform_unref)
*
* `GskTransform` is an object to describe transform matrices.
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
*
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
* Unlike `graphene_matrix_t`, `GskTransform` retains the steps in how
* a transform was constructed, and allows inspecting them. It is modeled
* after the way CSS describes transforms.
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
*
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
* `GskTransform` objects are immutable and cannot be changed after creation.
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
* This means code can safely expose them as properties of objects without
* having to worry about others changing them.
*/
#include "config.h"
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
#include "gsktransformprivate.h"
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
struct _GskTransformClass
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
{
gsize struct_size;
const char *type_name;
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
void (* finalize) (GskTransform *transform);
void (* to_matrix) (GskTransform *transform,
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
graphene_matrix_t *out_matrix);
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
void (* apply_2d) (GskTransform *transform,
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
float *out_xx,
float *out_yx,
float *out_xy,
float *out_yy,
float *out_dx,
float *out_dy);
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
void (* apply_affine) (GskTransform *transform,
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
float *out_scale_x,
float *out_scale_y,
float *out_dx,
float *out_dy);
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
void (* apply_translate) (GskTransform *transform,
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
float *out_dx,
float *out_dy);
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
void (* print) (GskTransform *transform,
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
GString *string);
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
GskTransform * (* apply) (GskTransform *transform,
GskTransform *apply_to);
transform: Implement gsk_transform_invert() And use it. And test it.
2019-03-03 16:57:11 +00:00
GskTransform * (* invert) (GskTransform *transform,
GskTransform *next);
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
/* both matrices have the same type */
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gboolean (* equal) (GskTransform *first_transform,
GskTransform *second_transform);
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
};
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
G_DEFINE_BOXED_TYPE (GskTransform, gsk_transform,
gsk_transform_ref,
gsk_transform_unref)
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
static gboolean
transform: Remove API to poke internals It is not interesting to users of GskTransform how it is made up internally. Users should just use the gsk_transform_to_*() APIs.
2019-03-01 05:48:30 +00:00
gsk_transform_is_identity (GskTransform *self);
GskTransform: Remove unused function from private header We only use this in gsktransform.c
2020-07-24 15:48:54 +00:00
static GskTransform *
gsk_transform_matrix_with_category (GskTransform *next,
const graphene_matrix_t*matrix,
GskTransformCategory category);
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
transform: Coalesce similar transforms If somebody does a transform like scale(5) scale(10) translate(1,1) translate(5,0) store it instead as scale(50) translate(6,1) This way, less memory is consumed and transforms are easier to read. In particular, this simplifies the typical transforms we do in GTK, which are just one translation after another.
2019-06-02 12:16:45 +00:00
static inline gboolean
gsk_transform_has_class (GskTransform *self,
const GskTransformClass *transform_class)
{
return self != NULL && self->transform_class == transform_class;
}
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
/*< private >
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
* gsk_transform_alloc:
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
* @transform_class: class structure for this self
transform: Store the category in the transform ... instead of computing it every time we need it. This should be faster and we want to use it a lot more prominently. Also, we have the struct memory available anyway.
2019-03-04 08:20:58 +00:00
* @category: The category of this transform. Will be used to initialize
Documentation fixes Mostly fixing up indentation of continuation lines, and other small cleanups.
2021-05-18 21:05:26 +00:00
* the result's category together with &next's category
docs: Reduce redundancy Remove a boatload of "or %NULL" from nullable parameters and return values. gi-docgen generates suitable text from the annotation that we don't need to duplicate. This adds a few missing nullable annotations too.
2021-05-21 00:45:06 +00:00
* @next: (transfer full) (nullable): Next transform to multiply with
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
*
gsk: Clean up docs syntax Replace leftover gtk-doc syntax (#Type) with backquotes.
2021-05-20 03:40:54 +00:00
* Returns: (transfer full): the newly created `GskTransform`
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
*/
static gpointer
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_transform_alloc (const GskTransformClass *transform_class,
transform: Store the category in the transform ... instead of computing it every time we need it. This should be faster and we want to use it a lot more prominently. Also, we have the struct memory available anyway.
2019-03-04 08:20:58 +00:00
GskTransformCategory category,
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
GskTransform *next)
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
{
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
GskTransform *self;
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
g_return_val_if_fail (transform_class != NULL, NULL);
Use atomic boxing instead of manual refcounting Artisanal, homegrown, locally sourced, vegan reference counting has been replaced by the appropriate API in GLib, which does small things like saturation and type checking.
2019-03-22 11:45:57 +00:00
self = g_atomic_rc_box_alloc0 (transform_class->struct_size);
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
self->transform_class = transform_class;
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
self->category = next ? MIN (category, next->category) : category;
GskTransform: Avoid a ref+unref pair If gsk_transform_is_identity() returned FALSE for the next transform, the previous code did a ref + unref pair, even though it was unneeded.
2020-12-18 09:10:56 +00:00
if (gsk_transform_is_identity (next))
gsk_transform_unref (next);
else
self->next = next;
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
return self;
}
/*** IDENTITY ***/
static void
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_identity_transform_finalize (GskTransform *transform)
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
{
}
static void
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_identity_transform_to_matrix (GskTransform *transform,
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
graphene_matrix_t *out_matrix)
{
graphene_matrix_init_identity (out_matrix);
}
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
static void
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
gsk_identity_transform_apply_2d (GskTransform *transform,
float *out_xx,
float *out_yx,
float *out_xy,
float *out_yy,
float *out_dx,
float *out_dy)
{
}
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
static void
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_identity_transform_apply_affine (GskTransform *transform,
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
float *out_scale_x,
float *out_scale_y,
float *out_dx,
float *out_dy)
{
}
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
static void
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
gsk_identity_transform_apply_translate (GskTransform *transform,
float *out_dx,
float *out_dy)
{
}
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
static void
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_identity_transform_print (GskTransform *transform,
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
GString *string)
{
gsk: Fix printing of identity transforms These need to print as "none", otherwise the parser does not accept them.
2019-06-03 16:19:57 +00:00
g_string_append (string, "none");
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
}
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
static GskTransform *
gsk_identity_transform_apply (GskTransform *transform,
GskTransform *apply_to)
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
{
transform: Implement gsk_transform_invert() And use it. And test it.
2019-03-03 16:57:11 +00:00
/* We do the following to make sure inverting a non-NULL transform
* will return a non-NULL transform.
*/
if (apply_to)
return apply_to;
else
return gsk_transform_new ();
}
static GskTransform *
gsk_identity_transform_invert (GskTransform *transform,
GskTransform *next)
{
/* We do the following to make sure inverting a non-NULL transform
* will return a non-NULL transform.
*/
if (next)
return next;
else
return gsk_transform_new ();
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
}
static gboolean
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_identity_transform_equal (GskTransform *first_transform,
GskTransform *second_transform)
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
{
return TRUE;
}
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
static const GskTransformClass GSK_IDENTITY_TRANSFORM_CLASS =
{
sizeof (GskTransform),
GskTransform: Fix identity transform class name
2020-12-18 09:12:28 +00:00
"GskIdentityTransform",
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_identity_transform_finalize,
gsk_identity_transform_to_matrix,
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
gsk_identity_transform_apply_2d,
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_identity_transform_apply_affine,
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
gsk_identity_transform_apply_translate,
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_identity_transform_print,
gsk_identity_transform_apply,
transform: Implement gsk_transform_invert() And use it. And test it.
2019-03-03 16:57:11 +00:00
gsk_identity_transform_invert,
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_identity_transform_equal,
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
};
transform: Remove API to poke internals It is not interesting to users of GskTransform how it is made up internally. Users should just use the gsk_transform_to_*() APIs.
2019-03-01 05:48:30 +00:00
/*<private>
* gsk_transform_is_identity:
docs: Reduce redundancy Remove a boatload of "or %NULL" from nullable parameters and return values. gi-docgen generates suitable text from the annotation that we don't need to duplicate. This adds a few missing nullable annotations too.
2021-05-21 00:45:06 +00:00
* @transform: (nullable): A transform
transform: Remove API to poke internals It is not interesting to users of GskTransform how it is made up internally. Users should just use the gsk_transform_to_*() APIs.
2019-03-01 05:48:30 +00:00
*
* Checks if the transform is a representation of the identity
* transform.
*
* This is different from a transform like `scale(2) scale(0.5)`
* which just results in an identity transform when simplified.
*
* Returns: %TRUE if this transform is a representation of
Documentation fixes Mostly fixing up indentation of continuation lines, and other small cleanups.
2021-05-18 21:05:26 +00:00
* the identity transform
transform: Remove API to poke internals It is not interesting to users of GskTransform how it is made up internally. Users should just use the gsk_transform_to_*() APIs.
2019-03-01 05:48:30 +00:00
**/
static gboolean
gsk_transform_is_identity (GskTransform *self)
{
return self == NULL ||
(self->transform_class == &GSK_IDENTITY_TRANSFORM_CLASS && gsk_transform_is_identity (self->next));
}
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
/*** MATRIX ***/
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
typedef struct _GskMatrixTransform GskMatrixTransform;
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
struct _GskMatrixTransform
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
{
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
GskTransform parent;
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
graphene_matrix_t matrix;
};
static void
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_matrix_transform_finalize (GskTransform *self)
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
{
}
static void
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_matrix_transform_to_matrix (GskTransform *transform,
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
graphene_matrix_t *out_matrix)
{
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
GskMatrixTransform *self = (GskMatrixTransform *) transform;
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
graphene_matrix_init_from_matrix (out_matrix, &self->matrix);
}
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
static void
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
gsk_matrix_transform_apply_2d (GskTransform *transform,
float *out_xx,
float *out_yx,
float *out_xy,
float *out_yy,
float *out_dx,
float *out_dy)
{
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
GskMatrixTransform *self = (GskMatrixTransform *) transform;
graphene_matrix_t mat;
graphene_matrix_init_from_2d (&mat,
*out_xx, *out_yx,
*out_xy, *out_yy,
*out_dx, *out_dy);
graphene_matrix_multiply (&self->matrix, &mat, &mat);
/* not using graphene_matrix_to_2d() because it may
* fail the is_2d() check due to improper rounding */
*out_xx = graphene_matrix_get_value (&mat, 0, 0);
*out_yx = graphene_matrix_get_value (&mat, 0, 1);
*out_xy = graphene_matrix_get_value (&mat, 1, 0);
*out_yy = graphene_matrix_get_value (&mat, 1, 1);
*out_dx = graphene_matrix_get_value (&mat, 3, 0);
*out_dy = graphene_matrix_get_value (&mat, 3, 1);
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
}
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
static void
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_matrix_transform_apply_affine (GskTransform *transform,
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
float *out_scale_x,
float *out_scale_y,
float *out_dx,
float *out_dy)
{
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
GskMatrixTransform *self = (GskMatrixTransform *) transform;
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
transform: Store the category in the transform ... instead of computing it every time we need it. This should be faster and we want to use it a lot more prominently. Also, we have the struct memory available anyway.
2019-03-04 08:20:58 +00:00
switch (transform->category)
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
{
transform: Make category public API Also rename it from GskMatrixCategory to GskTransformCategory.
2019-03-01 06:51:23 +00:00
case GSK_TRANSFORM_CATEGORY_UNKNOWN:
case GSK_TRANSFORM_CATEGORY_ANY:
case GSK_TRANSFORM_CATEGORY_3D:
case GSK_TRANSFORM_CATEGORY_2D:
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
default:
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
g_assert_not_reached ();
break;
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
transform: Make category public API Also rename it from GskMatrixCategory to GskTransformCategory.
2019-03-01 06:51:23 +00:00
case GSK_TRANSFORM_CATEGORY_2D_AFFINE:
Use getters for translation and scaling components Instead of accessing the matrix cells directly.
2019-03-22 13:54:29 +00:00
*out_dx += *out_scale_x * graphene_matrix_get_x_translation (&self->matrix);
*out_dy += *out_scale_y * graphene_matrix_get_y_translation (&self->matrix);
*out_scale_x *= graphene_matrix_get_x_scale (&self->matrix);
*out_scale_y *= graphene_matrix_get_y_scale (&self->matrix);
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
break;
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
transform: Make category public API Also rename it from GskMatrixCategory to GskTransformCategory.
2019-03-01 06:51:23 +00:00
case GSK_TRANSFORM_CATEGORY_2D_TRANSLATE:
Use getters for translation and scaling components Instead of accessing the matrix cells directly.
2019-03-22 13:54:29 +00:00
*out_dx += *out_scale_x * graphene_matrix_get_x_translation (&self->matrix);
*out_dy += *out_scale_y * graphene_matrix_get_y_translation (&self->matrix);
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
break;
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
transform: Make category public API Also rename it from GskMatrixCategory to GskTransformCategory.
2019-03-01 06:51:23 +00:00
case GSK_TRANSFORM_CATEGORY_IDENTITY:
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
break;
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
}
}
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
static void
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
gsk_matrix_transform_apply_translate (GskTransform *transform,
float *out_dx,
float *out_dy)
{
GskMatrixTransform *self = (GskMatrixTransform *) transform;
transform: Store the category in the transform ... instead of computing it every time we need it. This should be faster and we want to use it a lot more prominently. Also, we have the struct memory available anyway.
2019-03-04 08:20:58 +00:00
switch (transform->category)
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
{
transform: Make category public API Also rename it from GskMatrixCategory to GskTransformCategory.
2019-03-01 06:51:23 +00:00
case GSK_TRANSFORM_CATEGORY_UNKNOWN:
case GSK_TRANSFORM_CATEGORY_ANY:
case GSK_TRANSFORM_CATEGORY_3D:
case GSK_TRANSFORM_CATEGORY_2D:
case GSK_TRANSFORM_CATEGORY_2D_AFFINE:
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
default:
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
g_assert_not_reached ();
break;
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
transform: Make category public API Also rename it from GskMatrixCategory to GskTransformCategory.
2019-03-01 06:51:23 +00:00
case GSK_TRANSFORM_CATEGORY_2D_TRANSLATE:
Use getters for translation and scaling components Instead of accessing the matrix cells directly.
2019-03-22 13:54:29 +00:00
*out_dx += graphene_matrix_get_x_translation (&self->matrix);
*out_dy += graphene_matrix_get_y_translation (&self->matrix);
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
break;
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
transform: Make category public API Also rename it from GskMatrixCategory to GskTransformCategory.
2019-03-01 06:51:23 +00:00
case GSK_TRANSFORM_CATEGORY_IDENTITY:
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
break;
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
}
}
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
static void
string_append_double (GString *string,
double d)
{
char buf[G_ASCII_DTOSTR_BUF_SIZE];
g_ascii_formatd (buf, G_ASCII_DTOSTR_BUF_SIZE, "%g", d);
g_string_append (string, buf);
}
static void
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_matrix_transform_print (GskTransform *transform,
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
GString *string)
{
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
GskMatrixTransform *self = (GskMatrixTransform *) transform;
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
guint i;
float f[16];
g_string_append (string, "matrix3d(");
graphene_matrix_to_float (&self->matrix, f);
for (i = 0; i < 16; i++)
{
if (i > 0)
g_string_append (string, ", ");
string_append_double (string, f[i]);
}
g_string_append (string, ")");
}
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
static GskTransform *
gsk_matrix_transform_apply (GskTransform *transform,
GskTransform *apply_to)
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
{
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
GskMatrixTransform *self = (GskMatrixTransform *) transform;
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
return gsk_transform_matrix_with_category (apply_to,
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
&self->matrix,
transform: Store the category in the transform ... instead of computing it every time we need it. This should be faster and we want to use it a lot more prominently. Also, we have the struct memory available anyway.
2019-03-04 08:20:58 +00:00
transform->category);
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
}
transform: Implement gsk_transform_invert() And use it. And test it.
2019-03-03 16:57:11 +00:00
static GskTransform *
gsk_matrix_transform_invert (GskTransform *transform,
GskTransform *next)
{
GskMatrixTransform *self = (GskMatrixTransform *) transform;
graphene_matrix_t inverse;
if (!graphene_matrix_inverse (&self->matrix, &inverse))
{
gsk_transform_unref (next);
return NULL;
}
return gsk_transform_matrix_with_category (next,
&inverse,
transform: Store the category in the transform ... instead of computing it every time we need it. This should be faster and we want to use it a lot more prominently. Also, we have the struct memory available anyway.
2019-03-04 08:20:58 +00:00
transform->category);
transform: Implement gsk_transform_invert() And use it. And test it.
2019-03-03 16:57:11 +00:00
}
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
static gboolean
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_matrix_transform_equal (GskTransform *first_transform,
GskTransform *second_transform)
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
{
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
GskMatrixTransform *first = (GskMatrixTransform *) first_transform;
GskMatrixTransform *second = (GskMatrixTransform *) second_transform;
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
Use matrix equality in GskTransform Graphene has new API to check for equality, so we should use it instead of doing a byte-by-byte comparison.
2019-03-22 12:38:36 +00:00
if (graphene_matrix_equal_fast (&first->matrix, &second->matrix))
return TRUE;
return graphene_matrix_equal (&first->matrix, &second->matrix);
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
}
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
static const GskTransformClass GSK_TRANSFORM_TRANSFORM_CLASS =
{
sizeof (GskMatrixTransform),
"GskMatrixTransform",
gsk_matrix_transform_finalize,
gsk_matrix_transform_to_matrix,
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
gsk_matrix_transform_apply_2d,
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_matrix_transform_apply_affine,
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
gsk_matrix_transform_apply_translate,
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_matrix_transform_print,
gsk_matrix_transform_apply,
transform: Implement gsk_transform_invert() And use it. And test it.
2019-03-03 16:57:11 +00:00
gsk_matrix_transform_invert,
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_matrix_transform_equal,
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
};
GskTransform: Remove unused function from private header We only use this in gsktransform.c
2020-07-24 15:48:54 +00:00
static GskTransform *
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_transform_matrix_with_category (GskTransform *next,
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
const graphene_matrix_t *matrix,
transform: Make category public API Also rename it from GskMatrixCategory to GskTransformCategory.
2019-03-01 06:51:23 +00:00
GskTransformCategory category)
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
{
transform: Store the category in the transform ... instead of computing it every time we need it. This should be faster and we want to use it a lot more prominently. Also, we have the struct memory available anyway.
2019-03-04 08:20:58 +00:00
GskMatrixTransform *result = gsk_transform_alloc (&GSK_TRANSFORM_TRANSFORM_CLASS, category, next);
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
graphene_matrix_init_from_matrix (&result->matrix, matrix);
return &result->parent;
}
/**
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
* gsk_transform_matrix:
introspection: Stop using allow-none allow-none has been deprecated for a long time already. Instead use optional and nullable everywhere.
2021-05-19 11:24:34 +00:00
* @next: (nullable) (transfer full): the next transform
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
* @matrix: the matrix to multiply @next with
*
* Multiplies @next with the given @matrix.
*
transform: Fix docs to use "transform" instead of "matrix"
2020-06-13 06:25:51 +00:00
* Returns: The new transform
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
**/
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
GskTransform *
gsk_transform_matrix (GskTransform *next,
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
const graphene_matrix_t *matrix)
{
transform: Make category public API Also rename it from GskMatrixCategory to GskTransformCategory.
2019-03-01 06:51:23 +00:00
return gsk_transform_matrix_with_category (next, matrix, GSK_TRANSFORM_CATEGORY_UNKNOWN);
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
}
/*** TRANSLATE ***/
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
typedef struct _GskTranslateTransform GskTranslateTransform;
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
struct _GskTranslateTransform
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
{
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
GskTransform parent;
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
graphene_point3d_t point;
};
static void
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_translate_transform_finalize (GskTransform *self)
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
{
}
static void
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_translate_transform_to_matrix (GskTransform *transform,
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
graphene_matrix_t *out_matrix)
{
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
GskTranslateTransform *self = (GskTranslateTransform *) transform;
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
graphene_matrix_init_translate (out_matrix, &self->point);
}
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
static void
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
gsk_translate_transform_apply_2d (GskTransform *transform,
float *out_xx,
float *out_yx,
float *out_xy,
float *out_yy,
float *out_dx,
float *out_dy)
{
GskTranslateTransform *self = (GskTranslateTransform *) transform;
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
g_assert (self->point.z == 0.0);
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
*out_dx += *out_xx * self->point.x + *out_xy * self->point.y;
*out_dy += *out_yx * self->point.x + *out_yy * self->point.y;
}
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
static void
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_translate_transform_apply_affine (GskTransform *transform,
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
float *out_scale_x,
float *out_scale_y,
float *out_dx,
float *out_dy)
{
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
GskTranslateTransform *self = (GskTranslateTransform *) transform;
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
g_assert (self->point.z == 0.0);
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
*out_dx += *out_scale_x * self->point.x;
*out_dy += *out_scale_y * self->point.y;
}
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
static void
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
gsk_translate_transform_apply_translate (GskTransform *transform,
float *out_dx,
float *out_dy)
{
GskTranslateTransform *self = (GskTranslateTransform *) transform;
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
g_assert (self->point.z == 0.0);
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
*out_dx += self->point.x;
*out_dy += self->point.y;
}
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
static GskTransform *
gsk_translate_transform_apply (GskTransform *transform,
GskTransform *apply_to)
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
{
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
GskTranslateTransform *self = (GskTranslateTransform *) transform;
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
return gsk_transform_translate_3d (apply_to, &self->point);
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
}
transform: Implement gsk_transform_invert() And use it. And test it.
2019-03-03 16:57:11 +00:00
static GskTransform *
gsk_translate_transform_invert (GskTransform *transform,
GskTransform *next)
{
GskTranslateTransform *self = (GskTranslateTransform *) transform;
return gsk_transform_translate_3d (next, &GRAPHENE_POINT3D_INIT (-self->point.x, -self->point.y, -self->point.z));
}
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
static gboolean
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_translate_transform_equal (GskTransform *first_transform,
GskTransform *second_transform)
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
{
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
GskTranslateTransform *first = (GskTranslateTransform *) first_transform;
GskTranslateTransform *second = (GskTranslateTransform *) second_transform;
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
transform: Compare 3d points directly
2020-06-07 06:45:35 +00:00
return G_APPROX_VALUE (first->point.x, second->point.x, FLT_EPSILON) &&
G_APPROX_VALUE (first->point.y, second->point.y, FLT_EPSILON) &&
G_APPROX_VALUE (first->point.z, second->point.z, FLT_EPSILON);
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
}
static void
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_translate_transform_print (GskTransform *transform,
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
GString *string)
{
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
GskTranslateTransform *self = (GskTranslateTransform *) transform;
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
if (self->point.z == 0)
g_string_append (string, "translate(");
else
g_string_append (string, "translate3d(");
string_append_double (string, self->point.x);
g_string_append (string, ", ");
string_append_double (string, self->point.y);
if (self->point.z != 0)
{
g_string_append (string, ", ");
trasnform: Fix print statement We were printing the wrong variable.
2019-03-02 03:35:44 +00:00
string_append_double (string, self->point.z);
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
}
g_string_append (string, ")");
}
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
static const GskTransformClass GSK_TRANSLATE_TRANSFORM_CLASS =
{
sizeof (GskTranslateTransform),
"GskTranslateTransform",
gsk_translate_transform_finalize,
gsk_translate_transform_to_matrix,
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
gsk_translate_transform_apply_2d,
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_translate_transform_apply_affine,
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
gsk_translate_transform_apply_translate,
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_translate_transform_print,
gsk_translate_transform_apply,
transform: Implement gsk_transform_invert() And use it. And test it.
2019-03-03 16:57:11 +00:00
gsk_translate_transform_invert,
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_translate_transform_equal,
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
};
/**
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
* gsk_transform_translate:
introspection: Stop using allow-none allow-none has been deprecated for a long time already. Instead use optional and nullable everywhere.
2021-05-19 11:24:34 +00:00
* @next: (nullable) (transfer full): the next transform
transform: Fix docs to use "transform" instead of "matrix"
2020-06-13 06:25:51 +00:00
* @point: the point to translate the transform by
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
*
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
* Translates @next in 2-dimensional space by @point.
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
*
transform: Fix docs to use "transform" instead of "matrix"
2020-06-13 06:25:51 +00:00
* Returns: The new transform
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
**/
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
GskTransform *
gsk_transform_translate (GskTransform *next,
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
const graphene_point_t *point)
{
graphene_point3d_t point3d;
graphene_point3d_init (&point3d, point->x, point->y, 0);
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
return gsk_transform_translate_3d (next, &point3d);
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
}
/**
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
* gsk_transform_translate_3d:
introspection: Stop using allow-none allow-none has been deprecated for a long time already. Instead use optional and nullable everywhere.
2021-05-19 11:24:34 +00:00
* @next: (nullable) (transfer full): the next transform
transform: Fix docs to use "transform" instead of "matrix"
2020-06-13 06:25:51 +00:00
* @point: the point to translate the transform by
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
*
* Translates @next by @point.
*
transform: Fix docs to use "transform" instead of "matrix"
2020-06-13 06:25:51 +00:00
* Returns: The new transform
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
**/
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
GskTransform *
gsk_transform_translate_3d (GskTransform *next,
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
const graphene_point3d_t *point)
{
transform: Store the category in the transform ... instead of computing it every time we need it. This should be faster and we want to use it a lot more prominently. Also, we have the struct memory available anyway.
2019-03-04 08:20:58 +00:00
GskTranslateTransform *result;
GskTransform: Ignore identity transforms We often end up with e.g. a scale of 1 or a translation by 0/0. Ignore those transforms since they don't do anything.
2019-06-03 14:20:12 +00:00
if (graphene_point3d_equal (point, graphene_point3d_zero ()))
transform: Fix identity transform fast paths Returning an extra ref will cause leaks later.
2019-07-06 10:56:28 +00:00
return next;
GskTransform: Ignore identity transforms We often end up with e.g. a scale of 1 or a translation by 0/0. Ignore those transforms since they don't do anything.
2019-06-03 14:20:12 +00:00
transform: Coalesce similar transforms If somebody does a transform like scale(5) scale(10) translate(1,1) translate(5,0) store it instead as scale(50) translate(6,1) This way, less memory is consumed and transforms are easier to read. In particular, this simplifies the typical transforms we do in GTK, which are just one translation after another.
2019-06-02 12:16:45 +00:00
if (gsk_transform_has_class (next, &GSK_TRANSLATE_TRANSFORM_CLASS))
{
GskTranslateTransform *t = (GskTranslateTransform *) next;
GskTransform *r = gsk_transform_translate_3d (gsk_transform_ref (next->next),
&GRAPHENE_POINT3D_INIT(t->point.x + point->x,
t->point.y + point->y,
t->point.z + point->z));
gsk_transform_unref (next);
return r;
}
transform: Store the category in the transform ... instead of computing it every time we need it. This should be faster and we want to use it a lot more prominently. Also, we have the struct memory available anyway.
2019-03-04 08:20:58 +00:00
result = gsk_transform_alloc (&GSK_TRANSLATE_TRANSFORM_CLASS,
point->z == 0.0 ? GSK_TRANSFORM_CATEGORY_2D_TRANSLATE
: GSK_TRANSFORM_CATEGORY_3D,
next);
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
graphene_point3d_init_from_point (&result->point, point);
return &result->parent;
}
/*** ROTATE ***/
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
typedef struct _GskRotateTransform GskRotateTransform;
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
struct _GskRotateTransform
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
{
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
GskTransform parent;
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
float angle;
};
static void
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_rotate_transform_finalize (GskTransform *self)
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
{
}
GskTransform: Compute sin/deg of 90deg rotations manually Otherwise we might end up with inaccuracies, e.g. when gsk_transform_transform_bounds()'ing bounds.
2019-12-19 11:30:08 +00:00
static inline void
_sincos (float deg,
float *out_s,
float *out_c)
{
if (deg == 90.0)
{
*out_c = 0.0;
*out_s = 1.0;
}
else if (deg == 180.0)
{
*out_c = -1.0;
*out_s = 0.0;
}
else if (deg == 270.0)
{
*out_c = 0.0;
*out_s = -1.0;
}
else if (deg == 0.0)
{
*out_c = 1.0;
*out_s = 0.0;
}
else
{
build: Check for sincosf() sincosf() is really a GCC-specific function that may more may not be supported on non-GCC compilers, so we want to check for it, otherwise we use a fallback implementation, not unlike the one in demos/gtk-demo/gtkgears.c.
2020-01-08 09:05:44 +00:00
float angle = deg * M_PI / 180.0;
#ifdef HAVE_SINCOSF
sincosf (angle, out_s, out_c);
#else
*out_s = sinf (angle);
*out_c = cosf (angle);
#endif
GskTransform: Compute sin/deg of 90deg rotations manually Otherwise we might end up with inaccuracies, e.g. when gsk_transform_transform_bounds()'ing bounds.
2019-12-19 11:30:08 +00:00
}
}
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
static void
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_rotate_transform_to_matrix (GskTransform *transform,
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
graphene_matrix_t *out_matrix)
{
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
GskRotateTransform *self = (GskRotateTransform *) transform;
GskTransform: Compute sin/deg of 90deg rotations manually Otherwise we might end up with inaccuracies, e.g. when gsk_transform_transform_bounds()'ing bounds.
2019-12-19 11:30:08 +00:00
float c, s;
_sincos (self->angle, &s, &c);
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
transform: Split rotate() and rotate3d() class This is mainly for accuracy: We can guarantee the math we do for 2D rotations results in a 2D matrix.
2019-03-02 02:21:09 +00:00
graphene_matrix_init_from_2d (out_matrix,
c, s,
-s, c,
0, 0);
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
}
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
static void
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
gsk_rotate_transform_apply_2d (GskTransform *transform,
float *out_xx,
float *out_yx,
float *out_xy,
float *out_yy,
float *out_dx,
float *out_dy)
{
transform: Split rotate() and rotate3d() class This is mainly for accuracy: We can guarantee the math we do for 2D rotations results in a 2D matrix.
2019-03-02 02:21:09 +00:00
GskRotateTransform *self = (GskRotateTransform *) transform;
GskTransform: Compute sin/deg of 90deg rotations manually Otherwise we might end up with inaccuracies, e.g. when gsk_transform_transform_bounds()'ing bounds.
2019-12-19 11:30:08 +00:00
float s, c, xx, xy, yx, yy;
transform: Split rotate() and rotate3d() class This is mainly for accuracy: We can guarantee the math we do for 2D rotations results in a 2D matrix.
2019-03-02 02:21:09 +00:00
GskTransform: Compute sin/deg of 90deg rotations manually Otherwise we might end up with inaccuracies, e.g. when gsk_transform_transform_bounds()'ing bounds.
2019-12-19 11:30:08 +00:00
_sincos (self->angle, &s, &c);
transform: Split rotate() and rotate3d() class This is mainly for accuracy: We can guarantee the math we do for 2D rotations results in a 2D matrix.
2019-03-02 02:21:09 +00:00
xx = c * *out_xx + s * *out_xy;
yx = c * *out_yx + s * *out_yy;
xy = -s * *out_xx + c * *out_xy;
yy = -s * *out_yx + c * *out_yy;
*out_xx = xx;
*out_yx = yx;
*out_xy = xy;
*out_yy = yy;
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
}
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
static GskTransform *
gsk_rotate_transform_apply (GskTransform *transform,
GskTransform *apply_to)
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
{
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
GskRotateTransform *self = (GskRotateTransform *) transform;
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
transform: Split rotate() and rotate3d() class This is mainly for accuracy: We can guarantee the math we do for 2D rotations results in a 2D matrix.
2019-03-02 02:21:09 +00:00
return gsk_transform_rotate (apply_to, self->angle);
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
}
transform: Implement gsk_transform_invert() And use it. And test it.
2019-03-03 16:57:11 +00:00
static GskTransform *
gsk_rotate_transform_invert (GskTransform *transform,
GskTransform *next)
{
GskRotateTransform *self = (GskRotateTransform *) transform;
return gsk_transform_rotate (next, - self->angle);
}
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
static gboolean
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_rotate_transform_equal (GskTransform *first_transform,
GskTransform *second_transform)
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
{
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
GskRotateTransform *first = (GskRotateTransform *) first_transform;
GskRotateTransform *second = (GskRotateTransform *) second_transform;
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
Never compare floats for equality Floating point values cannot ever be compared for equality. GLib has a G_APPROX_VALUE macro that lets us compare two value within a provided precision, so we should use that instead.
2019-03-22 12:39:28 +00:00
return G_APPROX_VALUE (first->angle, second->angle, 0.01f);
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
}
static void
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_rotate_transform_print (GskTransform *transform,
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
GString *string)
{
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
GskRotateTransform *self = (GskRotateTransform *) transform;
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
transform: Split rotate() and rotate3d() class This is mainly for accuracy: We can guarantee the math we do for 2D rotations results in a 2D matrix.
2019-03-02 02:21:09 +00:00
g_string_append (string, "rotate(");
string_append_double (string, self->angle);
g_string_append (string, ")");
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
}
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
static const GskTransformClass GSK_ROTATE_TRANSFORM_CLASS =
{
sizeof (GskRotateTransform),
"GskRotateTransform",
gsk_rotate_transform_finalize,
gsk_rotate_transform_to_matrix,
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
gsk_rotate_transform_apply_2d,
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
NULL,
NULL,
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_rotate_transform_print,
gsk_rotate_transform_apply,
transform: Implement gsk_transform_invert() And use it. And test it.
2019-03-03 16:57:11 +00:00
gsk_rotate_transform_invert,
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_rotate_transform_equal,
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
};
GskTransform: Normalize angles Make sure all angles are in the [0..365) range
2019-12-19 11:29:33 +00:00
static inline float
normalize_angle (float angle)
{
if (angle >= 0 && angle < 360)
return angle;
gsk: Fix angle normalization I was getting assertions that normalize_angle() failed the result < 260 check. Doing some research on this it turns out to be a precision issue. If the incomming angle is very slightly below zero, then adding 360 to it may end up with exactly 360. I simplified the code a bit to avoid division and rounding, because in practice most angles will be "just outside" the 0-360 degree anyway. And i also added a workaround for the "result is 360" case by just setting it to 0.
2020-02-11 09:00:34 +00:00
while (angle >= 360)
angle -= 360;
while (angle < 0)
angle += 360;
GskTransform: Normalize angles Make sure all angles are in the [0..365) range
2019-12-19 11:29:33 +00:00
gsk: Fix angle normalization I was getting assertions that normalize_angle() failed the result < 260 check. Doing some research on this it turns out to be a precision issue. If the incomming angle is very slightly below zero, then adding 360 to it may end up with exactly 360. I simplified the code a bit to avoid division and rounding, because in practice most angles will be "just outside" the 0-360 degree anyway. And i also added a workaround for the "result is 360" case by just setting it to 0.
2020-02-11 09:00:34 +00:00
/* Due to precision issues we may end up with a result that is just
* past the allowed range when rounded. For example, something like
* -epsilon + 360 when rounded to a float may end up with 360.
* So, we handle these cases by returning the exact value 0.
*/
if (angle >= 360)
angle = 0;
GskTransform: Normalize angles Make sure all angles are in the [0..365) range
2019-12-19 11:29:33 +00:00
gsk: Fix angle normalization I was getting assertions that normalize_angle() failed the result < 260 check. Doing some research on this it turns out to be a precision issue. If the incomming angle is very slightly below zero, then adding 360 to it may end up with exactly 360. I simplified the code a bit to avoid division and rounding, because in practice most angles will be "just outside" the 0-360 degree anyway. And i also added a workaround for the "result is 360" case by just setting it to 0.
2020-02-11 09:00:34 +00:00
g_assert (angle < 360.0);
g_assert (angle >= 0.0);
GskTransform: add assertions to make normalize_angle() intentions clear 360deg should be expressed as 0deg and no angle should ever be negative.
2020-02-02 12:20:40 +00:00
gsk: Fix angle normalization I was getting assertions that normalize_angle() failed the result < 260 check. Doing some research on this it turns out to be a precision issue. If the incomming angle is very slightly below zero, then adding 360 to it may end up with exactly 360. I simplified the code a bit to avoid division and rounding, because in practice most angles will be "just outside" the 0-360 degree anyway. And i also added a workaround for the "result is 360" case by just setting it to 0.
2020-02-11 09:00:34 +00:00
return angle;
GskTransform: Normalize angles Make sure all angles are in the [0..365) range
2019-12-19 11:29:33 +00:00
}
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
/**
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
* gsk_transform_rotate:
introspection: Stop using allow-none allow-none has been deprecated for a long time already. Instead use optional and nullable everywhere.
2021-05-19 11:24:34 +00:00
* @next: (nullable) (transfer full): the next transform
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
* @angle: the rotation angle, in degrees (clockwise)
*
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
* Rotates @next @angle degrees in 2D - or in 3D-speak, around the z axis.
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
*
transform: Fix docs to use "transform" instead of "matrix"
2020-06-13 06:25:51 +00:00
* Returns: The new transform
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
*/
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
GskTransform *
gsk_transform_rotate (GskTransform *next,
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
float angle)
{
transform: Coalesce similar transforms If somebody does a transform like scale(5) scale(10) translate(1,1) translate(5,0) store it instead as scale(50) translate(6,1) This way, less memory is consumed and transforms are easier to read. In particular, this simplifies the typical transforms we do in GTK, which are just one translation after another.
2019-06-02 12:16:45 +00:00
GskRotateTransform *result;
GskTransform: Ignore identity transforms We often end up with e.g. a scale of 1 or a translation by 0/0. Ignore those transforms since they don't do anything.
2019-06-03 14:20:12 +00:00
if (angle == 0.0f)
transform: Fix identity transform fast paths Returning an extra ref will cause leaks later.
2019-07-06 10:56:28 +00:00
return next;
GskTransform: Ignore identity transforms We often end up with e.g. a scale of 1 or a translation by 0/0. Ignore those transforms since they don't do anything.
2019-06-03 14:20:12 +00:00
transform: Coalesce similar transforms If somebody does a transform like scale(5) scale(10) translate(1,1) translate(5,0) store it instead as scale(50) translate(6,1) This way, less memory is consumed and transforms are easier to read. In particular, this simplifies the typical transforms we do in GTK, which are just one translation after another.
2019-06-02 12:16:45 +00:00
if (gsk_transform_has_class (next, &GSK_ROTATE_TRANSFORM_CLASS))
{
GskTransform *r = gsk_transform_rotate (gsk_transform_ref (next->next),
((GskRotateTransform *) next)->angle + angle);
gsk_transform_unref (next);
return r;
}
result = gsk_transform_alloc (&GSK_ROTATE_TRANSFORM_CLASS,
GSK_TRANSFORM_CATEGORY_2D,
next);
transform: Split rotate() and rotate3d() class This is mainly for accuracy: We can guarantee the math we do for 2D rotations results in a 2D matrix.
2019-03-02 02:21:09 +00:00
GskTransform: Normalize angles Make sure all angles are in the [0..365) range
2019-12-19 11:29:33 +00:00
result->angle = normalize_angle (angle);
transform: Split rotate() and rotate3d() class This is mainly for accuracy: We can guarantee the math we do for 2D rotations results in a 2D matrix.
2019-03-02 02:21:09 +00:00
return &result->parent;
}
/*** ROTATE 3D ***/
typedef struct _GskRotate3dTransform GskRotate3dTransform;
struct _GskRotate3dTransform
{
GskTransform parent;
float angle;
graphene_vec3_t axis;
};
static void
gsk_rotate3d_transform_finalize (GskTransform *self)
{
}
static void
gsk_rotate3d_transform_to_matrix (GskTransform *transform,
graphene_matrix_t *out_matrix)
{
GskRotate3dTransform *self = (GskRotate3dTransform *) transform;
graphene_matrix_init_rotate (out_matrix, self->angle, &self->axis);
}
static GskTransform *
gsk_rotate3d_transform_apply (GskTransform *transform,
GskTransform *apply_to)
{
GskRotate3dTransform *self = (GskRotate3dTransform *) transform;
return gsk_transform_rotate_3d (apply_to, self->angle, &self->axis);
}
transform: Implement gsk_transform_invert() And use it. And test it.
2019-03-03 16:57:11 +00:00
static GskTransform *
gsk_rotate3d_transform_invert (GskTransform *transform,
GskTransform *next)
{
GskRotate3dTransform *self = (GskRotate3dTransform *) transform;
return gsk_transform_rotate_3d (next, - self->angle, &self->axis);
}
transform: Split rotate() and rotate3d() class This is mainly for accuracy: We can guarantee the math we do for 2D rotations results in a 2D matrix.
2019-03-02 02:21:09 +00:00
static gboolean
gsk_rotate3d_transform_equal (GskTransform *first_transform,
GskTransform *second_transform)
{
GskRotate3dTransform *first = (GskRotate3dTransform *) first_transform;
GskRotate3dTransform *second = (GskRotate3dTransform *) second_transform;
Never compare floats for equality Floating point values cannot ever be compared for equality. GLib has a G_APPROX_VALUE macro that lets us compare two value within a provided precision, so we should use that instead.
2019-03-22 12:39:28 +00:00
return G_APPROX_VALUE (first->angle, second->angle, 0.01f) &&
graphene_vec3_equal (&first->axis, &second->axis);
transform: Split rotate() and rotate3d() class This is mainly for accuracy: We can guarantee the math we do for 2D rotations results in a 2D matrix.
2019-03-02 02:21:09 +00:00
}
static void
gsk_rotate3d_transform_print (GskTransform *transform,
GString *string)
{
GskRotate3dTransform *self = (GskRotate3dTransform *) transform;
float f[3];
guint i;
g_string_append (string, "rotate3d(");
graphene_vec3_to_float (&self->axis, f);
for (i = 0; i < 3; i++)
{
string_append_double (string, f[i]);
g_string_append (string, ", ");
}
string_append_double (string, self->angle);
g_string_append (string, ")");
}
static const GskTransformClass GSK_ROTATE3D_TRANSFORM_CLASS =
{
sizeof (GskRotate3dTransform),
"GskRotate3dTransform",
gsk_rotate3d_transform_finalize,
gsk_rotate3d_transform_to_matrix,
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
NULL,
NULL,
NULL,
transform: Split rotate() and rotate3d() class This is mainly for accuracy: We can guarantee the math we do for 2D rotations results in a 2D matrix.
2019-03-02 02:21:09 +00:00
gsk_rotate3d_transform_print,
gsk_rotate3d_transform_apply,
transform: Implement gsk_transform_invert() And use it. And test it.
2019-03-03 16:57:11 +00:00
gsk_rotate3d_transform_invert,
transform: Split rotate() and rotate3d() class This is mainly for accuracy: We can guarantee the math we do for 2D rotations results in a 2D matrix.
2019-03-02 02:21:09 +00:00
gsk_rotate3d_transform_equal,
};
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
/**
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
* gsk_transform_rotate_3d:
introspection: Stop using allow-none allow-none has been deprecated for a long time already. Instead use optional and nullable everywhere.
2021-05-19 11:24:34 +00:00
* @next: (nullable) (transfer full): the next transform
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
* @angle: the rotation angle, in degrees (clockwise)
* @axis: The rotation axis
*
* Rotates @next @angle degrees around @axis.
*
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
* For a rotation in 2D space, use [method@Gsk.Transform.rotate]
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
*
transform: Fix docs to use "transform" instead of "matrix"
2020-06-13 06:25:51 +00:00
* Returns: The new transform
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
*/
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
GskTransform *
gsk_transform_rotate_3d (GskTransform *next,
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
float angle,
const graphene_vec3_t *axis)
{
transform: Split rotate() and rotate3d() class This is mainly for accuracy: We can guarantee the math we do for 2D rotations results in a 2D matrix.
2019-03-02 02:21:09 +00:00
GskRotate3dTransform *result;
GskTransform: Ignore identity transforms We often end up with e.g. a scale of 1 or a translation by 0/0. Ignore those transforms since they don't do anything.
2019-06-03 14:20:12 +00:00
transform: Split rotate() and rotate3d() class This is mainly for accuracy: We can guarantee the math we do for 2D rotations results in a 2D matrix.
2019-03-02 02:21:09 +00:00
if (graphene_vec3_get_x (axis) == 0.0 && graphene_vec3_get_y (axis) == 0.0)
return gsk_transform_rotate (next, angle);
GskTransform: Ignore identity transforms We often end up with e.g. a scale of 1 or a translation by 0/0. Ignore those transforms since they don't do anything.
2019-06-03 14:20:12 +00:00
if (angle == 0.0f)
transform: Fix identity transform fast paths Returning an extra ref will cause leaks later.
2019-07-06 10:56:28 +00:00
return next;
GskTransform: Ignore identity transforms We often end up with e.g. a scale of 1 or a translation by 0/0. Ignore those transforms since they don't do anything.
2019-06-03 14:20:12 +00:00
transform: Store the category in the transform ... instead of computing it every time we need it. This should be faster and we want to use it a lot more prominently. Also, we have the struct memory available anyway.
2019-03-04 08:20:58 +00:00
result = gsk_transform_alloc (&GSK_ROTATE3D_TRANSFORM_CLASS,
GSK_TRANSFORM_CATEGORY_3D,
next);
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
GskTransform: Normalize angles Make sure all angles are in the [0..365) range
2019-12-19 11:29:33 +00:00
result->angle = normalize_angle (angle);
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
graphene_vec3_init_from_vec3 (&result->axis, axis);
return &result->parent;
}
/*** SCALE ***/
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
typedef struct _GskScaleTransform GskScaleTransform;
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
struct _GskScaleTransform
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
{
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
GskTransform parent;
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
float factor_x;
float factor_y;
float factor_z;
};
static void
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_scale_transform_finalize (GskTransform *self)
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
{
}
static void
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_scale_transform_to_matrix (GskTransform *transform,
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
graphene_matrix_t *out_matrix)
{
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
GskScaleTransform *self = (GskScaleTransform *) transform;
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
graphene_matrix_init_scale (out_matrix, self->factor_x, self->factor_y, self->factor_z);
}
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
static void
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
gsk_scale_transform_apply_2d (GskTransform *transform,
float *out_xx,
float *out_yx,
float *out_xy,
float *out_yy,
float *out_dx,
float *out_dy)
{
GskScaleTransform *self = (GskScaleTransform *) transform;
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
g_assert (self->factor_z == 1.0);
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
*out_xx *= self->factor_x;
*out_yx *= self->factor_x;
*out_xy *= self->factor_y;
*out_yy *= self->factor_y;
}
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
static void
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_scale_transform_apply_affine (GskTransform *transform,
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
float *out_scale_x,
float *out_scale_y,
float *out_dx,
float *out_dy)
{
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
GskScaleTransform *self = (GskScaleTransform *) transform;
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
g_assert (self->factor_z == 1.0);
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
*out_scale_x *= self->factor_x;
*out_scale_y *= self->factor_y;
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
}
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
static GskTransform *
gsk_scale_transform_apply (GskTransform *transform,
GskTransform *apply_to)
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
{
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
GskScaleTransform *self = (GskScaleTransform *) transform;
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
return gsk_transform_scale_3d (apply_to, self->factor_x, self->factor_y, self->factor_z);
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
}
transform: Implement gsk_transform_invert() And use it. And test it.
2019-03-03 16:57:11 +00:00
static GskTransform *
gsk_scale_transform_invert (GskTransform *transform,
GskTransform *next)
{
GskScaleTransform *self = (GskScaleTransform *) transform;
return gsk_transform_scale_3d (next,
1.f / self->factor_x,
1.f / self->factor_y,
1.f / self->factor_z);
}
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
static gboolean
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_scale_transform_equal (GskTransform *first_transform,
GskTransform *second_transform)
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
{
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
GskScaleTransform *first = (GskScaleTransform *) first_transform;
GskScaleTransform *second = (GskScaleTransform *) second_transform;
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
transform: Properly compare scale transforms If the epsilon we use there is too much, we can run into rendering errors because the GPU will use the actual scale value.
2020-01-04 07:54:58 +00:00
return G_APPROX_VALUE (first->factor_x, second->factor_x, FLT_EPSILON) &&
G_APPROX_VALUE (first->factor_y, second->factor_y, FLT_EPSILON) &&
G_APPROX_VALUE (first->factor_z, second->factor_z, FLT_EPSILON);
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
}
static void
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_scale_transform_print (GskTransform *transform,
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
GString *string)
{
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
GskScaleTransform *self = (GskScaleTransform *) transform;
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
if (self->factor_z == 1.0)
{
g_string_append (string, "scale(");
string_append_double (string, self->factor_x);
if (self->factor_x != self->factor_y)
{
g_string_append (string, ", ");
string_append_double (string, self->factor_y);
}
g_string_append (string, ")");
}
else
{
g_string_append (string, "scale3d(");
string_append_double (string, self->factor_x);
g_string_append (string, ", ");
string_append_double (string, self->factor_y);
g_string_append (string, ", ");
string_append_double (string, self->factor_z);
g_string_append (string, ")");
}
}
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
static const GskTransformClass GSK_SCALE_TRANSFORM_CLASS =
{
sizeof (GskScaleTransform),
"GskScaleTransform",
gsk_scale_transform_finalize,
gsk_scale_transform_to_matrix,
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
gsk_scale_transform_apply_2d,
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_scale_transform_apply_affine,
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
NULL,
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_scale_transform_print,
gsk_scale_transform_apply,
transform: Implement gsk_transform_invert() And use it. And test it.
2019-03-03 16:57:11 +00:00
gsk_scale_transform_invert,
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_scale_transform_equal,
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
};
/**
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
* gsk_transform_scale:
introspection: Stop using allow-none allow-none has been deprecated for a long time already. Instead use optional and nullable everywhere.
2021-05-19 11:24:34 +00:00
* @next: (nullable) (transfer full): the next transform
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
* @factor_x: scaling factor on the X axis
* @factor_y: scaling factor on the Y axis
*
* Scales @next in 2-dimensional space by the given factors.
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
*
* Use [method@Gsk.Transform.scale_3d] to scale in all 3 dimensions.
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
*
transform: Fix docs to use "transform" instead of "matrix"
2020-06-13 06:25:51 +00:00
* Returns: The new transform
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
**/
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
GskTransform *
gsk_transform_scale (GskTransform *next,
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
float factor_x,
float factor_y)
{
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
return gsk_transform_scale_3d (next, factor_x, factor_y, 1.0);
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
}
/**
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
* gsk_transform_scale_3d:
introspection: Stop using allow-none allow-none has been deprecated for a long time already. Instead use optional and nullable everywhere.
2021-05-19 11:24:34 +00:00
* @next: (nullable) (transfer full): the next transform
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
* @factor_x: scaling factor on the X axis
* @factor_y: scaling factor on the Y axis
* @factor_z: scaling factor on the Z axis
*
* Scales @next by the given factors.
*
transform: Fix docs to use "transform" instead of "matrix"
2020-06-13 06:25:51 +00:00
* Returns: The new transform
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
**/
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
GskTransform *
gsk_transform_scale_3d (GskTransform *next,
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
float factor_x,
float factor_y,
float factor_z)
{
transform: Store the category in the transform ... instead of computing it every time we need it. This should be faster and we want to use it a lot more prominently. Also, we have the struct memory available anyway.
2019-03-04 08:20:58 +00:00
GskScaleTransform *result;
GskTransform: Ignore identity transforms We often end up with e.g. a scale of 1 or a translation by 0/0. Ignore those transforms since they don't do anything.
2019-06-03 14:20:12 +00:00
if (factor_x == 1 && factor_y == 1 && factor_z == 1)
transform: Fix identity transform fast paths Returning an extra ref will cause leaks later.
2019-07-06 10:56:28 +00:00
return next;
GskTransform: Ignore identity transforms We often end up with e.g. a scale of 1 or a translation by 0/0. Ignore those transforms since they don't do anything.
2019-06-03 14:20:12 +00:00
transform: Coalesce similar transforms If somebody does a transform like scale(5) scale(10) translate(1,1) translate(5,0) store it instead as scale(50) translate(6,1) This way, less memory is consumed and transforms are easier to read. In particular, this simplifies the typical transforms we do in GTK, which are just one translation after another.
2019-06-02 12:16:45 +00:00
if (gsk_transform_has_class (next, &GSK_SCALE_TRANSFORM_CLASS))
{
GskScaleTransform *scale = (GskScaleTransform *) next;
GskTransform *r = gsk_transform_scale_3d (gsk_transform_ref (next->next),
scale->factor_x * factor_x,
scale->factor_y * factor_y,
scale->factor_z * factor_z);
gsk_transform_unref (next);
return r;
}
transform: Store the category in the transform ... instead of computing it every time we need it. This should be faster and we want to use it a lot more prominently. Also, we have the struct memory available anyway.
2019-03-04 08:20:58 +00:00
result = gsk_transform_alloc (&GSK_SCALE_TRANSFORM_CLASS,
factor_z != 1.0 ? GSK_TRANSFORM_CATEGORY_3D
: GSK_TRANSFORM_CATEGORY_2D_AFFINE,
next);
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
result->factor_x = factor_x;
result->factor_y = factor_y;
result->factor_z = factor_z;
return &result->parent;
}
transform: Add perspective() This commit adds gsk_transform_perspective(), gtk_snapshot_perspective() and support for perspective() in the CSS syntax.
2019-03-05 08:27:30 +00:00
/*** PERSPECTIVE ***/
typedef struct _GskPerspectiveTransform GskPerspectiveTransform;
struct _GskPerspectiveTransform
{
GskTransform parent;
float depth;
};
static void
gsk_perspective_transform_finalize (GskTransform *self)
{
}
static void
gsk_perspective_transform_to_matrix (GskTransform *transform,
graphene_matrix_t *out_matrix)
{
GskPerspectiveTransform *self = (GskPerspectiveTransform *) transform;
float f[16] = { 1.f, 0.f, 0.f, 0.f,
0.f, 1.f, 0.f, 0.f,
transform: Treat 0 perspective transforms as identity This happens if you apply a perspective transform and its inverse, which is the negative version of the perspective.
2019-06-02 13:06:50 +00:00
0.f, 0.f, 1.f, self->depth ? -1.f / self->depth : 0.f,
transform: Add perspective() This commit adds gsk_transform_perspective(), gtk_snapshot_perspective() and support for perspective() in the CSS syntax.
2019-03-05 08:27:30 +00:00
0.f, 0.f, 0.f, 1.f };
graphene_matrix_init_from_float (out_matrix, f);
}
static GskTransform *
gsk_perspective_transform_apply (GskTransform *transform,
GskTransform *apply_to)
{
GskPerspectiveTransform *self = (GskPerspectiveTransform *) transform;
return gsk_transform_perspective (apply_to, self->depth);
}
static GskTransform *
gsk_perspective_transform_invert (GskTransform *transform,
GskTransform *next)
{
GskPerspectiveTransform *self = (GskPerspectiveTransform *) transform;
return gsk_transform_perspective (next, - self->depth);
}
static gboolean
gsk_perspective_transform_equal (GskTransform *first_transform,
GskTransform *second_transform)
{
GskPerspectiveTransform *first = (GskPerspectiveTransform *) first_transform;
GskPerspectiveTransform *second = (GskPerspectiveTransform *) second_transform;
Never compare floats for equality Floating point values cannot ever be compared for equality. GLib has a G_APPROX_VALUE macro that lets us compare two value within a provided precision, so we should use that instead.
2019-03-22 12:39:28 +00:00
return G_APPROX_VALUE (first->depth, second->depth, 0.001f);
transform: Add perspective() This commit adds gsk_transform_perspective(), gtk_snapshot_perspective() and support for perspective() in the CSS syntax.
2019-03-05 08:27:30 +00:00
}
static void
gsk_perspective_transform_print (GskTransform *transform,
GString *string)
{
GskPerspectiveTransform *self = (GskPerspectiveTransform *) transform;
g_string_append (string, "perspective(");
string_append_double (string, self->depth);
g_string_append (string, ")");
}
static const GskTransformClass GSK_PERSPECTIVE_TRANSFORM_CLASS =
{
sizeof (GskPerspectiveTransform),
"GskPerspectiveTransform",
gsk_perspective_transform_finalize,
gsk_perspective_transform_to_matrix,
NULL,
NULL,
NULL,
gsk_perspective_transform_print,
gsk_perspective_transform_apply,
gsk_perspective_transform_invert,
gsk_perspective_transform_equal,
};
/**
* gsk_transform_perspective:
introspection: Stop using allow-none allow-none has been deprecated for a long time already. Instead use optional and nullable everywhere.
2021-05-19 11:24:34 +00:00
* @next: (nullable) (transfer full): the next transform
transform: Add perspective() This commit adds gsk_transform_perspective(), gtk_snapshot_perspective() and support for perspective() in the CSS syntax.
2019-03-05 08:27:30 +00:00
* @depth: distance of the z=0 plane. Lower values give a more
Documentation fixes Mostly fixing up indentation of continuation lines, and other small cleanups.
2021-05-18 21:05:26 +00:00
* flattened pyramid and therefore a more pronounced
* perspective effect.
transform: Add perspective() This commit adds gsk_transform_perspective(), gtk_snapshot_perspective() and support for perspective() in the CSS syntax.
2019-03-05 08:27:30 +00:00
*
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
* Applies a perspective projection transform.
*
* This transform scales points in X and Y based on their Z value,
* scaling points with positive Z values away from the origin, and
transform: Add perspective() This commit adds gsk_transform_perspective(), gtk_snapshot_perspective() and support for perspective() in the CSS syntax.
2019-03-05 08:27:30 +00:00
* those with negative Z values towards the origin. Points
* on the z=0 plane are unchanged.
*
transform: Fix docs to use "transform" instead of "matrix"
2020-06-13 06:25:51 +00:00
* Returns: The new transform
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
*/
transform: Add perspective() This commit adds gsk_transform_perspective(), gtk_snapshot_perspective() and support for perspective() in the CSS syntax.
2019-03-05 08:27:30 +00:00
GskTransform *
gsk_transform_perspective (GskTransform *next,
float depth)
{
GskPerspectiveTransform *result;
transform: Fix identity transform fast paths Returning an extra ref will cause leaks later.
2019-07-06 10:56:28 +00:00
transform: Coalesce similar transforms If somebody does a transform like scale(5) scale(10) translate(1,1) translate(5,0) store it instead as scale(50) translate(6,1) This way, less memory is consumed and transforms are easier to read. In particular, this simplifies the typical transforms we do in GTK, which are just one translation after another.
2019-06-02 12:16:45 +00:00
if (gsk_transform_has_class (next, &GSK_PERSPECTIVE_TRANSFORM_CLASS))
{
GskTransform *r = gsk_transform_perspective (gsk_transform_ref (next->next),
((GskPerspectiveTransform *) next)->depth + depth);
gsk_transform_unref (next);
return r;
}
transform: Add perspective() This commit adds gsk_transform_perspective(), gtk_snapshot_perspective() and support for perspective() in the CSS syntax.
2019-03-05 08:27:30 +00:00
result = gsk_transform_alloc (&GSK_PERSPECTIVE_TRANSFORM_CLASS,
GSK_TRANSFORM_CATEGORY_ANY,
next);
result->depth = depth;
return &result->parent;
}
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
/*** PUBLIC API ***/
static void
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_transform_finalize (GskTransform *self)
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
{
self->transform_class->finalize (self);
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_transform_unref (self->next);
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
}
/**
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
* gsk_transform_ref:
introspection: Stop using allow-none allow-none has been deprecated for a long time already. Instead use optional and nullable everywhere.
2021-05-19 11:24:34 +00:00
* @self: (nullable): a `GskTransform`
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
*
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
* Acquires a reference on the given `GskTransform`.
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
*
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
* Returns: (transfer none): the `GskTransform` with an additional reference
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
*/
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
GskTransform *
gsk_transform_ref (GskTransform *self)
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
{
if (self == NULL)
return NULL;
Use atomic boxing instead of manual refcounting Artisanal, homegrown, locally sourced, vegan reference counting has been replaced by the appropriate API in GLib, which does small things like saturation and type checking.
2019-03-22 11:45:57 +00:00
return g_atomic_rc_box_acquire (self);
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
}
/**
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
* gsk_transform_unref:
introspection: Stop using allow-none allow-none has been deprecated for a long time already. Instead use optional and nullable everywhere.
2021-05-19 11:24:34 +00:00
* @self: (nullable): a `GskTransform`
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
*
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
* Releases a reference on the given `GskTransform`.
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
*
* If the reference was the last, the resources associated to the @self are
* freed.
*/
void
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_transform_unref (GskTransform *self)
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
{
if (self == NULL)
return;
Use atomic boxing instead of manual refcounting Artisanal, homegrown, locally sourced, vegan reference counting has been replaced by the appropriate API in GLib, which does small things like saturation and type checking.
2019-03-22 11:45:57 +00:00
g_atomic_rc_box_release_full (self, (GDestroyNotify) gsk_transform_finalize);
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
}
/**
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
* gsk_transform_print:
introspection: Stop using allow-none allow-none has been deprecated for a long time already. Instead use optional and nullable everywhere.
2021-05-19 11:24:34 +00:00
* @self: (nullable): a `GskTransform`
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
* @string: The string to print into
*
transform: Add gsk_transform_parse() It uses the new CSS parser.
2019-03-21 04:52:41 +00:00
* Converts @self into a human-readable string representation suitable
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
* for printing.
*
* The result of this function can later be parsed with
* [func@Gsk.Transform.parse].
*/
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
void
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_transform_print (GskTransform *self,
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
GString *string)
{
g_return_if_fail (string != NULL);
if (self == NULL)
{
g_string_append (string, "none");
return;
}
if (self->next != NULL)
{
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_transform_print (self->next, string);
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
g_string_append (string, " ");
}
self->transform_class->print (self, string);
}
/**
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
* gsk_transform_to_string:
gsk: Clean up docs syntax Replace leftover gtk-doc syntax (#Type) with backquotes.
2021-05-20 03:40:54 +00:00
* @self: (nullable): a `GskTransform`
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
*
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
* Converts a matrix into a string that is suitable for printing.
*
* The resulting string can be parsed with [func@Gsk.Transform.parse].
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
*
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
* This is a wrapper around [method@Gsk.Transform.print].
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
*
* Returns: A new string for @self
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
*/
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
char *
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_transform_to_string (GskTransform *self)
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
{
GString *string;
string = g_string_new ("");
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_transform_print (self, string);
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
return g_string_free (string, FALSE);
}
/**
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
* gsk_transform_to_matrix:
introspection: Stop using allow-none allow-none has been deprecated for a long time already. Instead use optional and nullable everywhere.
2021-05-19 11:24:34 +00:00
* @self: (nullable): a `GskTransform`
transform: Fix a wrong annotation It is (out caller-allocates), not (out) (caller-allocates).
2019-02-24 04:21:32 +00:00
* @out_matrix: (out caller-allocates): The matrix to set
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
*
* Computes the actual value of @self and stores it in @out_matrix.
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
*
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
* The previous value of @out_matrix will be ignored.
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
*/
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
void
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_transform_to_matrix (GskTransform *self,
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
graphene_matrix_t *out_matrix)
{
graphene_matrix_t m;
if (self == NULL)
{
graphene_matrix_init_identity (out_matrix);
return;
}
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_transform_to_matrix (self->next, out_matrix);
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
self->transform_class->to_matrix (self, &m);
graphene_matrix_multiply (&m, out_matrix, out_matrix);
}
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
/**
* gsk_transform_to_2d:
gsk: Clean up docs syntax Replace leftover gtk-doc syntax (#Type) with backquotes.
2021-05-20 03:40:54 +00:00
* @self: a 2D `GskTransform`
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
* @out_xx: (out): return location for the xx member
* @out_yx: (out): return location for the yx member
* @out_xy: (out): return location for the xy member
* @out_yy: (out): return location for the yy member
* @out_dx: (out): return location for the x0 member
* @out_dy: (out): return location for the y0 member
*
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
* Converts a `GskTransform` to a 2D transformation matrix.
*
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
* @self must be a 2D transformation. If you are not
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
* sure, use gsk_transform_get_category() >=
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
* %GSK_TRANSFORM_CATEGORY_2D to check.
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
*
* The returned values have the following layout:
*
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
* ```
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
* | xx yx | | a b 0 |
* | xy yy | = | c d 0 |
Cosmetics Fix up a variable name mismatch in GskTransform docs.
2019-06-03 16:15:16 +00:00
* | dx dy | | tx ty 1 |
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
* ```
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
*
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
* This function can be used to convert between a `GskTransform`
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
* and a matrix type from other 2D drawing libraries, in particular
* Cairo.
*/
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
void
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
gsk_transform_to_2d (GskTransform *self,
float *out_xx,
float *out_yx,
float *out_xy,
float *out_yy,
float *out_dx,
float *out_dy)
{
gsk: Optimize gsk_transform_to... Optimize these functions for the common cases.
2021-07-24 14:51:08 +00:00
*out_xx = 1.0f;
*out_yx = 0.0f;
*out_xy = 0.0f;
*out_yy = 1.0f;
*out_dx = 0.0f;
*out_dy = 0.0f;
if (self == NULL)
return;
if (G_UNLIKELY (self->category < GSK_TRANSFORM_CATEGORY_2D))
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
{
gsk: Optimize gsk_transform_to... Optimize these functions for the common cases.
2021-07-24 14:51:08 +00:00
char *s = gsk_transform_to_string (self);
g_warning ("Given transform \"%s\" is not a 2D transform.", s);
g_free (s);
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
return;
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
}
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
gsk_transform_to_2d (self->next,
out_xx, out_yx,
out_xy, out_yy,
out_dx, out_dy);
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
self->transform_class->apply_2d (self,
out_xx, out_yx,
out_xy, out_yy,
out_dx, out_dy);
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
}
/**
* gsk_transform_to_affine:
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
* @self: a `GskTransform`
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
* @out_scale_x: (out): return location for the scale
Documentation fixes Mostly fixing up indentation of continuation lines, and other small cleanups.
2021-05-18 21:05:26 +00:00
* factor in the x direction
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
* @out_scale_y: (out): return location for the scale
Documentation fixes Mostly fixing up indentation of continuation lines, and other small cleanups.
2021-05-18 21:05:26 +00:00
* factor in the y direction
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
* @out_dx: (out): return location for the translation
Documentation fixes Mostly fixing up indentation of continuation lines, and other small cleanups.
2021-05-18 21:05:26 +00:00
* in the x direction
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
* @out_dy: (out): return location for the translation
Documentation fixes Mostly fixing up indentation of continuation lines, and other small cleanups.
2021-05-18 21:05:26 +00:00
* in the y direction
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
*
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
* Converts a `GskTransform` to 2D affine transformation factors.
*
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
* @self must be a 2D transformation. If you are not
Documentation fixes Mostly fixing up indentation of continuation lines, and other small cleanups.
2021-05-18 21:05:26 +00:00
* sure, use
*
* gsk_transform_get_category() >= %GSK_TRANSFORM_CATEGORY_2D_AFFINE
*
* to check.
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
*/
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
void
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_transform_to_affine (GskTransform *self,
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
float *out_scale_x,
float *out_scale_y,
float *out_dx,
float *out_dy)
{
gsk: Optimize gsk_transform_to... Optimize these functions for the common cases.
2021-07-24 14:51:08 +00:00
*out_scale_x = 1.0f;
*out_scale_y = 1.0f;
*out_dx = 0.0f;
*out_dy = 0.0f;
GskTransform: Restructure _to_affine and _to_translation Make the simple cases more explicit.
2020-07-24 16:33:33 +00:00
if (self == NULL)
gsk: Optimize gsk_transform_to... Optimize these functions for the common cases.
2021-07-24 14:51:08 +00:00
return;
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
GskTransform: Restructure _to_affine and _to_translation Make the simple cases more explicit.
2020-07-24 16:33:33 +00:00
if (G_UNLIKELY (self->category < GSK_TRANSFORM_CATEGORY_2D_AFFINE))
{
char *s = gsk_transform_to_string (self);
g_warning ("Given transform \"%s\" is not an affine 2D transform.", s);
g_free (s);
return;
}
gsk: Optimize gsk_transform_to... Optimize these functions for the common cases.
2021-07-24 14:51:08 +00:00
gsk_transform_to_affine (self->next,
out_scale_x, out_scale_y,
out_dx, out_dy);
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
self->transform_class->apply_affine (self,
out_scale_x, out_scale_y,
out_dx, out_dy);
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
}
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
/**
* gsk_transform_to_translate:
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
* @self: a `GskTransform`
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
* @out_dx: (out): return location for the translation
Documentation fixes Mostly fixing up indentation of continuation lines, and other small cleanups.
2021-05-18 21:05:26 +00:00
* in the x direction
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
* @out_dy: (out): return location for the translation
Documentation fixes Mostly fixing up indentation of continuation lines, and other small cleanups.
2021-05-18 21:05:26 +00:00
* in the y direction
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
*
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
* Converts a `GskTransform` to a translation operation.
*
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
* @self must be a 2D transformation. If you are not
Documentation fixes Mostly fixing up indentation of continuation lines, and other small cleanups.
2021-05-18 21:05:26 +00:00
* sure, use
*
* gsk_transform_get_category() >= %GSK_TRANSFORM_CATEGORY_2D_TRANSLATE
*
* to check.
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
*/
transform: Redo querying API Make the API expect a tranform of the proper category instead of doing the check ourselves and returning TRUE/FALSE. The benefit is that the mai use case is switch (transform->category) statements and in those we know the category and don't need to check TRUE/FALSE. Using the wrong matrix will now cause a g_warning().
2019-03-04 18:33:04 +00:00
void
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
gsk_transform_to_translate (GskTransform *self,
float *out_dx,
float *out_dy)
{
gsk: Optimize gsk_transform_to... Optimize these functions for the common cases.
2021-07-24 14:51:08 +00:00
*out_dx = 0.0f;
*out_dy = 0.0f;
GskTransform: Restructure _to_affine and _to_translation Make the simple cases more explicit.
2020-07-24 16:33:33 +00:00
if (self == NULL)
gsk: Optimize gsk_transform_to... Optimize these functions for the common cases.
2021-07-24 14:51:08 +00:00
return;
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
GskTransform: Restructure _to_affine and _to_translation Make the simple cases more explicit.
2020-07-24 16:33:33 +00:00
if (G_UNLIKELY (self->category < GSK_TRANSFORM_CATEGORY_2D_TRANSLATE))
{
char *s = gsk_transform_to_string (self);
g_warning ("Given transform \"%s\" is not an affine 2D translation.", s);
g_free (s);
return;
}
gsk: Optimize gsk_transform_to... Optimize these functions for the common cases.
2021-07-24 14:51:08 +00:00
gsk_transform_to_translate (self->next, out_dx, out_dy);
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
gsk: Optimize gsk_transform_to... Optimize these functions for the common cases.
2021-07-24 14:51:08 +00:00
self->transform_class->apply_translate (self, out_dx, out_dy);
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
}
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
/**
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
* gsk_transform_transform:
introspection: Stop using allow-none allow-none has been deprecated for a long time already. Instead use optional and nullable everywhere.
2021-05-19 11:24:34 +00:00
* @next: (nullable) (transfer full): Transform to apply @other to
* @other: (nullable): Transform to apply
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
*
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
* Applies all the operations from @other to @next.
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
*
transform: Fix docs to use "transform" instead of "matrix"
2020-06-13 06:25:51 +00:00
* Returns: The new transform
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
*/
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
GskTransform *
gsk_transform_transform (GskTransform *next,
GskTransform *other)
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
{
if (other == NULL)
return next;
gsktransform: Make next == NULL case explicit This is handled with the is_identity() check as well but this way it's much more obvious what's happening
2020-06-06 07:05:49 +00:00
if (next == NULL)
return gsk_transform_ref (other);
transform: Add optimization for common case Transforming identity by an other transform does not mean we need to painstakingly apply the individual steps of other to construct a new transform, it means we can just return other. Or in math terms: I * B = B so just return B.
2020-02-19 23:02:19 +00:00
if (gsk_transform_is_identity (next))
{
transform: Don't crash for gsk_transform_transform (id, id) See attached tests
2020-02-21 17:30:13 +00:00
/* ref before unref to avoid catastrophe when other == next */
other = gsk_transform_ref (other);
transform: Add optimization for common case Transforming identity by an other transform does not mean we need to painstakingly apply the individual steps of other to construct a new transform, it means we can just return other. Or in math terms: I * B = B so just return B.
2020-02-19 23:02:19 +00:00
gsk_transform_unref (next);
transform: Don't crash for gsk_transform_transform (id, id) See attached tests
2020-02-21 17:30:13 +00:00
return other;
transform: Add optimization for common case Transforming identity by an other transform does not mean we need to painstakingly apply the individual steps of other to construct a new transform, it means we can just return other. Or in math terms: I * B = B so just return B.
2020-02-19 23:02:19 +00:00
}
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
next = gsk_transform_transform (next, other->next);
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
return other->transform_class->apply (other, next);
}
transform: Implement gsk_transform_invert() And use it. And test it.
2019-03-03 16:57:11 +00:00
/**
* gsk_transform_invert:
introspection: Stop using allow-none allow-none has been deprecated for a long time already. Instead use optional and nullable everywhere.
2021-05-19 11:24:34 +00:00
* @self: (nullable) (transfer full): Transform to invert
transform: Implement gsk_transform_invert() And use it. And test it.
2019-03-03 16:57:11 +00:00
*
* Inverts the given transform.
*
* If @self is not invertible, %NULL is returned.
* Note that inverting %NULL also returns %NULL, which is
* the correct inverse of %NULL. If you need to differentiate
* between those cases, you should check @self is not %NULL
* before calling this function.
*
Documentation fixes Mostly fixing up indentation of continuation lines, and other small cleanups.
2021-05-18 21:05:26 +00:00
* Returns: (nullable): The inverted transform
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
*/
transform: Implement gsk_transform_invert() And use it. And test it.
2019-03-03 16:57:11 +00:00
GskTransform *
gsk_transform_invert (GskTransform *self)
{
GskTransform *result = NULL;
GskTransform *cur;
for (cur = self; cur; cur = cur->next)
{
result = cur->transform_class->invert (cur, result);
if (result == NULL)
break;
}
gsk_transform_unref (self);
return result;
}
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
/**
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
* gsk_transform_equal:
GskTransform: Add nullable annotations to _equals() Both can be NULL here.
2020-12-20 20:22:15 +00:00
* @first: (nullable): the first transform
* @second: (nullable): the second transform
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
*
transform: Fix equal() docs This has nothing to do with matrices.
2020-06-07 07:37:20 +00:00
* Checks two transforms for equality.
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
*
Documentation fixes Mostly fixing up indentation of continuation lines, and other small cleanups.
2021-05-18 21:05:26 +00:00
* Returns: %TRUE if the two transforms perform the same operation
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
*/
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
gboolean
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
gsk_transform_equal (GskTransform *first,
GskTransform *second)
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
{
if (first == second)
return TRUE;
transform: Make sure the identity transform is equal to NULL
2020-02-21 17:25:05 +00:00
if (first == NULL)
return gsk_transform_is_identity (second);
if (second == NULL)
return gsk_transform_is_identity (first);
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
transform: Compare class before ->next
2019-12-31 13:03:42 +00:00
if (first->transform_class != second->transform_class)
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
return FALSE;
transform: Compare class before ->next
2019-12-31 13:03:42 +00:00
if (!gsk_transform_equal (first->next, second->next))
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
return FALSE;
return first->transform_class->equal (first, second);
}
transform: Make category public API Also rename it from GskMatrixCategory to GskTransformCategory.
2019-03-01 06:51:23 +00:00
/**
* gsk_transform_get_category:
gsk: Clean up docs syntax Replace leftover gtk-doc syntax (#Type) with backquotes.
2021-05-20 03:40:54 +00:00
* @self: (nullable): A `GskTransform`
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
*
transform: Make category public API Also rename it from GskMatrixCategory to GskTransformCategory.
2019-03-01 06:51:23 +00:00
* Returns the category this transform belongs to.
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
*
transform: Make category public API Also rename it from GskMatrixCategory to GskTransformCategory.
2019-03-01 06:51:23 +00:00
* Returns: The category of the transform
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
**/
transform: Make category public API Also rename it from GskMatrixCategory to GskTransformCategory.
2019-03-01 06:51:23 +00:00
GskTransformCategory
gsk: move transform structure to private.h This allows using the category without having to call a function from the hot paths of the renderer.
2021-02-11 00:04:20 +00:00
(gsk_transform_get_category) (GskTransform *self)
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
{
if (self == NULL)
transform: Make category public API Also rename it from GskMatrixCategory to GskTransformCategory.
2019-03-01 06:51:23 +00:00
return GSK_TRANSFORM_CATEGORY_IDENTITY;
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
transform: Store the category in the transform ... instead of computing it every time we need it. This should be faster and we want to use it a lot more prominently. Also, we have the struct memory available anyway.
2019-03-04 08:20:58 +00:00
return self->category;
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
}
/*
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
* gsk_transform_new: (constructor):
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
*
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
* Creates a new identity transform.
*
* This function is meant to be used by language
transform: Fix docs to use "transform" instead of "matrix"
2020-06-13 06:25:51 +00:00
* bindings. For C code, this is equivalent to using %NULL.
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
*
transform: Fix docs to use "transform" instead of "matrix"
2020-06-13 06:25:51 +00:00
* Returns: A new identity transform
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
*/
transform: Move to GSK The renaming of the prefix makes this a large patch.
2019-02-28 08:52:49 +00:00
GskTransform *
gsk_transform_new (void)
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
{
transform: Store the category in the transform ... instead of computing it every time we need it. This should be faster and we want to use it a lot more prominently. Also, we have the struct memory available anyway.
2019-03-04 08:20:58 +00:00
return gsk_transform_alloc (&GSK_IDENTITY_TRANSFORM_CLASS, GSK_TRANSFORM_CATEGORY_IDENTITY, NULL);
gtk: Add GtkTransform This is a new object (well, boxed type, but I'm calling it object) for dealing with transform in a more constructive way than graphene_matrix_t by keeping track of how the transform was created. This way, reasoning about the transform becomes easier, and we can create better ways to print it or transition from one transform to another one. An example of this is that while a 0 degree and a 360degree rotation are both the identity matrix, doing a transition between the two would cause a rotation.
2019-02-17 11:54:04 +00:00
}
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
/**
* gsk_transform_transform_bounds:
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
* @self: a `GskTransform`
* @rect: a `graphene_rect_t`
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
* @out_rect: (out caller-allocates): return location for the bounds
* of the transformed rectangle
*
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
* Transforms a `graphene_rect_t` using the given transform @self.
*
gsktransform: Fix documentation comment There is no @m.
2019-08-21 08:56:20 +00:00
* The result is the bounding box containing the coplanar quad.
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
*/
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
void
gsk_transform_transform_bounds (GskTransform *self,
const graphene_rect_t *rect,
graphene_rect_t *out_rect)
{
transform: Readd optimization This is the optimization from bbd4e2f60d9e060fbe3b0c2dcebcd1d6b7335a2a
2019-03-04 20:43:44 +00:00
switch (gsk_transform_get_category (self))
{
case GSK_TRANSFORM_CATEGORY_IDENTITY:
graphene_rect_init_from_rect (out_rect, rect);
break;
case GSK_TRANSFORM_CATEGORY_2D_TRANSLATE:
{
float dx, dy;
gsk_transform_to_translate (self, &dx, &dy);
gsk: Normalize when transforming bounds Bounds are assumed to be normalized, and transforms with negative scales or 3d rotations can make us get negative sizes. Fix by Benjamin Otte.
2020-09-28 15:23:51 +00:00
graphene_rect_init (out_rect,
rect->origin.x + dx,
rect->origin.y + dy,
rect->size.width,
rect->size.height);
transform: Readd optimization This is the optimization from bbd4e2f60d9e060fbe3b0c2dcebcd1d6b7335a2a
2019-03-04 20:43:44 +00:00
}
break;
transform: Add transform_bounds fast path for 2D_AFFINE transforms E.g. anything involving a scale. This is important when e.g. scrolling in the node list in the recorder, which scales every recorded node down to fit in the list.
2019-06-29 07:39:20 +00:00
case GSK_TRANSFORM_CATEGORY_2D_AFFINE:
{
float dx, dy, scale_x, scale_y;
gsk_transform_to_affine (self, &scale_x, &scale_y, &dx, &dy);
gsk: Normalize when transforming bounds Bounds are assumed to be normalized, and transforms with negative scales or 3d rotations can make us get negative sizes. Fix by Benjamin Otte.
2020-09-28 15:23:51 +00:00
graphene_rect_init (out_rect,
(rect->origin.x * scale_x) + dx,
(rect->origin.y * scale_y) + dy,
rect->size.width * scale_x,
rect->size.height * scale_y);
transform: Add transform_bounds fast path for 2D_AFFINE transforms E.g. anything involving a scale. This is important when e.g. scrolling in the node list in the recorder, which scales every recorded node down to fit in the list.
2019-06-29 07:39:20 +00:00
}
break;
transform: Readd optimization This is the optimization from bbd4e2f60d9e060fbe3b0c2dcebcd1d6b7335a2a
2019-03-04 20:43:44 +00:00
case GSK_TRANSFORM_CATEGORY_UNKNOWN:
case GSK_TRANSFORM_CATEGORY_ANY:
case GSK_TRANSFORM_CATEGORY_3D:
case GSK_TRANSFORM_CATEGORY_2D:
default:
{
graphene_matrix_t mat;
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
transform: Readd optimization This is the optimization from bbd4e2f60d9e060fbe3b0c2dcebcd1d6b7335a2a
2019-03-04 20:43:44 +00:00
gsk_transform_to_matrix (self, &mat);
Replace uses of graphene_matrix_transform_ Replace our uses of graphene_matrix_transform_point, _point3d and _bounds by our own versions that handle projective transforms correctly. This fixes render node bounds being incorrect for widgets involving projective transforms (e.g. testrevealer swing transformations), and also fixes picking on such widgets.
2020-09-01 12:54:01 +00:00
gsk_matrix_transform_bounds (&mat, rect, out_rect);
transform: Readd optimization This is the optimization from bbd4e2f60d9e060fbe3b0c2dcebcd1d6b7335a2a
2019-03-04 20:43:44 +00:00
}
break;
}
transform: Add more API In particular, add a per-category querying API for the matrix: - gsk_transform_to_translate() - gsk_transform_to_affine() - gsk_transform_to_2d() - gsk_transform_to_matrix() This way, code can use the relevant one for the given category.
2019-03-01 05:24:33 +00:00
}
transform: Add gsk_transform_parse() It uses the new CSS parser.
2019-03-21 04:52:41 +00:00
GskTransform: Add gsk_transform_transform_point() Equivalent of gsk_transform_transform_bounds() and graphene_matrix_transform_point() respectively.
2020-01-21 08:41:54 +00:00
/**
* gsk_transform_transform_point:
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
* @self: a `GskTransform`
* @point: a `graphene_point_t`
GskTransform: Add gsk_transform_transform_point() Equivalent of gsk_transform_transform_bounds() and graphene_matrix_transform_point() respectively.
2020-01-21 08:41:54 +00:00
* @out_point: (out caller-allocates): return location for
* the transformed point
*
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
* Transforms a `graphene_point_t` using the given transform @self.
GskTransform: Add gsk_transform_transform_point() Equivalent of gsk_transform_transform_bounds() and graphene_matrix_transform_point() respectively.
2020-01-21 08:41:54 +00:00
*/
void
gsk_transform_transform_point (GskTransform *self,
const graphene_point_t *point,
graphene_point_t *out_point)
{
switch (gsk_transform_get_category (self))
{
case GSK_TRANSFORM_CATEGORY_IDENTITY:
*out_point = *point;
break;
case GSK_TRANSFORM_CATEGORY_2D_TRANSLATE:
{
float dx, dy;
gsk_transform_to_translate (self, &dx, &dy);
out_point->x = point->x + dx;
out_point->y = point->y + dy;
}
break;
case GSK_TRANSFORM_CATEGORY_2D_AFFINE:
{
float dx, dy, scale_x, scale_y;
gsk_transform_to_affine (self, &scale_x, &scale_y, &dx, &dy);
out_point->x = (point->x * scale_x) + dx;
out_point->y = (point->y * scale_y) + dy;
}
break;
case GSK_TRANSFORM_CATEGORY_UNKNOWN:
case GSK_TRANSFORM_CATEGORY_ANY:
case GSK_TRANSFORM_CATEGORY_3D:
case GSK_TRANSFORM_CATEGORY_2D:
default:
{
graphene_matrix_t mat;
gsk_transform_to_matrix (self, &mat);
Replace uses of graphene_matrix_transform_ Replace our uses of graphene_matrix_transform_point, _point3d and _bounds by our own versions that handle projective transforms correctly. This fixes render node bounds being incorrect for widgets involving projective transforms (e.g. testrevealer swing transformations), and also fixes picking on such widgets.
2020-09-01 12:54:01 +00:00
gsk_matrix_transform_point (&mat, point, out_point);
GskTransform: Add gsk_transform_transform_point() Equivalent of gsk_transform_transform_bounds() and graphene_matrix_transform_point() respectively.
2020-01-21 08:41:54 +00:00
}
break;
}
}
transform: Add gsk_transform_parse() It uses the new CSS parser.
2019-03-21 04:52:41 +00:00
static guint
gsk_transform_parse_float (GtkCssParser *parser,
guint n,
gpointer data)
{
float *f = data;
double d;
if (!gtk_css_parser_consume_number (parser, &d))
return 0;
f[n] = d;
return 1;
}
static guint
gsk_transform_parse_scale (GtkCssParser *parser,
guint n,
gpointer data)
{
float *f = data;
double d;
if (!gtk_css_parser_consume_number (parser, &d))
return 0;
f[n] = d;
f[1] = d;
return 1;
}
gboolean
gsk_transform_parser_parse (GtkCssParser *parser,
GskTransform **out_transform)
{
const GtkCssToken *token;
GskTransform *transform = NULL;
float f[16] = { 0, };
GskTransform: NULL is a valid transform
2019-06-03 15:36:31 +00:00
gboolean parsed_something = FALSE;
transform: Add gsk_transform_parse() It uses the new CSS parser.
2019-03-21 04:52:41 +00:00
token = gtk_css_parser_get_token (parser);
if (gtk_css_token_is_ident (token, "none"))
{
gtk_css_parser_consume_token (parser);
*out_transform = NULL;
return TRUE;
}
while (TRUE)
{
if (gtk_css_token_is_function (token, "matrix"))
{
graphene_matrix_t matrix;
if (!gtk_css_parser_consume_function (parser, 6, 6, gsk_transform_parse_float, f))
goto fail;
graphene_matrix_init_from_2d (&matrix, f[0], f[1], f[2], f[3], f[4], f[5]);
transform = gsk_transform_matrix_with_category (transform,
&matrix,
GSK_TRANSFORM_CATEGORY_2D);
}
else if (gtk_css_token_is_function (token, "matrix3d"))
{
graphene_matrix_t matrix;
if (!gtk_css_parser_consume_function (parser, 16, 16, gsk_transform_parse_float, f))
goto fail;
graphene_matrix_init_from_float (&matrix, f);
transform = gsk_transform_matrix (transform, &matrix);
}
else if (gtk_css_token_is_function (token, "perspective"))
{
if (!gtk_css_parser_consume_function (parser, 1, 1, gsk_transform_parse_float, f))
goto fail;
transform = gsk_transform_perspective (transform, f[0]);
}
else if (gtk_css_token_is_function (token, "rotate") ||
gtk_css_token_is_function (token, "rotateZ"))
{
if (!gtk_css_parser_consume_function (parser, 1, 1, gsk_transform_parse_float, f))
goto fail;
transform = gsk_transform_rotate (transform, f[0]);
}
else if (gtk_css_token_is_function (token, "rotate3d"))
{
graphene_vec3_t axis;
if (!gtk_css_parser_consume_function (parser, 4, 4, gsk_transform_parse_float, f))
goto fail;
graphene_vec3_init (&axis, f[0], f[1], f[2]);
transform = gsk_transform_rotate_3d (transform, f[3], &axis);
}
else if (gtk_css_token_is_function (token, "rotateX"))
{
if (!gtk_css_parser_consume_function (parser, 1, 1, gsk_transform_parse_float, f))
goto fail;
transform = gsk_transform_rotate_3d (transform, f[0], graphene_vec3_x_axis ());
}
else if (gtk_css_token_is_function (token, "rotateY"))
{
if (!gtk_css_parser_consume_function (parser, 1, 1, gsk_transform_parse_float, f))
goto fail;
transform = gsk_transform_rotate_3d (transform, f[0], graphene_vec3_y_axis ());
}
else if (gtk_css_token_is_function (token, "scale"))
{
if (!gtk_css_parser_consume_function (parser, 1, 2, gsk_transform_parse_scale, f))
goto fail;
transform = gsk_transform_scale (transform, f[0], f[1]);
}
else if (gtk_css_token_is_function (token, "scale3d"))
{
if (!gtk_css_parser_consume_function (parser, 3, 3, gsk_transform_parse_float, f))
goto fail;
transform = gsk_transform_scale_3d (transform, f[0], f[1], f[2]);
}
else if (gtk_css_token_is_function (token, "scaleX"))
{
if (!gtk_css_parser_consume_function (parser, 1, 1, gsk_transform_parse_float, f))
goto fail;
transform = gsk_transform_scale (transform, f[0], 1.f);
}
else if (gtk_css_token_is_function (token, "scaleY"))
{
if (!gtk_css_parser_consume_function (parser, 1, 1, gsk_transform_parse_float, f))
goto fail;
transform = gsk_transform_scale (transform, 1.f, f[0]);
}
else if (gtk_css_token_is_function (token, "scaleZ"))
{
if (!gtk_css_parser_consume_function (parser, 1, 1, gsk_transform_parse_float, f))
goto fail;
transform = gsk_transform_scale_3d (transform, 1.f, 1.f, f[0]);
}
else if (gtk_css_token_is_function (token, "translate"))
{
f[1] = 0.f;
if (!gtk_css_parser_consume_function (parser, 1, 2, gsk_transform_parse_float, f))
goto fail;
transform = gsk_transform_translate (transform, &GRAPHENE_POINT_INIT (f[0], f[1]));
}
else if (gtk_css_token_is_function (token, "translate3d"))
{
if (!gtk_css_parser_consume_function (parser, 3, 3, gsk_transform_parse_float, f))
goto fail;
transform = gsk_transform_translate_3d (transform, &GRAPHENE_POINT3D_INIT (f[0], f[1], f[2]));
}
else if (gtk_css_token_is_function (token, "translateX"))
{
if (!gtk_css_parser_consume_function (parser, 1, 1, gsk_transform_parse_float, f))
goto fail;
transform = gsk_transform_translate (transform, &GRAPHENE_POINT_INIT (f[0], 0.f));
}
else if (gtk_css_token_is_function (token, "translateY"))
{
if (!gtk_css_parser_consume_function (parser, 1, 1, gsk_transform_parse_float, f))
goto fail;
transform = gsk_transform_translate (transform, &GRAPHENE_POINT_INIT (0.f, f[0]));
}
else if (gtk_css_token_is_function (token, "translateZ"))
{
if (!gtk_css_parser_consume_function (parser, 1, 1, gsk_transform_parse_float, f))
goto fail;
transform = gsk_transform_translate_3d (transform, &GRAPHENE_POINT3D_INIT (0.f, 0.f, f[0]));
}
else if (gtk_css_token_is_function (token, "skew"))
{
transform: Add skew parsing
2019-06-29 05:12:10 +00:00
graphene_matrix_t matrix;
if (!gtk_css_parser_consume_function (parser, 2, 2, gsk_transform_parse_float, f))
goto fail;
f[0] = f[0] / 180.0 * G_PI;
f[1] = f[1] / 180.0 * G_PI;
graphene_matrix_init_skew (&matrix, f[0], f[1]);
transform = gsk_transform_matrix (transform, &matrix);
transform: Add gsk_transform_parse() It uses the new CSS parser.
2019-03-21 04:52:41 +00:00
}
else if (gtk_css_token_is_function (token, "skewX"))
{
transform: Add skew parsing
2019-06-29 05:12:10 +00:00
graphene_matrix_t matrix;
if (!gtk_css_parser_consume_function (parser, 1, 1, gsk_transform_parse_float, f))
goto fail;
f[0] = f[0] / 180.0 * G_PI;
graphene_matrix_init_skew (&matrix, f[0], 0);
transform = gsk_transform_matrix (transform, &matrix);
transform: Add gsk_transform_parse() It uses the new CSS parser.
2019-03-21 04:52:41 +00:00
}
else if (gtk_css_token_is_function (token, "skewY"))
{
transform: Add skew parsing
2019-06-29 05:12:10 +00:00
graphene_matrix_t matrix;
if (!gtk_css_parser_consume_function (parser, 1, 1, gsk_transform_parse_float, f))
goto fail;
f[0] = f[0] / 180.0 * G_PI;
graphene_matrix_init_skew (&matrix, 0, f[0]);
transform = gsk_transform_matrix (transform, &matrix);
transform: Add gsk_transform_parse() It uses the new CSS parser.
2019-03-21 04:52:41 +00:00
}
else
{
break;
}
GskTransform: NULL is a valid transform
2019-06-03 15:36:31 +00:00
parsed_something = TRUE;
transform: Add gsk_transform_parse() It uses the new CSS parser.
2019-03-21 04:52:41 +00:00
token = gtk_css_parser_get_token (parser);
}
GskTransform: NULL is a valid transform
2019-06-03 15:36:31 +00:00
if (!parsed_something)
transform: Add gsk_transform_parse() It uses the new CSS parser.
2019-03-21 04:52:41 +00:00
{
gtk_css_parser_error_syntax (parser, "Expected a transform");
goto fail;
}
*out_transform = transform;
return TRUE;
fail:
gsk_transform_unref (transform);
*out_transform = NULL;
return FALSE;
}
/**
* gsk_transform_parse:
* @string: the string to parse
* @out_transform: (out): The location to put the transform in
*
* Parses the given @string into a transform and puts it in
docs: Improve gsk docs Convert link formats, add summaries, and make a few missing things show up in the docs.
2021-02-20 06:09:55 +00:00
* @out_transform.
*
* Strings printed via [method@Gsk.Transform.to_string]
transform: Add gsk_transform_parse() It uses the new CSS parser.
2019-03-21 04:52:41 +00:00
* can be read in again successfully using this function.
*
* If @string does not describe a valid transform, %FALSE is
* returned and %NULL is put in @out_transform.
*
* Returns: %TRUE if @string described a valid transform.
gsktransform: Convert docs
2021-02-25 12:00:35 +00:00
*/
transform: Add gsk_transform_parse() It uses the new CSS parser.
2019-03-21 04:52:41 +00:00
gboolean
gsk_transform_parse (const char *string,
GskTransform **out_transform)
{
GtkCssParser *parser;
GBytes *bytes;
gboolean result;
g_return_val_if_fail (string != NULL, FALSE);
g_return_val_if_fail (out_transform != NULL, FALSE);
bytes = g_bytes_new_static (string, strlen (string));
parser = gtk_css_parser_new_for_bytes (bytes, NULL, NULL, NULL, NULL, NULL);
result = gsk_transform_parser_parse (parser, out_transform);
if (result && !gtk_css_parser_has_token (parser, GTK_CSS_TOKEN_EOF))
{
g_clear_pointer (out_transform, gsk_transform_unref);
result = FALSE;
}
gtk_css_parser_unref (parser);
g_bytes_unref (bytes);
return result;
}
gsk: Add some private transform apis Some of the graphene_matrix_transform apis don't work correctly with projective matrices, so add our own variants.
2020-09-01 12:36:45 +00:00
/* Some of the graphene_matrix_transform apis yield unexpected
* results with projective matrices, since they silently drop
* the w component, so we provide working alternatives here.
*/
void
gsk_matrix_transform_point (const graphene_matrix_t *m,
const graphene_point_t *p,
graphene_point_t *res)
{
graphene_vec4_t vec4;
float w;
graphene_vec4_init (&vec4, p->x, p->y, 0.0f, 1.0f);
graphene_matrix_transform_vec4 (m, &vec4, &vec4);
w = graphene_vec4_get_w (&vec4);
res->x = graphene_vec4_get_x (&vec4) / w;
res->y = graphene_vec4_get_y (&vec4) / w;
}
void
gsk_matrix_transform_point3d (const graphene_matrix_t *m,
const graphene_point3d_t *p,
graphene_point3d_t *res)
{
graphene_vec4_t vec4;
float w;
graphene_vec4_init (&vec4, p->x, p->y, 0.0f, 1.0f);
graphene_matrix_transform_vec4 (m, &vec4, &vec4);
w = graphene_vec4_get_w (&vec4);
res->x = graphene_vec4_get_x (&vec4) / w;
res->y = graphene_vec4_get_y (&vec4) / w;
res->z = graphene_vec4_get_z (&vec4) / w;
}
void
Add gsk_matrix_transform_rect This is a projecting version of the corresponding graphene api. While we are at it, rewrite gsk_matrix_transform_bounds() to use gsk_matrix_transform_rect().
2020-09-04 16:51:11 +00:00
gsk_matrix_transform_rect (const graphene_matrix_t *m,
const graphene_rect_t *r,
graphene_quad_t *res)
gsk: Add some private transform apis Some of the graphene_matrix_transform apis don't work correctly with projective matrices, so add our own variants.
2020-09-01 12:36:45 +00:00
{
graphene_point_t ret[4];
graphene_rect_t rr;
graphene_rect_normalize_r (r, &rr);
#define TRANSFORM_POINT(matrix, rect, corner, out_p) do {\
graphene_vec4_t __s; \
graphene_point_t __p; \
float w; \
graphene_rect_get_ ## corner (rect, &__p); \
graphene_vec4_init (&__s, __p.x, __p.y, 0.f, 1.f); \
graphene_matrix_transform_vec4 (matrix, &__s, &__s); \
w = graphene_vec4_get_w (&__s); \
out_p.x = graphene_vec4_get_x (&__s) / w; \
out_p.y = graphene_vec4_get_y (&__s) / w; } while (0)
TRANSFORM_POINT (m, &rr, top_left, ret[0]);
TRANSFORM_POINT (m, &rr, top_right, ret[1]);
TRANSFORM_POINT (m, &rr, bottom_right, ret[2]);
TRANSFORM_POINT (m, &rr, bottom_left, ret[3]);
#undef TRANSFORM_POINT
Add gsk_matrix_transform_rect This is a projecting version of the corresponding graphene api. While we are at it, rewrite gsk_matrix_transform_bounds() to use gsk_matrix_transform_rect().
2020-09-04 16:51:11 +00:00
graphene_quad_init (res, &ret[0], &ret[1], &ret[2], &ret[3]);
}
void
gsk_matrix_transform_bounds (const graphene_matrix_t *m,
const graphene_rect_t *r,
graphene_rect_t *res)
{
graphene_quad_t q;
gsk: Add some private transform apis Some of the graphene_matrix_transform apis don't work correctly with projective matrices, so add our own variants.
2020-09-01 12:36:45 +00:00
Add gsk_matrix_transform_rect This is a projecting version of the corresponding graphene api. While we are at it, rewrite gsk_matrix_transform_bounds() to use gsk_matrix_transform_rect().
2020-09-04 16:51:11 +00:00
gsk_matrix_transform_rect (m, r, &q);
graphene_quad_bounds (&q, res);
gsk: Add some private transform apis Some of the graphene_matrix_transform apis don't work correctly with projective matrices, so add our own variants.
2020-09-01 12:36:45 +00:00
}
Reference in New Issue Copy Permalink