Commit Graph

68004 Commits

Author SHA1 Message Date
Matthias Clasen
78d06e58d0 Add gsk_path_measure_get_{path,tolerance}
These are just nice apis to have and avoid having to carry
these around as extra arguments in many places.

This was showing up as inconvenience in writing tests
for the measure apis.
2020-12-27 00:31:18 +01:00
Benjamin Otte
a5e13cc96b xxx path)_fill 2020-12-27 00:31:18 +01:00
Matthias Clasen
776dc54c8a Add gsk_path_get_stroke_bounds
A relatively cheap way to get bounds for the area
that would be affected by stroking a path.
2020-12-27 00:31:18 +01:00
Benjamin Otte
ee6879fba8 testsuite: Add tests for the dasher 2020-12-27 00:31:18 +01:00
Benjamin Otte
14f9395476 path: Add a foreach function that dashes a path 2020-12-27 00:31:18 +01:00
Benjamin Otte
1349cf2454 path: Deal with non-uniformness of progress parameter
The progress is non-uniform, so simple translation of progress doesn't work.
So check if larger and smaller values inch closer towards minimal distance.
2020-12-27 00:31:18 +01:00
Benjamin Otte
d47ebd388c path: Always decompose conics into at least 2 segments
Conics are evil in that their parameter skews towards the center, and if
it's a very flat conic (weight almost equal to 0), then we'd approximate
it with a single segment and not subdivide, which would cause the
parameter to be wildly off around 0.25 or 0.75.

And that would cause offset calculations to fail.
2020-12-27 00:31:18 +01:00
Matthias Clasen
61214221b3 testsuite Add curve tangent tests 2020-12-27 00:31:18 +01:00
Benjamin Otte
7eb4ed8f86 testsuite: Add a test for the conic that got us segment() 2020-12-27 00:31:18 +01:00
Benjamin Otte
92b472fec1 path: Add gsk_curve_segment()
Using split() twice with scaled t values does not work with conics.
2020-12-27 00:31:18 +01:00
Benjamin Otte
7a6b008479 testsuite: Add a test for gsk_curve_decompose() 2020-12-27 00:31:18 +01:00
Benjamin Otte
d8c26172a5 testuite: Add tests for gsk_curve_get_tangent() 2020-12-27 00:31:18 +01:00
Matthias Clasen
da835321a5 testuite: Add tests for gsk_curve_get_point()
Add a few tests for gsk_curve_get_point().

Since GskCurve is not public api, we add gskcurve.c
as source to the test binary.
2020-12-27 00:31:18 +01:00
Benjamin Otte
30635d1cac curve: Split eval() into get_point() and get_tangent()
That's more in line with the get_start/end_point/tangent() functions.

Plus, those calls are independent and we usually want one or the other.
2020-12-27 00:31:18 +01:00
Matthias Clasen
a60b72ba58 Add gsk_curve_get_{start,end}_tangent
Add a way to get the tangents at the start and end of the curve.
This will be used in stroking.
2020-12-27 00:31:18 +01:00
Benjamin Otte
36347892d8 testsuite: Add conics to the random paths 2020-12-27 00:31:18 +01:00
Benjamin Otte
924f0bc2c5 path: Add GskCurve
GskCurve is an abstraction for path operations. It's essentially a
collection of vfuncs per GskPathOperation.

GskStandardContour has been ported to use it where appropriate.
2020-12-27 00:31:18 +01:00
Benjamin Otte
23c5318de1 path: Introduce gskpathop
A gskpathop is a pointer to a graphene_point_t* with the low bits used
to encode the GskPathOperation. It's an easy way to introduce API for
operations.

So far it's just used to replace GskStandardOperation.
2020-12-27 00:31:18 +01:00
Benjamin Otte
678405bd22 WIP: css: Replace border rendering code with GskPath
The weight is wrong still, I need to compute the correct one to get real
45deg circle corners and not just roughly correct ones.
2020-12-27 00:31:18 +01:00
Benjamin Otte
946b76881b WIP: pathbuilder: Add gsk_path_builder_add_rounded_rect()
It works, but does not use a custom contour yet.
2020-12-27 00:31:18 +01:00
Benjamin Otte
061683344f path: Add conic curves
So far this just adds the API, if you use it, you'll get lots of
g_warnings().

This will be fixed in future commits.
2020-12-27 00:31:18 +01:00
Benjamin Otte
af9a85616d path: Rename to gtk_path_builder_add_segment()
It's about bulding paths, not about measuring them.
2020-12-27 00:31:18 +01:00
Benjamin Otte
0f332de3d3 path: Split contours into their own file
I'm not sure I want to keep all contours in one file, but for now that's
how it is.
2020-12-27 00:31:18 +01:00
Benjamin Otte
15e8f2bacc path: Make all private contour APIs take a GskContour
... instead of a path, index tuple.
2020-12-27 00:31:18 +01:00
Benjamin Otte
25f9eadb0d stroke: Add support for dashes
... and hook it up in the node parser and for Cairo rendering.
2020-12-27 00:31:18 +01:00
Matthias Clasen
ced2959cdd gsk: Implement parsing fill and stroke nodes
Make serialization and deserialization work for stroke and
fill nodes.
2020-12-27 00:31:18 +01:00
Benjamin Otte
3c363f00d5 path: Add flags to gsk_path_foreach()
This way we can default to the siplest possible foreach() output - like
cairo_copy_path_flat() decomposing everything into lines - and add flags
to get more and more fancy.

This will be useful to have conics automatically decomposed for Cairo
drawing or if we want to add more line types in the future.
2020-12-27 00:31:18 +01:00
Benjamin Otte
a9ad37ed89 testsuite: Add an in_fill() test 2020-12-27 00:31:18 +01:00
Matthias Clasen
acdf04287d Implement gsk_path_measure_in_fill
Implement this in the obvious way, using the decomposed form
of standard contours. Since the decomposed form is part of the
measure object, this api moves from gsk_path_in_fill to
gsk_path_measure_in_fill.
2020-12-27 00:31:18 +01:00
Benjamin Otte
34725c1207 testsuite: Add a parsing test
This test includes an implementation of a gsk_path_equal() func with
a tolerance that is necessary because parsing does not always work
100% exactly due to floating point rounding, so we can't just
compare the to_string() output.
2020-12-27 00:31:18 +01:00
Matthias Clasen
cb10cf07e6 path: Special-case rects and circles
Write out the commands for rects and circles in a special
way, and add code in the parser to recognize this, so we
can successfully round-trip these through the SVG path format.

The special way - for people who want to use it for debugging -
for now is that we use uppercase "Z" to close standard paths, but
lowercase "z" to close our special paths.

A test is included, but the random path serializations should take care
of it, too.
2020-12-27 00:31:18 +01:00
Matthias Clasen
a7ba2bde86 path: Fix serialization for circles
The svg A can not do a full circle, since it is a two point
parametrization - if the start and end point are the same,
it draws nothing. So, use two arcs.
2020-12-27 00:31:18 +01:00
Benjamin Otte
87b3cb1df8 testsuite: Add librsvg path tests 2020-12-27 00:31:18 +01:00
Matthias Clasen
2478963cbb path: Implement gsk_path_parse
Implement the SVG path syntax to read back the strings
that we generate when serializing paths. The tests for
this code are taken from librsvg.

This includes an elliptical arc implementation according
to the SVG spec. The code is mostly taken from librsvg,
but pretty directly follows the SVG spec implementation
notes. We don't export this, since the parametrization
is inconvenient. We do want an arc_to API, but
these are not the arcs we are looking for.
2020-12-27 00:31:18 +01:00
Matthias Clasen
abbbaef7e8 path: Implement SVG arcs
This is elliptical arc implementation according to the SVG spec.
The code is mostly taken from librsvg, but pretty directly
follows the SVG spec implementation notes.

We don't export this, since the parametrization is inconvenient.
We do want an arc_to API, but these are not the arcs we are
looking for.

It will be used in parsing SVG path syntax.
2020-12-27 00:31:18 +01:00
Matthias Clasen
69177416a0 stroke: Add miter limit
Add a miter limit to GskStroke. This will be needed to
fully implement line joins.

Also introduce the GSK_LINE_JOIN_MITER_CLIP value,
following SVG 2.0. cairo does not have it, so translate
it to plain miter when using cairo.
2020-12-27 00:31:18 +01:00
Matthias Clasen
ff47bfdbcf Documentation typo fixes 2020-12-27 00:31:18 +01:00
Benjamin Otte
120e449bdd testsuite: Add relative path functions
They're making the paths slightly weirder, but they test public API, so
woohoo!
2020-12-27 00:31:18 +01:00
Benjamin Otte
237c37f6c2 pathbuilder: Add relative path commands
And gsk_path_builder_get_current_point().

They will be needed by the string parser.
2020-12-27 00:31:18 +01:00
Benjamin Otte
683729c388 path: Add GSK_CIRCLE_POINT_INIT() to initialize points on the circle
This is just splitting out a commonly done operation into a macro.
2020-12-27 00:31:18 +01:00
Benjamin Otte
9fe71e7dae pathbuilder: Redo semantics for starting curves
We now always have a "current point" which is either the last point an
operation was made to, or (0, 0) if no drawing operation has
been made yet.

Adding a contour of any kind to the builder will always update the
current point to that contour's end point.
2020-12-27 00:31:18 +01:00
Benjamin Otte
c74970e6c1 xxx: demo 2020-12-27 00:31:18 +01:00
Benjamin Otte
5225d01504 path: Split GskPathBuilder into its own file
... and add missing API docs.
2020-12-27 00:31:18 +01:00
Benjamin Otte
916d950069 testsuite: Add a test using get_point() and get_closest_point() 2020-12-27 00:31:18 +01:00
Benjamin Otte
76d6fed248 testsuite: Add a test for get_point() 2020-12-27 00:31:18 +01:00
Benjamin Otte
7ff6235ae6 testsuite: Update create_random_path()
1. Allow specifying the max number of contours
2. Be smarter about creating the paths:
   With 10% chance, create a "weird" path like the empty one or only
   points or things like that.
   Otherwise create a bunch of contours, with 2/3 a standard contour,
   with 1/3 a predetermined one.
2020-12-27 00:31:18 +01:00
Benjamin Otte
9b957c26cb gtk-demo: Add cute maze demo 2020-12-27 00:31:18 +01:00
Benjamin Otte
445b2382e6 testsuite: Add tests for gsk_path_measure_get_closest_point() 2020-12-27 00:31:18 +01:00
Benjamin Otte
790a4201c3 path: Add gsk_path_measure_get_closest_point()
... and gsk_path_measure_get_closest_point_full().

Those 2 functions allow finding the closest point on a path to a given
point.
2020-12-27 00:31:18 +01:00
Benjamin Otte
00b9af65b5 spline: Use Skia's tolerance checks
This avoids measuring being too far off (it's still off, but it's less
than a percent now.
2020-12-27 00:31:18 +01:00