c8e57d63ec
cssstyelproperty: Rempove specified-type and computed-type props
...
This removes the necessity to keep a GType associated with style
properties and code can now make use of GtkCssValue completely.
2012-04-17 08:59:13 +02:00
75a5f04352
styleproperty: Remove default parse function
...
... and assert every style property brings its own.
2012-04-17 08:59:13 +02:00
a5d001b2ba
stylepropertyimpl: Add remaining parse functions
2012-04-17 08:59:12 +02:00
0fdc287339
styleproperty: Simplify compute_value function
...
The compute_value fallback path is only needed for custom properties,
the real style properties have custom compute functions if they need
them already.
2012-04-17 08:59:12 +02:00
7fbc583b88
css: Move shadow parse/print/compute funcs
...
They don't belong in the style funcs, as those are for custom
properties, and the shadow type is private.
2012-04-17 08:59:12 +02:00
dd144c2bad
stylepropertyimpl: Add parse funcs for enums
2012-04-17 08:59:12 +02:00
86a387f571
stylepropertyimpl: Add color parsing function
...
This is in preparation for removing the specified type and computed type
properties from GtkCssStyleProperty, which is in preparation for really
using GtkCssValue classes and not GTypes.
2012-04-17 08:59:12 +02:00
b5fc484c99
customproperty: Don't rely on querying style property
...
This way, we can remove type madness from GtkCssStyleProperty later.
2012-04-17 08:59:12 +02:00
bc9060a591
styleproperty: Make gtk_style_property_register() not be valist
...
Instead, make the caller create a GtkCssValue in advance.
2012-04-17 08:59:12 +02:00
ffe50c3b67
cssvalue: Add _gtk_css_value_equal()
...
For now, we return FALSE for all default css values, so this is not very
useful.
I also think of this as an optimization equal, not a guaranteed equal,
because we don't even have a notion of what "equal" means.
For example, for background-repeat, "repeat, repeat" and "repeat"
are functionally equivalent. But the cssvalue has no idea that it's used
for background-repeat.
As a more complicated example, "repeat, no-repeat" and "repeat" are
equal to what one sees as long as there's only one image listed
background-image-source. But once you start transition'ing to an image
with 2 sources, it's different...
2012-04-17 08:59:12 +02:00
3e601691d9
tests: Change expected errors to conform with output
...
This change isn't strictly correct, but I can't be bothered until we get
a tokenizer that can really differentiate between allowed values and
invalid syntax.
2012-04-17 08:59:12 +02:00
b0e764000f
cssparser: Use _gtk_css_parser_error_full()
2012-04-17 08:59:12 +02:00
e7acdec220
cssparser: Add _gtk_css_parser_error_full()
2012-04-17 08:59:11 +02:00
9b7640b898
styleproperty: Make _gtk_style_property_parse_value() return a CssValue
...
Also split out initial/inherit handling into a custom GtkCssValue class.
2012-04-17 08:59:11 +02:00
718ceaec45
css: Remove _gtk_css_style_property_is_specified_type()
...
That check is going to be refactored away.
2012-04-17 08:59:11 +02:00
04f5c8708a
cssvalue: Add a CSS value for arrays
...
Make the value auto-cycle its child values, so it can be used for
background properties.
2012-04-17 08:59:11 +02:00
df45983fcb
cssvalue: Remove _gtk_css_value_new_take_gvalue()
...
Having two constructors from GValues complicates refactorings, so I'd
rather not have them.
2012-04-17 08:59:11 +02:00
8e3e3d582f
cssvalue: Get rid of unused functions
2012-04-17 08:59:11 +02:00
5ac9ba714a
styleproperty: Make _gtk_style_property_query() take a GValue
...
... and don't make it return a GtkCssValue. We want to use this for
compat with the old GValue APIs after all...
2012-04-17 08:59:11 +02:00
58e4fdf911
styleproperty: Make print_func take a GtkCssValue
...
Also, constify GtkCssValue getters, so we can pass a const GtkCssValue
to the print_func.
2012-04-17 08:59:11 +02:00
b728cfd3e8
cssvalue: Make the structure vtable-based
...
Don't use real classes, just a vtable.
2012-04-17 08:59:11 +02:00
3cdb9c91ca
styleproperty: Add support for equal_func
2012-04-17 08:59:11 +02:00
d5a2392cc7
stylecontext: Don't clear cache when only state changes
...
This takes more memory, but changes to backdrop or active state are
quite expensive otherwise.
2012-04-17 08:59:11 +02:00
44187ca3b5
cssmatcher: Use quarks for classes
2012-04-17 08:59:10 +02:00
115c34498e
styleproperty: Add a function to compare values for equality
...
and default to never compare them as equal.
2012-04-17 08:59:10 +02:00
5bbf59b519
cssvalue: Constify a bunch of APIs
2012-04-17 08:59:10 +02:00
4a12717e4d
cssvalue: Remove unused functions
2012-04-17 08:59:10 +02:00
585a1fae4f
stylecontext: Really queue style changes
...
Instead of instantly applying a new style, just mark the context as
invalid. Only apply the new style at layout time.
2012-04-17 08:59:10 +02:00
eb537b60f4
stylecontext: Optimize the common case of "style didn't change"
2012-04-17 08:59:10 +02:00
ece9d2fd92
gtk: Make widget only invalidate widget positions
...
Instead of resetting them completely
2012-04-17 08:59:10 +02:00
e7a984b337
window: Set parent context instead of just resetting style
...
Otherwise the inherit properties won't inherit properly.
2012-04-17 08:59:10 +02:00
27c9352df3
widget: Replace reset_style() calls
...
... with finer grained _gtk_widget_invalidate_style_context() calls.
2012-04-17 08:59:10 +02:00
969daec353
widget: Don't go via path when resetting style
...
Instead, invalidate the path when the style context changes. And
invalidate the style context directly.
2012-04-17 08:59:09 +02:00
d589566f40
widget: Add _gtk_widget_invalidate_style_context()
...
This is supposed to replace gtk_widget_reset_style().
2012-04-17 08:59:09 +02:00
48a8ca8bbe
widget: Merge function into only caller
2012-04-17 08:59:09 +02:00
257d961a01
widget: Clear broken widget path in constructor
...
... where it belongs instead of fiddling with it in get_path().
2012-04-17 08:59:09 +02:00
198cf93f0f
stylecontext: On failure, exit the loop, don't try again
...
Because we will fail again. And then we try again. And then we fail
again. Ad infinitum.
2012-04-17 08:59:09 +02:00
15be680540
stylecontext: Invalidate when setting classes/state/regions
...
... and nobody has called style_context_save().
2012-04-17 08:59:09 +02:00
44d49c900d
stylecontext: Use _gtk_style_context_queue_invalidate() on self
2012-04-17 08:59:09 +02:00
39ff874a11
stylecontext: Introduce _gtk_style_context_queue_invalidate()
...
This is for only queueing invalidations instead of doing a full
invalidation cycle.
2012-04-17 08:59:09 +02:00
1a51ea3ea3
stylecontext: Keep track of children
2012-04-17 08:59:09 +02:00
7f511f2b33
widget: Don't set widget path on style context
...
Instead, make the style context use the widget's path. Saves a bunch of
memory.
2012-04-17 08:59:09 +02:00
e5ab48ac59
stylecontext: Notify widget directly of changes
...
This way, we don't need to g_signal_connect(), which saves a bunch of
memory and performance.
2012-04-17 08:59:08 +02:00
1aa4a196dc
stylecontext: Add setter for widget
...
... and actually set the widget on the style context. Note that this
function does not take a reference on the widget, which is a very good
reason to keep it private.
2012-04-17 08:59:08 +02:00
102d10016f
matcher: Add a superset matcher
...
This matcher always matches only on some relvant things and ignores the
rest. This allows you to match only on name and class, but ignore state
and parents/siblings for example.
2012-04-17 08:59:08 +02:00
5f5bf27287
matcher: Add a matcher that matches anything
...
Lesson learned: Adwaita uses all possible selector types and all of
those for parents, too. But so far no siblings.
2012-04-17 08:59:08 +02:00
d115441a57
reftests: Add a reftest for nth-child
...
Prime numbers are fun.
2012-04-17 08:59:08 +02:00
cbe40f27e1
tests: Add a parsing test for nth-child()
...
Now that we support the an+b notation, make sure we really do.
Maybe my generating script went a bit overboard with the space
possibilities though...
2012-04-17 08:59:08 +02:00
3bdde54aaf
selector: Rewrite position tracking
...
We now track the position as a (type,a,b) tuple where the numbers make
up the an + b formula from CSS3 nth-child.
Also, the get_sibling() and get_sibling_index() vfuncs were replaced by
a has_position() vfunc. This is mostly so that the matcher can always
return TRUE. And I need that for the everything matcher.
2012-04-17 08:59:08 +02:00
8dbe8c8349
syleprovider: Add a vfunc to get the changes
...
This way we can check what changes are even interesting for our matcher.
2012-04-17 08:59:08 +02:00
