Replace once_cell usage with std::sync::OnceLock
#1289
Conversation
andy128k
force-pushed
the
std_cells
branch
2 times, most recently
from
bd37ee0 to
e4513d2
Compare
|
I guess, this PR should be merged first. |
| @@ -160,8 +160,10 @@ unsafe extern "C" fn animation_get_size<T: PixbufAnimationImpl>( | |||
| } | |||
| } | |||
|
|
|||
| static STATIC_IMAGE_QUARK: Lazy<glib::Quark> = | |||
| Lazy::new(|| glib::Quark::from_str("gtk-rs-subclass-static-image")); | |||
| fn static_image_quark() -> glib::Quark { | |||
There was a problem hiding this comment.
I am not really a fan of this change tbh, as at the end of the day we don't get rid of once_cell from our dependencies tree. I would keep such Lazy usages
There was a problem hiding this comment.
May you explain? This PR literally removes usage of once_cell crate.
There was a problem hiding this comment.
Never mind, I got confused by the tests/examples deps pulling once_cell.
There was a problem hiding this comment.
I also don't like it much but as it's only internal it seems fine to me. We can switch to LazyLock at a later time once it's stable.
@andy128k can you create an issue about that?
There was a problem hiding this comment.
I actually don't think you need a function here at all, the quark is used only inside a single function which can define the static itself. See gtk-rs/gtk4-rs@3b8e9b0
andy128k
force-pushed
the
std_cells
branch
2 times, most recently
from
d53b865 to
dbf7540
Compare
But we have a |
Closes #30
Closes #1141
See also
gtk-rs/gir#1532gtk-rs/gir#1537This change does not fully remove
once_cell. Full removal will break a contract ofglib::Propertiesmacro.