support ordering magic methods for #[pyclass]
#3203
Conversation
davidhewitt
force-pushed
the
pyclass-eq
branch
2 times, most recently
from
c3b7a2c
to
a477688
Compare
|
As the new behavior may break users who were attempting to use |
I think this should happen only when that boiler plate reduction is in place. |
Completely agree. This PR is intended to be an addition only. I'd propose to then work on the opt-in implementation of |
|
(Pushed the wording adjustment as a new commit, when we're happy with what's going on here I'll squash.) |
|
Thanks, squashed! bors r=adamreichold |
3203: support ordering magic methods for `#[pyclass]` r=adamreichold a=davidhewitt Closes #2089 This adds `__lt__`, `__le__`, `__eq__`, `__ne__`, `__gt__`, and `__ge__` as per the Python implementations of what we call `__richcmp__`. There's a UI test confirming that the user cannot implement split forms and `__richcmp__` simultaneously. There's also a benchmark comparing implementing these split methods against using `__richcmp__`. I couldn't see a meaningful performance difference, so I'm tempted to deprecate `__richcmp__`, given that's not a magic method which exists in Python. Potentially we can provide options such as the opt-in `#[pyclass(eq, ord)]` to avoid boilerplate for people who don't want to implement six different methods. Co-authored-by: David Hewitt <1939362+davidhewitt@users.noreply.github.com>
|
Build failed: |
|
bors retry |
|
Build succeeded! The publicly hosted instance of bors-ng is deprecated and will go away soon. If you want to self-host your own instance, instructions are here. If you want to switch to GitHub's built-in merge queue, visit their help page. |
Closes #2089
This adds
__lt__,__le__,__eq__,__ne__,__gt__, and__ge__as per the Python implementations of what we call__richcmp__.There's a UI test confirming that the user cannot implement split forms and
__richcmp__simultaneously.There's also a benchmark comparing implementing these split methods against using
__richcmp__. I couldn't see a meaningful performance difference, so I'm tempted to deprecate__richcmp__, given that's not a magic method which exists in Python. Potentially we can provide options such as the opt-in#[pyclass(eq, ord)]to avoid boilerplate for people who don't want to implement six different methods.