Add [manual_slice_size_calculation]
#10601
Conversation
|
r? @llogiq (rustbot has picked a reviewer for you, use r? to override) |
rustbot
added
the
S-waiting-on-review
schubart
force-pushed
the
manual_slice_size_calculation
branch
from
b78ba71 to
b66aa09
Compare
| fn simplify<'tcx>( | ||
| cx: &LateContext<'tcx>, | ||
| expr1: &'tcx Expr<'tcx>, | ||
| expr2: &'tcx Expr<'tcx>, | ||
| ) -> Option<&'tcx Expr<'tcx>> { | ||
| simplify_half(cx, expr1, expr2).or_else(|| simplify_half(cx, expr2, expr1)) | ||
| } |
There was a problem hiding this comment.
I swear I've seen this pattern multiple times. Perhaps we should make a generic clippy_utils function of it.
There was a problem hiding this comment.
Such a helper could also fix some of the false negatives in my tests, e.g, size_of::<i32>() * 5 * slice.len(). The helper's semantics could be:
Check if this expression is a chain of multiplications (or additions or something more general) where two (or
n?) sub-expressions meet some condition.
I don't have the bandwidth to do it as part of this PR, though.
Done. I feel this kind of data flow analysis (where does a value come from, what do we know about it?) could be a whole new thing, applicable to a lot more than just this lint here. For example, in the following Java code Intellij warns that the List<String> strings = Collections.emptyList();
if (strings.isEmpty()) { // "Condition `strings.isEmpty()` is always `true`.
System.out.println("There are no strings");
}and in this code it warns that the list access will always fail: List<String> strings = foo();
if (strings.isEmpty()) {
System.out.println(strings.get(0)); // "The call to `get` always fails as an argument is out of bounds"
}Do we have anything like this in Clippy? |
|
We have, it's |
|
Thanks, I'll give it a shot. Let's not merge this yet. |
|
Done. Thanks for the suggestion, this is very nice! |
|
Great! @bors r+ |
|
☀️ Test successful - checks-action_dev_test, checks-action_remark_test, checks-action_test |
…Denton Avoid some manual slice length calculation No need for us to write the multiplication when `size_of_val` does exactly what we need. (rust-lang/rust-clippy#10601)
Avoid some manual slice length calculation No need for us to write the multiplication when `size_of_val` does exactly what we need. (rust-lang/rust-clippy#10601)
Fixes: #10518
changelog: new lint [
manual_slice_size_calculation]