Skip to content

#5270: Change int/uint range_rev to use (hi,lo] instead of [hi,lo). - #7684

Closed
pnkfelix wants to merge 3 commits into
rust-lang:masterfrom
pnkfelix:fsk-invert-range-rev-halfclosedness-issue5270-2ndpr
Closed

#5270: Change int/uint range_rev to use (hi,lo] instead of [hi,lo).#7684
pnkfelix wants to merge 3 commits into
rust-lang:masterfrom
pnkfelix:fsk-invert-range-rev-halfclosedness-issue5270-2ndpr

Conversation

@pnkfelix

Copy link
Copy Markdown
Contributor

Changes int/uint range_rev to iterate over range (hi,lo] instead of [hi,lo).

Fix#5270.

Also:

  • Adds unit tests for int/uint range functions
  • Updates the uses of range_rev to account for the new semantics. (Note that pretty much all of the updates there were strict improvements to the code in question; yay!)
  • Exposes new function, range_step_inclusive, which does the range [hi,lo], (at least when hi-lo is a multiple of the step parameter).
  • Special-cases when |step| == 1 removing unnecessary bounds-check. (I did not check whether LLVM was already performing this optimization; I figure it would be a net win to not leave that analysis to the compiler. If reviewer objects, I can easily remove that from the refactored code.)

(This pull request is a rebased version of PR #7524, which went stale due to recent unrelated changes to num libraries.)

… semantics.
Also added unit tests of range code to test refactoring. The
num-range-rev.rs test will need to be updated when the range_rev
semantics change.
@pnkfelix

Copy link
Copy Markdown
ContributorAuthor

hmm, seems like the make check failures are actually due to the refactoring I made, rather than the change to the range_rev semantics itself. Curiouser and curiouser.

@pnkfelix

Copy link
Copy Markdown
ContributorAuthor

From what I can tell, the make check failures are actually reproducible on master, without my refactoring. It is just that my refactoring makes them more reliably reproducible.

I spent some time tracking this down and the best I have come up with is Issue #7797, but I am not 100% certain that that is the same bug.

bors added a commit that referenced this pull request Jul 16, 2013
…ss-issue5270-2ndpr, r=cmr
Changes int/uint range_rev to iterate over range `(hi,lo]` instead of `[hi,lo)`.
Fix#5270.
Also:
* Adds unit tests for int/uint range functions
* Updates the uses of `range_rev` to account for the new semantics. (Note that pretty much all of the updates there were strict improvements to the code in question; yay!)
* Exposes new function, `range_step_inclusive`, which does the range `[hi,lo]`, (at least when `hi-lo` is a multiple of the `step` parameter).
* Special-cases when `|step| == 1` removing unnecessary bounds-check. (I did not check whether LLVM was already performing this optimization; I figure it would be a net win to not leave that analysis to the compiler. If reviewer objects, I can easily remove that from the refactored code.)
(This pull request is a rebased version of PR #7524, which went stale due to recent unrelated changes to num libraries.)
@borsbors closed this Jul 16, 2013
flip1995 pushed a commit to flip1995/rust that referenced this pull request Oct 7, 2021
…1995
fix for issue rust-lang#7683Fixesrust-lang#7683.
For Repeat [x; y] (x is the type and y is the times to repeat) . When y > 32, the compiler will report an error:
https://play.rust-lang.org/?version=stable&mode=debug&edition=2018&gist=7148558162685e91056e0550797ea74c
Because https://github.com/rust-lang/rust/blob/6cdd42f9f8dd4e5e5ba0aa816bc4c99ab8b102f9/library/std/src/primitive_docs.rs#L538
/// Arrays of sizes from 0 to 32 (inclusive) implement [`Default`] trait
/// if the element type allows it. As a stopgap, trait implementations are
/// statically generated up to size 32.
So here to detect this situation.
changelog: [`derivable_impls`]: No longer lints when arrays bigger than 32 elements are involved
U007D pushed a commit to U007D/rust-mos that referenced this pull request Aug 21, 2026
7684: Actually fix runnable order r=lnicola a=ivan770
Changes
![image](https://user-images.githubusercontent.com/14003886/107956771-51a6b200-6fa8-11eb-91aa-f4190d6a8ccd.png)
to
![image](https://user-images.githubusercontent.com/14003886/107956926-903c6c80-6fa8-11eb-8fb9-8bf7bb1879ac.png)
Sorry for missing this one in original PR. This was an issue before rust-lang/rust-analyzer@ee049b2, and I fixed it, yet *somehow* order got reversed (to a logical one)
Co-authored-by: ivan770 <leshenko.ivan770@gmail.com>
Sign up for freeto join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Change range_rev to be inclusive on the low end

2 participants

@pnkfelix@bors