Implement DerefMut for String #26241

SimonSapin · Jun 12, 2015

rust-highfive · Jun 12, 2015

Thanks for the pull request, and welcome! The Rust team is excited to review your changes, and you should hear from @aturon (or someone else) soon.

If any changes to this PR are deemed necessary, please add them as extra commits. This ensures that the reviewer can see what has changed since they last reviewed the code. The way Github handles out-of-date commits, this should also make it reasonably obvious what issues have or haven't been addressed. Large or tricky changes may require several passes of review and changes.

Please see the contribution instructions for more information.

SimonSapin · Jun 12, 2015

Is it also time to look at either deprecating or stabilizing OwnedAsciiExt and make_ascii_*case?

bluss · Jun 12, 2015

Those casing ascii methods are pretty broken in unicode context, but I guess we could allow arbitrary ascii -> ascii mappings. Since a grapheme can be multiple codepoints, this also edits / transforms graphemes that use at least one ascii codepoint, for example é (depending on normalization).

SimonSapin · Jun 12, 2015

@bluss Agreed, the AsciiExt methods are for very narrow use case like CSS property names (which are entirely within ASCII), which is why they’re not inherent or in the prelude. We also have to_lowercase and to_uppercase on char and str that do use the full Unicode mappings.

alexcrichton · Jun 12, 2015

I would be more uncomfortable with stabilizing make_ascii_{upper,lower}case (primarily in terms of naming), but I'm ok with this. I'm going to tag this with T-libs and I-needs-decision to ensure it gets some visibility.

SimonSapin · Jun 12, 2015

OK. make_ascii_*case is a motivation for this change, but changing or stabilising would be a separate decision.

Gankro · Jun 12, 2015

I say we just land DerefMut and call it a day. Other stuff is easy to impl outside std.

bluss · Jun 13, 2015

I don't want to complicate this issue, but we might want to add slicing of &mut str and .split_at_mut(mid), the latter being an important crutch for slices already to avoid needing unsafe sections.

SimonSapin · Jun 15, 2015

Added a commit:

Implement IndexMut for String and str.

... matching the existing Index impls.
There is no reason not to if String implement DerefMut.

The code removed in src/librustc/middle/effect.rs was added in #9750
to prevent things like s[0] = 0x80 where s: String,
but I belive became unnecessary when the Index(Mut) traits were introduced.

aturon · Jun 16, 2015

I'm 👍 on this change in general, but are we OK landing this insta-stable? (Since it's just filling out APIs according to vetted/stable convention, seems fine to me, but I wanted to check.)

SimonSapin · Jun 16, 2015

str::split_at_mut is unstable (like the existing str::split_at) but yes, the rest (DerefMut and IndexMut impls as well as str::slice_mut_unchecked) is insta-stable in the current form of this PR. I don’t mind changing it.

aturon · Jun 16, 2015

Hm, I was also forgetting that implementations of stable traits are instantly stable regardless. I'm happy with the PR as it stands.

alexcrichton · Jun 16, 2015

This PR is now entering its final comment period. There was some discussion at triage today about this and there was some hesitation about the power that DerefMut is adding, so we decided to FCP it instead of merging it immediately!

SimonSapin · Jun 16, 2015

After seeing #9750 again I’ve been pondering on whether this PR introduces a way to break the UTF-8 invariant in safe code. I believe it does not:

With x: &mut T, *T can only be assigned to directly when T is sized. Since str: !Sized, &mut str can not be assigned to directly.
The IndexMut impls in this PR all have Target=str, so from a &mut str they can only give another &mut str.
Although Index<usize, Target=u8> for str would make sense (and did sort of exist in Rust once upon a time: #9750) but is not implemented. The corresponding IndexMut would allow string mutation, but it’s not implemented either.
The above leaves as the only way to mutate &mut str from safe code using safe functions/methods like AsciiExt::make_ascii_lowercase that use unsafe code (such as transmute::<&mut str, &mut [u8]>) internally. It’s up to these functions/methods to preserve the invariant in the safe abstraction they provide.

bors · Jun 18, 2015

☔️ The latest upstream changes (presumably #26192) made this pull request unmergeable. Please resolve the merge conflicts.

SimonSapin · Jun 19, 2015

I’ll rebase once a decision is made on what to include or not.

alexcrichton · Jul 1, 2015

alexcrichton reviewed Jul 1, 2015

View changes

alexcrichton · Jul 1, 2015

src/libcollections/str.rs

@@ -806,6 +814,12 @@ impl str {
        core_str::StrExt::split_at(self, mid)
    }

+    /// Divide one mutable string slice into two at an index.
+    #[inline]
+    pub fn split_at_mut(&mut self, mid: usize) -> (&mut str, &mut str) {


Can you tag this with an explicit unstable tag?

alexcrichton · Jul 1, 2015

alexcrichton reviewed Jul 1, 2015

View changes

alexcrichton · Jul 1, 2015

src/libcollections/str.rs

+    /// Takes a bytewise mutable slice from a string.
+    ///
+    /// Same as `slice_unchecked`, but works with `&mut str` instead of `&str`.
+    #[stable(feature = "derefmut_for_string", since = "1.2.0")]


Can this enter as unstable?

alexcrichton · Jul 1, 2015

The consensus of the libs team is to move ahead with this PR, with a rebase and a few nits addressed this should be good to go, thanks @SimonSapin!

SimonSapin · Jul 1, 2015

Rebased and added a commit to address comments. This commits also adds #[unstable] to str::split_at, which existed before this PR but did not have any stability marker.

I’m getting an LLVM assertion error when building locally :( Could this be specific to stage1? I haven’t tried stage2 yet.

../../projects/rust/src/libstd/lib.rs:1:1: 1:1 note: expansion site
rustc: ../../../../../../../projects/rust/src/llvm/lib/CodeGen/AsmPrinter/DwarfUnit.cpp:525: bool isUnsignedDIType(llvm::DwarfDebug*, const llvm::DIType*): Assertion `T == dwarf::DW_TAG_typedef || T == dwarf::DW_TAG_const_type || T == dwarf::DW_TAG_volatile_type || T == dwarf::DW_TAG_restrict_type || T == dwarf::DW_TAG_enumeration_type' failed.
/home/simon/projects/rust/mk/tests.mk:394: recipe for target 'x86_64-unknown-linux-gnu/stage1/test/stdtest-x86_64-unknown-linux-gnu' failed
make: *** [x86_64-unknown-linux-gnu/stage1/test/stdtest-x86_64-unknown-linux-gnu] Aborted (core dumped)

SimonSapin · Jul 7, 2015

I’m still getting the same LLVM error after rebasing on today’s Rust master. Any suggestion what to do here?

alexcrichton · Jul 7, 2015

Unfortunately I don't know how to get past that error.

pnkfelix · Jul 13, 2015

This seems like it has been blocked for 6 days due to #26484. However this is unnecessary, assuming it does not add any new breakage to --enable-debug --enable-optimize, since we have no bot that tests that configuration (see also #27010).

pnkfelix · Jul 13, 2015

After discussion with @SimonSapin I am interpreting this statement from @alexcrichton 👍

The consensus of the libs team is to move ahead with this PR, with a rebase and a few nits addressed this should be good to go, thanks @SimonSapin!

as an implicit r=me for the PR.

pnkfelix · Jul 13, 2015

@bors r=alexcrichton 0375d9c

bors · Jul 13, 2015

🙀 0375d9c is not a valid commit SHA. Please try again with 23daca3.

SimonSapin · Jul 13, 2015

doener on IRC suggested that I might be hitting #26484. After running configure again with default flags (I’m not sure what I had before), make check-stage1-std succeeded locally.

SimonSapin · Jul 13, 2015

I just pushed a rebase.

pnkfelix · Jul 13, 2015

@bors r=alexcrichton eb99f0e

bors · Jul 13, 2015

⌛️ Testing commit eb99f0e with merge 13aed07...

bors · Jul 13, 2015

💔 Test failed - auto-mac-64-opt

SimonSapin · Jul 13, 2015

@alexcrichton With that last commit make check-stage1 (not just check-stage1-std) passes locally.

alexcrichton · Jul 13, 2015

@bors: r+ 3226858

bors · Jul 13, 2015

⌛️ Testing commit 3226858 with merge dd46cf8...

bors · Jul 14, 2015

☀️ Test successful - auto-linux-32-nopt-t, auto-linux-32-opt, auto-linux-64-nopt-t, auto-linux-64-opt, auto-linux-64-x-android-t, auto-mac-32-opt, auto-mac-64-nopt-t, auto-mac-64-opt, auto-win-gnu-32-nopt-t, auto-win-gnu-32-opt, auto-win-gnu-64-nopt-t, auto-win-gnu-64-opt

rust-highfive assigned aturon Jun 12, 2015

alexcrichton added I-needs-decision T-libs labels Jun 12, 2015

brson added the relnotes label Jun 12, 2015

SimonSapin force-pushed the SimonSapin:derefmut-for-string branch from a61076a to ef8ac71 Jun 15, 2015

alexcrichton added final-comment-period and removed I-needs-decision labels Jun 16, 2015

alexcrichton removed the final-comment-period label Jul 1, 2015

SimonSapin force-pushed the SimonSapin:derefmut-for-string branch from 285feb6 to 28357e5 Jul 1, 2015

SimonSapin force-pushed the SimonSapin:derefmut-for-string branch from 28357e5 to 23daca3 Jul 1, 2015

SimonSapin force-pushed the SimonSapin:derefmut-for-string branch from 23daca3 to eb99f0e Jul 13, 2015

bors added a commit that referenced this pull request Jul 13, 2015

Auto merge of #26241 - SimonSapin:derefmut-for-string, r=alexcrichton
See rust-lang/rfcs#1157

13aed07

Fix tests for changes in #26241.

Loading status checks…

3226858

SimonSapin force-pushed the SimonSapin:derefmut-for-string branch from 88fe063 to 3226858 Jul 13, 2015

bors added a commit that referenced this pull request Jul 13, 2015

Auto merge of #26241 - SimonSapin:derefmut-for-string, r=alexcrichton
See rust-lang/rfcs#1157

Loading status checks…

dd46cf8

bors merged commit 3226858 into rust-lang:master Jul 14, 2015
2 checks passed

2 checks passed

continuous-integration/travis-ci/pr The Travis CI build passed
Details

homu Test successful
Details

bors referenced this pull request Jul 14, 2015
Merged
Document Unicode complications when iterating "characters" #27012

SimonSapin deleted the SimonSapin:derefmut-for-string branch Jul 18, 2015

bluss referenced this pull request Sep 15, 2015
Merged
1.3 relnotes #28408

rust-lang/rust

Join GitHub today

Implement DerefMut for String #26241

Conversation

SimonSapin commented Jun 12, 2015

rust-highfive assigned aturon Jun 12, 2015

This comment has been minimized.

rust-highfive commented Jun 12, 2015

This comment has been minimized.

SimonSapin commented Jun 12, 2015

This comment has been minimized.

bluss commented Jun 12, 2015

This comment has been minimized.

SimonSapin commented Jun 12, 2015

This comment has been minimized.

alexcrichton commented Jun 12, 2015

alexcrichton added I-needs-decision T-libs labels Jun 12, 2015

This comment has been minimized.

SimonSapin commented Jun 12, 2015

This comment has been minimized.

Gankro commented Jun 12, 2015

brson added the relnotes label Jun 12, 2015

This comment has been minimized.

bluss commented Jun 13, 2015

This comment has been minimized.

SimonSapin commented Jun 15, 2015

SimonSapin force-pushed the SimonSapin:derefmut-for-string branch from a61076a to ef8ac71 Jun 15, 2015

This comment has been minimized.

aturon commented Jun 16, 2015

This comment has been minimized.

SimonSapin commented Jun 16, 2015

This comment has been minimized.

aturon commented Jun 16, 2015

alexcrichton added final-comment-period and removed I-needs-decision labels Jun 16, 2015

This comment has been minimized.

alexcrichton commented Jun 16, 2015

This comment has been minimized.

SimonSapin commented Jun 16, 2015

This comment has been minimized.

bors commented Jun 18, 2015

This comment has been minimized.

SimonSapin commented Jun 19, 2015

alexcrichton reviewed Jul 1, 2015 View changes

This comment has been minimized.

alexcrichton Jul 1, 2015

alexcrichton reviewed Jul 1, 2015 View changes

This comment has been minimized.

alexcrichton Jul 1, 2015

This comment has been minimized.

alexcrichton commented Jul 1, 2015

alexcrichton removed the final-comment-period label Jul 1, 2015

SimonSapin force-pushed the SimonSapin:derefmut-for-string branch from 285feb6 to 28357e5 Jul 1, 2015

This comment has been minimized.

SimonSapin commented Jul 1, 2015

SimonSapin force-pushed the SimonSapin:derefmut-for-string branch from 28357e5 to 23daca3 Jul 1, 2015

This comment has been minimized.

SimonSapin commented Jul 7, 2015

This comment has been minimized.

alexcrichton commented Jul 7, 2015

SimonSapin added some commits Jun 12, 2015

This comment has been minimized.

pnkfelix commented Jul 13, 2015

This comment has been minimized.

pnkfelix commented Jul 13, 2015

This comment has been minimized.

pnkfelix commented Jul 13, 2015

This comment has been minimized.

bors commented Jul 13, 2015

This comment has been minimized.

SimonSapin commented Jul 13, 2015

SimonSapin force-pushed the SimonSapin:derefmut-for-string branch from 23daca3 to eb99f0e Jul 13, 2015

This comment has been minimized.

SimonSapin commented Jul 13, 2015

This comment has been minimized.

pnkfelix commented Jul 13, 2015

This comment has been minimized.

bors commented Jul 13, 2015

bors added a commit that referenced this pull request Jul 13, 2015

This comment has been minimized.

bors commented Jul 13, 2015

SimonSapin force-pushed the SimonSapin:derefmut-for-string branch from `a61076a` to `ef8ac71` Jun 15, 2015

alexcrichton reviewed Jul 1, 2015

View changes

alexcrichton reviewed Jul 1, 2015

View changes

SimonSapin force-pushed the SimonSapin:derefmut-for-string branch from `285feb6` to `28357e5` Jul 1, 2015

SimonSapin force-pushed the SimonSapin:derefmut-for-string branch from `28357e5` to `23daca3` Jul 1, 2015

SimonSapin force-pushed the SimonSapin:derefmut-for-string branch from `23daca3` to `eb99f0e` Jul 13, 2015

SimonSapin force-pushed the SimonSapin:derefmut-for-string branch from `88fe063` to `3226858` Jul 13, 2015

bors merged commit `3226858` into rust-lang:master Jul 14, 2015
2 checks passed