diff options
| author | Eduard-Mihai Burtescu <edy.burt@gmail.com> | 2016-11-12 10:38:42 +0200 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2016-11-12 10:38:42 +0200 |
| commit | fb922772911704f1b30d046cef3f683eebf14dfd (patch) | |
| tree | 66ad3cc14ba8390c2a55d87f7d283e775041f8e6 /src | |
| parent | 220ff76e0b1b97c7deef11e320d0cb3b3032c73e (diff) | |
| parent | 4ce4900218bd8378e08cd7d1d4c57e370c0949a0 (diff) | |
| download | rust-fb922772911704f1b30d046cef3f683eebf14dfd.tar.gz rust-fb922772911704f1b30d046cef3f683eebf14dfd.zip | |
Rollup merge of #37692 - nbp:debug-lvalue-subslice, r=eddyb
Balance the debug output of Lvalue Subslice The current debug output for Lvalue Subslice is not balanced and does not respect the comment[1], which indicates that we use `slice[from:-to] in Python terms.`. In python terms slices which have a start but no end are written as `a[start:]`, so following the comment, I fixed the output accordingly. Grep-ing over the sources, I did not found any test cases checking this subslice debug output. Note, I have not yet tested this change yet, as I am still waiting for the end of LLVM compilation. [1] https://manishearth.github.io/rust-internals-docs/rustc/mir/enum.ProjectionElem.html
Diffstat (limited to 'src')
| -rw-r--r-- | src/librustc/mir/mod.rs | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/src/librustc/mir/mod.rs b/src/librustc/mir/mod.rs index 9d82006ac9f..b3cec6ec8ff 100644 --- a/src/librustc/mir/mod.rs +++ b/src/librustc/mir/mod.rs @@ -923,7 +923,7 @@ impl<'tcx> Debug for Lvalue<'tcx> { ProjectionElem::ConstantIndex { offset, min_length, from_end: true } => write!(fmt, "{:?}[-{:?} of {:?}]", data.base, offset, min_length), ProjectionElem::Subslice { from, to } if to == 0 => - write!(fmt, "{:?}[{:?}:", data.base, from), + write!(fmt, "{:?}[{:?}:]", data.base, from), ProjectionElem::Subslice { from, to } if from == 0 => write!(fmt, "{:?}[:-{:?}]", data.base, to), ProjectionElem::Subslice { from, to } => |