diff options
| author | Nicholas Nethercote <n.nethercote@gmail.com> | 2025-04-25 19:12:15 +1000 |
|---|---|---|
| committer | Nicholas Nethercote <n.nethercote@gmail.com> | 2025-04-26 06:35:44 +1000 |
| commit | 7ac2d1f1bdb2348b13eb4f5fb9c7d620e316034f (patch) | |
| tree | 310683cb906ab06430be2656138ac6e3cfd4a22f /compiler/rustc_mir_transform/src/errors.rs | |
| parent | e37c3674828a75ffdc2b94d9d273e9c02a9ee11f (diff) | |
| download | rust-7ac2d1f1bdb2348b13eb4f5fb9c7d620e316034f.tar.gz rust-7ac2d1f1bdb2348b13eb4f5fb9c7d620e316034f.zip | |
Improve HIR pretty-printing of if/else some more.
In the AST the "then" block is represented as a `Block`. In HIR the "then" block is represented as an `Expr` that happens to always be. `ExprKind::Block`. By deconstructing the `ExprKind::Block` to extract the block within, things print properly. For `issue-82392.rs`, note that we no longer print a type after the "then" block. This is good, it now matches how we don't print a type for the "else" block. (Well, we do print a type after the "else" block, but it's for the whole if/else.) Also tighten up some of the pattern matching -- these block expressions within if/else will never have labels.
Diffstat (limited to 'compiler/rustc_mir_transform/src/errors.rs')
0 files changed, 0 insertions, 0 deletions
