diff options
| author | bors[bot] <26634292+bors[bot]@users.noreply.github.com> | 2021-10-14 22:41:06 +0000 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2021-10-14 22:41:06 +0000 |
| commit | 3a79af7e277839f803bd0304f28dbe4971ffdad6 (patch) | |
| tree | 457e5a6a94502727d645ee7987c807636041caef /src/test/codegen | |
| parent | bfb8f73fb3f8a1a7fadfa31bda98a17549a91093 (diff) | |
| parent | bd9bab87ed72779c3bf407137928b75555f5d2a9 (diff) | |
| download | rust-3a79af7e277839f803bd0304f28dbe4971ffdad6.tar.gz rust-3a79af7e277839f803bd0304f28dbe4971ffdad6.zip | |
Merge #10491
10491: Support nested type on replace if let with match r=k-nasa a=k-nasa ## Why close: https://github.com/rust-analyzer/rust-analyzer/issues/8690 Now, Replacing if-let with match cant't output exhaustive patterns code. This was because the `else` conversion used specific types (ex. Option, Result) instead of wildcards. I thought it was more of a problem to generate non-exhaustive patterns than the benefits of using the concrete one. How about using wildcards in `else`? Is this change policy acceptable? ## What - using wildcards on `make_else_arm` - Change test cases Co-authored-by: k-nasa <htilcs1115@gmail.com>
Diffstat (limited to 'src/test/codegen')
0 files changed, 0 insertions, 0 deletions