|
|
|
|
|
by chrsm
2689 days ago
|
|
The real issue here is that it is not immediately obvious that this is actually an interface - one expects the second `err` to be *FormatError due to the signature - it happens to implement the error interface so it goes unnoticed. Shadowing in a different scope allows it to work properly as well. I admit that I've never actually run in to this issue in practice as I generally return custom errors as `error`, but I can see why it would confuse someone. |
|