defmt-print: Recover from decoding-errors
#598
Conversation
print/src/main.rs
Outdated
| Encoding::Raw => return Err(DecodeError::Malformed.into()), | ||
| // rzcobs encoding allows recovery from decoding-errors. therefore we stop | ||
| // decoding the current, corrupted, data and continue with new one | ||
| Encoding::Rzcobs => break, |
There was a problem hiding this comment.
I think this should be continue because may be more data cached in the stream_decoder. if there isn't an entire frame then the next iteration will hit the Eof case (which will make the program fetch more data from stdin)
print/src/main.rs
Outdated
| match stream_decoder.decode() { | ||
| Ok(frame) => forward_to_logger(&frame, location_info(&locs, &frame, ¤t_dir)), | ||
| Err(DecodeError::UnexpectedEof) => break, | ||
| Err(DecodeError::Malformed) => match table.encoding() { |
There was a problem hiding this comment.
could you move the 'can this encoding recover?' logic into a method on the Encoding enum itself? I think consumers should not be deciding on the Encoding variant but rather the defmt-decoder library should provide them that information.
… `recoverable() -> bool`
print/src/main.rs
Outdated
| match stream_decoder.decode() { | ||
| Ok(frame) => forward_to_logger(&frame, location_info(&locs, &frame, ¤t_dir)), | ||
| Err(DecodeError::UnexpectedEof) => break, | ||
| Err(DecodeError::Malformed) => match table.encoding().recoverable() { |
There was a problem hiding this comment.
(probably down to personal preference but if boolean is more common than match boolean IMO)
There was a problem hiding this comment.
I think it is more common to use if. But match saves one line 😋
|
bors r+ |
|
Build failed: |
|
bors retry |
|
Build succeeded: |
This PR uses the new
StreamDecoder-API indefmt-printand utilizes it to recover from decoding errors.Fixes #559