Struct CrcCorrector

pub struct CrcCorrector<const L: usize, W: Width> { /* private fields */ }

CRC Corrector

Associated const CrcCorrector::L refers to message length in bits including an appended CRC

Implementations

impl<const L: usize> CrcCorrector<L, u16>

pub const fn new(crc: Crc<u16, Table<1>>) -> Self

Construct CRC Corrector, including lookup table, for a given CRC algorithm and message length.

Panics

This function will panic if

• The message length is not a multiple of 8. (We expect byte data)
• The requested table length is too large
• The CRC algorithm cannot reliably perform single bit correction with the required message length

In the last case the problem is similar to a hash collision. Either choose a longer CRC (64 bit vs 32 bit vs 16 bit) or a different algorithm. Ultimately there is a limit to the message length a CRC can correct errors for.

pub fn correct( &self, data: &mut [u8], crc: u16, ) -> Result<Correction<u16>, Error>

Correct message with a single bit of corruption in either the provided message or the provided CRC. This method mutates the message to correct corruption but will not mutate it if correction fails.

If a correction is applied the CRC is validated again to ensure that the integrity of the data is okay. This isn’t strictly necessary, but guards against any bugs which would incorrectly ‘correct’ data.

Errors

This method will either return an error if correction could not be applied, or will mutate the data with a single bit correction and return an indication of which bit was the issue.

impl<const L: usize> CrcCorrector<L, u32>

pub const fn new(crc: Crc<u32, Table<1>>) -> Self

Construct CRC Corrector, including lookup table, for a given CRC algorithm and message length.

Panics

This function will panic if

• The message length is not a multiple of 8. (We expect byte data)
• The requested table length is too large
• The CRC algorithm cannot reliably perform single bit correction with the required message length

In the last case the problem is similar to a hash collision. Either choose a longer CRC (64 bit vs 32 bit vs 16 bit) or a different algorithm. Ultimately there is a limit to the message length a CRC can correct errors for.

pub fn correct( &self, data: &mut [u8], crc: u32, ) -> Result<Correction<u32>, Error>

Correct message with a single bit of corruption in either the provided message or the provided CRC. This method mutates the message to correct corruption but will not mutate it if correction fails.

If a correction is applied the CRC is validated again to ensure that the integrity of the data is okay. This isn’t strictly necessary, but guards against any bugs which would incorrectly ‘correct’ data.

Errors

This method will either return an error if correction could not be applied, or will mutate the data with a single bit correction and return an indication of which bit was the issue.

impl<const L: usize> CrcCorrector<L, u64>

pub const fn new(crc: Crc<u64, Table<1>>) -> Self

Construct CRC Corrector, including lookup table, for a given CRC algorithm and message length.

Panics

This function will panic if

• The message length is not a multiple of 8. (We expect byte data)
• The requested table length is too large
• The CRC algorithm cannot reliably perform single bit correction with the required message length

In the last case the problem is similar to a hash collision. Either choose a longer CRC (64 bit vs 32 bit vs 16 bit) or a different algorithm. Ultimately there is a limit to the message length a CRC can correct errors for.

pub fn correct( &self, data: &mut [u8], crc: u64, ) -> Result<Correction<u64>, Error>

Correct message with a single bit of corruption in either the provided message or the provided CRC. This method mutates the message to correct corruption but will not mutate it if correction fails.

If a correction is applied the CRC is validated again to ensure that the integrity of the data is okay. This isn’t strictly necessary, but guards against any bugs which would incorrectly ‘correct’ data.

Errors

This method will either return an error if correction could not be applied, or will mutate the data with a single bit correction and return an indication of which bit was the issue.

Trait Implementations

impl<const L: usize, W: Clone + Width> Clone for CrcCorrector<L, W>

fn clone(&self) -> CrcCorrector<L, W>

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.