Struct BlockIntegrityScanner 

pub struct BlockIntegrityScanner {
    pub registry: HashMap<String, ScanRecord>,
    pub config: ScannerConfig,
    pub stats: ScannerStats,
}

Scans stored blocks for integrity issues.

Callers first register_block each known block, then call scan_block (or scan_all) whenever they want to verify the live data against the registry.

Fields

registry: HashMap<String, ScanRecord>

Registry of all known blocks, keyed by CID string.

config: ScannerConfig

Configuration controlling which checks are performed.

stats: ScannerStats

Running statistics.

Implementations

impl BlockIntegrityScanner

pub fn new(config: ScannerConfig) -> Self

Creates a new scanner with the supplied configuration and an empty registry.

pub fn register_block( &mut self, cid: String, expected_size: u64, data: &[u8], now_secs: u64, )

Registers a block in the scanner’s registry.

The FNV-1a hash of data is computed and stored alongside the supplied metadata so that future scans can detect hash mismatches.

Parameters

• cid – Content identifier for the block.
• expected_size – Expected size of the block in bytes.
• data – Raw block data used to compute the stored hash.
• now_secs – Current Unix timestamp in seconds (caller-supplied so the scanner remains deterministic in tests).

pub fn scan_block( &mut self, cid: &str, actual_data: &[u8], scan_time_us: u64, ) -> ScanResult

Scans a single block and returns a ScanResult describing any integrity issues found.

The following checks are performed in order:

1. Missing block – If cid is not in the registry the result contains a single IntegrityIssue::MissingBlock and all other checks are skipped.
2. Size mismatch – If actual_data.len() as u64 != expected_size a IntegrityIssue::SizeViolation is added.
3. Hash verification – When verify_cid_hash is true, the FNV-1a hash of actual_data is compared with the stored hash; a difference adds a IntegrityIssue::CidMismatch.
4. Magic bytes – When magic_bytes is non-empty and actual_data does not start with those bytes, a IntegrityIssue::CorruptionMarker is added at offset = 0.
5. Max size – If actual_data.len() as u64 > max_size_bytes a IntegrityIssue::SizeViolation is added (using max_size_bytes as expected).

Stats are updated before returning.

pub fn scan_all( &mut self, blocks: &[(String, Vec<u8>)], scan_time_us: u64, ) -> Vec<ScanResult>

Scans all supplied blocks and returns one ScanResult per block.

Each element in blocks is a (cid, data) pair. The same scan_time_us value is recorded for every block in the batch.

pub fn healthy_blocks(&self) -> Vec<&ScanRecord>

Returns references to all ScanRecords in the registry.

This does not filter by scan history; it simply exposes the full contents of the registry at the time of the call.

pub fn stats(&self) -> &ScannerStats

Returns a reference to the scanner’s accumulated statistics.