pub fn is_encoded_binary(candidate: &str) -> boolExpand description
Conservative verdict for the confidence pipeline: does this generic
candidate decode to identifiable binary / serialized data? Real secrets
return false.
Memoized: a single match is scored on this twice (ML feature #41 in
ml_features and the generic-detector confidence penalty in
confidence::penalties), and a scan re-encounters the same token across
chunks. Without the cache every call re-decodes and re-parses the bytes.
Thread-local + bounded with wholesale eviction, mirroring
entropy::shannon_entropy. The verdict is a pure function of candidate,
so caching by content hash is always correct.