Module candidate_record 

Pure parsing of raw NUL-delimited changed-path records. What: Parse the NUL-delimited raw records of git diff-index and git diff-tree. Why: One listing process names every changed path with its mode, object and change kind, so no per-path Git process is ever needed. Pathnames are kept as the raw bytes Git printed; they are never decoded.

Git 2.56.0 (Documentation/diff-format.adoc, “RAW OUTPUT FORMAT”) prints, with -z: :<old mode> SP <new mode> SP <old oid> SP <new oid> SP <status> NUL <path> NUL.

In TS you’d write (pseudocode):

// const records = parseRawRecords(stdout); // [{ path, mode, change, object }]

Structs

CandidateRecord

What: One listed path. Vec<u8> is an owned byte list (sibling String would require UTF-8). Option<ObjectId> is “an object name or nothing”. Why: Git pathnames are arbitrary bytes on Unix and must round-trip exactly. A deleted path has no object on the candidate side, which the type states instead of carrying Git’s all-zero placeholder name.

Enums

CandidateChange

What: How a path differs from the baseline it was compared with. #[derive(...)] generates copying, debug printing and ==. Why: A deleted path has no content to read, and policies may treat new files differently from changed ones.

Functions

parse_raw_records

What: Parse complete raw -z output into records, in Git’s order. &[u8] borrows the output; the records own copies of what they keep. Why: Each record is two NUL-terminated tokens. Output that does not end with a NUL, or that holds an odd number of tokens, was cut short or is not raw output, and is refused whole: a partial list would leave paths unchecked.