cut Command Reference for CI Scripts

cut pulls specific columns (fields, characters, or bytes) from each line.

cut is the lightweight column extractor. Its main limitation versus awk: the delimiter is a single character and runs of whitespace are not collapsed.

Common flags/usage

• -f LIST: select fields (1, 2-4, 3-)
• -d CHAR: field delimiter (single char, default TAB)
• -c LIST: select character positions
• -b LIST: select byte positions
• --complement: invert the selection (GNU only)

Example

echo "${GITHUB_REF}" | cut -d/ -f3-       # refs/heads/main -> main
cut -d, -f1,3 results.csv
git log -1 --format='%H' | cut -c1-7      # short commit SHA

In CI

The delimiter must be a single character; you cannot split on a string or regex (use awk for that). cut does not collapse multiple spaces, so columns aligned with variable spacing come out wrong; pre-process with tr -s or use awk. --complement is GNU-only and missing on BSD/macOS.

Key takeaways

• cut splits on a single delimiter character only; use awk for strings or regex.
• It does not collapse repeated spaces, so squeeze whitespace first when needed.
• Field ranges like 2- and 1,3 select open-ended and discontinuous columns.

Related guides