[jayhemnani9910]

In the normal format a single added or deleted line still comes out as a range:

$ printf 'a\n' > x; printf 'a\nb\n' > y
$ diffutils diff x y
1a2,2
> b
$ diffutils diff y x
2,2d1
< b
$ /usr/bin/diff x y    # GNU diff 3.12
1a2
> b

GNU diff writes 1a2 and 2d1. -e has the same thing: 2,2d instead of 2d. #11 (for #10) fixed this for changed lines (1c1); added and deleted lines were left out.

This adds the one-line cases next to the existing ones in normal_diff.rs ((0, 1) and (1, 0)) and ed_diff.rs ((1, 0)).

Tests: test_single_line_added_or_deleted (normal) and test_single_line_deleted (ed) check the exact output. Both fail on main. cargo test: all pass (285 on main, 289 here; the two new tests run in both the lib and the bin target), including the permutation tests that feed the output to patch and ed. cargo fmt --check is clean, and cargo clippy --all-targets gives the same warnings as on main. I also compared the default and -e output with GNU diff 3.12 over 22 file pairs (single-line and multi-line adds and deletes, blank lines, missing newlines, empty files): the default format matches for all of them now except binary files. -e still writes its hunks top-down with already-shifted line numbers where GNU goes bottom-up; both scripts give the same file, so I left that alone.

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

An organization owner can choose the repository organization as the billing source for reviews requested by members with a paid Copilot license. Learn about Copilot code review billing. Review the Copilot code review billing settings.

[sylvestre]

could you please run the upstream test suite on it?
TESTS=basic ./tests/run-upstream-testsuite.sh
the whole suite would be even better, it would tell us if it makes more tests pass

[jayhemnani9910]

Ran the whole upstream suite on main (0019867) and on this branch: same on both, 7 pass, 22 fail, 4 skip, and no test changes its result. basic fails on both because of its -999999999999999999999999999999 option case, before it gets to the range output.

[jayhemnani9910]

Ran the full upstream suite (release builds) on main and on this branch: 7 passed, 22 failed, 4 skipped on both, and every test has the same result. basic fails on both at diff -9999999999999999999999999999 a b (unrecognized option), before it reaches anything this changes.