peczenyj / GDPR-IAB-TCFv2 / 25593411765

96%

main: 96%

Pull #111

github

feat(cli): short-help via -h and informative unknown-subcommand error

Two CLI usability changes, ripgrep-style.

1. `-h` and `--help` are now distinct.

   * `-h` prints a brief summary: synopsis line, common options (with
     short alias, long name, value placeholder, one-line description),
     and a pointer to `--help`. Available at the top level and per
     subcommand:

         iabtcfv2 -h                  # subcommand list + pointers
         iabtcfv2 dump -h             # short summary for dump
         iabtcfv2 validate -h         # short summary for validate

   * `--help` (and `iabtcfv2 help [<sub>]`) keep showing the full POD
     they always did. No behavior change for `--help`.

   * A bare `iabtcfv2` (no args, no flags) now prints the same short
     overview as `iabtcfv2 -h` rather than the full POD.

2. Unknown subcommand prints an informative error and exits 2 with a
   subcommand list and a pointer to `-h` / `--help`:

       iabtcfv2: unknown subcommand 'frobnicate'

       Available subcommands:
           dump       Parse TC strings and emit JSON.
           validate   Validate TC strings against ...

       Run 'iabtcfv2 -h' for a short overview, or 'iabtcfv2 --help' ...

   Replaces the previous one-line "Unknown subcommand: frobnicate"
   pod2usage output.

Implementation: a top-level %SUBCOMMANDS registry drives dispatch, the
short-help summaries, and the unknown-subcommand error message. New
subcommands plug in there with no other code edits. Niche flags
(--cmp-validator-timeout, --cmp-validator-network-ok, etc.) are
intentionally omitted from the `common` rows and remain documented in
the full POD only.

POD updated to describe the -h / --help split. Added 18 tests in
t/10-cli-iabtcfv2.t covering top-level short help, bare invocation,
both subcommand short helps, and the unknown-subcommand error path.

Tests: 17,620 t/ + xt/ tests pass; perltidy/critic/synopsis clean.

55 of 57 new or added lines in 1 file covered. (96.49%)

1157 of 1202 relevant lines covered (96.26%)

8836.18 hits per line