#!/usr/bin/env bash
#
# Check a commit message's Conventional Commits scope against plan/.
#
# An epic's id is its commit scope, so `feat(camo): ...` belongs to
# plan/camo.md. This catches the two ways that goes wrong: a typo, and a scope
# naming an epic nobody has written down.
#
# Valid ids are the filenames in plan/ and plan/complete/, where finished
# epics are archived. It used to read plan/README.md's tables as well, for an
# epic that was a row before it was a file; those tables are generated from
# the files now (scripts/gen-plan-readme.py), so a row that no file backs
# cannot exist and reading them would only find the same ids twice.
#
# It is not a rule that every commit belongs to an epic. A commit with no
# scope passes, and so does anything in NON_EPIC_SCOPES — build work,
# dependency bumps and repo chores are not epics and should not have to
# pretend otherwise. What fails is a scope that is neither.
#
# Run by prek at the commit-msg stage, with the message file as $1.

set -euo pipefail

# Scopes that are deliberately not epics. Keep this short: every entry is a
# commit that cannot be found from plan/.
NON_EPIC_SCOPES="plan deps scripts deploy ci"

# The component scopes this repo used before epics. They still pass so that
# adopting the hook does not reject ordinary work, but prefer the epic: `web`
# says which directory changed, `camo` says what the change was for. This list
# should shrink to nothing.
LEGACY_SCOPES="web api matches manifest todo"

msg_file="${1:?usage: check-commit-scope.sh <commit-msg-file>}"
repo_root="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)"
plan_dir="$repo_root/plan"

# First line that is not a comment or blank; git strips the comments later.
subject="$(grep -v '^#' "$msg_file" | grep -m1 . || true)"

# An empty message aborts the commit on git's side, so it is not this hook's
# to complain about.
[ -n "$subject" ] || exit 0

# type(scope)!: subject. No parentheses means no scope, which is allowed. The
# pattern is a variable because bash's [[ ]] parser trips on an unquoted ")".
subject_re='^[a-z]+\(([^)]*)\)!?:'
if [[ ! "$subject" =~ $subject_re ]]; then
  exit 0
fi
scopes="${BASH_REMATCH[1]}"

known="$NON_EPIC_SCOPES $LEGACY_SCOPES"

# Epics with a file, including the completed ones under plan/complete/. An
# epic that is finished keeps its id: a follow-up fix to it is still that
# epic's commit, and refusing the scope would only push the change under a
# vaguer one.
for f in "$plan_dir"/*.md "$plan_dir"/complete/*.md; do
  [ -e "$f" ] || continue
  id="$(basename "$f" .md)"
  [ "$id" = "README" ] && continue
  known="$known $id"
done

# Collapse whitespace so the membership test below has single separators.
known="$(printf '%s' "$known" | tr -s '[:space:]' ' ')"

bad=""
IFS=',' read -ra parts <<<"$scopes"
for scope in "${parts[@]}"; do
  scope="$(printf '%s' "$scope" | tr -d '[:space:]')"
  [ -n "$scope" ] || continue
  case " $known " in
  *" $scope "*) ;;
  *) bad="$bad $scope" ;;
  esac
done

[ -n "$bad" ] || exit 0

{
  echo "commit scope not in plan/:$bad"
  echo
  echo "A scope is an epic id, so that a commit is findable from the epic it"
  echo "belongs to. Use one of:"
  echo
  # shellcheck disable=SC2086
  printf '  %s\n' $known | sort -u
  echo
  echo "or write plan/<id>.md for the epic first, or drop the scope."
} >&2
exit 1