Phase 6, the last of the fresh-install spec. 159 scripts lose their .sh: 110 contracts, 47 Vicinae commands, 2 compositor contracts. A shebang and the executable bit already select the interpreter. The extension only ever added something that had to stay in sync, and the rename proved the point twice over in the space of an hour. The spec's stated risk was Vicinae's script discovery. One script was renamed and reloaded on its own before the other 46 followed; it came back as scripts:panama.capture and all 47 resolve. What the probe turned up instead is that the extension was never only a filename: Vicinae's command IDs embed it, so every ID changed. Nothing in this repository refers to them, so nothing breaks. The only trace is Vicinae's metadata.json, whose visited map had two Panama entries that are now orphaned -- two commands lost their usage ranking and will earn it back. Worth knowing before anyone renames these again on a machine that has a keybind pointing at one. Rewriting the references by exact filename missed two things it structurally could not see: a name built from a variable, settings-$page.sh, and a glob, -name '*.sh'. Both were in the contract that counts the generated commands, which promptly reported 47 expected and 0 found. The mechanical part of a rename is the part that looks finished. The three subcommands. panama doctor fronts a health check that already existed and already ran at the end of every install but could not be reached from a terminal. panama upgrade re-runs the installer from anywhere. panama test runs the suite, which had no entry point at all -- 121 files that were the main safety net in this repository and were invisible in it. Writing that runner found three tests nothing was running. calendar_agenda_bridge_test, home_assistant_bridge_test and kdeconnect_bridge_test are unittest suites without the executable bit, so no contract invoked them and the first draft of the runner skipped them silently. All three pass, and have passed unobserved for weeks. The runner collects *_test.py as well now, because a runner with a blind spot is worse than no runner for the same reason a dependency checker with one is: it reports PASS. Six worktrees pruned. Each was re-checked rather than trusted to the spec's list, and two needed it: panama-commands is not on feat/panama-commands but on feat/gnome-tweaks-parity, and fix/panama-displays-review reads [ahead 3] -- ahead of its remote, not of main, with every commit patch-equivalent to landed work. roadmap-completion stays; it has five commits that are genuinely unlanded. The branches are left alone: pruning a worktree costs nothing, deleting a branch is a decision. 121 contracts pass. Claude-Session: https://claude.ai/code/session_01NvgBuSWB5sE43yWmg21ozj
113 lines
6.6 KiB
Bash
Executable File
113 lines
6.6 KiB
Bash
Executable File
#!/usr/bin/env bash
|
|
|
|
# Getting a file back must never lose the file that was already there, and
|
|
# nothing here may touch how the machine boots.
|
|
#
|
|
# Four rules:
|
|
#
|
|
# 1. Restore sets the current version aside instead of overwriting it. A
|
|
# restore that destroys what you were about to compare against is how
|
|
# someone loses the work they were trying to save.
|
|
# 2. No rollback. snapper's rollback changes the btrfs default subvolume, and
|
|
# this system's fstab pins subvol= explicitly, which overrides it -- so a
|
|
# rollback would report success and change nothing after a reboot. A
|
|
# recovery feature that silently does nothing is worse than none.
|
|
# 3. Paths cannot escape the snapshot they came from.
|
|
# 4. Snapshot 0 is the live filesystem, not a snapshot, and can never be a
|
|
# target for reading or deleting.
|
|
#
|
|
# Read-only: it reads snapshot state and exercises refusals. It never creates,
|
|
# deletes, or restores anything.
|
|
|
|
set -uo pipefail
|
|
|
|
repo_dir="$(cd "$(dirname "${BASH_SOURCE[0]}")/../.." && pwd)"
|
|
helper="$repo_dir/config/dot/quickshell/scripts/panama-snapshots"
|
|
service="$repo_dir/config/dot/quickshell/services/Snapshots.qml"
|
|
page="$repo_dir/config/dot/quickshell/modules/settings/SnapshotsPage.qml"
|
|
|
|
fail() {
|
|
printf 'snapshots contract: %s\n' "$1" >&2
|
|
exit 1
|
|
}
|
|
|
|
for path in "$helper" "$service" "$page"; do
|
|
[[ -r "$path" ]] || fail "missing $path"
|
|
done
|
|
[[ -x "$helper" ]] || fail 'panama-snapshots is not executable'
|
|
|
|
# ── 1. Restore keeps what was there ─────────────────────────────────────────
|
|
restore_body="$(sed -n '/^def restore/,/^def /p' "$helper")"
|
|
[[ -n "$restore_body" ]] || fail 'restore is missing'
|
|
grep -q 'before-restore' <<<"$restore_body" \
|
|
|| fail 'restore does not set the current version aside'
|
|
grep -q 'os.rename(destination, kept)' <<<"$restore_body" \
|
|
|| fail 'the current version is not moved before the snapshot copy is written'
|
|
# The move must happen BEFORE the copy, or there is nothing left to move.
|
|
rename_line="$(grep -n 'os.rename(destination, kept)' <<<"$restore_body" | head -1 | cut -d: -f1)"
|
|
copy_line="$(grep -n 'shutil.copy' <<<"$restore_body" | head -1 | cut -d: -f1)"
|
|
[[ -n "$rename_line" && -n "$copy_line" && "$rename_line" -lt "$copy_line" ]] \
|
|
|| fail 'the snapshot copy is written before the current version is set aside'
|
|
|
|
# ── 2. No rollback ──────────────────────────────────────────────────────────
|
|
grep -qE '"rollback"|set-default|btrfs subvolume set-default|undochange' "$helper" \
|
|
&& fail 'the helper reaches for rollback, which this system fstab would silently ignore'
|
|
grep -qiE 'rollback' "$(dirname "$page")/$(basename "$page")" \
|
|
| grep -v '^\s*//' >/dev/null 2>&1
|
|
page_code="$(grep -vE '^\s*//' "$page")"
|
|
grep -qi 'rollback' <<<"$page_code" \
|
|
&& fail 'the page offers rollback'
|
|
|
|
# ── 3. Paths cannot escape ──────────────────────────────────────────────────
|
|
grep -q 'def safe_relative' "$helper" || fail 'there is no path containment check'
|
|
command -v jq >/dev/null 2>&1 || { printf 'snapshots contract: SKIP (no jq)\n'; exit 0; }
|
|
|
|
snapshot="$("$helper" snapshot 2>/dev/null)" || fail 'snapshot failed'
|
|
config="$(jq -r '.configs[0].name // ""' <<<"$snapshot")"
|
|
number="$(jq -r '.configs[0].snapshots[0].number // 0' <<<"$snapshot")"
|
|
|
|
if [[ -n "$config" && "$number" != "0" ]]; then
|
|
refusal() { "$helper" "$@" 2>/dev/null | jq -r '.error // ""'; }
|
|
for bad in "../../etc" "../.." "gib/../../../etc"; do
|
|
answer="$(refusal browse "$config" "$number" "$bad")"
|
|
[[ "$answer" == "That path is not inside the snapshot." ]] \
|
|
|| fail "browsing \"$bad\" was not refused by the containment check: $answer"
|
|
done
|
|
answer="$(refusal restore "$config" "$number" "../../etc/passwd")"
|
|
[[ "$answer" == "That path is not inside the snapshot." ]] \
|
|
|| fail "restoring \"../../etc/passwd\" was not refused: $answer"
|
|
|
|
# ── 4. The live filesystem is not a snapshot ────────────────────────────
|
|
# The REASON again: with the guard removed, snapshot 0 fails anyway because
|
|
# its directory does not exist -- so a test that accepts any error passes
|
|
# with the guard deleted and proves nothing.
|
|
for answer in "$(refusal browse "$config" 0 "")" "$(refusal delete "$config" 0)"; do
|
|
[[ "$answer" == "That is the current state, not a snapshot." ]] \
|
|
|| fail "snapshot 0 was rejected for the wrong reason, so the live filesystem is not actually guarded: $answer"
|
|
done
|
|
fi
|
|
|
|
# ── Shape ───────────────────────────────────────────────────────────────────
|
|
jq -e '(.configs | type == "array") and (.unprotected | type == "array") and (.space | type == "object")' \
|
|
<<<"$snapshot" >/dev/null || fail 'the snapshot is missing configs, unprotected, or space'
|
|
jq -e '[.configs[] | has("name") and has("subvolume") and has("timelineEnabled") and has("limits")] | all' \
|
|
<<<"$snapshot" >/dev/null || fail 'a configuration is missing its name, subvolume, timeline flag, or limits'
|
|
jq -e '[.configs[].snapshots[]? | .number > 0] | all' <<<"$snapshot" >/dev/null \
|
|
|| fail 'the live filesystem is listed as a snapshot'
|
|
|
|
# ── Destructive actions are confirmed ───────────────────────────────────────
|
|
grep -q 'confirmingDelete' "$page" || fail 'the page deletes a snapshot without confirming'
|
|
grep -q 'confirmingRestore' "$page" || fail 'the page restores without confirming'
|
|
grep -q 'keeping whatever is there now' "$page" \
|
|
|| fail 'the page does not say that restoring keeps the current version'
|
|
|
|
# ── Space is reported honestly ──────────────────────────────────────────────
|
|
# Per-snapshot size needs btrfs quotas, which cost performance on every write.
|
|
# A made-up number would be worse than saying it is not measured.
|
|
grep -q 'Not measured' "$page" \
|
|
|| fail 'the page reports a per-snapshot size it cannot actually measure'
|
|
|
|
printf 'snapshots contract: PASS (%d volume(s), %d snapshot(s), no rollback)\n' \
|
|
"$(jq '.configs | length' <<<"$snapshot")" \
|
|
"$(jq '[.configs[].snapshots[]?] | length' <<<"$snapshot")"
|