Skip to content

ci: DRIFT is advisory here — python is the reference, not a port #91

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
anthmFS merged 1 commit into from
Aug 13, 2026
Merged

ci: DRIFT is advisory here — python is the reference, not a port #91

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
36 changes: 34 additions & 2 deletions scripts/run-ci.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -94,6 +94,38 @@ python3 -m pip freeze 2>/dev/null \
| grep -iE '^(fastapi|starlette|pydantic|pydantic-core|anyio|uvicorn|httpx|requests)==' \
| sed 's/^/ /' || true

# DRIFT — ADVISORY here, and only here.
#
# For the nine PORTS, DRIFT is a hard gate: it asks whether the port still matches
# the reference. That question has a right answer and a port can be wrong.
#
# This repo IS the reference. python_signatures.json is DERIVED from this tree, so
# python cannot drift from itself — a difference after regeneration means only that
# porting-sdk's committed snapshot has not caught up yet. Failing python's CI for
# that made adding public API here require a lockstep change in another repo, which
# inverted the dependency: the source of truth was blocked on its own derivative.
#
# So it reports and does not fail. The delta is printed because it is genuinely
# useful — it is exactly the new surface, and it is what a maintainer regenerates
# downstream after this merges. Nothing is weakened for the ports: they still check
# against the reference with a hard gate, and the reference is still enumerated on
# every run by the SIGNATURES gate above (which SEMVER-DIFF depends on, and which
# still fails loudly if this tree cannot be enumerated at all).
drift_gate() {
if (cd "$PORTING_SDK_DIR" && git diff --quiet -- python_signatures.json 2>/dev/null); then
echo " oracle snapshot matches this tree."
return 0
fi
local stat
stat="$(cd "$PORTING_SDK_DIR" && git diff --shortstat -- python_signatures.json 2>/dev/null)"
echo " ADVISORY: this tree's public surface differs from porting-sdk's committed"
echo " oracle (${stat# }). That is expected when you add or change public API."
echo " Not a failure: python is the reference, the oracle is derived from it."
echo " Say so in your PR ('changes public surface') and a maintainer regenerates"
echo " it downstream — see .github/CONTRIBUTING.md."
return 0
}

# FMT — ruff format. LOCAL applies; CI --check.
fmt_gate() {
if [ -n "${CI:-}" ]; then
Expand Down Expand Up @@ -207,8 +239,8 @@ sched_gate SIGNATURES desc="regenerate python_signatures.json (reference oracle)
--signalwire-python "$PORT_ROOT/signalwire" \
--out "$PORTING_SDK_DIR/python_signatures.json"

sched_gate DRIFT deps=SIGNATURES desc="python_signatures.json unchanged after regen" \
-- bash -c "cd '$PORTING_SDK_DIR' && git diff --quiet -- python_signatures.json"
sched_gate DRIFT deps=SIGNATURES desc="oracle snapshot vs this tree (ADVISORY — python is the reference)" \
--fn drift_gate

sched_gate SEMVER-DIFF deps=SIGNATURES desc="version bump matches surface change vs python_signatures.baseline.json (the reference is not exempt)" \
-- python3 "$PORTING_SDK_DIR/scripts/semver_diff.py" --port python --repo "$PORT_ROOT"
Expand Down