Files
computer-playbook/cli/meta/ci/validate.py
T
kevinveenbirkenbachandClaude Opus 5 8434a31611 fix(ci): answer a role-only selection by existence, not by variant None
Rows are indexed by `(name, variant)` and a variant is an int. A token that
names only a role carries no variant, so the lookup ran with `None` and could
never hit, and every such token was reported as matching no discovered row.

That is the form `collapse_to_roles` produces, so it is the form a retriggered
priority line arrives in. Run 34348821577 shows the result: all twenty-one
priority entries carried

    priority: '<role>' matches no discovered row
              (no row in this run's mode and lifecycle envelope)

while baserow, checkmk, decidim and the rest went on to deploy fine. The
warning was not wrong about one entry and right about twenty; it was blind. A
role that genuinely has no row in the envelope looked exactly like a role with a
full set, which is how priority work goes missing without anyone being able to
see which entry did.

A token that narrows nothing beyond the role name is now answered by existence:
it asks for the role in whatever axes the rotation picks, so any discovered row
satisfies it. `Pin.pinned` already draws that line - "whether the token narrows
anything at all beyond the role name" - and pinned tokens keep the old path,
`check_pins` included.

Three tests come with it. There was already one for a bare role that matches
nothing, and none for a bare role that matches something, which is why the
lookup could stay broken. One of the new ones uses variant 4 so a future
implementation cannot pass by quietly guessing variant 0. Checked against HEAD:
the same input returns the spurious warning there, so the tests fail without
this change.

This restores the signal, not the lost jobs. The other silent path, `priority`
being truncated to the job budget in cli/meta/ci/chunks.py, is untouched.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
2026-09-10 01:44:59 +02:00

177 lines
6.6 KiB
Python

"""Check a run's selection inputs against the branch before anything deploys.
Usage:
python -m cli.meta.ci.validate [--whitelist "..."] [--priority "..."]
[--modes auto] [--tor auto] [--distros "..."] [--filesystem "..."]
[--lifecycles "..."]
A selection token names a row that has to exist *on this branch*
(:mod:`utils.github.variant.selection`). Nothing else in the chain can tell
the operator that early: the matrix builder raises on the first bad token, one
per chunk job, so a stale list costs one full CI round per entry and reports
the second problem only after the first is fixed. Every problem is collected
here instead, and the exit code gates the run.
What makes a token bad, in the order this checks it:
* it does not parse, or names a mode or onion state that does not exist;
* it names a role the discovery query does not return -- a typo, a role
outside the run's lifecycle envelope, or a role that no longer exists;
* it pins a variant the role does not declare (any more): the usual case is a
list carried over from an older run whose variants have since been renumbered;
* it pins a mode the row cannot take, an onion state the row, the mode or the
run's own tor axis rules out, or a distro or filesystem the run's own pool
does not hold.
A bare role name that matches nothing is reported too, but as a warning: the
diff-derived whitelist legitimately names roles the envelope filters out.
"""
from __future__ import annotations
import argparse
import sys
from cli.meta.ci import query
from utils.cache.applications import get_variants
from utils.github.variant import axes, pools, selection, tor
from utils.roles.display import display_names
def problems(
tokens: str,
*,
modes: tuple[str, ...],
tor_mode: str,
distros: tuple[str, ...],
filesystems: tuple[str, ...],
lifecycles: str,
label: str,
) -> tuple[list[str], list[str]]:
"""Every reason the tokens of one input cannot deploy on this branch.
A token that narrows nothing beyond the role name is answered by existence:
it asks for the role in whatever axes the rotation picks, so any discovered
row satisfies it. Rows are keyed by ``(name, variant)`` and a variant is an
int, so looking such a token up variant-first can only miss, which reported
every collapsed priority entry as matching nothing and left a role that
really had no row indistinguishable from one with a full set.
Args:
tokens: the raw ``whitelist``/``priority`` value.
modes: the run's selected deploy modes.
tor_mode: the run's tor axis.
distros: the distro pool the run draws from.
filesystems: the filesystem pool the run draws from.
lifecycles: the run's lifecycle envelope.
label: the input's name, for the messages.
Returns:
``(errors, warnings)``.
"""
pins = selection.parse_list(display_names().decode_list(tokens))
if not pins:
return [], []
rows = {
(row["name"], row["variant"]): query.row_modes(row, modes)
for row in query.discover_rows(
modes, whitelist=selection.names(pins), lifecycles=lifecycles
)
}
discovered_apps = {app for app, _variant in rows}
declared = get_variants()
errors: list[str] = []
warnings: list[str] = []
for pin in pins:
token = selection.describe(pin)
if not pin.pinned:
if pin.app not in discovered_apps:
warnings.append(
f"{label}: {token!r} matches no discovered row "
f"(no row in this run's mode and lifecycle envelope)"
)
continue
variants = pin.variants or (None,)
for variant in variants:
offered = rows.get((pin.app, variant))
if offered is None:
count = len(declared.get(pin.app) or [])
reason = (
f"role declares {count} variant(s)"
if pin.variants
else "no row in this run's mode and lifecycle envelope"
)
message = f"{label}: {token!r} matches no discovered row ({reason})"
(errors if pin.pinned else warnings).append(message)
continue
try:
axes.check_pins(
pin.app,
"" if variant is None else str(variant),
offered,
pin_mode=pin.mode,
pin_tor=pin.tor,
pin_distro=pin.distro,
pin_filesystem=pin.filesystem,
capable=tor.tor_capable(pin.app, variant, declared),
tor_mode=tor_mode,
distros=distros,
filesystems=filesystems,
)
except SystemExit as refusal:
errors.append(f"{label}: {token!r} {str(refusal).split(': ', 1)[-1]}")
return errors, warnings
def main(argv: list[str] | None = None) -> int:
parser = argparse.ArgumentParser(
description="Validate a run's selection inputs against this branch."
)
parser.add_argument("--whitelist", default="")
parser.add_argument("--priority", default="")
parser.add_argument("--modes", default=query.ALL_MODES)
parser.add_argument("--tor", default=None)
parser.add_argument("--distros", default="")
parser.add_argument("--filesystem", default="")
parser.add_argument("--lifecycles", default="")
args = parser.parse_args(argv)
modes = query.resolve_modes(args.modes)
tor_mode = tor.resolve_tor_mode(args.tor)
distros = pools.resolve_distros(args.distros)
filesystems = pools.resolve_filesystems(args.filesystem)
errors: list[str] = []
warnings: list[str] = []
for label, tokens in (("whitelist", args.whitelist), ("priority", args.priority)):
found, warned = problems(
tokens,
modes=modes,
tor_mode=tor_mode,
distros=distros,
filesystems=filesystems,
lifecycles=args.lifecycles,
label=label,
)
errors += found
warnings += warned
for warning in warnings:
print(f"::warning::{warning}")
for error in errors:
print(f"::error::{error}", file=sys.stderr)
if errors:
print(
f"\n{len(errors)} unusable selection(s). Every one of them would abort a "
f"chunk's discovery, so the run is refused here instead.",
file=sys.stderr,
)
return 1
print("Selection inputs are valid for this branch.")
return 0
if __name__ == "__main__":
sys.exit(main())