01b6eb1804
CI / build (push) Successful in 17s
CI / helm (push) Successful in 22s
CI / lint (push) Successful in 28s
CI / typecheck (push) Successful in 47s
CI / benchmark-regression (push) Has been skipped
CI / quality (push) Successful in 3m49s
CI / security (push) Successful in 4m11s
CI / unit_tests (push) Successful in 9m19s
CI / docker (push) Successful in 1m22s
CI / coverage (push) Successful in 12m35s
CI / e2e_tests (push) Successful in 16m17s
CI / integration_tests (push) Successful in 25m5s
CI / status-check (push) Successful in 2s
CI / benchmark-publish (push) Successful in 28m31s
## Summary Add M6 parallel-scaling coverage for 10+ concurrent subplans: - **15-subplan parallel scenario** with explicit peak-concurrency bound checks (`max_parallel=10`) and thread-safe concurrency tracking via `_build_executor()`. - **Deep hierarchical decomposition** coverage (4+ levels) with adjusted leaf condition that only stops early when hitting `max_depth` or when the workset is trivially small (`min_files_per_subplan`). - **Non-progress guard** in `_build_hierarchy` to prevent pathological recursion when clustering cannot meaningfully split the file set. - **Small-project regression test** (< 50 files) verifying decomposition depth does not increase unexpectedly with the relaxed leaf condition. - **ASV benchmark** for 15-subplan parallel execution with `max_parallel=10` to track scaling behavior. ### Removed from this PR The `_build_hierarchy` child-linkage correctness fix (returning `node_id` from recursive calls instead of using `nodes[-1].node_id`) has been **removed** per review feedback — it is a separate bug fix and will be submitted as an independent issue/PR per CONTRIBUTING.md §Atomic Commits. ## Approach - **Concurrency tracking:** The `_build_executor()` closure in step definitions detects `context.concurrency_counter` / `context.concurrency_lock` and performs thread-safe peak tracking in a try/finally block. - **Leaf condition:** Replaced the `max_files_per_subplan` / `max_tokens_per_subplan` leaf check with a `min_files_per_subplan` check to allow deeper decomposition for large projects. Added a non-progress guard so clustering that cannot split the file set terminates immediately rather than recursing to `max_depth`. - **Deterministic IDs:** `_ids_for_count()` preserves legacy fixed IDs for the first 5 subplans and generates additional deterministic IDs for scale scenarios. ## Validation ### Passing - `nox -s lint` — all checks passed - `nox -s typecheck` — 0 errors, 0 warnings - `nox -s unit_tests` — 12,988 scenarios passed, 0 failed - `nox -s coverage_report` — 97% (passes `--fail-under=97`) Closes #855 Reviewed-on: #1201 Co-authored-by: Brent E. Edwards <brent.edwards@cleverthis.com> Co-committed-by: Brent E. Edwards <brent.edwards@cleverthis.com>
311 lines
9.7 KiB
Python
311 lines
9.7 KiB
Python
"""Robot Framework helper for bug #647 regression tests.
|
|
|
|
Runs `plan tree`, `plan explain`, and `plan correct` through real CLI
|
|
subprocesses with real DI/container wiring (no mocks).
|
|
"""
|
|
|
|
from __future__ import annotations
|
|
|
|
import os
|
|
import sys
|
|
import tempfile
|
|
from collections.abc import Callable
|
|
from contextlib import suppress
|
|
from dataclasses import dataclass
|
|
from pathlib import Path
|
|
from typing import NoReturn
|
|
|
|
_SRC = str(Path(__file__).resolve().parents[1] / "src")
|
|
if _SRC not in sys.path:
|
|
sys.path.insert(0, _SRC)
|
|
|
|
_ROBOT = str(Path(__file__).resolve().parent)
|
|
if _ROBOT not in sys.path:
|
|
sys.path.insert(0, _ROBOT)
|
|
|
|
from helper_e2e_common import run_cli # noqa: E402
|
|
from ulid import ULID # noqa: E402
|
|
|
|
from cleveragents.application.container import reset_container # noqa: E402
|
|
from cleveragents.application.services import decision_service as ds_mod # noqa: E402
|
|
from cleveragents.application.services.plan_lifecycle_service import ( # noqa: E402
|
|
PlanLifecycleService,
|
|
)
|
|
from cleveragents.config.settings import Settings # noqa: E402
|
|
from cleveragents.domain.models.core.decision import ( # noqa: E402
|
|
ContextSnapshot,
|
|
DecisionType,
|
|
ResourceRef,
|
|
)
|
|
from cleveragents.infrastructure.database.unit_of_work import UnitOfWork # noqa: E402
|
|
|
|
|
|
@dataclass(frozen=True)
|
|
class DecisionContext:
|
|
"""Data required to run and verify one CLI command."""
|
|
|
|
plan_id: str
|
|
decision_id: str
|
|
database_url: str
|
|
db_path: str
|
|
uow: UnitOfWork
|
|
|
|
|
|
def _fail(message: str) -> NoReturn:
|
|
"""Exit with expected regression-failure status."""
|
|
print(f"FAIL: {message}", file=sys.stderr)
|
|
sys.exit(1)
|
|
|
|
|
|
def _fail_unexpected(message: str) -> NoReturn:
|
|
"""Exit with unexpected-infrastructure-failure status."""
|
|
print(f"UNEXPECTED: {message}", file=sys.stderr)
|
|
sys.exit(2)
|
|
|
|
|
|
def _setup_decisions() -> DecisionContext:
|
|
"""Create a real sqlite database and seed a small decision tree."""
|
|
fd, db_path = tempfile.mkstemp(prefix="cr647_", suffix=".db")
|
|
os.close(fd)
|
|
database_url = f"sqlite:///{db_path}"
|
|
os.environ["CLEVERAGENTS_DATABASE_URL"] = database_url
|
|
|
|
reset_container()
|
|
uow = UnitOfWork(database_url)
|
|
uow.init_database()
|
|
|
|
settings = Settings(database_url=database_url, async_enabled=False)
|
|
decision_svc = ds_mod.DecisionService(
|
|
settings=settings,
|
|
unit_of_work=uow,
|
|
)
|
|
|
|
lifecycle_svc = PlanLifecycleService(settings=settings, unit_of_work=uow)
|
|
action_name = f"local/cr647-{ULID()!s}"
|
|
lifecycle_svc.create_action(
|
|
name=action_name,
|
|
description="Container.resolve regression action",
|
|
definition_of_done="Regression guard setup complete",
|
|
strategy_actor="openai/gpt-4",
|
|
execution_actor="openai/gpt-4",
|
|
)
|
|
plan = lifecycle_svc.use_action(action_name)
|
|
plan_id = plan.identity.plan_id
|
|
root = decision_svc.record_decision(
|
|
plan_id=plan_id,
|
|
decision_type=DecisionType.PROMPT_DEFINITION,
|
|
question="What should we build?",
|
|
chosen_option="A REST API",
|
|
alternatives_considered=["GraphQL API", "gRPC service"],
|
|
confidence_score=0.95,
|
|
rationale="REST fits requirements",
|
|
context_snapshot=ContextSnapshot(
|
|
hot_context_hash="sha256:root",
|
|
hot_context_ref="store://root",
|
|
relevant_resources=[
|
|
ResourceRef(resource_id=str(ULID()), path="src/main.py"),
|
|
],
|
|
actor_state_ref="checkpoint://root",
|
|
),
|
|
)
|
|
child = decision_svc.record_decision(
|
|
plan_id=plan_id,
|
|
decision_type=DecisionType.STRATEGY_CHOICE,
|
|
question="Which framework should we use?",
|
|
chosen_option="FastAPI",
|
|
parent_decision_id=root.decision_id,
|
|
alternatives_considered=["Flask", "Django"],
|
|
confidence_score=0.90,
|
|
rationale="FastAPI fits async service requirements",
|
|
context_snapshot=ContextSnapshot(
|
|
hot_context_hash="sha256:child",
|
|
hot_context_ref="store://child",
|
|
relevant_resources=[
|
|
ResourceRef(resource_id=str(ULID()), path="src/api.py"),
|
|
],
|
|
actor_state_ref="checkpoint://child",
|
|
),
|
|
)
|
|
return DecisionContext(
|
|
plan_id=plan_id,
|
|
decision_id=child.decision_id,
|
|
database_url=database_url,
|
|
db_path=db_path,
|
|
uow=uow,
|
|
)
|
|
|
|
|
|
def _cleanup(ctx: DecisionContext) -> None:
|
|
"""Best-effort cleanup that does not mask later teardown steps."""
|
|
with suppress(Exception):
|
|
ctx.uow.engine.dispose()
|
|
with suppress(Exception):
|
|
reset_container()
|
|
with suppress(Exception):
|
|
os.environ.pop("CLEVERAGENTS_DATABASE_URL", None)
|
|
for suffix in ("", "-journal", "-wal", "-shm"):
|
|
with suppress(OSError):
|
|
Path(ctx.db_path + suffix).unlink(missing_ok=True)
|
|
# Keep reset last to avoid stale singleton recreation during cleanup.
|
|
with suppress(Exception):
|
|
Settings.reset()
|
|
|
|
|
|
def _assert_tree_output(ctx: DecisionContext, output: str) -> None:
|
|
"""Require tree output to include the seeded decision id."""
|
|
if ctx.decision_id not in output:
|
|
_fail(
|
|
"Expected plan tree output to include seeded decision data.\n"
|
|
f"Output:\n{output}"
|
|
)
|
|
|
|
|
|
def _assert_explain_output(ctx: DecisionContext, output: str) -> None:
|
|
"""Require explain output to include seeded id and content."""
|
|
if ctx.decision_id not in output:
|
|
_fail(
|
|
"Expected plan explain output to include seeded decision id.\n"
|
|
f"Output:\n{output}"
|
|
)
|
|
if "FastAPI" not in output:
|
|
_fail(
|
|
"Expected plan explain output to include child chosen decision content "
|
|
"(FastAPI).\n"
|
|
f"Output:\n{output}"
|
|
)
|
|
|
|
|
|
def _assert_correct_output(ctx: DecisionContext, output: str) -> None:
|
|
"""Require correct output to reference the seeded decision id and mode."""
|
|
if ctx.decision_id not in output:
|
|
_fail(
|
|
"Expected plan correct output to include seeded decision id.\n"
|
|
f"Output:\n{output}"
|
|
)
|
|
# Validate command-specific content: dry-run mode or revert indication.
|
|
lowered = output.lower()
|
|
if "revert" not in lowered and "dry" not in lowered:
|
|
_fail(
|
|
"Expected plan correct output to reference revert mode or dry-run.\n"
|
|
f"Output:\n{output}"
|
|
)
|
|
|
|
|
|
def _run_and_verify(
|
|
*,
|
|
label: str,
|
|
cli_args_factory: Callable[[DecisionContext], list[str]],
|
|
positive_assertion: Callable[[DecisionContext, str], None],
|
|
) -> None:
|
|
"""Run one command end-to-end and enforce regression invariants."""
|
|
ctx = _setup_decisions()
|
|
try:
|
|
args = cli_args_factory(ctx)
|
|
result = run_cli(
|
|
*args,
|
|
workspace=str(Path.cwd()),
|
|
env_extra={"CLEVERAGENTS_DATABASE_URL": ctx.database_url},
|
|
timeout=120,
|
|
)
|
|
output = (result.stdout or "") + (result.stderr or "")
|
|
lowered = output.lower()
|
|
|
|
if result.returncode != 0:
|
|
if "attributeerror" in lowered and "has no attribute 'resolve'" in lowered:
|
|
_fail(
|
|
f"{label} unexpectedly reproduced resolve() crash.\n"
|
|
f"rc={result.returncode}\n{output}"
|
|
)
|
|
_fail_unexpected(
|
|
f"{label} failed with unexpected non-zero exit.\n"
|
|
f"rc={result.returncode}\n{output}"
|
|
)
|
|
|
|
if "attributeerror" in lowered or "has no attribute 'resolve'" in lowered:
|
|
_fail(
|
|
f"{label} printed resolve() crash text despite exit 0.\n"
|
|
f"Output:\n{output}"
|
|
)
|
|
|
|
positive_assertion(ctx, output)
|
|
print(f"{label}-ok")
|
|
finally:
|
|
_cleanup(ctx)
|
|
|
|
|
|
def plan_tree_crash() -> None:
|
|
"""Regression guard for ``plan tree`` DI-container path."""
|
|
_run_and_verify(
|
|
label="plan-tree-crash",
|
|
cli_args_factory=lambda ctx: ["plan", "tree", ctx.plan_id, "--format", "json"],
|
|
positive_assertion=_assert_tree_output,
|
|
)
|
|
|
|
|
|
def plan_explain_crash() -> None:
|
|
"""Regression guard for ``plan explain`` DI-container path."""
|
|
_run_and_verify(
|
|
label="plan-explain-crash",
|
|
cli_args_factory=lambda ctx: [
|
|
"plan",
|
|
"explain",
|
|
ctx.decision_id,
|
|
"--show-context",
|
|
"--show-reasoning",
|
|
"--format",
|
|
"json",
|
|
],
|
|
positive_assertion=_assert_explain_output,
|
|
)
|
|
|
|
|
|
def plan_correct_crash() -> None:
|
|
"""Regression guard for ``plan correct`` DI-container path."""
|
|
_run_and_verify(
|
|
label="plan-correct-crash",
|
|
cli_args_factory=lambda ctx: [
|
|
"plan",
|
|
"correct",
|
|
ctx.decision_id,
|
|
"--mode",
|
|
"revert",
|
|
"--guidance",
|
|
"Use Django instead",
|
|
# NOTE: ``--plan`` is required by implementation when there is no
|
|
# active-plan CLI context, though spec examples show it omitted.
|
|
# See spec-vs-implementation drift tracked separately.
|
|
"--plan",
|
|
ctx.plan_id,
|
|
"--dry-run",
|
|
"--format",
|
|
"json",
|
|
],
|
|
positive_assertion=_assert_correct_output,
|
|
)
|
|
|
|
|
|
_COMMANDS: dict[str, Callable[[], None]] = {
|
|
"plan-tree-crash": plan_tree_crash,
|
|
"plan-explain-crash": plan_explain_crash,
|
|
"plan-correct-crash": plan_correct_crash,
|
|
}
|
|
|
|
|
|
def main() -> int:
|
|
"""Dispatch helper command from Robot ``Run Process``."""
|
|
if len(sys.argv) < 2:
|
|
print(f"Usage: helper_container_resolve_crash.py <{'|'.join(_COMMANDS)}>")
|
|
return 1
|
|
|
|
handler = _COMMANDS.get(sys.argv[1])
|
|
if handler is None:
|
|
print(f"Unknown command: {sys.argv[1]}")
|
|
return 1
|
|
|
|
handler()
|
|
return 0
|
|
|
|
|
|
if __name__ == "__main__":
|
|
sys.exit(main())
|