marius/Open_Ultra_SAST
1
0
Fork 0
mirror of https://github.com/norandom/OpenUltraSAST.git synced 2026-10-16 03:49:26 +00:00
Code Issues Packages Projects Releases Wiki Activity

chore(p0): drop never-enabled hunter tools, test-only functions and receiver delivery

Browse Source 
Off every active line (PATHFINDER-2026-10-13/03-unified-proposal.md, cuts).
Deleted: the four hunter tools no caller enables (read_definition,
entry_points, flows, obligations) with their schemas and dispatch, the dead
mapping branch of find_refs; 20 functions whose only caller was one test
(mapping task records, compute_ranking_metrics, build_quick_hunter_prompt,
validate_evidence_transition, should_fuse, harness processors,
rank_targets_with_model, model/report slice rendering, taxonomy
hierarchical/related, config_value.verdict, taint.verdict, roles.load_roles,
features.language_of and build_for_scan, folds.leave_one_*,
features.recorded_case, search BatchExecutor and the three __main__ shims,
delivery.put_input); the receiver delivery mode of the plane. push_inputs.py
moved to benchmarks/push. memory.purge stays: the live-S3 test cleanup uses
it. Net: 2,730 lines deleted, 207 added. 473 targeted tests passed; 3,282
tests collect. breadth.toml records the removals.

Co-Authored-By: Claude Fable 5.1 <[email protected]>
This commit is contained in:
Marius Ciepluch
2026-10-13 11:06:09 +02:00
co-authored by Claude Fable 5.1
parent f5a904b7f2
commit f63703a604
60 changed files with 212 additions and 2576 deletions
Download Patch File Download Diff File Expand all files Collapse all files
tests/test_learn_schema.py
+1 -57
View File
@@ -10,14 +10,10 @@ from typing import Any
import pytest
from openultrasast.findings import quick_scan_findings
from openultrasast.learn import schema
from openultrasast.learn.features import Delta, Part, build_for_scan, record
from openultrasast.learn.features import Part, record
from openultrasast.learn.schema import EXCLUDED_FIELDS, FEATURES, FeatureRecordError, validate_record
from openultrasast.mapping import analyze_entry_points, attach_reachability_hints
from openultrasast.plane.tasks.features import case_features
from openultrasast.preprocess import preprocess_repository
from openultrasast.rank import rank_targets
ROOT = Path(__file__).resolve().parents[1]
APP = """import os
@@ -32,22 +28,6 @@ def {outer}():
"""
def _scan(root: Path) -> list[dict[str, Any]]:
_, targets = preprocess_repository(root)
assert targets and all((root / t.path).stat().st_size > 0 for t in targets), "the fixture was read"
targets = attach_reachability_hints(targets, analyze_entry_points(root, targets))
findings = quick_scan_findings(root, targets, rank_targets(targets))
assert findings, "the fixture's os.system fires a quick rule"
engine = {"questions": 3, "completed": 3, "degradations": [], "findings": []}
return build_for_scan(root, findings, engine)
def _repo(root: Path, path: str, inner: str, outer: str) -> Path:
(root / path).parent.mkdir(parents=True, exist_ok=True)
(root / path).write_text(APP.format(inner=inner, outer=outer))
return root
# --- the allow-list ----------------------------------------------------------------------------------------------
@@ -63,35 +43,10 @@ def test_features_are_allow_listed(tmp_path: Path) -> None:
mechanisms = tomllib.loads((ROOT / "benchmarks" / "pairs" / "mechanisms.toml").read_text())["mechanism"]
assert tuple(m["id"] for m in mechanisms) == schema.MECHANISMS, "the mechanism buckets are the closed vocabulary"
assert set(schema.RULE_TAG_MECHANISM.values()) <= set(schema.MECHANISMS)
records = _scan(_repo(tmp_path, "app/one.py", "alpha", "beta"))
assert records
for rec in records:
assert set(rec["x"]) == {s.name for s in schema.features_for("static")}, "every emitted key is in FEATURES"
assert rec["profile"] == "static" and not any(k.startswith(("verify.", "agree.", "ms.")) for k in rec["x"])
assert {s.name for s in schema.features_for("static")} < set(names)
assert [s.name for s in FEATURES if s.prior] == ["qr.prior_hits"]
def test_record_is_name_invariant(tmp_path: Path) -> None:
"""Two repositories that differ only in repository name, paths, file and function names and commit ids give the
same features: nothing identifies a repository."""
first = _scan(_repo(tmp_path / "acme-shop", "app/one.py", "alpha", "beta"))
second = _scan(_repo(tmp_path / "other-thing", "lib/sub/two.py", "gamma", "delta"))
assert [r["x"] for r in first] == [r["x"] for r in second]
assert [r["instruments"] for r in first] == [r["instruments"] for r in second]
assert first[0]["candidate"] != second[0]["candidate"], "the key differs; it lives only in the envelope"
assert first[0]["x"]["qr.enabled_hits"] >= 1 and first[0]["x"]["facts.entry_distance"] is not None
root = tmp_path / "acme-shop"
_, targets = preprocess_repository(root)
findings = quick_scan_findings(root, targets, rank_targets(targets))
changed = {"app/one.py": [(5, 5)]}
one = build_for_scan(root, findings, None, base=Delta(root, changed, pin="a" * 40))
two = build_for_scan(root, findings, None, base=Delta(root, changed, pin="b" * 40))
assert len(one) == 1 and one[0]["unit"] == "delta" and one[0]["base"] == "a" * 40
assert one[0]["x"]["delta.novelty"] == "new" and one[0]["x"]["delta.changed_lines"] == 1
assert [r["x"] for r in one] == [r["x"] for r in two], "a commit id is envelope, never a feature"
# --- label fields ------------------------------------------------------------------------------------------------
AGREED = {
@@ -165,17 +120,6 @@ def test_missing_is_not_zero() -> None:
assert unread["instruments"]["engine"]["state"] == "failed" and unread["x"]["eng.findings"] is None, "no question is not zero"
def test_build_for_scan_marks_an_engine_that_did_not_run_or_read(tmp_path: Path) -> None:
root = _repo(tmp_path / "r", "app/one.py", "alpha", "beta")
_, targets = preprocess_repository(root)
findings = quick_scan_findings(root, targets, rank_targets(targets))
unread = build_for_scan(root, findings, {"questions": 0, "findings": []})
absent = build_for_scan(root, findings, None)
assert {r["instruments"]["engine"]["state"] for r in unread} == {"failed"}
assert {r["instruments"]["engine"]["state"] for r in absent} == {"none"}
assert all(r["x"]["eng.findings"] is None for r in (*unread, *absent))
def test_validate_record_refuses_values_outside_their_type() -> None:
good = _plane()[0]
validate_record(good)