fix: harden preprocessing state and child capabilities

This commit is contained in:
2026-08-11 14:30:16 +02:00
parent a5916f6177
commit ee5ac381b3
12 changed files with 323 additions and 104 deletions
+8
View File
@@ -11,6 +11,14 @@ from tht.ports.vector import VectorStoreError
from tht.vectorstore.embeddings import EmbeddingsError
@pytest.fixture(autouse=True)
def _child_capability_for_pipeline_unit_tests(monkeypatch):
# These tests exercise pipeline result/JSON behavior; process-boundary
# authorization is covered by test_workspace_writer_lock.py.
import tht.cli.preprocess_cmd as command
monkeypatch.setattr(command, "_require_writer_capability", lambda **kwargs: None)
def test_preprocess_evidence_json_is_pristine(monkeypatch, tmp_path):
import tht.cli.preprocess_cmd as command
@@ -5,12 +5,21 @@ from datetime import UTC, datetime
from pathlib import Path
from types import SimpleNamespace
import pytest
from typer.testing import CliRunner
from tht.cli import app
from tht.memory import MemoryRecord, save_registry
@pytest.fixture(autouse=True)
def _child_capability_for_vector_unit_tests(monkeypatch):
import tht.cli.preprocess_cmd as preprocess
import tht.cli.vector_cmd as vector
monkeypatch.setattr(preprocess, "_require_writer_capability", lambda **kwargs: None)
monkeypatch.setattr(vector, "_require_writer_capability", lambda **kwargs: None)
class _FakeEmbedder:
def embed_documents(self, documents):
return [[0.1] * 4 for _ in documents]
+7 -1
View File
@@ -24,6 +24,12 @@ from tht.mschema.models import (
from tht.mschema.render import to_mschema_text, to_schema_dict
@pytest.fixture(autouse=True)
def _child_capability_for_schema_unit_tests(monkeypatch):
import tht.cli.schema_cmd as command
monkeypatch.setattr(command, "_require_writer_capability", lambda **kwargs: None)
def _physical():
return PhysicalSchema(
database="d", schema="s", introspected_at=datetime(2026, 1, 1),
@@ -605,7 +611,7 @@ def test_fresh_process_human_warning_cardinality_is_one_across_failure_and_write
[sys.executable, "-c", probe, "schema", "suggest-fks", "--write", "-c", str(cfg)],
check=True, capture_output=True, text=True,
)
assert json.loads(response.stdout) == {"warnings": 1, "exit": 0}
assert json.loads(response.stdout) == {"warnings": 1, "exit": 1}
physical.unlink()
for branch in branches:
+20 -1
View File
@@ -1,8 +1,12 @@
from __future__ import annotations
import os, stat
import os
import pytest
from tht.workspace_writer_lock import WorkspaceWriterConflict, verify_workspace_writer_fds
def test_verifier_rejects_missing_capability():
with pytest.raises(WorkspaceWriterConflict): verify_workspace_writer_fds(env={})
@@ -14,3 +18,18 @@ def test_verifier_checks_fd_identity(tmp_path):
cap=verify_workspace_writer_fds(writer_fd=w, root_fd=r, env=env)
assert cap.inode == st.st_ino
finally: os.close(w); os.close(r)
def test_verifier_rejects_unrelated_lock(tmp_path):
root = tmp_path / "root"; other = tmp_path / "other"
root.mkdir(mode=0o700); other.mkdir(mode=0o700)
expected = root / "writer.lock"; forged = other / "forged.lock"
expected.touch(mode=0o600); forged.touch(mode=0o600)
root_fd = os.open(root, os.O_RDONLY); forged_fd = os.open(forged, os.O_RDWR)
try:
st = os.fstat(root_fd)
env = {"THOTH_WORKSPACE_ID": "abc-workspace", "THOTH_WORKSPACE_REVISION": "a" * 40, "THOTH_WORKSPACE_DEVICE": str(st.st_dev), "THOTH_WORKSPACE_INODE": str(st.st_ino)}
with pytest.raises(WorkspaceWriterConflict):
verify_workspace_writer_fds(writer_fd=forged_fd, root_fd=root_fd, env=env)
finally:
os.close(forged_fd); os.close(root_fd)