Skip to content
Open
Show file tree
Hide file tree
Changes from all commits
Commits
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
38 changes: 23 additions & 15 deletions pyrit/datasets/seed_datasets/remote/remote_dataset_loader.py
Original file line number Diff line number Diff line change
Expand Up @@ -449,26 +449,31 @@ async def _fetch_zip_from_url_async(
def _download_and_parse() -> dict[str, list[dict[str, Any]]]:
zip_path: Path
temp_to_clean: Path | None = None
if cache and cache_path.exists():
zip_path = cache_path
else:
if cache:
cache_dir.mkdir(parents=True, exist_ok=True)
try:
if cache and cache_path.exists():
zip_path = cache_path
else:
with tempfile.NamedTemporaryFile(delete=False, suffix=".zip") as tmp:
zip_path = Path(tmp.name)
if cache:
cache_dir.mkdir(parents=True, exist_ok=True)
with tempfile.NamedTemporaryFile(
delete=False,
dir=cache_dir,
suffix=".zip.part",
) as tmp:
zip_path = Path(tmp.name)
else:
with tempfile.NamedTemporaryFile(delete=False, suffix=".zip") as tmp:
zip_path = Path(tmp.name)
temp_to_clean = zip_path

logger.info(f"Downloading zip archive from {source}")
with requests.get(source, stream=True) as response:
response.raise_for_status()
with zip_path.open("wb") as fh:
for chunk in response.iter_content(chunk_size=1 << 16):
if chunk:
fh.write(chunk)
logger.info(f"Downloading zip archive from {source}")
with requests.get(source, stream=True) as response:
response.raise_for_status()
with zip_path.open("wb") as fh:
for chunk in response.iter_content(chunk_size=1 << 16):
if chunk:
fh.write(chunk)

try:
results: dict[str, list[dict[str, Any]]] = {}
with zipfile.ZipFile(zip_path) as zf:
members = set(zf.namelist())
Expand All @@ -485,6 +490,9 @@ def _download_and_parse() -> dict[str, list[dict[str, Any]]]:
"list[dict[str, Any]]",
self.FILE_TYPE_HANDLERS[file_type]["read"](text),
)
if cache and temp_to_clean is not None:
zip_path.replace(cache_path)
temp_to_clean = None
return results
finally:
if temp_to_clean is not None:
Expand Down
37 changes: 37 additions & 0 deletions tests/unit/datasets/test_remote_dataset_loader.py
Original file line number Diff line number Diff line change
Expand Up @@ -256,6 +256,43 @@ async def test_caches_zip_on_disk(self, tmp_path, monkeypatch):
cached = list((tmp_path / "seed-prompt-entries").glob("*.zip"))
assert len(cached) == 1

async def test_interrupted_download_does_not_poison_cache(self, tmp_path, monkeypatch):
monkeypatch.setattr(
"pyrit.datasets.seed_datasets.remote.remote_dataset_loader.DB_DATA_PATH",
tmp_path,
)
zip_bytes = self._make_zip_bytes({"x.json": '[{"k": "v"}]'})

def interrupted_chunks():
yield zip_bytes[:20]
raise RuntimeError("connection dropped")

interrupted_response = self._mock_streaming_response(b"")
interrupted_response.iter_content.return_value = interrupted_chunks()
successful_response = self._mock_streaming_response(zip_bytes)

with patch(
"pyrit.datasets.seed_datasets.remote.remote_dataset_loader.requests.get",
side_effect=[interrupted_response, successful_response],
) as mock_get:
loader = ConcreteRemoteLoader()
with pytest.raises(RuntimeError, match="connection dropped"):
await loader._fetch_zip_from_url_async(source=self.SOURCE, inner_files=["x.json"], cache=True)

cache_dir = tmp_path / "seed-prompt-entries"
assert list(cache_dir.iterdir()) == []

result = await loader._fetch_zip_from_url_async(
source=self.SOURCE,
inner_files=["x.json"],
cache=True,
)

assert result == {"x.json": [{"k": "v"}]}
assert mock_get.call_count == 2
assert len(list(cache_dir.glob("*.zip"))) == 1
assert not list(cache_dir.glob("*.part"))

async def test_cache_false_does_not_persist_zip(self, tmp_path, monkeypatch):
monkeypatch.setattr(
"pyrit.datasets.seed_datasets.remote.remote_dataset_loader.DB_DATA_PATH",
Expand Down