QTZPOC-20

QTZPOC-20: Refactor ModelImport status engine for lazy initialization, add byte-counted size validation to bundle extraction, optimize MLflow run ID lookups, and ensure workflow cancellation propagates correctly.

tests/laborious/utils/bundle/test_reader.py

@@ -437,7 +437,7 @@ def test_gate_5_wraps_an_os_error_during_unpacking(work_dir, reader, monkeypatch
     def _raise_os_error(*_args, **_kwargs):
         raise OSError('disk full')
 
-    monkeypatch.setattr('laborious.utils.bundle.reader.shutil.copyfileobj', _raise_os_error)
+    monkeypatch.setattr(zipfile.ZipExtFile, 'read', _raise_os_error)
 
     with pytest.raises(BundleGateError, match='could not be unpacked') as error:
         _open(reader, work_dir, payload)
@@ -463,6 +463,52 @@ def test_extract_member_rechecks_containment_on_the_resolved_path(work_dir, read
     assert error.value.gate == 5
 
 
+def test_extract_member_rejects_a_stream_larger_than_its_declared_size(work_dir, reader, monkeypatch):
+    """A decompressed stream longer than the entry's declared `file_size` is rejected mid-write,
+    instead of trusting the central-directory metadata gate 4 already checked in aggregate.
+
+    A single crafted DEFLATE chunk can make `ZipExtFile.read()` return far more bytes than
+    `file_size` declares before its own EOF/CRC bookkeeping catches up, so the stand-in here
+    forces exactly that response shape rather than fighting zipfile's own CRC validation."""
+    root = work_dir / 'root'
+    root.mkdir()
+    buffer = io.BytesIO()
+    with zipfile.ZipFile(buffer, 'w') as archive:
+        archive.writestr('payload.bin', b'x' * 10)
+
+    call_count = {'n': 0}
+
+    def _oversized_read(self, *_args, **_kwargs):
+        call_count['n'] += 1
+        return b'x' * 1000 if call_count['n'] == 1 else b''
+
+    monkeypatch.setattr(zipfile.ZipExtFile, 'read', _oversized_read)
+
+    with zipfile.ZipFile(io.BytesIO(buffer.getvalue())) as archive:
+        entry = archive.infolist()[0]
+        with pytest.raises(BundleGateError, match='more than its declared') as error:
+            reader._extract_member(archive, entry, root, root.resolve())
+
+    assert error.value.gate == 5
+
+
+def test_extract_member_writes_a_well_formed_member_in_full(work_dir, reader):
+    """A member whose decompressed size matches its declared `file_size` is extracted byte for
+    byte, exactly as before the byte-counting check was added."""
+    root = work_dir / 'root'
+    root.mkdir()
+    payload = b'y' * 100
+    buffer = io.BytesIO()
+    with zipfile.ZipFile(buffer, 'w') as archive:
+        archive.writestr('payload.bin', payload)
+
+    with zipfile.ZipFile(io.BytesIO(buffer.getvalue())) as archive:
+        entry = archive.infolist()[0]
+        reader._extract_member(archive, entry, root, root.resolve())
+
+    assert (root / 'payload.bin').read_bytes() == payload
+
+
 # ------------------------------------------------------------------- gate 6