test: Add tests for corrupt FileCache

Author: martin-kokos

src/poetry/utils/cache.py

@@ -4,6 +4,7 @@
 import dataclasses
 import hashlib
 import json
+import logging
 import shutil
 import time
 
@@ -18,6 +19,8 @@
 MAX_DATE = 9999999999
 T = TypeVar("T")
 
+logger = logging.getLogger(__name__)
+
 
 def decode(string: bytes, encodings: list[str] | None = None) -> str:
     """
@@ -172,7 +175,14 @@ def _get_payload(self, key: str) -> T | None:
             return None
 
         with open(path, "rb") as f:
-            payload = self._deserialize(f.read())
+            file_content = f.read()
+
+        try:
+            payload = self._deserialize(file_content)
+        except (json.JSONDecodeError, ValueError):
+            self.forget(key)
+            logger.warning("Corrupt cache file was detected and cleaned up.")
+            return None
 
         if payload.expired:
             self.forget(key)

tests/utils/test_cache.py

@@ -1,5 +1,7 @@
 from __future__ import annotations
 
+import shutil
+
 from typing import TYPE_CHECKING
 from typing import Any
 from typing import TypeVar
@@ -192,3 +194,63 @@ def test_cachy_compatibility(
 
     assert cachy_file_cache.get("key3") == test_str
     assert cachy_file_cache.get("key4") == test_obj
+
+
+def test_missing_cache_file(
+    poetry_file_cache: FileCache, mocker: MockerFixture
+) -> None:
+    poetry_file_cache.put("key1", "value")
+
+    key1_path = (
+        poetry_file_cache.path
+        / "81/74/09/96/87/a2/66/21/8174099687a26621f4e2cdd7cc03b3dacedb3fb962255b1aafd033cabe831530"  # noqa: E501
+    )
+    assert key1_path.exists()
+    key1_path.unlink()  # corrupt cache by removing a key file
+
+    assert poetry_file_cache.get("key1") is None
+
+
+def test_missing_cache_path(
+    poetry_file_cache: FileCache, mocker: MockerFixture
+) -> None:
+    poetry_file_cache.put("key1", "value")
+
+    key1_partial_path = poetry_file_cache.path / "81/74/09/96/87/a2/"
+    assert key1_partial_path.exists()
+    shutil.rmtree(
+        key1_partial_path
+    )  # corrupt cache by removing a subdirectory containting a key file
+
+    assert poetry_file_cache.get("key1") is None
+
+
+@pytest.mark.parametrize(
+    "corrupt_payload",
+    [
+        "",  # empty file
+        b"\x00",  # null
+        "99999999",  # truncated file
+        '999999a999"value"',  # corrupt lifetime
+        b'9999999999"va\xd8\x00"',  # invalid unicode
+        "fil3systemFa!led",  # garbage file
+    ],
+)
+def test_detect_corrupted_cache_key_file(
+    corrupt_payload: str | bytes, poetry_file_cache: FileCache, mocker: MockerFixture
+) -> None:
+    poetry_file_cache.put("key1", "value")
+
+    key1_path = (
+        poetry_file_cache.path
+        / "81/74/09/96/87/a2/66/21/8174099687a26621f4e2cdd7cc03b3dacedb3fb962255b1aafd033cabe831530"  # noqa: E501
+    )
+    assert key1_path.exists()
+
+    # original content: 9999999999"value"
+
+    write_modes = {str: "w", bytes: "wb"}
+    with open(key1_path, write_modes[type(corrupt_payload)]) as f:
+        f.write(corrupt_payload)  # write corrupt data
+
+    assert poetry_file_cache.get("key1") is None