skipping artifact paths

.gitignore

@@ -50,6 +50,7 @@ coverage.xml
 *.py,cover
 .hypothesis/
 .pytest_cache/
+!tests/notebooks/artifact_folder/__pycache__/
 
 # Translations
 *.mo

jupyter_cache/cache/main.py

@@ -23,19 +23,33 @@
 
 CACHE_LIMIT_KEY = "cache_limit"
 DEFAULT_CACHE_LIMIT = 1000
+ARTIFACT_SKIP_PATTERNS = ["__pycache__"]
 
 
 class NbArtifacts(NbArtifactsAbstract):
     """Container for artefacts of a notebook execution."""
 
-    def __init__(self, paths: List[str], in_folder, check_existence=True):
+    def __init__(
+        self,
+        paths: List[str],
+        in_folder,
+        check_existence=True,
+        skip_patterns=ARTIFACT_SKIP_PATTERNS,
+    ):
         """Initiate NbArtifacts
 
         :param paths: list of paths
         :param check_existence: check the paths exist
         :param in_folder: The folder that all paths should be in (or subfolder).
+        :param skip_patterns: Exclude paths that contain one of these patterns
         :raises IOError: if check_existence and file does not exist
         """
+
+        def path_not_in_skip_artifacts(path):
+            # TODO: This should probably be a more-specific regex in the future
+            return all(pattern not in str(path) for pattern in skip_patterns)
+
+        paths = list(filter(path_not_in_skip_artifacts, paths))
         self.paths = [Path(p).absolute() for p in paths]
         self.in_folder = Path(in_folder).absolute()
         to_relative_paths(self.paths, self.in_folder, check_existence=check_existence)

tests/test_cache.py

@@ -146,6 +146,32 @@ def test_artifacts(tmp_path):
         assert path.joinpath("artifact_folder").exists()
 
 
+def test_artifacts_skip_patterns(tmp_path):
+    cache = JupyterCacheBase(str(tmp_path))
+    with pytest.raises(IOError):
+        cache.cache_notebook_file(
+            path=os.path.join(NB_PATH, "basic.ipynb"),
+            uri="basic.ipynb",
+            artifacts=(os.path.join(NB_PATH),),
+            check_validity=False,
+        )
+    cache.cache_notebook_file(
+        path=os.path.join(NB_PATH, "basic.ipynb"),
+        uri="basic.ipynb",
+        artifacts=(
+            os.path.join(NB_PATH, "artifact_folder", "artifact.txt"),
+            os.path.join(NB_PATH, "artifact_folder", "__pycache__"),
+        ),
+        check_validity=False,
+    )
+
+    # __pycache__ is ignored and not saved as artifact in cache
+    bundle = cache.get_cache_bundle(1)
+    assert {str(p) for p in bundle.artifacts.relative_paths} == {
+        "artifact_folder/artifact.txt"
+    }
+
+
 # jupyter_client/session.py:371: DeprecationWarning:
 # Session._key_changed is deprecated in traitlets: use @observe and @unobserve instead
 @pytest.mark.filterwarnings("ignore")