Update find cache and rm tests

cloudbuild/run_tests.sh

@@ -56,13 +56,15 @@ case "$TEST_SUITE" in
     # - test_core_versioned.py: HNS buckets do not support versioning.
     # - test_core.py::test_sign: Current Cloud Build auth setup does not support this.
     # - test_core.py::test_mv_file_cache: Integration test only applicable for regional buckets.
+    # - test_core.py::test_rm_wildcards_non_recursive: HNS buckets have different behavior for non-recursive wildcard deletion.
     pytest "${ARGS[@]}" gcsfs/ \
       --deselect gcsfs/tests/test_extended_gcsfs.py \
       --deselect gcsfs/tests/test_zonal_file.py \
       --deselect gcsfs/tests/test_extended_gcsfs_unit.py \
       --deselect gcsfs/tests/test_core_versioned.py \
       --deselect gcsfs/tests/test_core.py::test_sign \
-      --deselect gcsfs/tests/test_core.py::test_mv_file_cache
+      --deselect gcsfs/tests/test_core.py::test_mv_file_cache \
+      --deselect gcsfs/tests/test_core.py::test_rm_wildcards_non_recursive
     ;;
 
   "zonal-core")
@@ -107,6 +109,7 @@ case "$TEST_SUITE" in
     # - test_array fails due to CRC32C TypeError with array objects.
     # - test_sign fails because it requires a private key
     # - test_mv_file_cache: Integration test only applicable for regional buckets.
+    # - test_rm_wildcards_non_recursive: HNS buckets have different behavior for non-recursive wildcard deletion.
     ZONAL_DESELECTS+=(
       "--deselect=gcsfs/tests/test_core.py::test_flush"
       "--deselect=gcsfs/tests/test_core.py::test_write_blocks"
@@ -115,6 +118,7 @@ case "$TEST_SUITE" in
       "--deselect=gcsfs/tests/test_core.py::test_array"
       "--deselect=gcsfs/tests/test_core.py::test_sign"
       "--deselect=gcsfs/tests/test_core.py::test_mv_file_cache"
+      "--deselect=gcsfs/tests/test_core.py::test_rm_wildcards_non_recursive"
     )
 
     pytest "${ARGS[@]}" "${ZONAL_DESELECTS[@]}" gcsfs/tests/test_core.py

docs/source/hns_buckets.rst

@@ -56,7 +56,8 @@ While ``gcsfs`` aims to abstract the differences via the ``fsspec`` API, you sho
 2. **``mkdir`` behavior:** Previously, in a flat namespace, calling ``mkdir`` on a path could only ensure the underlying bucket exists. With HNS enabled, calling ``mkdir`` will create an actual folder resource in GCS. Furthermore, if you want to create nested folders (eg: bucket/a/b/c/d), pass ``create_parents=True``, it will physically create all intermediate folder resources along the specified path.
 3. **No mixing or toggling:** You cannot toggle HNS on an existing flat-namespace bucket. You must create a new HNS bucket and migrate your data.
 4. **Object naming:** Object names in HNS cannot end with a slash (``/``) unless without the creation of physical folder resources.
-5. **Rename Operation Benchmarks**
+5. **Non-Recursive Wildcard Deletions:** When using wildcard deletions without recursion (e.g., ``gcs.rm("bucket/dir/*", recursive=False)``), if the pattern matches a non-empty directory, HNS buckets will raise an error (typically surfaced as an ``OSError`` due to a failed precondition). In contrast, standard flat-namespace buckets silently ignore non-empty directories under the same circumstances.
+6. **Rename Operation Benchmarks**
 
 The following benchmarks show the time taken (in seconds) to rename a directory containing a large number of files (spread across 256 folders and 8 levels) in a standard Regional bucket versus an HNS bucket (can be replicated using `gcsfs/tests/perf/microbenchmarks/rename`):
 

gcsfs/extended_gcsfs.py

@@ -811,7 +811,7 @@ async def _expand_path_with_details(
             raise ValueError("maxdepth must be at least 1")
 
         if isinstance(path, str):
-            out = await self._expand_path_with_details(
+            return await self._expand_path_with_details(
                 [path], recursive, maxdepth, detail=detail
             )
         else:
@@ -1000,6 +1000,8 @@ async def _find(
         For buckets with flat structure, it falls back to the parent implementation.
         """
         path = self._strip_protocol(path)
+        if maxdepth is not None and maxdepth < 1:
+            raise ValueError("maxdepth must be at least 1")
         bucket, _, _ = self.split_path(path)
 
         is_hns = await self._is_bucket_hns_enabled(bucket)
@@ -1017,14 +1019,16 @@ async def _find(
 
         # Hybrid approach for HNS enabled buckets
         # 1. Fetch all files from super find() method by passing withdirs as False.
+        # We pass maxdepth as None here to ensure we fetch all files for caching,
+        # and then filter by maxdepth at the end of this method.
         files_task = asyncio.create_task(
             super()._find(
                 path,
                 withdirs=False,  # Fetch files only
                 detail=True,  # Get full details for merging and populating cache
                 prefix=prefix,
                 versions=versions,
-                maxdepth=maxdepth,
+                maxdepth=None,
                 update_cache=False,  # Defer caching until merging files and folders
                 **kwargs,
             )

gcsfs/tests/integration/test_extended_hns.py

@@ -918,6 +918,66 @@ def test_rm_empty_folder_cache_invalidation(self, gcs_hns):
             sibling_dir in gcsfs.dircache
         ), "Sibling directory cache should not be invalidated."
 
+    def test_rm_wildcard_with_empty_folders(self, gcs_hns):
+        """Test deleting files and empty folders using wildcards on HNS bucket."""
+        gcsfs = gcs_hns
+        base_dir = f"{TEST_HNS_BUCKET}/rm_wildcard_empty_{uuid.uuid4().hex}"
+
+        # Setup: Some files, some empty folders, some non-empty folders
+        empty_dir1 = f"{base_dir}/empty_1"
+        empty_dir2 = f"{base_dir}/empty_2"
+        empty_dir3 = f"{base_dir}/one_more_empty_dir"
+        other_dir = f"{base_dir}/other_dir"
+        file1 = f"{base_dir}/file1.txt"
+
+        gcsfs.mkdir(empty_dir1, create_parents=True)
+        gcsfs.mkdir(empty_dir2)
+        gcsfs.touch(file1)
+        gcsfs.mkdir(empty_dir3)
+        gcsfs.touch(f"{other_dir}/file2.txt")
+
+        # 1. Remove only matching empty folders using wildcard
+        gcsfs.rm(f"{base_dir}/empty_*", recursive=True)
+
+        assert not gcsfs.exists(empty_dir1)
+        assert not gcsfs.exists(empty_dir2)
+        assert gcsfs.exists(file1)
+        assert gcsfs.exists(other_dir)
+
+        # 2. Remove remaining using a broad wildcard
+        gcsfs.rm(f"{base_dir}/*", recursive=True)
+
+        assert not gcsfs.exists(file1)
+        assert not gcsfs.exists(other_dir)
+        assert not gcsfs.exists(empty_dir3)
+
+        # Verify base_dir is now empty (HNS might still have the base_dir itself)
+        assert gcsfs.ls(base_dir) == []
+
+        gcsfs.rm(base_dir, recursive=True)
+
+        assert not gcsfs.exists(base_dir)
+
+    def test_rm_wildcards_non_recursive(self, gcs_hns):
+        """Test 'base_dir/*' (non-recursive) raises an error on non-empty directories for HNS buckets."""
+        gcsfs = gcs_hns
+        base_dir = f"{TEST_HNS_BUCKET}/test_rm_asterisk_{uuid.uuid4().hex}"
+        files = [
+            f"{base_dir}/other.txt",
+            f"{base_dir}/subdir/nested.txt",
+        ]
+        for f in files:
+            gcsfs.touch(f)
+
+        # 3. Test 'base_dir/*' (non-recursive)
+        # For HNS, this matches `other.txt` and `subdir`. Because `subdir` is not empty,
+        # delete_folder raises FailedPrecondition, which currently wraps to OSError.
+        with pytest.raises(OSError, match="Pre condition failed"):
+            gcsfs.rm(f"{base_dir}/*")
+
+        assert not gcsfs.exists(files[0])
+        assert gcsfs.exists(files[1])  # subdir file still exists
+
 
 @pytest.fixture()
 def test_structure(gcs_hns):
@@ -1052,6 +1112,28 @@ def test_find_updates_dircache_without_prefix(
         }
         assert not empty_dir_listing
 
+    def test_find_maxdepth_updates_cache(self, gcs_hns, test_structure):
+        """Test that find with maxdepth updates cache for deeper objects."""
+        base_dir = test_structure["base_dir"]
+        gcs_hns.invalidate_cache()
+        assert not gcs_hns.dircache
+
+        # Run find with maxdepth=1
+        # This should return root_file, empty_dir, dir_with_files
+        # But it should also cache objects within dir_with_files, etc.
+        result = gcs_hns.find(base_dir, maxdepth=1)
+        assert test_structure["root_file"] in result
+        assert test_structure["file1"] not in result
+
+        # Verify that the cache is populated for deeper objects even if not returned
+        # from find due to maxdepth filter
+        assert test_structure["dir_with_files"] in gcs_hns.dircache
+        dir_with_files_cache = {
+            d["name"] for d in gcs_hns.dircache[test_structure["dir_with_files"]]
+        }
+        assert test_structure["file1"] in dir_with_files_cache
+        assert test_structure["nested_dir"] in dir_with_files_cache
+
     def test_find_does_not_update_dircache_with_prefix(self, gcs_hns, test_structure):
         """Test that find() does NOT populate the dircache when a prefix is given."""
         base_dir = test_structure["base_dir"]

gcsfs/tests/test_core.py

@@ -354,6 +354,87 @@ def test_rm_chunked_batch(gcs):
         assert fn not in files_removed
 
 
+def test_rm_wildcards_in_directory(gcs):
+    base_dir = f"{TEST_BUCKET}/test_rm_complex_{uuid.uuid4().hex}"
+    files = [
+        f"{base_dir}/file1.txt",
+        f"{base_dir}/file2.txt",
+        f"{base_dir}/other.txt",
+        f"{base_dir}/a1.dat",
+        f"{base_dir}/b1.dat",
+        f"{base_dir}/subdir/nested.txt",
+    ]
+    for f in files:
+        gcs.touch(f)
+
+    # 1. Test '?' wildcard (non-recursive)
+    gcs.rm(f"{base_dir}/file?.txt")
+    assert not gcs.exists(files[0])
+    assert not gcs.exists(files[1])
+    assert gcs.exists(files[2])
+    assert gcs.exists(files[5])
+
+    # 2. Test list of patterns with '*' wildcards (non-recursive)
+    gcs.rm([f"{base_dir}/a*", f"{base_dir}/b*"])
+    assert not gcs.exists(files[3])
+    assert not gcs.exists(files[4])
+    assert gcs.exists(files[2])
+    assert gcs.exists(files[5])  # subdir file still exists
+
+    # 3. Test non-matching pattern
+    with pytest.raises(FileNotFoundError):
+        gcs.rm(f"{base_dir}/non_existent*")
+
+    # 4. Test recursive wildcard cleanup
+    gcs.rm(f"{base_dir}/**", recursive=True)
+    assert not gcs.exists(base_dir)
+
+
+def test_rm_wildcard_bucket_non_recursive(gcs):
+    """Tests non-recursive rm with a wildcard at the bucket root."""
+    # Test 'bucket/*' (non-recursive)
+    new_bucket = f"gcsfs-test-rm-{uuid.uuid4().hex}"
+    gcs.mkdir(new_bucket)
+    try:
+        gcs.touch(f"{new_bucket}/file1")
+        gcs.touch(f"{new_bucket}/subdir/file2")
+        gcs.rm(f"{new_bucket}/*")
+        assert not gcs.exists(f"{new_bucket}/file1")
+        assert gcs.exists(f"{new_bucket}/subdir/file2")
+    finally:
+        gcs.rm(new_bucket, recursive=True)
+
+
+def test_rm_wildcard_bucket_recursive(gcs):
+    """Tests recursive rm with a wildcard at the bucket root."""
+    # Test 'bucket/*' (recursive)
+    new_bucket = f"gcsfs-test-rm-recursive-{uuid.uuid4().hex}"
+    gcs.mkdir(new_bucket)
+    try:
+        gcs.touch(f"{new_bucket}/file1")
+        gcs.touch(f"{new_bucket}/subdir/file2")
+        gcs.rm(f"{new_bucket}/*", recursive=True)
+        assert not gcs.exists(f"{new_bucket}/file1")
+        assert not gcs.exists(f"{new_bucket}/subdir/file2")
+    finally:
+        if gcs.exists(new_bucket):
+            gcs.rm(new_bucket, recursive=True)
+
+
+def test_rm_wildcards_non_recursive(gcs):
+    base_dir = f"{TEST_BUCKET}/test_rm_asterisk_{uuid.uuid4().hex}"
+    files = [
+        f"{base_dir}/other.txt",
+        f"{base_dir}/subdir/nested.txt",
+    ]
+    for f in files:
+        gcs.touch(f)
+
+    gcs.rm(f"{base_dir}/*")
+    assert not gcs.exists(files[0])
+    assert gcs.exists(files[1])  # subdir file still exists
+
+
 def test_file_access(gcs):
     fn = TEST_BUCKET + "/nested/file1"
     data = b"hello\n"

gcsfs/tests/test_extended_hns_gcsfs.py

@@ -1116,7 +1116,7 @@ def test_hns_find_withdirs_maxdepth(self, gcs_hns, gcs_hns_mocks):
             _, call_kwargs = mocks["super_find"].call_args
             assert call_kwargs.get("withdirs") is False
             assert call_kwargs.get("detail") is True
-            assert call_kwargs.get("maxdepth") == 1
+            assert call_kwargs.get("maxdepth") is None
 
             # Assert that list_folders was called with the correct request
             expected_folder_id = "find_test/"