Files
dvc/tests/unit/remote/test_base.py
T
Peter Rowlands (변기호) 74bd8137ba remote: finish separating RemoteTree methods (#3946)
* remote: move upload()/download() into tree

* remote: move path_info/path_cls into tree

* remote.azure: finish moving methods into tree

* remote.gs: finish moving methods into tree

* remote.hdfs: finish moving methods to tree

* remote.http: finish moving methods into tree

* remote.oss: finish moving methods into tree

* remote.s3: finish moving methods into tree

* remote.ssh: finish moving methods into tree

* remote.gdrive: finish moving methods into tree

* tests: update remote unit tests

* tests: update func tests

* remote: use walk_files() for all remotes

* list_cache_paths() now uses tree.walk_files() for all remotes except
  local/ssh

* fix DS warnings

* bugfixes
2020-06-04 23:22:31 +03:00

147 lines
4.2 KiB
Python

import math
import mock
import pytest
from dvc.path_info import PathInfo
from dvc.remote.base import BaseRemote, RemoteCmdError, RemoteMissingDepsError
class _CallableOrNone:
"""Helper for testing if object is callable() or None."""
def __eq__(self, other):
return other is None or callable(other)
CallableOrNone = _CallableOrNone()
REMOTE_CLS = BaseRemote
def test_missing_deps(dvc):
requires = {"missing": "missing"}
with mock.patch.object(REMOTE_CLS, "REQUIRES", requires):
with pytest.raises(RemoteMissingDepsError):
REMOTE_CLS(dvc, {})
def test_cmd_error(dvc):
config = {}
cmd = "sed 'hello'"
ret = "1"
err = "sed: expression #1, char 2: extra characters after command"
with mock.patch.object(
REMOTE_CLS.TREE_CLS,
"remove",
side_effect=RemoteCmdError("base", cmd, ret, err),
):
with pytest.raises(RemoteCmdError):
REMOTE_CLS(dvc, config).tree.remove("file")
@mock.patch.object(BaseRemote, "_cache_checksums_traverse")
@mock.patch.object(BaseRemote, "_cache_object_exists")
def test_cache_exists(object_exists, traverse, dvc):
remote = BaseRemote(dvc, {})
# remote does not support traverse
remote.CAN_TRAVERSE = False
with mock.patch.object(
remote, "cache_checksums", return_value=list(range(256))
):
checksums = set(range(1000))
remote.cache_exists(checksums)
object_exists.assert_called_with(checksums, None, None)
traverse.assert_not_called()
remote.CAN_TRAVERSE = True
# large remote, small local
object_exists.reset_mock()
traverse.reset_mock()
with mock.patch.object(
remote, "cache_checksums", return_value=list(range(256))
):
checksums = list(range(1000))
remote.cache_exists(checksums)
# verify that _cache_paths_with_max() short circuits
# before returning all 256 remote checksums
max_checksums = math.ceil(
remote._max_estimation_size(checksums)
/ pow(16, remote.TRAVERSE_PREFIX_LEN)
)
assert max_checksums < 256
object_exists.assert_called_with(
frozenset(range(max_checksums, 1000)), None, None
)
traverse.assert_not_called()
# large remote, large local
object_exists.reset_mock()
traverse.reset_mock()
remote.JOBS = 16
with mock.patch.object(
remote, "cache_checksums", return_value=list(range(256))
):
checksums = list(range(1000000))
remote.cache_exists(checksums)
object_exists.assert_not_called()
traverse.assert_called_with(
256 * pow(16, remote.TRAVERSE_PREFIX_LEN),
set(range(256)),
None,
None,
)
@mock.patch.object(
BaseRemote, "cache_checksums", return_value=[],
)
@mock.patch.object(
BaseRemote, "path_to_checksum", side_effect=lambda x: x,
)
def test_cache_checksums_traverse(path_to_checksum, cache_checksums, dvc):
remote = BaseRemote(dvc, {})
remote.tree.path_info = PathInfo("foo")
# parallel traverse
size = 256 / remote.JOBS * remote.LIST_OBJECT_PAGE_SIZE
list(remote._cache_checksums_traverse(size, {0}))
for i in range(1, 16):
cache_checksums.assert_any_call(
prefix=f"{i:03x}", progress_callback=CallableOrNone
)
for i in range(1, 256):
cache_checksums.assert_any_call(
prefix=f"{i:02x}", progress_callback=CallableOrNone
)
# default traverse (small remote)
size -= 1
cache_checksums.reset_mock()
list(remote._cache_checksums_traverse(size - 1, {0}))
cache_checksums.assert_called_with(
prefix=None, progress_callback=CallableOrNone
)
def test_cache_checksums(dvc):
remote = BaseRemote(dvc, {})
remote.tree.path_info = PathInfo("foo")
with mock.patch.object(
remote, "list_cache_paths", return_value=["12/3456", "bar"]
):
checksums = list(remote.cache_checksums())
assert checksums == ["123456"]
@pytest.mark.parametrize(
"checksum, result",
[(None, False), ("", False), ("3456.dir", True), ("3456", False)],
)
def test_is_dir_checksum(checksum, result):
assert BaseRemote.is_dir_checksum(checksum) == result