def helper_resolve_all_files(self, dataset, kwargs):
"""Helper method to populate the DatasetFileConnection"""
manifest = Manifest(dataset, get_logged_in_username())
if "after" in kwargs:
after_index = int(base64.b64decode(kwargs["after"]))
else:
after_index = 0
# Generate naive cursors
edges, indexes = manifest.list(first=kwargs.get("first"),
after_index=after_index)
cursors = [
base64.b64encode("{}".format(x).encode("UTF-8")).decode("UTF-8")
for x in indexes
]
edge_objs = []
for edge, cursor in zip(edges, cursors):
create_data = {
"owner": self.owner,
"name": self.name,
"key": edge['key'],
"_file_info": edge
}
edge_objs.append(
DatasetFileConnection.Edge(node=DatasetFile(**create_data),
cursor=cursor))
has_previous_page = False
has_next_page = len(edges) > 0
start_cursor = None
end_cursor = None
if cursors:
start_cursor = cursors[0]
end_cursor = cursors[-1]
if indexes[-1] == len(manifest.manifest) - 1:
has_next_page = False
if kwargs.get("after"):
if int(base64.b64decode(kwargs["after"])) > 0:
has_previous_page = True
page_info = graphene.relay.PageInfo(
has_next_page=has_next_page,
has_previous_page=has_previous_page,
start_cursor=start_cursor,
end_cursor=end_cursor)
return DatasetFileConnection(edges=edge_objs, page_info=page_info)
def test_complete_dataset_upload_transaction_simple(
self, mock_config_file_background_tests):
im = InventoryManager(mock_config_file_background_tests[0])
ds = im.create_dataset('default',
'default',
"new-ds",
storage_type="gigantum_object_v1",
description="100")
m = Manifest(ds, 'default')
helper_append_file(m.cache_mgr.cache_root, m.dataset_revision,
"test1.txt", "fake content!")
helper_append_file(m.cache_mgr.cache_root, m.dataset_revision,
"test2.txt", "moar fake content!")
dl_kwargs = {
'dispatcher': Dispatcher,
'logged_in_username': "******",
'logged_in_email': "*****@*****.**",
'dataset_owner': "default",
'dataset_name': "new-ds",
'config_file': mock_config_file_background_tests[0]
}
assert len(m.manifest) == 0
gtmcore.dispatcher.dataset_jobs.complete_dataset_upload_transaction(
**dl_kwargs)
m = Manifest(ds, 'default')
# make sure manifest got updated
assert len(m.manifest) == 2
assert 'test1.txt' in m.manifest
assert 'test2.txt' in m.manifest
assert m.manifest['test1.txt']['b'] == '13'
assert len(m.manifest['test1.txt']['h']) == 128
assert 'manifest-' in m.manifest['test1.txt']['fn']
assert m.manifest['test2.txt']['b'] == '18'
assert len(m.manifest['test2.txt']['h']) == 128
assert 'manifest-' in m.manifest['test2.txt']['fn']
assert m.manifest['test2.txt']['h'] != m.manifest['test1.txt']['h']
# Make sure activity created
assert len(ds.git.log()) == 6
assert "_GTM_ACTIVITY_START_" in ds.git.log()[0]['message']
assert "Uploaded 2 new file(s)." in ds.git.log()[0]['message']
def generate_bg_hash_job_list(filenames: List[str],
manifest: Manifest,
dispatcher_obj: Dispatcher) -> List[BackgroundHashJob]:
"""Method to generate batches of files to be hashed, ensuring files aren't added to a batch once it is
larger than MAX_JOB_BYTES
Args:
filenames: list of files to be hashed
manifest: the Manifest instance
dispatcher_obj: the Dispatcher instance
Returns:
list
"""
num_cores = manifest.get_num_hashing_cpus()
file_lists: List[List] = [list() for _ in range(num_cores)]
size_sums = [0 for _ in range(num_cores)]
revision_dir = manifest.current_revision_dir
for filename in filenames:
index = size_sums.index(min(size_sums))
file_lists[index].append(filename)
size_sums[index] += os.path.getsize(os.path.join(revision_dir, filename))
if all(fs > MAX_JOB_BYTES for fs in size_sums):
# 1GB of data to hash already in every job. Add another.
file_lists.append(list())
size_sums.append(0)
# Prune Jobs back if there are lots of cores but not lots of work
file_lists = [x for x in file_lists if x != []]
size_sums = [x for x in size_sums if x != 0]
# Prep hashing jobs
return [BackgroundHashJob(dispatcher_obj, fl, ss) for ss, fl in zip(size_sums, file_lists)]
def _get_dataset_file_info(self, dataset) -> dict:
"""helper method to iterate over the manifest and get file info for the overview page
Returns:
None
"""
m = Manifest(dataset, get_logged_in_username())
count = 0
total_bytes = 0
file_type_distribution: OrderedDict = OrderedDict()
for key in m.manifest:
item = m.manifest[key]
if key[-1] == '/':
# Skip directories
continue
filename = os.path.basename(key)
if filename[0] == '.':
# Skip hidden files
continue
if '.' not in filename:
# Skip files without an extension
continue
# Count file type distribution
_, ext = os.path.splitext(filename)
if ext:
file_type = ext
if file_type in file_type_distribution:
file_type_distribution[file_type] += 1
else:
file_type_distribution[file_type] = 1
# Count total file size
total_bytes += int(item['b'])
# Count files
count += 1
# Format the output for file type distribution
formatted_file_type_info: List[str] = list()
file_type_distribution = OrderedDict(
sorted(file_type_distribution.items(),
key=itemgetter(1),
reverse=True))
for file_type in file_type_distribution:
percentage = float(
file_type_distribution[file_type]) / float(count)
formatted_file_type_info.append(f"{percentage:.2f}|{file_type}")
self._dataset_file_info = {
'num_files': count,
'total_bytes': total_bytes,
'local_bytes': count,
'file_type_distribution': formatted_file_type_info
}
return self._dataset_file_info
def _push_dataset_objects(self, dataset: Dataset, logged_in_username: str,
feedback_callback: Callable, access_token,
id_token) -> None:
dataset.backend.set_default_configuration(logged_in_username,
access_token, id_token)
m = Manifest(dataset, logged_in_username)
iom = IOManager(dataset, m)
iom.push_objects(status_update_fn=feedback_callback)
iom.manifest.link_revision()
def test_move_dataset_file(self, fixture_working_dir, snapshot):
im = InventoryManager(fixture_working_dir[0])
ds = im.create_dataset('default',
'default',
"dataset-move",
storage_type="gigantum_object_v1",
description="testing move")
m = Manifest(ds, 'default')
revision = m.dataset_revision
helper_append_file(m.cache_mgr.cache_root, revision, "test1.txt",
"asdfasdghndfdf")
m.sweep_all_changes()
revision = m.dataset_revision
cr = m.cache_mgr.cache_root
assert os.path.exists(os.path.join(cr, revision, "test1.txt")) is True
query = """
mutation myMutation {
moveDatasetFile(input: {datasetOwner: "default", datasetName: "dataset-move",
srcPath: "test1.txt", dstPath: "test1-renamed.txt"}) {
updatedEdges {
node {
id
key
isDir
isLocal
size
}
}
}
}
"""
result = fixture_working_dir[2].execute(query)
assert 'errors' not in result
snapshot.assert_match(result)
revision = m.dataset_revision
cr = m.cache_mgr.cache_root
assert os.path.exists(os.path.join(cr, revision, "test1.txt")) is False
assert os.path.exists(os.path.join(cr, revision,
"test1-renamed.txt")) is True
def mutate_and_get_payload(cls,
root,
info,
dataset_owner,
dataset_name,
keys,
client_mutation_id=None):
logged_in_username = get_logged_in_username()
ds = InventoryManager().load_dataset(logged_in_username,
dataset_owner,
dataset_name,
author=get_logged_in_author())
ds.namespace = dataset_owner
m = Manifest(ds, logged_in_username)
with ds.lock():
m.delete(keys)
return DeleteDatasetFiles(success=True)
def test_delete_dataset_while_linked(self, mock_config_file):
inv_manager = InventoryManager(mock_config_file[0])
auth = GitAuthor(name="test", email="*****@*****.**")
lb = inv_manager.create_labbook("test",
"test",
"labbook1",
description="my first labbook")
ds = inv_manager.create_dataset("test",
"test",
"dataset1",
"gigantum_object_v1",
description="my first dataset",
author=auth)
ds_root_dir = ds.root_dir
lb_root_dir = lb.root_dir
assert os.path.exists(ds_root_dir) is True
assert os.path.exists(lb_root_dir) is True
# Link dataset
inv_manager.link_dataset_to_labbook(f"{ds_root_dir}/.git", "test",
"dataset1", lb)
m = Manifest(ds, 'test')
helper_append_file(m.cache_mgr.cache_root, m.dataset_revision,
"test1.txt", "asdfasdf")
helper_append_file(m.cache_mgr.cache_root, m.dataset_revision,
"test2.txt", "dfg")
assert os.path.exists(
os.path.join(m.cache_mgr.cache_root, m.dataset_revision,
"test1.txt")) is True
assert os.path.exists(
os.path.join(m.cache_mgr.cache_root, m.dataset_revision,
"test2.txt")) is True
dataset_delete_job = inv_manager.delete_dataset(
"test", "test", "dataset1")
assert os.path.exists(ds_root_dir) is False
assert os.path.exists(lb_root_dir) is True
assert os.path.exists(m.cache_mgr.cache_root) is True
assert dataset_delete_job.namespace == "test"
assert dataset_delete_job.name == "dataset1"
assert dataset_delete_job.cache_root == m.cache_mgr.cache_root
jobs.clean_dataset_file_cache("test",
dataset_delete_job.namespace,
dataset_delete_job.name,
dataset_delete_job.cache_root,
config_file=mock_config_file[0])
assert os.path.exists(m.cache_mgr.cache_root) is True
cache_base, _ = m.cache_mgr.cache_root.rsplit(os.path.sep, 1)
assert os.path.exists(cache_base) is True
def mutate_and_get_payload(cls,
root,
info,
dataset_owner,
dataset_name,
src_path,
dst_path,
client_mutation_id=None):
logged_in_username = get_logged_in_username()
ds = InventoryManager().load_dataset(logged_in_username,
dataset_owner,
dataset_name,
author=get_logged_in_author())
ds.namespace = dataset_owner
m = Manifest(ds, logged_in_username)
with ds.lock():
edge_data = m.move(src_path, dst_path)
file_edges = list()
for edge_dict in edge_data:
file_edges.append(
DatasetFile(owner=dataset_owner,
name=dataset_name,
key=edge_dict['key'],
is_dir=edge_dict['is_dir'],
is_favorite=edge_dict['is_favorite'],
modified_at=edge_dict['modified_at'],
is_local=edge_dict['is_local'],
size=str(edge_dict['size'])))
cursors = [
base64.b64encode("{}".format(cnt).encode("UTF-8")).decode("UTF-8")
for cnt, x in enumerate(file_edges)
]
edge_objs = [
DatasetFileConnection.Edge(node=e, cursor=c)
for e, c in zip(file_edges, cursors)
]
return MoveDatasetFile(updated_edges=edge_objs)
def test_sync__dataset(self, mock_config_file):
def update_feedback(msg: str,
has_failures: Optional[bool] = None,
failure_detail: Optional[str] = None,
percent_complete: Optional[float] = None):
"""Method to update the job's metadata and provide feedback to the UI"""
assert has_failures is None or has_failures is False
assert failure_detail is None
def dispatch_query_mock(self, job_key):
JobStatus = namedtuple("JobStatus", ['status', 'meta'])
return JobStatus(status='finished',
meta={'completed_bytes': '100'})
def dispatch_mock(self, method_reference, kwargs, metadata, persist):
return "afakejobkey"
username = '******'
im = InventoryManager(mock_config_file[0])
ds = im.create_dataset(username, username, 'dataset-1',
'gigantum_object_v1')
m = Manifest(ds, username)
wf = DatasetWorkflow(ds)
iom = IOManager(ds, m)
assert len(glob.glob(f'{iom.push_dir}/*')) == 0
wf.publish(username=username, feedback_callback=update_feedback)
# Put a file into the dataset that needs to be pushed
helper_append_file(m.cache_mgr.cache_root, m.dataset_revision,
"test1.txt", "asdfadfsdf")
m.sweep_all_changes()
assert len(glob.glob(f'{iom.push_dir}/*')) == 1
with patch.object(Dispatcher, 'dispatch_task', dispatch_mock):
with patch.object(Dispatcher, 'query_task', dispatch_query_mock):
wf.sync(username=username, feedback_callback=update_feedback)
assert os.path.exists(wf.remote)
assert len(glob.glob(f'{iom.push_dir}/*')) == 0
def helper_resolve_all_files(self, dataset, kwargs):
"""Helper method to populate the DatasetFileConnection"""
manifest = Manifest(dataset, get_logged_in_username())
# Generate naive cursors
# TODO: Use manifest pagination interface
edges = manifest.list()
cursors = [base64.b64encode("{}".format(cnt).encode("UTF-8")).decode("UTF-8") for cnt, x in enumerate(edges)]
# Process slicing and cursor args
lbc = ListBasedConnection(edges, cursors, kwargs)
lbc.apply()
edge_objs = []
for edge, cursor in zip(lbc.edges, lbc.cursors):
create_data = {"owner": self.owner,
"name": self.name,
"key": edge['key'],
"_file_info": edge}
edge_objs.append(DatasetFileConnection.Edge(node=DatasetFile(**create_data), cursor=cursor))
return DatasetFileConnection(edges=edge_objs, page_info=lbc.page_info)
def _helper_local_bytes(dataset):
"""Helper to compute total size of a dataset on disk"""
m = Manifest(dataset, get_logged_in_username())
total_size = 0
for dirpath, dirnames, filenames in os.walk(m.current_revision_dir):
for f in filenames:
if f == '.smarthash':
continue
fp = os.path.join(dirpath, f)
total_size += os.path.getsize(fp)
return total_size
def test_make_directory_error(self, fixture_working_dir, snapshot):
im = InventoryManager(fixture_working_dir[0])
ds = im.create_dataset('default',
'default',
"dataset-dir",
storage_type="gigantum_object_v1",
description="testing move")
m = Manifest(ds, 'default')
# Test where parent dir doesnt exists (because you need to create the parent)
query = """
mutation myMutation {
makeDatasetDirectory(input: {datasetOwner: "default", datasetName: "dataset-dir",
key: "test_dir1/test_dir2/"}) {
newDatasetFileEdge {
node {
id
key
isDir
isLocal
size
}
}
}
}
"""
result = fixture_working_dir[2].execute(query)
assert 'errors' in result
assert 'Parent directory' in result['errors'][0]['message']
# Test where missing trailing slash
query = """
mutation myMutation {
makeDatasetDirectory(input: {datasetOwner: "default", datasetName: "dataset-dir",
key: "test_dir1"}) {
newDatasetFileEdge {
node {
id
key
isDir
isLocal
size
}
}
}
}
"""
result = fixture_working_dir[2].execute(query)
assert 'errors' in result
assert 'Provided relative path must end in' in result['errors'][0][
'message']
def mutate_and_get_payload(cls,
root,
info,
dataset_owner,
dataset_name,
key,
client_mutation_id=None):
logged_in_username = get_logged_in_username()
ds = InventoryManager().load_dataset(logged_in_username,
dataset_owner,
dataset_name,
author=get_logged_in_author())
ds.namespace = dataset_owner
m = Manifest(ds, logged_in_username)
if key[-1] != '/':
raise ValueError(
"Provided relative path must end in `/` to indicate it is a directory"
)
with ds.lock():
file_info = m.create_directory(key)
create_data = {
'owner': dataset_owner,
'name': dataset_name,
'key': file_info['key'],
'_file_info': file_info
}
# TODO: Fix cursor implementation, this currently doesn't make sense
cursor = base64.b64encode(f"{0}".encode('utf-8'))
return MakeDatasetDirectory(
new_dataset_file_edge=DatasetFileConnection.Edge(
node=DatasetFile(**create_data), cursor=cursor))
def delete_dataset(self, username: str, owner: str,
dataset_name: str) -> None:
"""Delete a Dataset from this Gigantum working directory.
Args:
username: Active username
owner: Namespace in which to place this Dataset
dataset_name: Name of the Datasets
Returns:
None
"""
ds = self.load_dataset(username, owner, dataset_name)
# Delete dataset contents from file cache
m = Manifest(ds, username)
shutil.rmtree(m.cache_mgr.cache_root, ignore_errors=True)
# Delete dataset repository from working dir
shutil.rmtree(ds.root_dir, ignore_errors=True)
def test_delete_dataset(self, mock_config_file):
inv_manager = InventoryManager(mock_config_file[0])
auth = GitAuthor(name="test", email="*****@*****.**")
ds = inv_manager.create_dataset("test", "test", "dataset1", "gigantum_object_v1",
description="my first dataset",
author=auth)
root_dir = ds.root_dir
assert os.path.exists(root_dir) is True
m = Manifest(ds, 'test')
helper_append_file(m.cache_mgr.cache_root, m.dataset_revision, "test1.txt", "asdfasdf")
helper_append_file(m.cache_mgr.cache_root, m.dataset_revision, "test2.txt", "dfg")
assert os.path.exists(os.path.join(m.cache_mgr.cache_root, m.dataset_revision, "test1.txt")) is True
assert os.path.exists(os.path.join(m.cache_mgr.cache_root, m.dataset_revision, "test2.txt")) is True
inv_manager.delete_dataset("test", "test", "dataset1")
assert os.path.exists(root_dir) is False
assert os.path.exists(m.cache_mgr.cache_root) is False
cache_base, _ = m.cache_mgr.cache_root.rsplit(os.path.sep, 1)
assert os.path.exists(cache_base) is True
def test_delete_dataset_files(self, fixture_working_dir, snapshot):
im = InventoryManager(fixture_working_dir[0])
ds = im.create_dataset('default',
'default',
"dataset-delete",
storage_type="gigantum_object_v1",
description="testing delete")
m = Manifest(ds, 'default')
os.makedirs(
os.path.join(m.cache_mgr.cache_root, m.dataset_revision,
"other_dir"))
helper_append_file(m.cache_mgr.cache_root, m.dataset_revision,
"test1.txt", "asdfadfsdf")
helper_append_file(m.cache_mgr.cache_root, m.dataset_revision,
"test2.txt", "fdsfgfd")
helper_append_file(m.cache_mgr.cache_root, m.dataset_revision,
"test3.txt", "ghgdsr")
helper_append_file(m.cache_mgr.cache_root, m.dataset_revision,
"other_dir/test3.txt", "hhgf")
helper_append_file(m.cache_mgr.cache_root, m.dataset_revision,
"other_dir/test1.txt", "jkjghfg")
m.sweep_all_changes()
revision = m.dataset_revision
assert os.path.exists(
os.path.join(m.cache_mgr.cache_root, revision,
"test1.txt")) is True
assert os.path.exists(
os.path.join(m.cache_mgr.cache_root, revision,
"test2.txt")) is True
assert os.path.exists(
os.path.join(m.cache_mgr.cache_root, revision,
"test3.txt")) is True
assert os.path.exists(
os.path.join(m.cache_mgr.cache_root, revision, "other_dir",
"test3.txt")) is True
assert os.path.exists(
os.path.join(m.cache_mgr.cache_root, revision, "other_dir",
"test1.txt")) is True
query = """
mutation myMutation {
deleteDatasetFiles(input: {datasetOwner: "default", datasetName: "dataset-delete",
keys: ["test1.txt"]}) {
success
}
}
"""
result = fixture_working_dir[2].execute(query)
assert 'errors' not in result
assert result['data']['deleteDatasetFiles']['success'] is True
revision = m.dataset_revision
assert os.path.exists(
os.path.join(m.cache_mgr.cache_root, revision,
"test1.txt")) is False
assert os.path.exists(
os.path.join(m.cache_mgr.cache_root, revision,
"test2.txt")) is True
assert os.path.exists(
os.path.join(m.cache_mgr.cache_root, revision,
"test3.txt")) is True
assert os.path.exists(
os.path.join(m.cache_mgr.cache_root, revision, "other_dir",
"test3.txt")) is True
assert os.path.exists(
os.path.join(m.cache_mgr.cache_root, revision, "other_dir",
"test1.txt")) is True
query = """
mutation myMutation {
deleteDatasetFiles(input: {datasetOwner: "default", datasetName: "dataset-delete",
keys: ["test3.txt", "other_dir/"]}) {
success
}
}
"""
result = fixture_working_dir[2].execute(query)
assert 'errors' not in result
assert result['data']['deleteDatasetFiles']['success'] is True
revision = m.dataset_revision
assert os.path.exists(
os.path.join(m.cache_mgr.cache_root, revision,
"test1.txt")) is False
assert os.path.exists(
os.path.join(m.cache_mgr.cache_root, revision,
"test2.txt")) is True
assert os.path.exists(
os.path.join(m.cache_mgr.cache_root, revision,
"test3.txt")) is False
assert os.path.exists(
os.path.join(m.cache_mgr.cache_root, revision, "other_dir",
"test3.txt")) is False
assert os.path.exists(
os.path.join(m.cache_mgr.cache_root, revision, "other_dir",
"test1.txt")) is False
def test_update_dataset_link(self, fixture_working_dir, snapshot):
im = InventoryManager(fixture_working_dir[0])
lb = im.create_labbook('default', 'default', 'test-lb',
'testing dataset links')
ds = im.create_dataset('default',
'default',
"dataset100",
storage_type="gigantum_object_v1",
description="100")
manifest = Manifest(ds, 'default')
helper_append_file(manifest.cache_mgr.cache_root,
manifest.dataset_revision, "test1.txt", "12345")
manifest.sweep_all_changes()
# Fake publish to a local bare repo
_MOCK_create_remote_repo2(ds, 'default', None, None)
assert os.path.exists(os.path.join(lb.root_dir,
'.gitmodules')) is False
overview_query = """
{
labbook(owner: "default", name:"test-lb")
{
linkedDatasets{
name
overview {
localBytes
totalBytes
}
}
}
}
"""
query = """
mutation myMutation($lo: String!, $ln: String!, $do: String!, $dn: String!,
$a: String!, $du: String) {
modifyDatasetLink(input: {labbookOwner: $lo, labbookName: $ln, datasetOwner: $do, datasetName: $dn,
action: $a, datasetUrl: $du}) {
newLabbookEdge {
node {
id
name
description
linkedDatasets {
name
}
}
}
}
}
"""
variables = {
"lo": "default",
"ln": "test-lb",
"do": "default",
"dn": "dataset100",
"a": "link",
"du": ds.remote
}
result = fixture_working_dir[2].execute(query,
variable_values=variables)
assert "errors" not in result
snapshot.assert_match(result)
assert os.path.exists(os.path.join(lb.root_dir, '.gitmodules')) is True
dataset_submodule_dir = os.path.join(lb.root_dir, '.gigantum',
'datasets', 'default',
'dataset100')
assert os.path.exists(dataset_submodule_dir) is True
assert os.path.exists(os.path.join(dataset_submodule_dir,
'.gigantum')) is True
assert os.path.exists(
os.path.join(dataset_submodule_dir, 'test_file.dat')) is False
with open(os.path.join(lb.root_dir, '.gitmodules'), 'rt') as mf:
data = mf.read()
assert len(data) > 0
# check overview
result = fixture_working_dir[2].execute(overview_query)
assert "errors" not in result
assert result['data']['labbook']['linkedDatasets'][0]['overview'][
'localBytes'] == '5'
assert result['data']['labbook']['linkedDatasets'][0]['overview'][
'totalBytes'] == '5'
# Make change to published dataset
git_dir = os.path.join(tempfile.gettempdir(),
'test_update_dataset_link_mutation')
try:
os.makedirs(git_dir)
call_subprocess(['git', 'clone', ds.remote],
cwd=git_dir,
check=True)
with open(os.path.join(git_dir, ds.name, 'test_file.dat'),
'wt') as tf:
tf.write("Test File Contents")
call_subprocess(['git', 'add', 'test_file.dat'],
cwd=os.path.join(git_dir, ds.name),
check=True)
call_subprocess(['git', 'commit', '-m', 'editing repo'],
cwd=os.path.join(git_dir, ds.name),
check=True)
call_subprocess(['git', 'push'],
cwd=os.path.join(git_dir, ds.name),
check=True)
query = """
mutation myMutation($lo: String!, $ln: String!, $do: String!, $dn: String!,
$a: String!) {
modifyDatasetLink(input: {labbookOwner: $lo, labbookName: $ln, datasetOwner: $do, datasetName: $dn,
action: $a}) {
newLabbookEdge {
node {
id
name
description
linkedDatasets {
name
}
}
}
}
}
"""
variables = {
"lo": "default",
"ln": "test-lb",
"do": "default",
"dn": "dataset100",
"a": "update"
}
result = fixture_working_dir[2].execute(query,
variable_values=variables)
assert "errors" not in result
snapshot.assert_match(result)
# verify change is reflected
assert os.path.exists(
os.path.join(dataset_submodule_dir, 'test_file.dat')) is True
# Verify activity record
assert "Updated Dataset `default/dataset100` link to version" in lb.git.log(
)[0]['message']
finally:
if os.path.exists(git_dir):
shutil.rmtree(git_dir)
def test_push_objects(self, mock_config_file, mock_dataset_head):
im = InventoryManager(mock_config_file[0])
ds = im.create_dataset('default',
'default',
"dataset100",
storage_type="gigantum_object_v1",
description="100")
manifest = Manifest(ds, 'default')
iom = IOManager(ds, manifest)
revision = manifest.dataset_revision
os.makedirs(
os.path.join(manifest.cache_mgr.cache_root, revision, "other_dir"))
helper_append_file(manifest.cache_mgr.cache_root, revision,
"test1.txt", "test content 1")
helper_append_file(manifest.cache_mgr.cache_root, revision,
"test2.txt", "test content 2")
manifest.sweep_all_changes()
obj_to_push = iom.objects_to_push()
assert len(obj_to_push) == 2
_, obj1 = obj_to_push[0].object_path.rsplit('/', 1)
_, obj2 = obj_to_push[1].object_path.rsplit('/', 1)
with aioresponses() as mocked_responses:
mocked_responses.put(
f'https://api.gigantum.com/object-v1/{ds.namespace}/{ds.name}/{obj1}',
payload={
"presigned_url": f"https://dummyurl.com/{obj1}?params=1",
"namespace": ds.namespace,
"key_id": "hghghg",
"obj_id": obj1,
"dataset": ds.name
},
status=200)
mocked_responses.put(f"https://dummyurl.com/{obj1}?params=1",
payload={},
status=200)
mocked_responses.put(
f'https://api.gigantum.com/object-v1/{ds.namespace}/{ds.name}/{obj2}',
payload={
"presigned_url": f"https://dummyurl.com/{obj2}?params=1",
"namespace": ds.namespace,
"key_id": "hghghg",
"obj_id": obj2,
"dataset": ds.name
},
status=200)
mocked_responses.put(f"https://dummyurl.com/{obj2}?params=1",
payload={},
status=200)
job_kwargs = {
'objs': obj_to_push,
'logged_in_username': "******",
'access_token': "faketoken",
'id_token': "faketoken",
'dataset_owner': ds.namespace,
'dataset_name': ds.name,
'config_file': ds.client_config.config_file,
}
gtmcore.dispatcher.dataset_jobs.push_dataset_objects(**job_kwargs)
def test_verify_contents_linked_dataset(self, mock_dataset_with_local_dir):
class JobMock():
def __init__(self):
self.meta = dict()
def save_meta(self):
pass
CURRENT_JOB = JobMock()
def get_current_job_mock():
return CURRENT_JOB
with patch('gtmcore.dispatcher.jobs.get_current_job',
side_effect=get_current_job_mock):
ds = mock_dataset_with_local_dir[0]
im = InventoryManager()
ds.backend.update_from_remote(ds, lambda x: print(x))
m = Manifest(ds, 'tester')
assert len(m.manifest.keys()) == 4
assert os.path.isfile(
os.path.join(m.cache_mgr.cache_root, m.dataset_revision,
'test1.txt'))
assert os.path.isfile(
os.path.join(m.cache_mgr.cache_root, m.dataset_revision,
'test2.txt'))
assert os.path.isfile(
os.path.join(m.cache_mgr.cache_root, m.dataset_revision,
'subdir', 'test3.txt'))
modified_items = ds.backend.verify_contents(ds, lambda x: print(x))
assert len(modified_items) == 0
lb = im.create_labbook("tester", "tester", 'test-labbook')
im.link_dataset_to_labbook(f"{ds.root_dir}/.git", "tester",
ds.name, lb)
dataset_dir = os.path.join(lb.root_dir, '.gigantum', 'datasets',
'tester', ds.name)
ds = im.load_dataset_from_directory(dataset_dir)
test_dir = os.path.join(mock_dataset_with_local_dir[1],
"local_data", "test_dir")
with open(os.path.join(test_dir, 'test1.txt'), 'wt') as tf:
tf.write("This file got changed in the filesystem")
kwargs = {
'logged_in_username': "******",
'access_token': "asdf",
'id_token': "1234",
'dataset_owner': "tester",
'dataset_name': 'dataset-1',
'labbook_owner': "tester",
'labbook_name': 'test-labbook'
}
jobs.verify_dataset_contents(**kwargs)
job = gtmcore.dispatcher.jobs.get_current_job()
assert 'modified_keys' in job.meta
assert job.meta['modified_keys'] == ["test1.txt"]
assert 'Validating contents of 3 files.' in job.meta['feedback']
def test_pull_objects(self, mock_config_file, mock_dataset_head):
im = InventoryManager(mock_config_file[0])
ds = im.create_dataset('default',
'default',
"dataset100",
storage_type="gigantum_object_v1",
description="100")
m = Manifest(ds, 'default')
iom = IOManager(ds, m)
os.makedirs(
os.path.join(m.cache_mgr.cache_root, m.dataset_revision,
"other_dir"))
helper_append_file(m.cache_mgr.cache_root, m.dataset_revision,
"test1.txt", "asdfadfsdf")
helper_append_file(m.cache_mgr.cache_root, m.dataset_revision,
"test2.txt", "fdsfgfd")
m.sweep_all_changes()
obj_to_push = iom.objects_to_push()
assert len(obj_to_push) == 2
_, obj_id_1 = obj_to_push[0].object_path.rsplit('/', 1)
_, obj_id_2 = obj_to_push[1].object_path.rsplit('/', 1)
obj1_target = obj_to_push[0].object_path
obj2_target = obj_to_push[1].object_path
obj1_source = os.path.join('/tmp', uuid.uuid4().hex)
obj2_source = os.path.join('/tmp', uuid.uuid4().hex)
assert os.path.exists(obj1_target) is True
assert os.path.exists(obj2_target) is True
helper_compress_file(obj1_target, obj1_source)
helper_compress_file(obj2_target, obj2_source)
assert os.path.isfile(obj1_target) is False
assert os.path.isfile(obj2_target) is False
assert os.path.isfile(obj1_source) is True
assert os.path.isfile(obj2_source) is True
# Clear out from linked dir
os.remove(
os.path.join(m.cache_mgr.cache_root, m.dataset_revision,
'test1.txt'))
os.remove(
os.path.join(m.cache_mgr.cache_root, m.dataset_revision,
'test2.txt'))
with patch.object(Configuration, 'find_default_config',
lambda self: mock_config_file[0]):
with aioresponses() as mocked_responses:
mocked_responses.get(
f'https://api.gigantum.com/object-v1/{ds.namespace}/{ds.name}/{obj_id_1}',
payload={
"presigned_url":
f"https://dummyurl.com/{obj_id_1}?params=1",
"namespace": ds.namespace,
"obj_id": obj_id_1,
"dataset": ds.name
},
status=200)
with open(obj1_source, 'rb') as data1:
mocked_responses.get(
f"https://dummyurl.com/{obj_id_1}?params=1",
body=data1.read(),
status=200,
content_type='application/octet-stream')
mocked_responses.get(
f'https://api.gigantum.com/object-v1/{ds.namespace}/{ds.name}/{obj_id_2}',
payload={
"presigned_url":
f"https://dummyurl.com/{obj_id_2}?params=1",
"namespace": ds.namespace,
"obj_id": obj_id_2,
"dataset": ds.name
},
status=200)
with open(obj2_source, 'rb') as data2:
mocked_responses.get(
f"https://dummyurl.com/{obj_id_2}?params=1",
body=data2.read(),
status=200,
content_type='application/octet-stream')
dl_kwargs = {
'logged_in_username': "******",
'access_token': "asdf",
'id_token': "1234",
'dataset_owner': "default",
'dataset_name': "dataset100",
'labbook_owner': None,
'labbook_name': None,
'keys': ["test1.txt"]
}
gtmcore.dispatcher.dataset_jobs.pull_objects(**dl_kwargs)
# Manually link since this is disabled by default in the job (because in real use, multiple jobs run
# in parallel and you only want to link once.
m.link_revision()
assert os.path.isfile(obj1_target) is True
assert os.path.isfile(obj2_target) is False
decompressor = snappy.StreamDecompressor()
with open(obj1_source, 'rb') as dd:
source1 = decompressor.decompress(dd.read())
source1 += decompressor.flush()
with open(obj1_target, 'rt') as dd:
dest1 = dd.read()
assert source1.decode("utf-8") == dest1
# Download other file
dl_kwargs = {
'logged_in_username': "******",
'access_token': "asdf",
'id_token': "1234",
'dataset_owner': "default",
'dataset_name': "dataset100",
'labbook_owner': None,
'labbook_name': None,
'keys': ["test2.txt"]
}
gtmcore.dispatcher.dataset_jobs.pull_objects(**dl_kwargs)
# Manually link since this is disabled by default in the job (because in real use, multiple jobs run
# in parallel and you only want to link once.
m.link_revision()
assert os.path.isfile(obj1_target) is True
assert os.path.isfile(obj2_target) is True
with open(obj1_source, 'rb') as dd:
source1 = decompressor.decompress(dd.read())
source1 += decompressor.flush()
with open(obj1_target, 'rt') as dd:
dest1 = dd.read()
assert source1.decode("utf-8") == dest1
with open(obj2_source, 'rb') as dd:
source1 = decompressor.decompress(dd.read())
source1 += decompressor.flush()
with open(obj2_target, 'rt') as dd:
dest1 = dd.read()
assert source1.decode("utf-8") == dest1
def mutate_and_get_payload(cls,
root,
info,
labbook_owner,
labbook_name,
dataset_owner,
dataset_name,
action,
dataset_url=None,
client_mutation_id=None):
logged_in_username = get_logged_in_username()
im = InventoryManager()
lb = im.load_labbook(logged_in_username,
labbook_owner,
labbook_name,
author=get_logged_in_author())
with lb.lock():
if action == 'link':
if dataset_url:
remote_domain = cls._get_remote_domain(
dataset_url, dataset_owner, dataset_name)
if remote_domain:
# Make sure git creds are configured for the remote
admin_service = None
for remote in lb.client_config.config['git'][
'remotes']:
if remote_domain == remote:
admin_service = lb.client_config.config['git'][
'remotes'][remote]['admin_service']
break
if "HTTP_AUTHORIZATION" in info.context.headers.environ:
token = parse_token(info.context.headers.
environ["HTTP_AUTHORIZATION"])
else:
raise ValueError(
"Authorization header not provided."
" Must have a valid session to query for collaborators"
)
mgr = GitLabManager(remote_domain, admin_service,
token)
mgr.configure_git_credentials(remote_domain,
logged_in_username)
else:
# Link to local dataset
ds = im.load_dataset(logged_in_username, dataset_owner,
dataset_name)
dataset_url = f"{ds.root_dir}/.git"
# Link the dataset to the labbook
ds = im.link_dataset_to_labbook(dataset_url, dataset_owner,
dataset_name, lb)
ds.namespace = dataset_owner
# Preload the dataloader
info.context.dataset_loader.prime(
f"{get_logged_in_username()}&{dataset_owner}&{dataset_name}",
ds)
# Relink the revision
m = Manifest(ds, logged_in_username)
m.link_revision()
elif action == 'unlink':
im.unlink_dataset_from_labbook(dataset_owner, dataset_name, lb)
elif action == 'update':
ds = im.update_linked_dataset_reference(
dataset_owner, dataset_name, lb)
m = Manifest(ds, logged_in_username)
m.force_reload()
info.context.dataset_loader.prime(
f"{get_logged_in_username()}&{dataset_owner}&{dataset_name}",
ds)
else:
raise ValueError(
"Unsupported action. Use `link`, `unlink`, or `update`")
info.context.labbook_loader.prime(
f"{get_logged_in_username()}&{labbook_owner}&{labbook_name}",
lb)
edge = LabbookConnection.Edge(node=Labbook(owner=labbook_owner,
name=labbook_name),
cursor=base64.b64encode(
f"{0}".encode('utf-8')))
return ModifyDatasetLink(new_labbook_edge=edge)
def create_dataset(self,
username: str,
owner: str,
dataset_name: str,
storage_type: str,
description: Optional[str] = None,
author: Optional[GitAuthor] = None) -> Dataset:
"""Create a new Dataset in this Gigantum working directory.
Args:
username: Active username
owner: Namespace in which to place this Dataset
dataset_name: Name of the Dataset
storage_type: String identifying the type of Dataset to instantiate
description: Optional brief description of Dataset
author: Optional Git Author
Returns:
Newly created LabBook instance
"""
dataset = Dataset(config_file=self.config_file,
author=author,
namespace=owner)
if storage_type not in SUPPORTED_STORAGE_BACKENDS:
raise ValueError(
f"Unsupported Dataset storage type: {storage_type}")
try:
build_info = Configuration(self.config_file).config['build_info']
except KeyError:
logger.warning("Could not obtain build_info from config")
build_info = None
# Build data file contents
dataset._data = {
"schema": DATASET_CURRENT_SCHEMA,
"id": uuid.uuid4().hex,
"name": dataset_name,
"storage_type": storage_type,
"description": description or '',
"created_on": datetime.datetime.utcnow().isoformat(),
"build_info": build_info
}
dataset._validate_gigantum_data()
logger.info("Creating new Dataset on disk for {}/{}/{}".format(
username, owner, dataset_name))
# lock while creating initial directory
with dataset.lock(
lock_key=f"new_dataset_lock|{username}|{owner}|{dataset_name}"
):
# Verify or Create user subdirectory
# Make sure you expand a user dir string
starting_dir = os.path.expanduser(
dataset.client_config.config["git"]["working_directory"])
user_dir = os.path.join(starting_dir, username)
if not os.path.isdir(user_dir):
os.makedirs(user_dir)
# Create owner dir - store LabBooks in working dir > logged in user > owner
owner_dir = os.path.join(user_dir, owner)
if not os.path.isdir(owner_dir):
os.makedirs(owner_dir)
# Create `datasets` subdir in the owner dir
owner_dir = os.path.join(owner_dir, "datasets")
else:
owner_dir = os.path.join(owner_dir, "datasets")
# Verify name not already in use
if os.path.isdir(os.path.join(owner_dir, dataset_name)):
raise ValueError(
f"Dataset `{dataset_name}` already exists locally. Choose a new Dataset name"
)
# Create Dataset subdirectory
new_root_dir = os.path.join(owner_dir, dataset_name)
os.makedirs(new_root_dir)
dataset._set_root_dir(new_root_dir)
# Init repository
dataset.git.initialize()
# Create Directory Structure
dirs = [
'manifest', 'metadata', '.gigantum',
os.path.join('.gigantum', 'favorites'),
os.path.join('.gigantum', 'activity'),
os.path.join('.gigantum', 'activity', 'log')
]
for d in dirs:
p = os.path.join(dataset.root_dir, d, '.gitkeep')
os.makedirs(os.path.dirname(p), exist_ok=True)
with open(p, 'w') as gk:
gk.write(
"This file is necessary to keep this directory tracked by Git"
" and archivable by compression tools. Do not delete or modify!"
)
dataset._save_gigantum_data()
# Create an empty storage.json file
dataset.backend_config = {}
# Create .gitignore default file
shutil.copyfile(
os.path.join(resource_filename('gtmcore', 'dataset'),
'gitignore.default'),
os.path.join(dataset.root_dir, ".gitignore"))
# Commit
dataset.git.add_all()
# NOTE: this string is used to indicate there are no more activity records to get. Changing the string will
# break activity paging.
# TODO: Improve method for detecting the first activity record
dataset.git.commit(f"Creating new empty Dataset: {dataset_name}")
# Create Activity Record
adr = ActivityDetailRecord(ActivityDetailType.DATASET,
show=False,
importance=0)
adr.add_value('text/plain',
f"Created new Dataset: {username}/{dataset_name}")
ar = ActivityRecord(
ActivityType.DATASET,
message=f"Created new Dataset: {username}/{dataset_name}",
show=True,
importance=255,
linked_commit=dataset.git.commit_hash)
ar.add_detail_object(adr)
store = ActivityStore(dataset)
store.create_activity_record(ar)
# Initialize file cache and link revision
m = Manifest(dataset, username)
m.link_revision()
return dataset
def test_complete_dataset_upload_transaction_failure(
self, mock_config_file_background_tests):
im = InventoryManager(mock_config_file_background_tests[0])
ds = im.create_dataset('default',
'default',
"new-ds",
storage_type="gigantum_object_v1",
description="100")
m = Manifest(ds, 'default')
dispatcher_obj = Dispatcher()
helper_write_big_file(m.cache_mgr.cache_root, m.dataset_revision,
"test1.dat", "12")
helper_write_big_file(m.cache_mgr.cache_root, m.dataset_revision,
"test2.dat", "23")
helper_append_file(m.cache_mgr.cache_root, m.dataset_revision,
"zztest3.txt", "fake content 3")
helper_append_file(m.cache_mgr.cache_root, m.dataset_revision,
"zztest4.txt", "fake content 4")
helper_append_file(m.cache_mgr.cache_root, m.dataset_revision,
"zztest5.txt", "fake content 5")
helper_append_file(m.cache_mgr.cache_root, m.dataset_revision,
"zztest6.txt", "fake content 6")
job_kwargs = {
'dispatcher': Dispatcher,
'logged_in_username': "******",
'logged_in_email': "*****@*****.**",
'dataset_owner': "default",
'dataset_name': "new-ds",
'config_file': mock_config_file_background_tests[0]
}
job_metadata = {
'dataset': f"default|default|new-ds",
'method': 'complete_dataset_upload_transaction'
}
assert len(m.manifest) == 0
job_key = dispatcher_obj.dispatch_task(
gtmcore.dispatcher.dataset_jobs.
complete_dataset_upload_transaction,
kwargs=job_kwargs,
metadata=job_metadata)
time.sleep(3)
# Remove files to make them fail
os.remove(
os.path.join(m.cache_mgr.cache_root, m.dataset_revision,
"zztest4.txt"))
os.remove(
os.path.join(m.cache_mgr.cache_root, m.dataset_revision,
"zztest5.txt"))
cnt = 0
while cnt < 120:
job_status = dispatcher_obj.query_task(job_key)
if job_status.status == 'finished':
break
time.sleep(1)
cnt += 1
assert cnt < 119
m = Manifest(ds, 'default')
assert len(m.manifest) == 4
assert 'test1.dat' in m.manifest
assert 'test2.dat' in m.manifest
assert 'zztest3.txt' in m.manifest
assert 'zztest6.txt' in m.manifest
assert 'zztest5.txt' not in m.manifest
assert 'zztest4.txt' not in m.manifest
assert job_status.meta['has_failures'] is True
assert 'The following files failed to hash. Try re-uploading the files again:\nzztest4.txt \nzztest5.txt' ==\
job_status.meta['failure_detail']
assert 'An error occurred while processing some files. Check details and re-upload.' == \
job_status.meta['feedback']
def test_complete_dataset_upload_transaction_all_types(
self, mock_config_file_background_tests):
im = InventoryManager(mock_config_file_background_tests[0])
ds = im.create_dataset('default',
'default',
"new-ds",
storage_type="gigantum_object_v1",
description="100")
m = Manifest(ds, 'default')
helper_append_file(m.cache_mgr.cache_root, m.dataset_revision,
"test1.txt", "fake content 1")
helper_append_file(m.cache_mgr.cache_root, m.dataset_revision,
"test2.txt", "fake content 2")
helper_append_file(m.cache_mgr.cache_root, m.dataset_revision,
"test3.txt", "fake content 3")
dl_kwargs = {
'dispatcher': Dispatcher,
'logged_in_username': "******",
'logged_in_email': "*****@*****.**",
'dataset_owner': "default",
'dataset_name': "new-ds",
'config_file': mock_config_file_background_tests[0]
}
assert len(m.manifest) == 0
gtmcore.dispatcher.dataset_jobs.complete_dataset_upload_transaction(
**dl_kwargs)
m = Manifest(ds, 'default')
# make sure manifest got updated
assert len(m.manifest) == 3
assert 'test1.txt' in m.manifest
assert 'test2.txt' in m.manifest
assert 'test3.txt' in m.manifest
hash1 = m.manifest['test1.txt']['h']
# Make sure activity created
assert len(ds.git.log()) == 6
assert "_GTM_ACTIVITY_START_" in ds.git.log()[0]['message']
assert "Uploaded 3 new file(s)." in ds.git.log()[0]['message']
helper_append_file(m.cache_mgr.cache_root, m.dataset_revision,
"test1.txt", "fake content changed")
helper_append_file(m.cache_mgr.cache_root, m.dataset_revision,
"test4.txt", "fake content 4")
os.remove(
os.path.join(m.cache_mgr.cache_root, m.dataset_revision,
"test3.txt"))
gtmcore.dispatcher.dataset_jobs.complete_dataset_upload_transaction(
**dl_kwargs)
m = Manifest(ds, 'default')
# make sure manifest got updated
assert len(m.manifest) == 3
assert 'test1.txt' in m.manifest
assert 'test2.txt' in m.manifest
assert 'test4.txt' in m.manifest
assert hash1 != m.manifest['test1.txt']['h']
# Make sure activity created
assert len(ds.git.log()) == 8
assert "_GTM_ACTIVITY_START_" in ds.git.log()[0]['message']
assert "Uploaded 1 new file(s). Uploaded 1 modified file(s). 1 deleted file(s)." in ds.git.log(
)[0]['message']
def test_download_dataset_files(self, mock_config_file_background_tests,
mock_dataset_head):
def dispatch_query_mock(self, job_key):
JobStatus = namedtuple("JobStatus", ['status', 'meta'])
return JobStatus(status='finished',
meta={'completed_bytes': '500'})
def dispatch_mock(self, method_reference, kwargs, metadata, persist):
with aioresponses() as mocked_responses:
mocked_responses.get(
f'https://api.gigantum.com/object-v1/{ds.namespace}/{ds.name}/{obj_id_1}',
payload={
"presigned_url":
f"https://dummyurl.com/{obj_id_1}?params=1",
"namespace": ds.namespace,
"obj_id": obj_id_1,
"dataset": ds.name
},
status=200)
with open(obj1_source, 'rb') as data1:
mocked_responses.get(
f"https://dummyurl.com/{obj_id_1}?params=1",
body=data1.read(),
status=200,
content_type='application/octet-stream')
gtmcore.dispatcher.dataset_jobs.pull_objects(**kwargs)
return "afakejobkey"
im = InventoryManager(mock_config_file_background_tests[0])
ds = im.create_dataset('default',
'default',
"dataset100",
storage_type="gigantum_object_v1",
description="100")
m = Manifest(ds, 'default')
iom = IOManager(ds, m)
helper_append_file(m.cache_mgr.cache_root, m.dataset_revision,
"test1.txt", "asdfadfsdf")
m.sweep_all_changes()
obj_to_push = iom.objects_to_push()
assert len(obj_to_push) == 1
_, obj_id_1 = obj_to_push[0].object_path.rsplit('/', 1)
obj1_target = obj_to_push[0].object_path
obj1_source = os.path.join('/tmp', uuid.uuid4().hex)
assert os.path.exists(obj1_target) is True
helper_compress_file(obj1_target, obj1_source)
assert os.path.isfile(obj1_target) is False
assert os.path.isfile(obj1_source) is True
# Clear out from linked dir
os.remove(
os.path.join(m.cache_mgr.cache_root, m.dataset_revision,
'test1.txt'))
with patch.object(Configuration, 'find_default_config',
lambda self: mock_config_file_background_tests[0]):
with patch.object(Dispatcher, 'dispatch_task', dispatch_mock):
with patch.object(Dispatcher, 'query_task',
dispatch_query_mock):
dl_kwargs = {
'logged_in_username': "******",
'access_token': "asdf",
'id_token': "1234",
'dataset_owner': "default",
'dataset_name': "dataset100",
'labbook_owner': None,
'labbook_name': None,
'keys': ["test1.txt"],
'config_file': mock_config_file_background_tests[0]
}
gtmcore.dispatcher.dataset_jobs.download_dataset_files(
**dl_kwargs)
assert os.path.isfile(obj1_target) is True
decompressor = snappy.StreamDecompressor()
with open(obj1_source, 'rb') as dd:
source1 = decompressor.decompress(dd.read())
source1 += decompressor.flush()
with open(obj1_target, 'rt') as dd:
dest1 = dd.read()
assert source1.decode("utf-8") == dest1
def test_download_dataset_files_file_fail(
self, mock_config_file_background_tests):
def dispatch_query_mock(self, job_key):
# mock the job actually running and returning status
JobStatus = namedtuple("JobStatus", ['status', 'meta'])
return JobStatus(status='finished',
meta={
'completed_bytes': '0',
'failure_keys': 'test1.txt'
})
def dispatch_mock(self, method_reference, kwargs, metadata, persist):
gtmcore.dispatcher.dataset_jobs.pull_objects(**kwargs)
return "afakejobkey"
im = InventoryManager(mock_config_file_background_tests[0])
ds = im.create_dataset('default',
'default',
"dataset100",
storage_type="gigantum_object_v1",
description="100")
m = Manifest(ds, 'default')
iom = IOManager(ds, m)
helper_append_file(m.cache_mgr.cache_root, m.dataset_revision,
"test1.txt", "asdfadfsdf")
m.sweep_all_changes()
obj_to_push = iom.objects_to_push()
assert len(obj_to_push) == 1
_, obj_id_1 = obj_to_push[0].object_path.rsplit('/', 1)
obj1_target = obj_to_push[0].object_path
obj1_source = os.path.join('/tmp', uuid.uuid4().hex)
assert os.path.exists(obj1_target) is True
helper_compress_file(obj1_target, obj1_source)
assert os.path.isfile(obj1_target) is False
assert os.path.isfile(obj1_source) is True
# Clear out from linked dir
os.remove(
os.path.join(m.cache_mgr.cache_root, m.dataset_revision,
'test1.txt'))
with patch.object(Configuration, 'find_default_config',
lambda self: mock_config_file_background_tests[0]):
with patch.object(Dispatcher, 'dispatch_task', dispatch_mock):
with patch.object(Dispatcher, 'query_task',
dispatch_query_mock):
dl_kwargs = {
'logged_in_username': "******",
'access_token': "asdf",
'id_token': "1234",
'dataset_owner': "default",
'dataset_name': "dataset100",
'labbook_owner': None,
'labbook_name': None,
'keys': ["test1.txt"],
'config_file': mock_config_file_background_tests[0]
}
with pytest.raises(IOError):
gtmcore.dispatcher.dataset_jobs.download_dataset_files(
**dl_kwargs)
assert os.path.isfile(obj1_target) is False
def test_add_file(self, mock_create_dataset):
"""Test adding a new file to a labbook"""
class DummyContext(object):
def __init__(self, file_handle):
self.dataset_loader = None
self.files = {'uploadChunk': file_handle}
client = Client(mock_create_dataset[3],
middleware=[DataloaderMiddleware()])
# Create file to upload
test_file = os.path.join(tempfile.gettempdir(), "myValidFile.dat")
est_size = 9000000
try:
os.remove(test_file)
except:
pass
with open(test_file, 'wb') as tf:
tf.write(os.urandom(est_size))
new_file_size = os.path.getsize(tf.name)
# Get upload params
chunk_size = 4194000
file_info = os.stat(test_file)
file_size = int(file_info.st_size / 1000)
total_chunks = int(math.ceil(file_info.st_size / chunk_size))
ds = InventoryManager(mock_create_dataset[0]).load_dataset(
'default', 'default', 'dataset1')
fsc = HostFilesystemCache(ds, 'default')
target_file = os.path.join(fsc.current_revision_dir, "myValidFile.dat")
txid = "000-unitest-transaction"
with open(test_file, 'rb') as tf:
# Check for file to exist (shouldn't yet)
assert os.path.exists(target_file) is False
for chunk_index in range(total_chunks):
# Upload a chunk
chunk = io.BytesIO()
chunk.write(tf.read(chunk_size))
chunk.seek(0)
file = FileStorage(chunk)
query = f"""
mutation addDatasetFile{{
addDatasetFile(input:{{owner:"default",
datasetName: "dataset1",
filePath: "myValidFile.dat",
transactionId: "{txid}",
chunkUploadParams:{{
uploadId: "fdsfdsfdsfdfs",
chunkSize: {chunk_size},
totalChunks: {total_chunks},
chunkIndex: {chunk_index},
fileSizeKb: {file_size},
filename: "{os.path.basename(test_file)}"
}}
}}) {{
newDatasetFileEdge {{
node{{
id
key
isDir
size
}}
}}
}}
}}
"""
r = client.execute(query, context_value=DummyContext(file))
assert 'errors' not in r
# So, these will only be populated once the last chunk is uploaded. Will be None otherwise.
assert r['data']['addDatasetFile']['newDatasetFileEdge']['node'][
'isDir'] is False
assert r['data']['addDatasetFile']['newDatasetFileEdge']['node'][
'key'] == 'myValidFile.dat'
assert r['data']['addDatasetFile']['newDatasetFileEdge']['node'][
'size'] == f"{new_file_size}"
# When done uploading, file should exist in the labbook
assert os.path.exists(target_file)
assert os.path.isfile(target_file)
complete_query = f"""
mutation completeQuery {{
completeDatasetUploadTransaction(input: {{
owner: "default",
datasetName: "dataset1",
transactionId: "{txid}"
}}) {{
success
}}
}}
"""
r = client.execute(complete_query, context_value=DummyContext(file))
assert 'errors' not in r
m = Manifest(ds, 'default')
status = m.status()
assert len(status.created) == 0
assert len(status.modified) == 0
assert len(status.deleted) == 0
assert 'Uploaded 1 new file(s)' in ds.git.log()[0]['message']
def test_make_directory(self, fixture_working_dir, snapshot):
im = InventoryManager(fixture_working_dir[0])
ds = im.create_dataset('default',
'default',
"dataset-dir",
storage_type="gigantum_object_v1",
description="testing move")
m = Manifest(ds, 'default')
m.link_revision()
query = """
mutation myMutation {
makeDatasetDirectory(input: {datasetOwner: "default", datasetName: "dataset-dir",
key: "test_dir1/"}) {
newDatasetFileEdge {
node {
id
key
isDir
isLocal
size
}
}
}
}
"""
result = fixture_working_dir[2].execute(query)
assert 'errors' not in result
assert result['data']['makeDatasetDirectory']['newDatasetFileEdge'][
'node']['key'] == 'test_dir1/'
assert result['data']['makeDatasetDirectory']['newDatasetFileEdge'][
'node']['isDir'] is True
assert result['data']['makeDatasetDirectory']['newDatasetFileEdge'][
'node']['isLocal'] is True
assert result['data']['makeDatasetDirectory']['newDatasetFileEdge'][
'node']['size'] == '0'
assert os.path.isdir(
os.path.join(m.cache_mgr.current_revision_dir,
"test_dir1")) is True
query = """
mutation myMutation {
makeDatasetDirectory(input: {datasetOwner: "default", datasetName: "dataset-dir",
key: "test_dir1/test_dir2/"}) {
newDatasetFileEdge {
node {
id
key
isDir
isLocal
size
}
}
}
}
"""
result = fixture_working_dir[2].execute(query)
assert 'errors' not in result
assert result['data']['makeDatasetDirectory']['newDatasetFileEdge'][
'node']['key'] == 'test_dir1/test_dir2/'
assert result['data']['makeDatasetDirectory']['newDatasetFileEdge'][
'node']['isDir'] is True
assert result['data']['makeDatasetDirectory']['newDatasetFileEdge'][
'node']['isLocal'] is True
assert result['data']['makeDatasetDirectory']['newDatasetFileEdge'][
'node']['size'] == '0'
assert os.path.isdir(
os.path.join(m.cache_mgr.current_revision_dir,
"test_dir1")) is True
assert os.path.isdir(
os.path.join(m.cache_mgr.current_revision_dir, "test_dir1",
"test_dir2")) is True
def _push_dataset_objects(self, logged_in_username: str,
feedback_callback: Callable, access_token,
id_token) -> None:
"""Method to schedule a push operta
Args:
logged_in_username:
feedback_callback:
access_token:
id_token:
Returns:
"""
dispatcher_obj = Dispatcher()
try:
self.dataset.backend.set_default_configuration(
logged_in_username, access_token, id_token)
m = Manifest(self.dataset, logged_in_username)
iom = IOManager(self.dataset, m)
obj_batches, total_bytes, num_files = iom.compute_push_batches()
if obj_batches:
# Schedule jobs for batches
bg_jobs = list()
for objs in obj_batches:
job_kwargs = {
'objs': objs,
'logged_in_username': logged_in_username,
'access_token': access_token,
'id_token': id_token,
'dataset_owner': self.dataset.namespace,
'dataset_name': self.dataset.name,
'config_file': self.dataset.client_config.config_file,
}
job_metadata = {
'dataset':
f"{logged_in_username}|{self.dataset.namespace}|{self.dataset.name}",
'method': 'pull_objects'
}
feedback_callback(
f"Preparing to upload {num_files} files. Please wait..."
)
job_key = dispatcher_obj.dispatch_task(
method_reference=gtmcore.dispatcher.dataset_jobs.
push_dataset_objects,
kwargs=job_kwargs,
metadata=job_metadata,
persist=True)
bg_jobs.append(
BackgroundUploadJob(dispatcher_obj, objs, job_key))
logger.info(
f"Schedule dataset object upload job for"
f" {logged_in_username}/{self.dataset.namespace}/{self.dataset.name} with"
f" {len(objs)} objects to upload")
while sum([(x.is_complete or x.is_failed)
for x in bg_jobs]) != len(bg_jobs):
# Refresh all job statuses and update status feedback
[j.refresh_status() for j in bg_jobs]
total_completed_bytes = sum(
[j.completed_bytes for j in bg_jobs])
if total_completed_bytes > 0:
pc = (float(total_completed_bytes) /
float(total_bytes)) * 100
feedback_callback(
f"Please wait - Uploading {num_files} files ({format_size(total_completed_bytes)}"
f" of {format_size(total_bytes)}) - {round(pc)}% complete",
percent_complete=pc)
time.sleep(1)
# if you get here, all jobs are done or failed.
# Remove all the push files so they can be regenerated if needed
for f in glob.glob(f'{iom.push_dir}/*'):
os.remove(f)
# Aggregate failures if they exist
failure_keys: List[str] = list()
for j in bg_jobs:
if j.is_failed:
# Background job hard failed. Assume entire batch should get re-uploaded
for obj in j.objs:
failure_keys.append(
f"{obj.dataset_path} at {obj.revision[0:8]}")
m.queue_to_push(obj.object_path, obj.dataset_path,
obj.revision)
else:
for obj in j.get_failed_objects():
# Some individual objects failed
failure_keys.append(
f"{obj.dataset_path} at {obj.revision[0:8]}")
m.queue_to_push(obj.object_path, obj.dataset_path,
obj.revision)
# Set final status for UI
if len(failure_keys) == 0:
feedback_callback(f"Upload complete!",
percent_complete=100,
has_failures=False)
else:
failure_str = "\n".join(failure_keys)
failure_detail_str = f"Files that failed to upload:\n{failure_str}"
feedback_callback("",
percent_complete=100,
has_failures=True,
failure_detail=failure_detail_str)
# Finish up by linking everything just in case
iom.manifest.link_revision()
if len(failure_keys) > 0:
# If any downloads failed, exit non-zero to the UI knows there was an error
raise IOError(
f"{len(failure_keys)} file(s) failed to upload. Check message detail for more information"
" and try to sync again.")
except Exception as err:
logger.exception(err)
raise
