class ImportService(object):
"""
Service for importing TVB entities into system.
It supports TVB exported H5 files as input, but it should also handle H5 files
generated outside of TVB, as long as they respect the same structure.
"""
def __init__(self):
self.logger = get_logger(__name__)
self.user_id = None
self.files_helper = FilesHelper()
self.created_projects = []
def _download_and_unpack_project_zip(self, uploaded, uq_file_name,
temp_folder):
if isinstance(uploaded, FieldStorage) or isinstance(uploaded, Part):
if not uploaded.file:
raise ImportException(
"Please select the archive which contains the project structure."
)
with open(uq_file_name, 'wb') as file_obj:
self.files_helper.copy_file(uploaded.file, file_obj)
else:
shutil.copy2(uploaded, uq_file_name)
try:
self.files_helper.unpack_zip(uq_file_name, temp_folder)
except FileStructureException as excep:
self.logger.exception(excep)
raise ImportException(
"Bad ZIP archive provided. A TVB exported project is expected!"
)
@staticmethod
def _compute_unpack_path():
"""
:return: the name of the folder where to expand uploaded zip
"""
now = datetime.now()
date_str = "%d-%d-%d_%d-%d-%d_%d" % (now.year, now.month, now.day,
now.hour, now.minute, now.second,
now.microsecond)
uq_name = "%s-ImportProject" % date_str
return os.path.join(TvbProfile.current.TVB_TEMP_FOLDER, uq_name)
@transactional
def import_project_structure(self, uploaded, user_id):
"""
Execute import operations:
1. check if ZIP or folder
2. find all project nodes
3. for each project node:
- create project
- create all operations and groups
- import all images
- create all dataTypes
"""
self.user_id = user_id
self.created_projects = []
# Now compute the name of the folder where to explode uploaded ZIP file
temp_folder = self._compute_unpack_path()
uq_file_name = temp_folder + ".zip"
try:
self._download_and_unpack_project_zip(uploaded, uq_file_name,
temp_folder)
self._import_projects_from_folder(temp_folder)
except Exception as excep:
self.logger.exception(
"Error encountered during import. Deleting projects created during this operation."
)
# Remove project folders created so far.
# Note that using the project service to remove the projects will not work,
# because we do not have support for nested transaction.
# Removing from DB is not necessary because in transactional env a simple exception throw
# will erase everything to be inserted.
for project in self.created_projects:
project_path = os.path.join(TvbProfile.current.TVB_STORAGE,
FilesHelper.PROJECTS_FOLDER,
project.name)
shutil.rmtree(project_path)
raise ImportException(str(excep))
finally:
# Now delete uploaded file
if os.path.exists(uq_file_name):
os.remove(uq_file_name)
# Now delete temporary folder where uploaded ZIP was exploded.
if os.path.exists(temp_folder):
shutil.rmtree(temp_folder)
def _import_projects_from_folder(self, temp_folder):
"""
Process each project from the uploaded pack, to extract names.
"""
project_roots = []
for root, _, files in os.walk(temp_folder):
if FilesHelper.TVB_PROJECT_FILE in files:
project_roots.append(root)
for temp_project_path in project_roots:
update_manager = ProjectUpdateManager(temp_project_path)
update_manager.run_all_updates()
project = self.__populate_project(temp_project_path)
# Populate the internal list of create projects so far, for cleaning up folders, in case of failure
self.created_projects.append(project)
# Ensure project final folder exists on disk
project_path = self.files_helper.get_project_folder(project)
shutil.move(
os.path.join(temp_project_path, FilesHelper.TVB_PROJECT_FILE),
project_path)
# Now import project operations with their results
self.import_project_operations(project, temp_project_path)
# Import images and move them from temp into target
self._store_imported_images(project, temp_project_path,
project.name)
def _load_datatypes_from_operation_folder(self, src_op_path,
operation_entity,
datatype_group):
"""
Loads datatypes from operation folder
:returns: Datatype entities as dict {original_path: Dt instance}
"""
all_datatypes = {}
for file_name in os.listdir(src_op_path):
if file_name.endswith(FilesHelper.TVB_STORAGE_FILE_EXTENSION):
h5_file = os.path.join(src_op_path, file_name)
try:
file_update_manager = FilesUpdateManager()
file_update_manager.upgrade_file(h5_file)
datatype = self.load_datatype_from_file(
h5_file, operation_entity.id, datatype_group,
operation_entity.fk_launched_in)
all_datatypes[h5_file] = datatype
except IncompatibleFileManagerException:
os.remove(h5_file)
self.logger.warning(
"Incompatible H5 file will be ignored: %s" % h5_file)
self.logger.exception("Incompatibility details ...")
return all_datatypes
def _store_imported_datatypes_in_db(self, project, all_datatypes):
# type: (Project, dict) -> int
sorted_dts = sorted(
all_datatypes.items(),
key=lambda dt_item: dt_item[1].create_date or datetime.now())
count = 0
for dt_path, datatype in sorted_dts:
datatype_already_in_tvb = dao.get_datatype_by_gid(datatype.gid)
if not datatype_already_in_tvb:
self.store_datatype(datatype, dt_path)
count += 1
else:
AlgorithmService.create_link([datatype_already_in_tvb.id],
project.id)
file_path = h5.h5_file_for_index(datatype).path
h5_class = H5File.h5_class_from_file(file_path)
reference_list = h5_class(file_path).gather_references()
for _, reference_gid in reference_list:
if not reference_gid:
continue
ref_index = dao.get_datatype_by_gid(reference_gid.hex)
if ref_index is None:
os.remove(file_path)
dao.remove_entity(datatype.__class__, datatype.id)
raise MissingReferenceException(
'Imported file depends on datatypes that do not exist. Please upload '
'those first!')
return count
def _store_imported_images(self, project, temp_project_path, project_name):
"""
Import all images from project
"""
images_root = os.path.join(temp_project_path,
FilesHelper.IMAGES_FOLDER)
target_images_path = self.files_helper.get_images_folder(project_name)
for root, _, files in os.walk(images_root):
for metadata_file in files:
if metadata_file.endswith(FilesHelper.TVB_FILE_EXTENSION):
self._import_image(root, metadata_file, project.id,
target_images_path)
def _retrieve_operations_in_order(self, project, import_path):
# type: (Project, str) -> list[Operation2ImportData]
retrieved_operations = []
for root, _, files in os.walk(import_path):
if OPERATION_XML in files:
# Previous Operation format for uploading previous versions of projects
operation_file_path = os.path.join(root, OPERATION_XML)
operation, operation_xml_parameters = self.__build_operation_from_file(
project, operation_file_path)
operation.import_file = operation_file_path
self.logger.debug("Found operation in old XML format: " +
str(operation))
retrieved_operations.append(
Operation2ImportData(
operation,
root,
info_from_xml=operation_xml_parameters))
else:
# We strive for the new format with ViewModelH5
main_view_model = None
dt_paths = []
all_view_model_files = []
for file in files:
if file.endswith(FilesHelper.TVB_STORAGE_FILE_EXTENSION):
h5_file = os.path.join(root, file)
try:
h5_class = H5File.h5_class_from_file(h5_file)
if h5_class is ViewModelH5:
all_view_model_files.append(h5_file)
if not main_view_model:
view_model = h5.load_view_model_from_file(
h5_file)
if type(view_model
) in VIEW_MODEL2ADAPTER.keys():
main_view_model = view_model
else:
file_update_manager = FilesUpdateManager()
file_update_manager.upgrade_file(h5_file)
dt_paths.append(h5_file)
except Exception:
self.logger.warning(
"Unreadable H5 file will be ignored: %s" %
h5_file)
if main_view_model is not None:
alg = VIEW_MODEL2ADAPTER[type(main_view_model)]
operation = Operation(main_view_model.gid.hex,
project.fk_admin,
project.id,
alg.id,
status=STATUS_FINISHED,
user_group=main_view_model.
generic_attributes.operation_tag,
start_date=datetime.now(),
completion_date=datetime.now())
operation.create_date = main_view_model.create_date
self.logger.debug(
"Found main ViewModel to create operation for it: " +
str(operation))
retrieved_operations.append(
Operation2ImportData(operation, root, main_view_model,
dt_paths, all_view_model_files))
elif len(dt_paths) > 0:
alg = dao.get_algorithm_by_module(TVB_IMPORTER_MODULE,
TVB_IMPORTER_CLASS)
default_adapter = ABCAdapter.build_adapter(alg)
view_model = default_adapter.get_view_model_class()()
view_model.data_file = dt_paths[0]
vm_path = h5.store_view_model(view_model, root)
all_view_model_files.append(vm_path)
operation = Operation(view_model.gid.hex,
project.fk_admin,
project.id,
alg.id,
status=STATUS_FINISHED,
start_date=datetime.now(),
completion_date=datetime.now())
self.logger.debug(
"Found no ViewModel in folder, so we default to " +
str(operation))
retrieved_operations.append(
Operation2ImportData(operation, root, view_model,
dt_paths, all_view_model_files,
True))
return sorted(retrieved_operations,
key=lambda op_data: op_data.order_field)
def import_project_operations(self, project, import_path):
"""
This method scans provided folder and identify all operations that needs to be imported
"""
imported_operations = []
ordered_operations = self._retrieve_operations_in_order(
project, import_path)
success_no = 0
for operation_data in ordered_operations:
if operation_data.is_old_form:
operation_entity, datatype_group = self.__import_operation(
operation_data.operation)
new_op_folder = self.files_helper.get_project_folder(
project, str(operation_entity.id))
try:
operation_datatypes = self._load_datatypes_from_operation_folder(
operation_data.operation_folder, operation_entity,
datatype_group)
# Create and store view_model from operation
view_model = self._get_new_form_view_model(
operation_entity, operation_data.info_from_xml)
h5.store_view_model(view_model, new_op_folder)
operation_entity.view_model_gid = view_model.gid.hex
dao.store_entity(operation_entity)
self._store_imported_datatypes_in_db(
project, operation_datatypes)
imported_operations.append(operation_entity)
success_no = success_no + 1
except MissingReferenceException:
operation_entity.status = STATUS_ERROR
dao.store_entity(operation_entity)
elif operation_data.main_view_model is not None:
operation_entity = dao.store_entity(operation_data.operation)
dt_group = None # TODO
# Store the DataTypes in db
dts = {}
for dt_path in operation_data.dt_paths:
dt = self.load_datatype_from_file(dt_path,
operation_entity.id,
dt_group, project.id)
if isinstance(dt, BurstConfiguration):
dao.store_entity(dt)
else:
dts[dt_path] = dt
try:
stored_dts_count = self._store_imported_datatypes_in_db(
project, dts)
if stored_dts_count > 0 or not operation_data.is_self_generated:
imported_operations.append(operation_entity)
new_op_folder = self.files_helper.get_project_folder(
project, str(operation_entity.id))
for h5_file in operation_data.all_view_model_files:
shutil.move(h5_file, new_op_folder)
else:
# In case all Dts under the current operation were Links and the ViewModel is dummy,
# don't keep the Operation empty in DB
dao.remove_entity(Operation, operation_entity.id)
except MissingReferenceException:
operation_entity.status = STATUS_ERROR
dao.store_entity(operation_entity)
else:
self.logger.warning(
"Folder %s will be ignored, as we could not find a serialized "
"operation or DTs inside!" %
operation_data.operation_folder)
self.logger.warning(
"Project has been only partially imported because of some missing dependent datatypes. "
+ "%d files were successfully imported from a total of %d!" %
(success_no, len(ordered_operations)))
return imported_operations
@staticmethod
def _get_new_form_view_model(operation, xml_parameters):
# type (Operation) -> ViewModel
ad = ABCAdapter.build_adapter(operation.algorithm)
view_model = ad.get_view_model_class()()
if xml_parameters:
params = json.loads(xml_parameters)
declarative_attrs = type(view_model).declarative_attrs
for param in params:
new_param_name = param
if param[0] == "_":
new_param_name = param[1:]
new_param_name = new_param_name.lower()
if new_param_name in declarative_attrs:
setattr(view_model, new_param_name, params[param])
return view_model
def _import_image(self, src_folder, metadata_file, project_id,
target_images_path):
"""
Create and store a image entity.
"""
figure_dict = XMLReader(os.path.join(src_folder,
metadata_file)).read_metadata()
actual_figure = os.path.join(
src_folder,
os.path.split(figure_dict['file_path'])[1])
if not os.path.exists(actual_figure):
self.logger.warning("Expected to find image path %s .Skipping" %
actual_figure)
return
figure_dict['fk_user_id'] = self.user_id
figure_dict['fk_project_id'] = project_id
figure_entity = manager_of_class(ResultFigure).new_instance()
figure_entity = figure_entity.from_dict(figure_dict)
stored_entity = dao.store_entity(figure_entity)
# Update image meta-data with the new details after import
figure = dao.load_figure(stored_entity.id)
shutil.move(actual_figure, target_images_path)
self.logger.debug("Store imported figure")
self.files_helper.write_image_metadata(figure)
def load_datatype_from_file(self,
current_file,
op_id,
datatype_group=None,
current_project_id=None):
# type: (str, int, DataTypeGroup, int) -> HasTraitsIndex
"""
Creates an instance of datatype from storage / H5 file
:returns: DatatypeIndex
"""
self.logger.debug("Loading DataType from file: %s" % current_file)
h5_class = H5File.h5_class_from_file(current_file)
if h5_class is BurstConfigurationH5:
if current_project_id is None:
op_entity = dao.get_operationgroup_by_id(op_id)
current_project_id = op_entity.fk_launched_in
h5_file = BurstConfigurationH5(current_file)
burst = BurstConfiguration(current_project_id)
burst.fk_simulation = op_id
# burst.fk_operation_group = TODO
# burst.fk_metric_operation_group = TODO
h5_file.load_into(burst)
result = burst
else:
datatype, generic_attributes = h5.load_with_links(current_file)
index_class = h5.REGISTRY.get_index_for_datatype(
datatype.__class__)
datatype_index = index_class()
datatype_index.fill_from_has_traits(datatype)
datatype_index.fill_from_generic_attributes(generic_attributes)
# Add all the required attributes
if datatype_group is not None:
datatype_index.fk_datatype_group = datatype_group.id
datatype_index.fk_from_operation = op_id
associated_file = h5.path_for_stored_index(datatype_index)
if os.path.exists(associated_file):
datatype_index.disk_size = FilesHelper.compute_size_on_disk(
associated_file)
result = datatype_index
return result
def store_datatype(self, datatype, current_file=None):
"""This method stores data type into DB"""
try:
self.logger.debug("Store datatype: %s with Gid: %s" %
(datatype.__class__.__name__, datatype.gid))
# Now move storage file into correct folder if necessary
if current_file is not None:
final_path = h5.path_for_stored_index(datatype)
if final_path != current_file:
shutil.move(current_file, final_path)
return dao.store_entity(datatype)
except MissingDataSetException as e:
self.logger.exception(e)
error_msg = "Datatype %s has missing data and could not be imported properly." % (
datatype, )
raise ImportException(error_msg)
except IntegrityError as excep:
self.logger.exception(excep)
error_msg = "Could not import data with gid: %s. There is already a one with " \
"the same name or gid." % datatype.gid
raise ImportException(error_msg)
def __populate_project(self, project_path):
"""
Create and store a Project entity.
"""
self.logger.debug("Creating project from path: %s" % project_path)
project_dict = self.files_helper.read_project_metadata(project_path)
project_entity = manager_of_class(Project).new_instance()
project_entity = project_entity.from_dict(project_dict, self.user_id)
try:
self.logger.debug("Storing imported project")
return dao.store_entity(project_entity)
except IntegrityError as excep:
self.logger.exception(excep)
error_msg = (
"Could not import project: %s with gid: %s. There is already a "
"project with the same name or gid.") % (project_entity.name,
project_entity.gid)
raise ImportException(error_msg)
def __build_operation_from_file(self, project, operation_file):
"""
Create Operation entity from metadata file.
"""
operation_dict = XMLReader(operation_file).read_metadata()
operation_entity = manager_of_class(Operation).new_instance()
return operation_entity.from_dict(operation_dict, dao, self.user_id,
project.gid)
@staticmethod
def __import_operation(operation_entity):
"""
Store a Operation entity.
"""
operation_entity = dao.store_entity(operation_entity)
operation_group_id = operation_entity.fk_operation_group
datatype_group = None
if operation_group_id is not None:
try:
datatype_group = dao.get_datatypegroup_by_op_group_id(
operation_group_id)
except SQLAlchemyError:
# If no dataType group present for current op. group, create it.
operation_group = dao.get_operationgroup_by_id(
operation_group_id)
datatype_group = DataTypeGroup(
operation_group, operation_id=operation_entity.id)
datatype_group.state = UploadAlgorithmCategoryConfig.defaultdatastate
datatype_group = dao.store_entity(datatype_group)
return operation_entity, datatype_group
def import_simulator_configuration_zip(self, zip_file):
# Now compute the name of the folder where to explode uploaded ZIP file
temp_folder = self._compute_unpack_path()
uq_file_name = temp_folder + ".zip"
if isinstance(zip_file, FieldStorage) or isinstance(zip_file, Part):
if not zip_file.file:
raise ServicesBaseException(
"Could not process the given ZIP file...")
with open(uq_file_name, 'wb') as file_obj:
self.files_helper.copy_file(zip_file.file, file_obj)
else:
shutil.copy2(zip_file, uq_file_name)
try:
self.files_helper.unpack_zip(uq_file_name, temp_folder)
return temp_folder
except FileStructureException as excep:
raise ServicesBaseException(
"Could not process the given ZIP file..." + str(excep))
# # Sort all h5 files based on their creation date stored in the files themselves
# sorted_h5_files = sorted(h5_files, key=lambda h5_path: _get_create_date_for_sorting(h5_path) or datetime.now())
# return sorted_h5_files
#
#
# def _get_create_date_for_sorting(h5_file):
# storage_manager = HDF5StorageManager(os.path.dirname(h5_file), os.path.basename(h5_file))
# root_metadata = storage_manager.get_metadata()
# create_date_str = str(root_metadata['Create_date'], 'utf-8')
# create_date = datetime.strptime(create_date_str.replace('datetime:', ''), '%Y-%m-%d %H:%M:%S.%f')
# return create_date
class ImportService(object):
"""
Service for importing TVB entities into system.
It supports TVB exported H5 files as input, but it should also handle H5 files
generated outside of TVB, as long as they respect the same structure.
"""
def __init__(self):
self.logger = get_logger(__name__)
self.user_id = None
self.files_helper = FilesHelper()
self.created_projects = []
def _download_and_unpack_project_zip(self, uploaded, uq_file_name,
temp_folder):
if isinstance(uploaded, FieldStorage) or isinstance(uploaded, Part):
if not uploaded.file:
raise ProjectImportException(
"Please select the archive which contains the project structure."
)
with open(uq_file_name, 'wb') as file_obj:
self.files_helper.copy_file(uploaded.file, file_obj)
else:
shutil.copy2(uploaded, uq_file_name)
try:
self.files_helper.unpack_zip(uq_file_name, temp_folder)
except FileStructureException as excep:
self.logger.exception(excep)
raise ProjectImportException(
"Bad ZIP archive provided. A TVB exported project is expected!"
)
@staticmethod
def _compute_unpack_path():
"""
:return: the name of the folder where to expand uploaded zip
"""
now = datetime.now()
date_str = "%d-%d-%d_%d-%d-%d_%d" % (now.year, now.month, now.day,
now.hour, now.minute, now.second,
now.microsecond)
uq_name = "%s-ImportProject" % date_str
return os.path.join(TvbProfile.current.TVB_TEMP_FOLDER, uq_name)
@transactional
def import_project_structure(self, uploaded, user_id):
"""
Execute import operations:
1. check if ZIP or folder
2. find all project nodes
3. for each project node:
- create project
- create all operations
- import all images
- create all dataTypes
"""
self.user_id = user_id
self.created_projects = []
# Now compute the name of the folder where to explode uploaded ZIP file
temp_folder = self._compute_unpack_path()
uq_file_name = temp_folder + ".zip"
try:
self._download_and_unpack_project_zip(uploaded, uq_file_name,
temp_folder)
self._import_projects_from_folder(temp_folder)
except Exception as excep:
self.logger.exception(
"Error encountered during import. Deleting projects created during this operation."
)
# Remove project folders created so far.
# Note that using the project service to remove the projects will not work,
# because we do not have support for nested transaction.
# Removing from DB is not necessary because in transactional env a simple exception throw
# will erase everything to be inserted.
for project in self.created_projects:
project_path = os.path.join(TvbProfile.current.TVB_STORAGE,
FilesHelper.PROJECTS_FOLDER,
project.name)
shutil.rmtree(project_path)
raise ProjectImportException(str(excep))
finally:
# Now delete uploaded file
if os.path.exists(uq_file_name):
os.remove(uq_file_name)
# Now delete temporary folder where uploaded ZIP was exploded.
if os.path.exists(temp_folder):
shutil.rmtree(temp_folder)
@staticmethod
def _load_burst_info_from_json(project_path):
bursts_dict = {}
dt_mappings_dict = {}
bursts_file = os.path.join(project_path, BURST_INFO_FILE)
if os.path.isfile(bursts_file):
with open(bursts_file) as f:
bursts_info_dict = json.load(f)
bursts_dict = bursts_info_dict[BURSTS_DICT_KEY]
dt_mappings_dict = bursts_info_dict[DT_BURST_MAP]
os.remove(bursts_file)
return bursts_dict, dt_mappings_dict
@staticmethod
def _import_bursts(project_entity, bursts_dict):
"""
Re-create old bursts, but keep a mapping between the id it has here and the old-id it had
in the project where they were exported, so we can re-add the datatypes to them.
"""
burst_ids_mapping = {}
for old_burst_id in bursts_dict:
burst_information = BurstInformation.load_from_dict(
bursts_dict[old_burst_id])
burst_entity = model.BurstConfiguration(project_entity.id)
burst_entity.from_dict(burst_information.data)
burst_entity = dao.store_entity(burst_entity)
burst_ids_mapping[int(old_burst_id)] = burst_entity.id
# We don't need the data in dictionary form anymore, so update it with new BurstInformation object
bursts_dict[old_burst_id] = burst_information
return burst_ids_mapping
def _import_projects_from_folder(self, temp_folder):
"""
Process each project from the uploaded pack, to extract names.
"""
project_roots = []
for root, _, files in os.walk(temp_folder):
if FilesHelper.TVB_PROJECT_FILE in files:
project_roots.append(root)
for project_path in project_roots:
update_manager = ProjectUpdateManager(project_path)
update_manager.run_all_updates()
project_entity = self.__populate_project(project_path)
# Compute the path where to store files of the imported project
new_project_path = os.path.join(TvbProfile.current.TVB_STORAGE,
FilesHelper.PROJECTS_FOLDER,
project_entity.name)
if project_path != new_project_path:
shutil.move(project_path, new_project_path)
self.created_projects.append(project_entity)
# Keep a list with all burst that were imported since we will want to also add the workflow
# steps after we are finished with importing the operations and datatypes. We need to first
# stored bursts since we need to know which new id's they have for operations parent_burst.
bursts_dict, dt_mappings_dict = self._load_burst_info_from_json(
new_project_path)
burst_ids_mapping = self._import_bursts(project_entity,
bursts_dict)
# Now import project operations
self.import_project_operations(project_entity, new_project_path,
dt_mappings_dict, burst_ids_mapping)
# Import images
self._store_imported_images(project_entity)
# Now we can finally import workflow related entities
self.import_workflows(project_entity, bursts_dict,
burst_ids_mapping)
def import_workflows(self, project, bursts_dict, burst_ids_mapping):
"""
Import the workflow entities for all bursts imported in the project.
:param project: the current
:param bursts_dict: a dictionary that holds all the required information in order to
import the bursts from the new project
:param burst_ids_mapping: a dictionary of the form {old_burst_id : new_burst_id} so we
know what burst to link each workflow to
"""
for burst_id in bursts_dict:
workflows_info = bursts_dict[burst_id].get_workflows()
for one_wf_info in workflows_info:
# Use the new burst id when creating the workflow
workflow_entity = model.Workflow(
project.id, burst_ids_mapping[int(burst_id)])
workflow_entity.from_dict(one_wf_info.data)
workflow_entity = dao.store_entity(workflow_entity)
wf_steps_info = one_wf_info.get_workflow_steps()
view_steps_info = one_wf_info.get_view_steps()
self.import_workflow_steps(workflow_entity, wf_steps_info,
view_steps_info)
def import_workflow_steps(self, workflow, wf_steps, view_steps):
"""
Import all workflow steps for the given workflow. We create both wf_steps and view_steps
in the same method, since if a wf_step has to be omited for some reason, we also need to
omit that view step.
:param workflow: a model.Workflow entity from which we need to add workflow steps
:param wf_steps: a list of WorkflowStepInformation entities, from which we will rebuild
the workflow steps
:param view_steps: a list of WorkflowViewStepInformation entities, from which we will
rebuild the workflow view steps
"""
for wf_step in wf_steps:
try:
algorithm = wf_step.get_algorithm()
if algorithm is None:
# The algorithm is invalid for some reason. Just remove also the view step.
position = wf_step.index()
for entry in view_steps:
if entry.index() == position:
view_steps.remove(entry)
continue
wf_step_entity = model.WorkflowStep(algorithm.id)
wf_step_entity.from_dict(wf_step.data)
wf_step_entity.fk_workflow = workflow.id
wf_step_entity.fk_operation = wf_step.get_operation_id()
dao.store_entity(wf_step_entity)
except Exception:
# only log exception and ignore this as it is not very important:
self.logger.exception("Could not restore WorkflowStep: %s" %
wf_step.get_algorithm().name)
for view_step in view_steps:
try:
algorithm = view_step.get_algorithm()
if algorithm is None:
continue
view_step_entity = model.WorkflowStepView(algorithm.id)
view_step_entity.from_dict(view_step.data)
view_step_entity.fk_workflow = workflow.id
view_step_entity.fk_portlet = view_step.get_portlet().id
dao.store_entity(view_step_entity)
except Exception:
# only log exception and ignore this as it is not very important:
self.logger.exception("Could not restore WorkflowViewStep " +
view_step.get_algorithm().name)
@staticmethod
def _append_tmp_to_folders_containing_operations(import_path):
"""
Find folders containing operations and rename them, return the renamed paths
"""
pths = []
for root, _, files in os.walk(import_path):
if FilesHelper.TVB_OPERARATION_FILE in files:
# Found an operation folder - append TMP to its name
tmp_op_folder = root + 'tmp'
os.rename(root, tmp_op_folder)
operation_file_path = os.path.join(
tmp_op_folder, FilesHelper.TVB_OPERARATION_FILE)
pths.append(operation_file_path)
return pths
def _load_operations_from_paths(self, project, op_paths):
"""
Load operations from paths containing them.
:returns: Operations ordered by start/creation date to be sure data dependency is resolved correct
"""
def by_time(op):
return op.start_date or op.create_date or datetime.now()
operations = []
for operation_file_path in op_paths:
operation = self.__build_operation_from_file(
project, operation_file_path)
operation.import_file = operation_file_path
operations.append(operation)
operations.sort(key=by_time)
return operations
def _load_datatypes_from_operation_folder(self, op_path, operation_entity,
datatype_group):
"""
Loads datatypes from operation folder
:returns: Datatypes ordered by creation date (to solve any dependencies)
"""
all_datatypes = []
for file_name in os.listdir(op_path):
if file_name.endswith(FilesHelper.TVB_STORAGE_FILE_EXTENSION):
h5_file = os.path.join(op_path, file_name)
try:
file_update_manager = FilesUpdateManager()
file_update_manager.upgrade_file(h5_file)
datatype = self.load_datatype_from_file(
op_path, file_name, operation_entity.id,
datatype_group)
all_datatypes.append(datatype)
except IncompatibleFileManagerException:
os.remove(h5_file)
self.logger.warning(
"Incompatible H5 file will be ignored: %s" % h5_file)
self.logger.exception("Incompatibility details ...")
all_datatypes.sort(key=lambda dt_date: dt_date.create_date)
for dt in all_datatypes:
self.logger.debug("Import order %s: %s" % (dt.type, dt.gid))
return all_datatypes
def _store_imported_datatypes_in_db(self, project, all_datatypes,
dt_burst_mappings, burst_ids_mapping):
def by_time(dt):
return dt.create_date or datetime.now()
if burst_ids_mapping is None:
burst_ids_mapping = {}
if dt_burst_mappings is None:
dt_burst_mappings = {}
all_datatypes.sort(key=by_time)
for datatype in all_datatypes:
old_burst_id = dt_burst_mappings.get(datatype.gid)
if old_burst_id is not None:
datatype.fk_parent_burst = burst_ids_mapping[old_burst_id]
datatype_allready_in_tvb = dao.get_datatype_by_gid(datatype.gid)
if not datatype_allready_in_tvb:
# Compute disk size. Similar to ABCAdapter._capture_operation_results.
# No need to close the h5 as we have not written to it.
associated_file = os.path.join(
datatype.storage_path, datatype.get_storage_file_name())
datatype.disk_size = FilesHelper.compute_size_on_disk(
associated_file)
self.store_datatype(datatype)
else:
FlowService.create_link([datatype_allready_in_tvb.id],
project.id)
def _store_imported_images(self, project):
"""
Import all images from project
"""
images_root = self.files_helper.get_images_folder(project.name)
# for file_name in os.listdir(images_root):
for root, _, files in os.walk(images_root):
for file_name in files:
if file_name.endswith(FilesHelper.TVB_FILE_EXTENSION):
self._populate_image(os.path.join(root, file_name),
project.id)
def import_project_operations(self,
project,
import_path,
dt_burst_mappings=None,
burst_ids_mapping=None):
"""
This method scans provided folder and identify all operations that needs to be imported
"""
op_paths = self._append_tmp_to_folders_containing_operations(
import_path)
operations = self._load_operations_from_paths(project, op_paths)
imported_operations = []
datatypes = []
# Here we process each operation found
for operation in operations:
self.logger.debug("Importing operation " + str(operation))
old_operation_folder, _ = os.path.split(operation.import_file)
operation_entity, datatype_group = self.__import_operation(
operation)
# Rename operation folder with the ID of the stored operation
new_operation_path = FilesHelper().get_operation_folder(
project.name, operation_entity.id)
if old_operation_folder != new_operation_path:
# Delete folder of the new operation, otherwise move will fail
shutil.rmtree(new_operation_path)
shutil.move(old_operation_folder, new_operation_path)
operation_datatypes = self._load_datatypes_from_operation_folder(
new_operation_path, operation_entity, datatype_group)
imported_operations.append(operation_entity)
datatypes.extend(operation_datatypes)
self._store_imported_datatypes_in_db(project, datatypes,
dt_burst_mappings,
burst_ids_mapping)
return imported_operations
def _populate_image(self, file_name, project_id):
"""
Create and store a image entity.
"""
figure_dict = XMLReader(file_name).read_metadata()
new_path = os.path.join(
os.path.split(file_name)[0],
os.path.split(figure_dict['file_path'])[1])
if not os.path.exists(new_path):
self.logger.warn("Expected to find image path %s .Skipping" %
new_path)
op = dao.get_operation_by_gid(figure_dict['fk_from_operation'])
figure_dict['fk_op_id'] = op.id if op is not None else None
figure_dict['fk_user_id'] = self.user_id
figure_dict['fk_project_id'] = project_id
figure_entity = manager_of_class(model.ResultFigure).new_instance()
figure_entity = figure_entity.from_dict(figure_dict)
stored_entity = dao.store_entity(figure_entity)
# Update image meta-data with the new details after import
figure = dao.load_figure(stored_entity.id)
self.logger.debug("Store imported figure")
self.files_helper.write_image_metadata(figure)
def load_datatype_from_file(self,
storage_folder,
file_name,
op_id,
datatype_group=None,
move=True):
"""
Creates an instance of datatype from storage / H5 file
:returns: datatype
"""
self.logger.debug("Loading datatType from file: %s" % file_name)
storage_manager = HDF5StorageManager(storage_folder, file_name)
meta_dictionary = storage_manager.get_metadata()
meta_structure = DataTypeMetaData(meta_dictionary)
# Now try to determine class and instantiate it
class_name = meta_structure[DataTypeMetaData.KEY_CLASS_NAME]
class_module = meta_structure[DataTypeMetaData.KEY_MODULE]
datatype = __import__(class_module, globals(), locals(), [class_name])
datatype = getattr(datatype, class_name)
type_instance = manager_of_class(datatype).new_instance()
# Now we fill data into instance
type_instance.type = str(type_instance.__class__.__name__)
type_instance.module = str(type_instance.__module__)
# Fill instance with meta data
type_instance.load_from_metadata(meta_dictionary)
#Add all the required attributes
if datatype_group is not None:
type_instance.fk_datatype_group = datatype_group.id
type_instance.set_operation_id(op_id)
# Now move storage file into correct folder if necessary
current_file = os.path.join(storage_folder, file_name)
new_file = type_instance.get_storage_file_path()
if new_file != current_file and move:
shutil.move(current_file, new_file)
return type_instance
def store_datatype(self, datatype):
"""This method stores data type into DB"""
try:
self.logger.debug("Store datatype: %s with Gid: %s" %
(datatype.__class__.__name__, datatype.gid))
return dao.store_entity(datatype)
except MissingDataSetException:
self.logger.error(
"Datatype %s has missing data and could not be imported properly."
% (datatype, ))
os.remove(datatype.get_storage_file_path())
except IntegrityError as excep:
self.logger.exception(excep)
error_msg = "Could not import data with gid: %s. There is already a one with " \
"the same name or gid." % datatype.gid
# Delete file if can't be imported
os.remove(datatype.get_storage_file_path())
raise ProjectImportException(error_msg)
def __populate_project(self, project_path):
"""
Create and store a Project entity.
"""
self.logger.debug("Creating project from path: %s" % project_path)
project_dict = self.files_helper.read_project_metadata(project_path)
project_entity = manager_of_class(model.Project).new_instance()
project_entity = project_entity.from_dict(project_dict, self.user_id)
try:
self.logger.debug("Storing imported project")
return dao.store_entity(project_entity)
except IntegrityError as excep:
self.logger.exception(excep)
error_msg = (
"Could not import project: %s with gid: %s. There is already a "
"project with the same name or gid.") % (project_entity.name,
project_entity.gid)
raise ProjectImportException(error_msg)
def __build_operation_from_file(self, project, operation_file):
"""
Create Operation entity from metadata file.
"""
operation_dict = XMLReader(operation_file).read_metadata()
operation_entity = manager_of_class(model.Operation).new_instance()
return operation_entity.from_dict(operation_dict, dao, self.user_id,
project.gid)
@staticmethod
def __import_operation(operation_entity):
"""
Store a Operation entity.
"""
operation_entity = dao.store_entity(operation_entity)
operation_group_id = operation_entity.fk_operation_group
datatype_group = None
if operation_group_id is not None:
try:
datatype_group = dao.get_datatypegroup_by_op_group_id(
operation_group_id)
except SQLAlchemyError:
# If no dataType group present for current op. group, create it.
operation_group = dao.get_operationgroup_by_id(
operation_group_id)
datatype_group = model.DataTypeGroup(
operation_group, operation_id=operation_entity.id)
datatype_group.state = ADAPTERS['Upload']['defaultdatastate']
datatype_group = dao.store_entity(datatype_group)
return operation_entity, datatype_group
def load_burst_entity(self, json_burst, project_id):
"""
Load BurstConfiguration from JSON (possibly exported from a different machine).
Nothing gets persisted in DB or on disk.
:param json_burst: Burst JSON export
:param project_id: Current project ID (it will be used if the user later starts this simulation)
:return: BurstConfiguration filled from JSON
"""
burst_information = BurstInformation.load_from_dict(json_burst)
burst_entity = model.BurstConfiguration(project_id)
burst_entity.from_dict(burst_information.data)
burst_entity.prepare_after_load()
burst_entity.reset_tabs()
workflow_info = burst_information.get_workflows()[0]
workflow_entity = model.Workflow(project_id, None)
workflow_entity.from_dict(workflow_info.data)
view_steps = workflow_info.get_view_steps()
analyze_steps = workflow_info.get_workflow_steps()
for view_step in view_steps:
try:
algorithm = view_step.get_algorithm()
portlet = view_step.get_portlet()
view_step_entity = model.WorkflowStepView(
algorithm.id, portlet_id=portlet.id)
view_step_entity.from_dict(view_step.data)
view_step_entity.workflow = workflow_entity
## For each visualize step, also load all of the analyze steps.
analyzers = []
for an_step in analyze_steps:
if (an_step.data["tab_index"] != view_step_entity.tab_index
or an_step.data["index_in_tab"] !=
view_step_entity.index_in_tab):
continue
algorithm = an_step.get_algorithm()
wf_step_entity = model.WorkflowStep(algorithm.id)
wf_step_entity.from_dict(an_step.data)
wf_step_entity.workflow = workflow_entity
analyzers.append(wf_step_entity)
portlet = PortletConfiguration(portlet.id)
portlet.set_visualizer(view_step_entity)
portlet.set_analyzers(analyzers)
burst_entity.set_portlet(view_step_entity.tab_index,
view_step_entity.index_in_tab,
portlet)
except Exception:
# only log exception and ignore this step from loading
self.logger.exception("Could not restore Workflow Step " +
view_step.get_algorithm().name)
return burst_entity
class ImportService(object):
"""
Service for importing TVB entities into system.
It supports TVB exported H5 files as input, but it should also handle H5 files
generated outside of TVB, as long as they respect the same structure.
"""
def __init__(self):
self.logger = get_logger(__name__)
self.user_id = None
self.files_helper = FilesHelper()
self.created_projects = []
def _download_and_unpack_project_zip(self, uploaded, uq_file_name,
temp_folder):
if isinstance(uploaded, FieldStorage) or isinstance(uploaded, Part):
if not uploaded.file:
raise ProjectImportException(
"Please select the archive which contains the project structure."
)
with open(uq_file_name, 'wb') as file_obj:
self.files_helper.copy_file(uploaded.file, file_obj)
else:
shutil.copy2(uploaded, uq_file_name)
try:
self.files_helper.unpack_zip(uq_file_name, temp_folder)
except FileStructureException as excep:
self.logger.exception(excep)
raise ProjectImportException(
"Bad ZIP archive provided. A TVB exported project is expected!"
)
@staticmethod
def _compute_unpack_path():
"""
:return: the name of the folder where to expand uploaded zip
"""
now = datetime.now()
date_str = "%d-%d-%d_%d-%d-%d_%d" % (now.year, now.month, now.day,
now.hour, now.minute, now.second,
now.microsecond)
uq_name = "%s-ImportProject" % date_str
return os.path.join(TvbProfile.current.TVB_TEMP_FOLDER, uq_name)
@transactional
def import_project_structure(self, uploaded, user_id):
"""
Execute import operations:
1. check if ZIP or folder
2. find all project nodes
3. for each project node:
- create project
- create all operations
- import all images
- create all dataTypes
"""
self.user_id = user_id
self.created_projects = []
# Now compute the name of the folder where to explode uploaded ZIP file
temp_folder = self._compute_unpack_path()
uq_file_name = temp_folder + ".zip"
try:
self._download_and_unpack_project_zip(uploaded, uq_file_name,
temp_folder)
self._import_projects_from_folder(temp_folder)
except Exception as excep:
self.logger.exception(
"Error encountered during import. Deleting projects created during this operation."
)
# Remove project folders created so far.
# Note that using the project service to remove the projects will not work,
# because we do not have support for nested transaction.
# Removing from DB is not necessary because in transactional env a simple exception throw
# will erase everything to be inserted.
for project in self.created_projects:
project_path = os.path.join(TvbProfile.current.TVB_STORAGE,
FilesHelper.PROJECTS_FOLDER,
project.name)
shutil.rmtree(project_path)
raise ProjectImportException(str(excep))
finally:
# Now delete uploaded file
if os.path.exists(uq_file_name):
os.remove(uq_file_name)
# Now delete temporary folder where uploaded ZIP was exploded.
if os.path.exists(temp_folder):
shutil.rmtree(temp_folder)
@staticmethod
def _load_burst_info_from_json(project_path):
bursts_dict = {}
dt_mappings_dict = {}
bursts_file = os.path.join(project_path, BURST_INFO_FILE)
if os.path.isfile(bursts_file):
with open(bursts_file) as f:
bursts_info_dict = json.load(f)
bursts_dict = bursts_info_dict[BURSTS_DICT_KEY]
dt_mappings_dict = bursts_info_dict[DT_BURST_MAP]
os.remove(bursts_file)
return bursts_dict, dt_mappings_dict
@staticmethod
def _import_bursts(project_entity, bursts_dict):
"""
Re-create old bursts, but keep a mapping between the id it has here and the old-id it had
in the project where they were exported, so we can re-add the datatypes to them.
"""
burst_ids_mapping = {}
# for old_burst_id in bursts_dict:
# burst_information = BurstInformation.load_from_dict(bursts_dict[old_burst_id])
# burst_entity = BurstConfiguration(project_entity.id)
# burst_entity.from_dict(burst_information.data)
# burst_entity = dao.store_entity(burst_entity)
# burst_ids_mapping[int(old_burst_id)] = burst_entity.id
# We don't need the data in dictionary form anymore, so update it with new BurstInformation object
# bursts_dict[old_burst_id] = burst_information
return burst_ids_mapping
def _import_projects_from_folder(self, temp_folder):
"""
Process each project from the uploaded pack, to extract names.
"""
project_roots = []
for root, _, files in os.walk(temp_folder):
if FilesHelper.TVB_PROJECT_FILE in files:
project_roots.append(root)
for project_path in project_roots:
update_manager = ProjectUpdateManager(project_path)
update_manager.run_all_updates()
project_entity = self.__populate_project(project_path)
# Compute the path where to store files of the imported project
new_project_path = os.path.join(TvbProfile.current.TVB_STORAGE,
FilesHelper.PROJECTS_FOLDER,
project_entity.name)
if project_path != new_project_path:
shutil.move(project_path, new_project_path)
self.created_projects.append(project_entity)
# Keep a list with all burst that were imported since we will want to also add the workflow
# steps after we are finished with importing the operations and datatypes. We need to first
# stored bursts since we need to know which new id's they have for operations parent_burst.
# bursts_dict, dt_mappings_dict = self._load_burst_info_from_json(new_project_path)
# burst_ids_mapping = self._import_bursts(project_entity, bursts_dict)
# Now import project operations
self.import_project_operations(project_entity, new_project_path)
# Import images
self._store_imported_images(project_entity)
@staticmethod
def _append_tmp_to_folders_containing_operations(import_path):
"""
Find folders containing operations and rename them, return the renamed paths
"""
pths = []
for root, _, files in os.walk(import_path):
if FilesHelper.TVB_OPERARATION_FILE in files:
# Found an operation folder - append TMP to its name
tmp_op_folder = root + 'tmp'
os.rename(root, tmp_op_folder)
operation_file_path = os.path.join(
tmp_op_folder, FilesHelper.TVB_OPERARATION_FILE)
pths.append(operation_file_path)
return pths
def _load_operations_from_paths(self, project, op_paths):
"""
Load operations from paths containing them.
:returns: Operations ordered by start/creation date to be sure data dependency is resolved correct
"""
def by_time(op):
return op.start_date or op.create_date or datetime.now()
operations = []
for operation_file_path in op_paths:
operation = self.__build_operation_from_file(
project, operation_file_path)
operation.import_file = operation_file_path
operations.append(operation)
operations.sort(key=by_time)
return operations
def _load_datatypes_from_operation_folder(self, op_path, operation_entity,
datatype_group):
"""
Loads datatypes from operation folder
:returns: Datatypes ordered by creation date (to solve any dependencies)
"""
all_datatypes = []
for file_name in os.listdir(op_path):
if file_name.endswith(FilesHelper.TVB_STORAGE_FILE_EXTENSION):
h5_file = os.path.join(op_path, file_name)
try:
file_update_manager = FilesUpdateManager()
file_update_manager.upgrade_file(h5_file)
datatype = self.load_datatype_from_file(
op_path, file_name, operation_entity.id,
datatype_group)
all_datatypes.append(datatype)
except IncompatibleFileManagerException:
os.remove(h5_file)
self.logger.warning(
"Incompatible H5 file will be ignored: %s" % h5_file)
self.logger.exception("Incompatibility details ...")
all_datatypes.sort(key=lambda dt_date: dt_date.create_date)
for dt in all_datatypes:
self.logger.debug("Import order %s: %s" % (dt.type, dt.gid))
return all_datatypes
def _store_imported_datatypes_in_db(self, project, all_datatypes,
dt_burst_mappings, burst_ids_mapping):
def by_time(dt):
return dt.create_date or datetime.now()
if burst_ids_mapping is None:
burst_ids_mapping = {}
if dt_burst_mappings is None:
dt_burst_mappings = {}
all_datatypes.sort(key=by_time)
for datatype in all_datatypes:
old_burst_id = dt_burst_mappings.get(datatype.gid)
if old_burst_id is not None:
datatype.fk_parent_burst = burst_ids_mapping[old_burst_id]
datatype_allready_in_tvb = dao.get_datatype_by_gid(datatype.gid)
if not datatype_allready_in_tvb:
self.store_datatype(datatype)
else:
AlgorithmService.create_link([datatype_allready_in_tvb.id],
project.id)
def _store_imported_images(self, project):
"""
Import all images from project
"""
images_root = self.files_helper.get_images_folder(project.name)
# for file_name in os.listdir(images_root):
for root, _, files in os.walk(images_root):
for file_name in files:
if file_name.endswith(FilesHelper.TVB_FILE_EXTENSION):
self._populate_image(os.path.join(root, file_name),
project.id)
def import_project_operations(self,
project,
import_path,
dt_burst_mappings=None,
burst_ids_mapping=None):
"""
This method scans provided folder and identify all operations that needs to be imported
"""
op_paths = self._append_tmp_to_folders_containing_operations(
import_path)
operations = self._load_operations_from_paths(project, op_paths)
imported_operations = []
# Here we process each operation found
for operation in operations:
self.logger.debug("Importing operation " + str(operation))
old_operation_folder, _ = os.path.split(operation.import_file)
operation_entity, datatype_group = self.__import_operation(
operation)
# Rename operation folder with the ID of the stored operation
new_operation_path = FilesHelper().get_operation_folder(
project.name, operation_entity.id)
if old_operation_folder != new_operation_path:
# Delete folder of the new operation, otherwise move will fail
shutil.rmtree(new_operation_path)
shutil.move(old_operation_folder, new_operation_path)
operation_datatypes = self._load_datatypes_from_operation_folder(
new_operation_path, operation_entity, datatype_group)
self._store_imported_datatypes_in_db(project, operation_datatypes,
dt_burst_mappings,
burst_ids_mapping)
imported_operations.append(operation_entity)
return imported_operations
def _populate_image(self, file_name, project_id):
"""
Create and store a image entity.
"""
figure_dict = XMLReader(file_name).read_metadata()
new_path = os.path.join(
os.path.split(file_name)[0],
os.path.split(figure_dict['file_path'])[1])
if not os.path.exists(new_path):
self.logger.warning("Expected to find image path %s .Skipping" %
new_path)
op = dao.get_operation_by_gid(figure_dict['fk_from_operation'])
figure_dict['fk_op_id'] = op.id if op is not None else None
figure_dict['fk_user_id'] = self.user_id
figure_dict['fk_project_id'] = project_id
figure_entity = manager_of_class(ResultFigure).new_instance()
figure_entity = figure_entity.from_dict(figure_dict)
stored_entity = dao.store_entity(figure_entity)
# Update image meta-data with the new details after import
figure = dao.load_figure(stored_entity.id)
self.logger.debug("Store imported figure")
self.files_helper.write_image_metadata(figure)
def load_datatype_from_file(self,
storage_folder,
file_name,
op_id,
datatype_group=None,
move=True,
final_storage=None):
"""
Creates an instance of datatype from storage / H5 file
:returns: DatatypeIndex
"""
self.logger.debug("Loading DataType from file: %s" % file_name)
datatype, generic_attributes = h5.load_with_references(
os.path.join(storage_folder, file_name))
index_class = h5.REGISTRY.get_index_for_datatype(datatype.__class__)
datatype_index = index_class()
datatype_index.fill_from_has_traits(datatype)
datatype_index.fill_from_generic_attributes(generic_attributes)
# Add all the required attributes
if datatype_group is not None:
datatype_index.fk_datatype_group = datatype_group.id
datatype_index.fk_from_operation = op_id
associated_file = h5.path_for_stored_index(datatype_index)
if os.path.exists(associated_file):
datatype_index.disk_size = FilesHelper.compute_size_on_disk(
associated_file)
# Now move storage file into correct folder if necessary
if move and final_storage is not None:
current_file = os.path.join(storage_folder, file_name)
h5_type = h5.REGISTRY.get_h5file_for_datatype(datatype.__class__)
final_path = h5.path_for(final_storage, h5_type, datatype.gid)
if final_path != current_file and move:
shutil.move(current_file, final_path)
return datatype_index
def store_datatype(self, datatype):
"""This method stores data type into DB"""
try:
self.logger.debug("Store datatype: %s with Gid: %s" %
(datatype.__class__.__name__, datatype.gid))
return dao.store_entity(datatype)
except MissingDataSetException:
self.logger.error(
"Datatype %s has missing data and could not be imported properly."
% (datatype, ))
os.remove(datatype.get_storage_file_path())
except IntegrityError as excep:
self.logger.exception(excep)
error_msg = "Could not import data with gid: %s. There is already a one with " \
"the same name or gid." % datatype.gid
# Delete file if can't be imported
os.remove(datatype.get_storage_file_path())
raise ProjectImportException(error_msg)
def __populate_project(self, project_path):
"""
Create and store a Project entity.
"""
self.logger.debug("Creating project from path: %s" % project_path)
project_dict = self.files_helper.read_project_metadata(project_path)
project_entity = manager_of_class(Project).new_instance()
project_entity = project_entity.from_dict(project_dict, self.user_id)
try:
self.logger.debug("Storing imported project")
return dao.store_entity(project_entity)
except IntegrityError as excep:
self.logger.exception(excep)
error_msg = (
"Could not import project: %s with gid: %s. There is already a "
"project with the same name or gid.") % (project_entity.name,
project_entity.gid)
raise ProjectImportException(error_msg)
def __build_operation_from_file(self, project, operation_file):
"""
Create Operation entity from metadata file.
"""
operation_dict = XMLReader(operation_file).read_metadata()
operation_entity = manager_of_class(Operation).new_instance()
return operation_entity.from_dict(operation_dict, dao, self.user_id,
project.gid)
@staticmethod
def __import_operation(operation_entity):
"""
Store a Operation entity.
"""
operation_entity = dao.store_entity(operation_entity)
operation_group_id = operation_entity.fk_operation_group
datatype_group = None
if operation_group_id is not None:
try:
datatype_group = dao.get_datatypegroup_by_op_group_id(
operation_group_id)
except SQLAlchemyError:
# If no dataType group present for current op. group, create it.
operation_group = dao.get_operationgroup_by_id(
operation_group_id)
datatype_group = DataTypeGroup(
operation_group, operation_id=operation_entity.id)
datatype_group.state = UploadAlgorithmCategoryConfig.defaultdatastate
datatype_group = dao.store_entity(datatype_group)
return operation_entity, datatype_group
def import_simulator_configuration_zip(self, zip_file):
# Now compute the name of the folder where to explode uploaded ZIP file
temp_folder = self._compute_unpack_path()
uq_file_name = temp_folder + ".zip"
if isinstance(zip_file, FieldStorage) or isinstance(zip_file, Part):
if not zip_file.file:
raise ServicesBaseException(
"Could not process the given ZIP file...")
with open(uq_file_name, 'wb') as file_obj:
self.files_helper.copy_file(zip_file.file, file_obj)
else:
shutil.copy2(zip_file, uq_file_name)
try:
self.files_helper.unpack_zip(uq_file_name, temp_folder)
return temp_folder
except FileStructureException as excep:
raise ServicesBaseException(
"Could not process the given ZIP file..." + str(excep))
class FigureService:
"""
Service layer for Figure entities.
"""
_TYPE_PNG = "png"
_TYPE_SVG = "svg"
_BRANDING_BAR_PNG = os.path.join(os.path.dirname(__file__), "resources",
"branding_bar.png")
_BRANDING_BAR_SVG = os.path.join(os.path.dirname(__file__), "resources",
"branding_bar.svg")
_DEFAULT_SESSION_NAME = "Default"
_DEFAULT_IMAGE_FILE_NAME = "snapshot."
def __init__(self):
self.logger = get_logger(self.__class__.__module__)
self.file_helper = FilesHelper()
def _write_png(self, store_path, export_data):
img_data = base64.b64decode(export_data) # decode the image
final_image = Image.open(
StringIO(img_data)) # place it in a PIL stream
branding_bar = Image.open(
FigureService._BRANDING_BAR_PNG) # place the branding bar over
final_image.paste(branding_bar,
(0, final_image.size[1] - branding_bar.size[1]),
branding_bar)
final_image.save(store_path) # store to disk as PNG
def _write_svg(self, store_path, export_data):
dom = xml.dom.minidom.parseString(export_data)
figureSvg = dom.getElementsByTagName('svg')[
0] # get the original image
dom = xml.dom.minidom.parse(FigureService._BRANDING_BAR_SVG)
try:
width = float(figureSvg.getAttribute('width').replace('px', ''))
height = float(figureSvg.getAttribute('height').replace('px', ''))
except ValueError: # defaults when dimensions are not given
width = 1024
height = 768
figureSvg.setAttribute("width", str(width))
figureSvg.setAttribute("height", str(height))
finalSvg = dom.createElement('svg') # prepare the final svg
brandingSvg = dom.getElementsByTagName('svg')[
0] # get the branding bar
brandingSvg.setAttribute("y",
str(height)) # position it below the figure
height += float(brandingSvg.getAttribute('height').replace(
'px', '')) # increase original height with branding bar's height
finalSvg.setAttribute("width",
str(width)) # same width as original figure
finalSvg.setAttribute("height", str(height))
finalSvg.appendChild(figureSvg) # add the image
finalSvg.appendChild(brandingSvg) # and the branding bar
# Generate path where to store image
with open(store_path, 'w') as dest:
finalSvg.writexml(dest) # store to disk
def _image_path(self, project_name, img_type):
"Generate path where to store image"
images_folder = self.file_helper.get_images_folder(project_name)
file_name = FigureService._DEFAULT_IMAGE_FILE_NAME + img_type
return utils.get_unique_file_name(images_folder, file_name)
@staticmethod
def _generate_image_name(project, user, operation, image_name):
if not image_name:
if operation is not None:
# create a name based on the operation that created the image
# e.g. TVB-Algo-Name-354
image_name = operation.algorithm.displayname.replace(' ', '-')
else:
# default to a generic name
image_name = "figure"
figure_count = dao.get_figure_count(project.id, user.id) + 1
return 'TVB-%s-%s' % (image_name, figure_count)
def store_result_figure(self,
project,
user,
img_type,
export_data,
image_name=None,
operation_id=None):
"""
Store into a file, Result Image and reference in DB.
"""
store_path, file_name = self._image_path(project.name, img_type)
if img_type == FigureService._TYPE_PNG: # PNG file from canvas
self._write_png(store_path, export_data)
elif img_type == FigureService._TYPE_SVG: # SVG file from svg viewer
self._write_svg(store_path, export_data)
if operation_id:
operation = dao.get_operation_by_id(operation_id)
else:
operation = None
operation_id = None
image_name = self._generate_image_name(project, user, operation,
image_name)
# Store entity into DB
entity = model.ResultFigure(operation_id, user.id, project.id,
FigureService._DEFAULT_SESSION_NAME,
image_name, file_name, img_type)
entity = dao.store_entity(entity)
# Load instance from DB to have lazy fields loaded
figure = dao.load_figure(entity.id)
# Write image meta data to disk
self.file_helper.write_image_metadata(figure)
if operation:
# Force writing operation meta data on disk.
# This is important later for operation import
self.file_helper.write_operation_metadata(operation)
def retrieve_result_figures(self,
project,
user,
selected_session_name='all_sessions'):
"""
Retrieve from DB all the stored Displayer previews that belongs to the specified session. The
previews are for current user and project; grouped by session.
"""
result, previews_info = dao.get_previews(project.id, user.id,
selected_session_name)
for name in result:
for figure in result[name]:
figures_folder = self.file_helper.get_images_folder(
project.name)
figure_full_path = os.path.join(figures_folder,
figure.file_path)
# Compute the path
figure.file_path = utils.path2url_part(figure_full_path)
return result, previews_info
@staticmethod
def load_figure(figure_id):
"""
Loads a stored figure by its id.
"""
return dao.load_figure(figure_id)
def edit_result_figure(self, figure_id, **data):
"""
Retrieve and edit a previously stored figure.
"""
figure = dao.load_figure(figure_id)
figure.session_name = data['session_name']
figure.name = data['name']
dao.store_entity(figure)
# Load instance from DB to have lazy fields loaded.
figure = dao.load_figure(figure_id)
# Store figure meta data in an XML attached to the image.
self.file_helper.write_image_metadata(figure)
def remove_result_figure(self, figure_id):
"""
Remove figure from DB and file storage.
"""
figure = dao.load_figure(figure_id)
# Delete all figure related files from disk.
figures_folder = self.file_helper.get_images_folder(
figure.project.name)
path2figure = os.path.join(figures_folder, figure.file_path)
if os.path.exists(path2figure):
os.remove(path2figure)
self.file_helper.remove_image_metadata(figure)
# Remove figure reference from DB.
result = dao.remove_entity(model.ResultFigure, figure_id)
return result
class FigureService:
"""
Service layer for Figure entities.
"""
_TYPE_PNG = "png"
_TYPE_SVG = "svg"
_BRANDING_BAR_PNG = os.path.join(os.path.dirname(__file__), "resources", "branding_bar.png")
_BRANDING_BAR_SVG = os.path.join(os.path.dirname(__file__), "resources", "branding_bar.svg")
_DEFAULT_SESSION_NAME = "Default"
_DEFAULT_IMAGE_FILE_NAME = "snapshot."
def __init__(self):
self.logger = get_logger(self.__class__.__module__)
self.file_helper = FilesHelper()
def _write_png(self, store_path, export_data):
img_data = base64.b64decode(export_data) # decode the image
final_image = Image.open(StringIO(img_data)) # place it in a PIL stream
branding_bar = Image.open(FigureService._BRANDING_BAR_PNG) # place the branding bar over
final_image.paste(branding_bar, (0, final_image.size[1] - branding_bar.size[1]), branding_bar)
final_image.save(store_path) # store to disk as PNG
def _write_svg(self, store_path, export_data):
dom = xml.dom.minidom.parseString(export_data)
figureSvg = dom.getElementsByTagName('svg')[0] # get the original image
dom = xml.dom.minidom.parse(FigureService._BRANDING_BAR_SVG)
try:
width = float(figureSvg.getAttribute('width').replace('px', ''))
height = float(figureSvg.getAttribute('height').replace('px', ''))
except ValueError: # defaults when dimensions are not given
width = 1024
height = 768
figureSvg.setAttribute("width", str(width))
figureSvg.setAttribute("height", str(height))
finalSvg = dom.createElement('svg') # prepare the final svg
brandingSvg = dom.getElementsByTagName('svg')[0] # get the branding bar
brandingSvg.setAttribute("y", str(height)) # position it below the figure
height += float(brandingSvg.getAttribute('height').replace('px', '')) # increase original height with branding bar's height
finalSvg.setAttribute("width", str(width)) # same width as original figure
finalSvg.setAttribute("height", str(height))
finalSvg.appendChild(figureSvg) # add the image
finalSvg.appendChild(brandingSvg) # and the branding bar
# Generate path where to store image
with open(store_path, 'w') as dest:
finalSvg.writexml(dest) # store to disk
def _image_path(self, project_name, img_type):
"Generate path where to store image"
images_folder = self.file_helper.get_images_folder(project_name)
file_name = FigureService._DEFAULT_IMAGE_FILE_NAME + img_type
return utils.get_unique_file_name(images_folder, file_name)
@staticmethod
def _generate_image_name(project, user, operation, image_name):
if not image_name:
if operation is not None:
# create a name based on the operation that created the image
# e.g. TVB-Algo-Name-354
image_name = operation.algorithm.name.replace(' ', '-')
else:
# default to a generic name
image_name = "figure"
figure_count = dao.get_figure_count(project.id, user.id) + 1
return 'TVB-%s-%s' % (image_name, figure_count)
def store_result_figure(self, project, user, img_type, export_data, image_name=None, operation_id=None):
"""
Store into a file, Result Image and reference in DB.
"""
store_path, file_name = self._image_path(project.name, img_type)
if img_type == FigureService._TYPE_PNG: # PNG file from canvas
self._write_png(store_path, export_data)
elif img_type == FigureService._TYPE_SVG: # SVG file from svg viewer
self._write_svg(store_path, export_data)
if operation_id:
operation = dao.get_operation_by_id(operation_id)
else:
operation = None
operation_id = None
image_name = self._generate_image_name(project, user, operation, image_name)
# Store entity into DB
entity = model.ResultFigure(operation_id, user.id, project.id, FigureService._DEFAULT_SESSION_NAME,
image_name, file_name, img_type)
entity = dao.store_entity(entity)
# Load instance from DB to have lazy fields loaded
figure = dao.load_figure(entity.id)
# Write image meta data to disk
self.file_helper.write_image_metadata(figure)
if operation:
# Force writing operation meta data on disk.
# This is important later for operation import
self.file_helper.write_operation_metadata(operation)
def retrieve_result_figures(self, project, user, selected_session_name='all_sessions'):
"""
Retrieve from DB all the stored Displayer previews that belongs to the specified session. The
previews are for current user and project; grouped by session.
"""
result, previews_info = dao.get_previews(project.id, user.id, selected_session_name)
for name in result:
for figure in result[name]:
figures_folder = self.file_helper.get_images_folder(project.name)
figure_full_path = os.path.join(figures_folder, figure.file_path)
# Compute the path
figure.file_path = utils.path2url_part(figure_full_path)
return result, previews_info
@staticmethod
def load_figure(figure_id):
"""
Loads a stored figure by its id.
"""
return dao.load_figure(figure_id)
def edit_result_figure(self, figure_id, **data):
"""
Retrieve and edit a previously stored figure.
"""
figure = dao.load_figure(figure_id)
figure.session_name = data['session_name']
figure.name = data['name']
dao.store_entity(figure)
# Load instance from DB to have lazy fields loaded.
figure = dao.load_figure(figure_id)
# Store figure meta data in an XML attached to the image.
self.file_helper.write_image_metadata(figure)
def remove_result_figure(self, figure_id):
"""
Remove figure from DB and file storage.
"""
figure = dao.load_figure(figure_id)
# Delete all figure related files from disk.
figures_folder = self.file_helper.get_images_folder(figure.project.name)
path2figure = os.path.join(figures_folder, figure.file_path)
if os.path.exists(path2figure):
os.remove(path2figure)
self.file_helper.remove_image_metadata(figure)
# Remove figure reference from DB.
result = dao.remove_entity(model.ResultFigure, figure_id)
return result
class ImportService(object):
"""
Service for importing TVB entities into system.
It supports TVB exported H5 files as input, but it should also handle H5 files
generated outside of TVB, as long as they respect the same structure.
"""
def __init__(self):
self.logger = get_logger(__name__)
self.user_id = None
self.files_helper = FilesHelper()
self.created_projects = []
def _download_and_unpack_project_zip(self, uploaded, uq_file_name, temp_folder):
if isinstance(uploaded, FieldStorage) or isinstance(uploaded, Part):
if not uploaded.file:
raise ProjectImportException("Please select the archive which contains the project structure.")
with open(uq_file_name, 'wb') as file_obj:
self.files_helper.copy_file(uploaded.file, file_obj)
else:
shutil.copy2(uploaded, uq_file_name)
try:
self.files_helper.unpack_zip(uq_file_name, temp_folder)
except FileStructureException as excep:
self.logger.exception(excep)
raise ProjectImportException("Bad ZIP archive provided. A TVB exported project is expected!")
@staticmethod
def _compute_unpack_path():
"""
:return: the name of the folder where to expand uploaded zip
"""
now = datetime.now()
date_str = "%d-%d-%d_%d-%d-%d_%d" % (now.year, now.month, now.day, now.hour,
now.minute, now.second, now.microsecond)
uq_name = "%s-ImportProject" % date_str
return os.path.join(TvbProfile.current.TVB_TEMP_FOLDER, uq_name)
@transactional
def import_project_structure(self, uploaded, user_id):
"""
Execute import operations:
1. check if ZIP or folder
2. find all project nodes
3. for each project node:
- create project
- create all operations
- import all images
- create all dataTypes
"""
self.user_id = user_id
self.created_projects = []
# Now compute the name of the folder where to explode uploaded ZIP file
temp_folder = self._compute_unpack_path()
uq_file_name = temp_folder + ".zip"
try:
self._download_and_unpack_project_zip(uploaded, uq_file_name, temp_folder)
self._import_projects_from_folder(temp_folder)
except Exception as excep:
self.logger.exception("Error encountered during import. Deleting projects created during this operation.")
# Remove project folders created so far.
# Note that using the project service to remove the projects will not work,
# because we do not have support for nested transaction.
# Removing from DB is not necessary because in transactional env a simple exception throw
# will erase everything to be inserted.
for project in self.created_projects:
project_path = os.path.join(TvbProfile.current.TVB_STORAGE, FilesHelper.PROJECTS_FOLDER, project.name)
shutil.rmtree(project_path)
raise ProjectImportException(str(excep))
finally:
# Now delete uploaded file
if os.path.exists(uq_file_name):
os.remove(uq_file_name)
# Now delete temporary folder where uploaded ZIP was exploded.
if os.path.exists(temp_folder):
shutil.rmtree(temp_folder)
@staticmethod
def _load_burst_info_from_json(project_path):
bursts_dict = {}
dt_mappings_dict = {}
bursts_file = os.path.join(project_path, BURST_INFO_FILE)
if os.path.isfile(bursts_file):
with open(bursts_file) as f:
bursts_info_dict = json.load(f)
bursts_dict = bursts_info_dict[BURSTS_DICT_KEY]
dt_mappings_dict = bursts_info_dict[DT_BURST_MAP]
os.remove(bursts_file)
return bursts_dict, dt_mappings_dict
@staticmethod
def _import_bursts(project_entity, bursts_dict):
"""
Re-create old bursts, but keep a mapping between the id it has here and the old-id it had
in the project where they were exported, so we can re-add the datatypes to them.
"""
burst_ids_mapping = {}
for old_burst_id in bursts_dict:
burst_information = BurstInformation.load_from_dict(bursts_dict[old_burst_id])
burst_entity = model.BurstConfiguration(project_entity.id)
burst_entity.from_dict(burst_information.data)
burst_entity = dao.store_entity(burst_entity)
burst_ids_mapping[int(old_burst_id)] = burst_entity.id
# We don't need the data in dictionary form anymore, so update it with new BurstInformation object
bursts_dict[old_burst_id] = burst_information
return burst_ids_mapping
def _import_projects_from_folder(self, temp_folder):
"""
Process each project from the uploaded pack, to extract names.
"""
project_roots = []
for root, _, files in os.walk(temp_folder):
if FilesHelper.TVB_PROJECT_FILE in files:
project_roots.append(root)
for project_path in project_roots:
update_manager = ProjectUpdateManager(project_path)
update_manager.run_all_updates()
project_entity = self.__populate_project(project_path)
# Compute the path where to store files of the imported project
new_project_path = os.path.join(TvbProfile.current.TVB_STORAGE,
FilesHelper.PROJECTS_FOLDER, project_entity.name)
if project_path != new_project_path:
shutil.move(project_path, new_project_path)
self.created_projects.append(project_entity)
# Keep a list with all burst that were imported since we will want to also add the workflow
# steps after we are finished with importing the operations and datatypes. We need to first
# stored bursts since we need to know which new id's they have for operations parent_burst.
bursts_dict, dt_mappings_dict = self._load_burst_info_from_json(new_project_path)
burst_ids_mapping = self._import_bursts(project_entity, bursts_dict)
# Now import project operations
self.import_project_operations(project_entity, new_project_path, dt_mappings_dict, burst_ids_mapping)
# Import images
self._store_imported_images(project_entity)
# Now we can finally import workflow related entities
self.import_workflows(project_entity, bursts_dict, burst_ids_mapping)
def import_workflows(self, project, bursts_dict, burst_ids_mapping):
"""
Import the workflow entities for all bursts imported in the project.
:param project: the current
:param bursts_dict: a dictionary that holds all the required information in order to
import the bursts from the new project
:param burst_ids_mapping: a dictionary of the form {old_burst_id : new_burst_id} so we
know what burst to link each workflow to
"""
for burst_id in bursts_dict:
workflows_info = bursts_dict[burst_id].get_workflows()
for one_wf_info in workflows_info:
# Use the new burst id when creating the workflow
workflow_entity = model.Workflow(project.id, burst_ids_mapping[int(burst_id)])
workflow_entity.from_dict(one_wf_info.data)
workflow_entity = dao.store_entity(workflow_entity)
wf_steps_info = one_wf_info.get_workflow_steps()
view_steps_info = one_wf_info.get_view_steps()
self.import_workflow_steps(workflow_entity, wf_steps_info, view_steps_info)
def import_workflow_steps(self, workflow, wf_steps, view_steps):
"""
Import all workflow steps for the given workflow. We create both wf_steps and view_steps
in the same method, since if a wf_step has to be omited for some reason, we also need to
omit that view step.
:param workflow: a model.Workflow entity from which we need to add workflow steps
:param wf_steps: a list of WorkflowStepInformation entities, from which we will rebuild
the workflow steps
:param view_steps: a list of WorkflowViewStepInformation entities, from which we will
rebuild the workflow view steps
"""
for wf_step in wf_steps:
try:
algorithm = wf_step.get_algorithm()
if algorithm is None:
# The algorithm is invalid for some reason. Just remove also the view step.
position = wf_step.index()
for entry in view_steps:
if entry.index() == position:
view_steps.remove(entry)
continue
wf_step_entity = model.WorkflowStep(algorithm.id)
wf_step_entity.from_dict(wf_step.data)
wf_step_entity.fk_workflow = workflow.id
wf_step_entity.fk_operation = wf_step.get_operation_id()
dao.store_entity(wf_step_entity)
except Exception:
# only log exception and ignore this as it is not very important:
self.logger.exception("Could not restore WorkflowStep: %s" % wf_step.get_algorithm().name)
for view_step in view_steps:
try:
algorithm = view_step.get_algorithm()
if algorithm is None:
continue
view_step_entity = model.WorkflowStepView(algorithm.id)
view_step_entity.from_dict(view_step.data)
view_step_entity.fk_workflow = workflow.id
view_step_entity.fk_portlet = view_step.get_portlet().id
dao.store_entity(view_step_entity)
except Exception:
# only log exception and ignore this as it is not very important:
self.logger.exception("Could not restore WorkflowViewStep " + view_step.get_algorithm().name)
@staticmethod
def _append_tmp_to_folders_containing_operations(import_path):
"""
Find folders containing operations and rename them, return the renamed paths
"""
pths = []
for root, _, files in os.walk(import_path):
if FilesHelper.TVB_OPERARATION_FILE in files:
# Found an operation folder - append TMP to its name
tmp_op_folder = root + 'tmp'
os.rename(root, tmp_op_folder)
operation_file_path = os.path.join(tmp_op_folder, FilesHelper.TVB_OPERARATION_FILE)
pths.append(operation_file_path)
return pths
def _load_operations_from_paths(self, project, op_paths):
"""
Load operations from paths containing them.
:returns: Operations ordered by start/creation date to be sure data dependency is resolved correct
"""
def by_time(op):
return op.start_date or op.create_date or datetime.now()
operations = []
for operation_file_path in op_paths:
operation = self.__build_operation_from_file(project, operation_file_path)
operation.import_file = operation_file_path
operations.append(operation)
operations.sort(key=by_time)
return operations
def _load_datatypes_from_operation_folder(self, op_path, operation_entity, datatype_group):
"""
Loads datatypes from operation folder
:returns: Datatypes ordered by creation date (to solve any dependencies)
"""
all_datatypes = []
for file_name in os.listdir(op_path):
if file_name.endswith(FilesHelper.TVB_STORAGE_FILE_EXTENSION):
h5_file = os.path.join(op_path, file_name)
try:
file_update_manager = FilesUpdateManager()
file_update_manager.upgrade_file(h5_file)
datatype = self.load_datatype_from_file(op_path, file_name, operation_entity.id, datatype_group)
all_datatypes.append(datatype)
except IncompatibleFileManagerException:
os.remove(h5_file)
self.logger.warning("Incompatible H5 file will be ignored: %s" % h5_file)
self.logger.exception("Incompatibility details ...")
all_datatypes.sort(key=lambda dt_date: dt_date.create_date)
for dt in all_datatypes:
self.logger.debug("Import order %s: %s" % (dt.type, dt.gid))
return all_datatypes
def _store_imported_datatypes_in_db(self, project, all_datatypes, dt_burst_mappings, burst_ids_mapping):
def by_time(dt):
return dt.create_date or datetime.now()
if burst_ids_mapping is None:
burst_ids_mapping = {}
if dt_burst_mappings is None:
dt_burst_mappings = {}
all_datatypes.sort(key=by_time)
for datatype in all_datatypes:
old_burst_id = dt_burst_mappings.get(datatype.gid)
if old_burst_id is not None:
datatype.fk_parent_burst = burst_ids_mapping[old_burst_id]
datatype_allready_in_tvb = dao.get_datatype_by_gid(datatype.gid)
if not datatype_allready_in_tvb:
# Compute disk size. Similar to ABCAdapter._capture_operation_results.
# No need to close the h5 as we have not written to it.
associated_file = os.path.join(datatype.storage_path, datatype.get_storage_file_name())
datatype.disk_size = FilesHelper.compute_size_on_disk(associated_file)
self.store_datatype(datatype)
else:
FlowService.create_link([datatype_allready_in_tvb.id], project.id)
def _store_imported_images(self, project):
"""
Import all images from project
"""
images_root = self.files_helper.get_images_folder(project.name)
# for file_name in os.listdir(images_root):
for root, _, files in os.walk(images_root):
for file_name in files:
if file_name.endswith(FilesHelper.TVB_FILE_EXTENSION):
self._populate_image(os.path.join(root, file_name), project.id)
def import_project_operations(self, project, import_path, dt_burst_mappings=None, burst_ids_mapping=None):
"""
This method scans provided folder and identify all operations that needs to be imported
"""
op_paths = self._append_tmp_to_folders_containing_operations(import_path)
operations = self._load_operations_from_paths(project, op_paths)
imported_operations = []
datatypes = []
# Here we process each operation found
for operation in operations:
self.logger.debug("Importing operation " + str(operation))
old_operation_folder, _ = os.path.split(operation.import_file)
operation_entity, datatype_group = self.__import_operation(operation)
# Rename operation folder with the ID of the stored operation
new_operation_path = FilesHelper().get_operation_folder(project.name, operation_entity.id)
if old_operation_folder != new_operation_path:
# Delete folder of the new operation, otherwise move will fail
shutil.rmtree(new_operation_path)
shutil.move(old_operation_folder, new_operation_path)
operation_datatypes = self._load_datatypes_from_operation_folder(new_operation_path, operation_entity,
datatype_group)
imported_operations.append(operation_entity)
datatypes.extend(operation_datatypes)
self._store_imported_datatypes_in_db(project, datatypes, dt_burst_mappings, burst_ids_mapping)
return imported_operations
def _populate_image(self, file_name, project_id):
"""
Create and store a image entity.
"""
figure_dict = XMLReader(file_name).read_metadata()
new_path = os.path.join(os.path.split(file_name)[0], os.path.split(figure_dict['file_path'])[1])
if not os.path.exists(new_path):
self.logger.warn("Expected to find image path %s .Skipping" % new_path)
op = dao.get_operation_by_gid(figure_dict['fk_from_operation'])
figure_dict['fk_op_id'] = op.id if op is not None else None
figure_dict['fk_user_id'] = self.user_id
figure_dict['fk_project_id'] = project_id
figure_entity = manager_of_class(model.ResultFigure).new_instance()
figure_entity = figure_entity.from_dict(figure_dict)
stored_entity = dao.store_entity(figure_entity)
# Update image meta-data with the new details after import
figure = dao.load_figure(stored_entity.id)
self.logger.debug("Store imported figure")
self.files_helper.write_image_metadata(figure)
def load_datatype_from_file(self, storage_folder, file_name, op_id, datatype_group=None, move=True):
"""
Creates an instance of datatype from storage / H5 file
:returns: datatype
"""
self.logger.debug("Loading datatType from file: %s" % file_name)
storage_manager = HDF5StorageManager(storage_folder, file_name)
meta_dictionary = storage_manager.get_metadata()
meta_structure = DataTypeMetaData(meta_dictionary)
# Now try to determine class and instantiate it
class_name = meta_structure[DataTypeMetaData.KEY_CLASS_NAME]
class_module = meta_structure[DataTypeMetaData.KEY_MODULE]
datatype = __import__(class_module, globals(), locals(), [class_name])
datatype = getattr(datatype, class_name)
type_instance = manager_of_class(datatype).new_instance()
# Now we fill data into instance
type_instance.type = str(type_instance.__class__.__name__)
type_instance.module = str(type_instance.__module__)
# Fill instance with meta data
type_instance.load_from_metadata(meta_dictionary)
#Add all the required attributes
if datatype_group is not None:
type_instance.fk_datatype_group = datatype_group.id
type_instance.set_operation_id(op_id)
# Now move storage file into correct folder if necessary
current_file = os.path.join(storage_folder, file_name)
new_file = type_instance.get_storage_file_path()
if new_file != current_file and move:
shutil.move(current_file, new_file)
return type_instance
def store_datatype(self, datatype):
"""This method stores data type into DB"""
try:
self.logger.debug("Store datatype: %s with Gid: %s" % (datatype.__class__.__name__, datatype.gid))
return dao.store_entity(datatype)
except MissingDataSetException:
self.logger.error("Datatype %s has missing data and could not be imported properly." % (datatype,))
os.remove(datatype.get_storage_file_path())
except IntegrityError as excep:
self.logger.exception(excep)
error_msg = "Could not import data with gid: %s. There is already a one with " \
"the same name or gid." % datatype.gid
# Delete file if can't be imported
os.remove(datatype.get_storage_file_path())
raise ProjectImportException(error_msg)
def __populate_project(self, project_path):
"""
Create and store a Project entity.
"""
self.logger.debug("Creating project from path: %s" % project_path)
project_dict = self.files_helper.read_project_metadata(project_path)
project_entity = manager_of_class(model.Project).new_instance()
project_entity = project_entity.from_dict(project_dict, self.user_id)
try:
self.logger.debug("Storing imported project")
return dao.store_entity(project_entity)
except IntegrityError as excep:
self.logger.exception(excep)
error_msg = ("Could not import project: %s with gid: %s. There is already a "
"project with the same name or gid.") % (project_entity.name, project_entity.gid)
raise ProjectImportException(error_msg)
def __build_operation_from_file(self, project, operation_file):
"""
Create Operation entity from metadata file.
"""
operation_dict = XMLReader(operation_file).read_metadata()
operation_entity = manager_of_class(model.Operation).new_instance()
return operation_entity.from_dict(operation_dict, dao, self.user_id, project.gid)
@staticmethod
def __import_operation(operation_entity):
"""
Store a Operation entity.
"""
operation_entity = dao.store_entity(operation_entity)
operation_group_id = operation_entity.fk_operation_group
datatype_group = None
if operation_group_id is not None:
try:
datatype_group = dao.get_datatypegroup_by_op_group_id(operation_group_id)
except SQLAlchemyError:
# If no dataType group present for current op. group, create it.
operation_group = dao.get_operationgroup_by_id(operation_group_id)
datatype_group = model.DataTypeGroup(operation_group, operation_id=operation_entity.id)
datatype_group.state = ADAPTERS['Upload']['defaultdatastate']
datatype_group = dao.store_entity(datatype_group)
return operation_entity, datatype_group
def load_burst_entity(self, json_burst, project_id):
"""
Load BurstConfiguration from JSON (possibly exported from a different machine).
Nothing gets persisted in DB or on disk.
:param json_burst: Burst JSON export
:param project_id: Current project ID (it will be used if the user later starts this simulation)
:return: BurstConfiguration filled from JSON
"""
burst_information = BurstInformation.load_from_dict(json_burst)
burst_entity = model.BurstConfiguration(project_id)
burst_entity.from_dict(burst_information.data)
burst_entity.prepare_after_load()
burst_entity.reset_tabs()
workflow_info = burst_information.get_workflows()[0]
workflow_entity = model.Workflow(project_id, None)
workflow_entity.from_dict(workflow_info.data)
view_steps = workflow_info.get_view_steps()
analyze_steps = workflow_info.get_workflow_steps()
for view_step in view_steps:
try:
algorithm = view_step.get_algorithm()
portlet = view_step.get_portlet()
view_step_entity = model.WorkflowStepView(algorithm.id, portlet_id=portlet.id)
view_step_entity.from_dict(view_step.data)
view_step_entity.workflow = workflow_entity
## For each visualize step, also load all of the analyze steps.
analyzers = []
for an_step in analyze_steps:
if (an_step.data["tab_index"] != view_step_entity.tab_index
or an_step.data["index_in_tab"] != view_step_entity.index_in_tab):
continue
algorithm = an_step.get_algorithm()
wf_step_entity = model.WorkflowStep(algorithm.id)
wf_step_entity.from_dict(an_step.data)
wf_step_entity.workflow = workflow_entity
analyzers.append(wf_step_entity)
portlet = PortletConfiguration(portlet.id)
portlet.set_visualizer(view_step_entity)
portlet.set_analyzers(analyzers)
burst_entity.set_portlet(view_step_entity.tab_index, view_step_entity.index_in_tab, portlet)
except Exception:
# only log exception and ignore this step from loading
self.logger.exception("Could not restore Workflow Step " + view_step.get_algorithm().name)
return burst_entity
class FigureService:
"""
Service layer for Figure entities.
"""
_TYPE_PNG = "png"
_TYPE_SVG = "svg"
_BRANDING_BAR_PNG = os.path.join(os.path.dirname(__file__), "resources", "branding_bar.png")
_BRANDING_BAR_SVG = os.path.join(os.path.dirname(__file__), "resources", "branding_bar.svg")
_DEFAULT_SESSION_NAME = "Default"
_DEFAULT_IMAGE_FILE_NAME = "snapshot."
def __init__(self):
self.logger = get_logger(self.__class__.__module__)
self.file_helper = FilesHelper()
def store_result_figure(self, project, user, img_type, operation_id, export_data):
"""
Store into a file, Result Image and reference in DB.
"""
# Generate path where to store image
store_path = self.file_helper.get_images_folder(project.name, operation_id)
store_path = utils.get_unique_file_name(store_path, FigureService._DEFAULT_IMAGE_FILE_NAME + img_type)[0]
file_path = os.path.split(store_path)[1]
if img_type == FigureService._TYPE_PNG: # PNG file from canvas
imgData = base64.b64decode(export_data) # decode the image
fakeImgFile = StringIO(imgData) # PIL.Image only opens from file, so fake one
origImg = Image.open(fakeImgFile)
brandingBar = Image.open(FigureService._BRANDING_BAR_PNG)
finalSize = (origImg.size[0], # original width
origImg.size[1] + brandingBar.size[1]) # original height + brandingBar height
finalImg = Image.new("RGBA", finalSize)
finalImg.paste(origImg, (0, 0)) # add the original image
finalImg.paste(brandingBar, (0, origImg.size[1])) # add the branding bar, below the original
# the extra width will be discarded
finalImg.save(store_path) # store to disk
elif img_type == FigureService._TYPE_SVG: # SVG file from svg viewer
dom = xml.dom.minidom.parseString(export_data)
figureSvg = dom.getElementsByTagName('svg')[0] # get the original image
dom = xml.dom.minidom.parse(FigureService._BRANDING_BAR_SVG)
brandingSvg = dom.getElementsByTagName('svg')[0] # get the branding bar
brandingSvg.setAttribute("y", figureSvg.getAttribute("height")) # position it below the figure
finalSvg = dom.createElement('svg') # prepare the final svg
width = figureSvg.getAttribute('width').replace('px', '') # same width as original figure
finalSvg.setAttribute("width", width)
height = float(figureSvg.getAttribute('height').replace('px', '')) # increase original height with
height += float(brandingSvg.getAttribute('height').replace('px', '')) # branding bar's height
finalSvg.setAttribute("height", str(height))
finalSvg.appendChild(figureSvg) # add the image
finalSvg.appendChild(brandingSvg) # and the branding bar
# Generate path where to store image
dest = open(store_path, 'w')
finalSvg.writexml(dest) # store to disk
dest.close()
operation = dao.get_operation_by_id(operation_id)
file_name = 'TVB-%s-%s' % (operation.algorithm.name.replace(' ', '-'), operation_id) # e.g. TVB-Algo-Name-352
# Store entity into DB
entity = model.ResultFigure(operation_id, user.id, project.id, FigureService._DEFAULT_SESSION_NAME,
file_name, file_path, img_type)
entity = dao.store_entity(entity)
# Load instance from DB to have lazy fields loaded
figure = dao.load_figure(entity.id)
# Write image meta data to disk
self.file_helper.write_image_metadata(figure)
# Force writing operation meta data on disk.
# This is important later for operation import
self.file_helper.write_operation_metadata(operation)
def retrieve_result_figures(self, project, user, selected_session_name='all_sessions'):
"""
Retrieve from DB all the stored Displayer previews that belongs to the specified session. The
previews are for current user and project; grouped by session.
"""
result, previews_info = dao.get_previews(project.id, user.id, selected_session_name)
for name in result:
for figure in result[name]:
figures_folder = self.file_helper.get_images_folder(project.name, figure.operation.id)
figure_full_path = os.path.join(figures_folder, figure.file_path)
# Compute the path
figure.file_path = utils.path2url_part(figure_full_path)
return result, previews_info
@staticmethod
def load_figure(figure_id):
"""
Loads a stored figure by its id.
"""
return dao.load_figure(figure_id)
def edit_result_figure(self, figure_id, **data):
"""
Retrieve and edit a previously stored figure.
"""
figure = dao.load_figure(figure_id)
figure.session_name = data['session_name']
figure.name = data['name']
dao.store_entity(figure)
# Load instance from DB to have lazy fields loaded.
figure = dao.load_figure(figure_id)
# Store figure meta data in an XML attached to the image.
self.file_helper.write_image_metadata(figure)
def remove_result_figure(self, figure_id):
"""
Remove figure from DB and file storage.
"""
figure = dao.load_figure(figure_id)
# Delete all figure related files from disk.
figures_folder = self.file_helper.get_images_folder(figure.project.name, figure.operation.id)
path2figure = os.path.join(figures_folder, figure.file_path)
if os.path.exists(path2figure):
os.remove(path2figure)
self.file_helper.remove_image_metadata(figure)
# Remove figure reference from DB.
result = dao.remove_entity(model.ResultFigure, figure_id)
return result
