def obj_create(self, bundle, request=None, **kwargs):
"""
Set creator and update full text.
"""
# Because users may have authenticated via headers the request.user may
# not be a full User instance. To be sure, we fetch one.
user = UserProxy.objects.get(id=request.user.id)
bundle = super(DatasetResource, self).obj_create(bundle, request=request, creator=user, **kwargs)
if "columns" in request.GET:
columns = request.GET["columns"].split(",")
else:
columns = None
if "typed_columns" in request.GET:
if not columns:
raise BadRequest(_('The "columns" argument must also be specified when specifying "typed_columns".'))
typed_columns = [True if c.lower() == "true" else False for c in request.GET["typed_columns"].split(",")]
if len(typed_columns) != len(columns):
raise BadRequest(
_(
'The "typed_columns" argument must be a comma-separated list of True/False values with the same number of values as the "columns" argument.'
)
)
else:
typed_columns = None
if "column_types" in request.GET:
if not columns:
raise BadRequest(_('The "columns" argument must also be specified when specifying "column_types".'))
column_types = [None if c.lower() == "" else c.lower() for c in request.GET["column_types"].split(",")]
if len(column_types) != len(columns):
raise BadRequest(
_(
'The "column_types" argument must be a comma-separated list of types with the same number of values as the "columns" argument.'
)
)
else:
column_types = None
if columns:
bundle.obj.column_schema = make_column_schema(columns, typed_columns, column_types)
bundle.obj.save()
# After ALL changes have been made to the object and its relations, update its full text in Solr.
bundle.obj.update_full_text()
return bundle
def import_data(self, user, upload, external_id_field_index=None):
"""
Import data into this ``Dataset`` from a given ``DataUpload``.
"""
self.lock()
try:
if upload.imported:
raise DataImportError(
_('This file has already been imported.'))
task_type = get_import_task_type_for_upload(upload)
if not task_type:
# This is normally caught on the client.
raise DataImportError(
_('This file type is not supported for data import.'))
if self.column_schema:
# This is normally caught on the client.
if upload.columns != [c['name'] for c in self.column_schema]:
raise DataImportError(
_('The columns in this file do not match those in the dataset.'
))
else:
self.column_schema = make_column_schema(
upload.columns, types=upload.guessed_types)
if self.sample_data is None:
self.sample_data = upload.sample_data
# If this is the first import and the API hasn't been used, save that information
if self.initial_upload is None and self.row_count is None:
self.initial_upload = upload
self.current_task = TaskStatus.objects.create(
task_name=task_type.name,
task_description=_('Import data from %(filename)s into %(slug)s.') \
% {'filename': upload.filename, 'slug': self.slug},
creator=user
)
self.save()
task_type.apply_async(
args=[self.slug, upload.id],
kwargs={'external_id_field_index': external_id_field_index},
task_id=self.current_task.id)
except:
self.unlock()
raise
def import_data(self, user, upload, external_id_field_index=None, schema_overrides = {}):
"""
Import data into this ``Dataset`` from a given ``DataUpload``.
"""
self.lock()
try:
if upload.imported:
raise DataImportError(_('This file has already been imported.'))
task_type = get_import_task_type_for_upload(upload)
if not task_type:
# This is normally caught on the client.
raise DataImportError(_('This file type is not supported for data import.'))
if self.column_schema:
# This is normally caught on the client.
if upload.columns != [c['name'] for c in self.column_schema]:
raise DataImportError(_('The columns in this file do not match those in the dataset.'))
else:
self.column_schema = make_column_schema(upload.columns, types=upload.guessed_types, overrides=schema_overrides)
if self.sample_data is None:
self.sample_data = upload.sample_data
# If this is the first import and the API hasn't been used, save that information
if self.initial_upload is None and self.row_count is None:
self.initial_upload = upload
self.current_task = TaskStatus.objects.create(
task_name=task_type.name,
task_description=_('Import data from %(filename)s into %(slug)s.') \
% {'filename': upload.filename, 'slug': self.slug},
creator=user
)
self.save()
task_type.apply_async(
args=[self.slug, upload.id],
kwargs={ 'external_id_field_index': external_id_field_index },
task_id=self.current_task.id
)
except:
self.unlock()
raise
def obj_create(self, bundle, request=None, **kwargs):
"""
Set creator and update full text.
"""
bundle = super(DatasetResource, self).obj_create(bundle, request=request, creator=request.user, **kwargs)
if 'columns' in request.GET:
columns = request.GET['columns'].split(',')
else:
columns = None
if 'typed_columns' in request.GET:
if not columns:
raise BadRequest('The "columns" argument must also be specified when specifying "typed_columns".')
typed_columns = [True if c.lower() == 'true' else False for c in request.GET['typed_columns'].split(',')]
if len(typed_columns) != len(columns):
raise BadRequest('The "typed_columns" argument must be a comma-separated list of True/False values with the same number of values as the "columns" argument.')
else:
typed_columns = None
if 'column_types' in request.GET:
if not columns:
raise BadRequest('The "columns" argument must also be specified when specifying "column_types".')
column_types = [None if c.lower() == '' else c.lower() for c in request.GET['column_types'].split(',')]
if len(column_types) != len(columns):
raise BadRequest('The "column_types" argument must be a comma-separated list of types with the same number of values as the "columns" argument.')
else:
column_types = None
if columns:
bundle.obj.column_schema = make_column_schema(columns, typed_columns, column_types)
bundle.obj.save()
# After ALL changes have been made to the object and its relations, update its full text in Solr.
bundle.obj.update_full_text()
return bundle
def obj_create(self, bundle, request=None, **kwargs):
"""
Set creator and update full text.
"""
# Because users may have authenticated via headers the request.user may
# not be a full User instance. To be sure, we fetch one.
user = UserProxy.objects.get(id=request.user.id)
bundle = super(DatasetResource, self).obj_create(bundle,
request=request,
creator=user,
**kwargs)
if 'columns' in request.GET:
columns = request.GET['columns'].split(',')
else:
columns = None
if 'typed_columns' in request.GET:
if not columns:
raise BadRequest(
'The "columns" argument must also be specified when specifying "typed_columns".'
)
typed_columns = [
True if c.lower() == 'true' else False
for c in request.GET['typed_columns'].split(',')
]
if len(typed_columns) != len(columns):
raise BadRequest(
'The "typed_columns" argument must be a comma-separated list of True/False values with the same number of values as the "columns" argument.'
)
else:
typed_columns = None
if 'column_types' in request.GET:
if not columns:
raise BadRequest(
'The "columns" argument must also be specified when specifying "column_types".'
)
column_types = [
None if c.lower() == '' else c.lower()
for c in request.GET['column_types'].split(',')
]
if len(column_types) != len(columns):
raise BadRequest(
'The "column_types" argument must be a comma-separated list of types with the same number of values as the "columns" argument.'
)
else:
column_types = None
if columns:
bundle.obj.column_schema = make_column_schema(
columns, typed_columns, column_types)
bundle.obj.save()
# After ALL changes have been made to the object and its relations, update its full text in Solr.
bundle.obj.update_full_text()
return bundle
