def _projects_filter_for_non_privileged_users(user: User,
queryset: QuerySet,
project_relation: str,
action: str = 'create'):
if not user.is_anonymous and (user.is_admin or user.is_superuser):
return queryset.distinct()
# Construct the public projects filter field lookup.
project_filter = project_relation + 'is_public'
# Filter the object list into two querysets;
# One where the related Projects are public and one where they are private
public_objects = queryset.filter(**{project_filter: True}).distinct()
private_objects = queryset.filter(**{project_filter: False}).distinct()
# In case of an anonymous user, only return the public objects
if user.is_anonymous:
unpublished_exclude = project_relation + 'publishingstatus__status'
queryset = public_objects.exclude(
**{
unpublished_exclude: PublishingStatus.STATUS_UNPUBLISHED
}).distinct()
# Otherwise, check to which objects the user has (change) permission
elif private_objects.exists():
include_user_owned = hasattr(queryset.model, 'user')
if action == 'list':
# The view permission is new, and previously only the change
# permission existed. To avoid adding new view permissions for
# all the objects, we also check if a user has change
# permissions, which implicitly implies view permissions.
change_permission = type(
private_objects[0])._meta.db_table.replace('_', '.change_')
change_filter = user.get_permission_filter(change_permission,
project_relation,
include_user_owned)
change_objects = private_objects.filter(change_filter).distinct()
# Check if user has view permission on the queryset
view_permission = change_permission.replace('.change_', '.view_')
view_filter = user.get_permission_filter(view_permission,
project_relation,
include_user_owned)
view_objects = private_objects.filter(view_filter).distinct()
private_objects = (change_objects | view_objects).distinct()
else:
permission = type(private_objects[0])._meta.db_table.replace(
'_', '.change_')
filter_ = user.get_permission_filter(permission, project_relation,
include_user_owned)
private_objects = private_objects.filter(filter_).distinct()
queryset = public_objects | private_objects
return queryset.distinct()
def _filter_queryset(self, queryset: models.QuerySet, prefix: str):
"""Internal logic to filter a queryset on request-parameters for a temporal slice."""
if not self.is_versioned or self.slice_value == "*":
# allow this method to be called unconditionally on objects,
# and allow ?geldigOp=* to return all data
return queryset
# Either take a given ?geldigOp=yyyy-mm-dd OR ?geldigOp=NOW()
slice_value = self.slice_value or now()
range_fields = self.slice_range_fields or self.default_range_fields
if range_fields is not None:
# start <= value AND (end IS NULL OR value < end)
start, end = map(to_snake_case, range_fields)
return queryset.filter(
Q(**{f"{prefix}{start}__lte": slice_value})
& (
Q(**{f"{prefix}{end}__isnull": True})
| Q(**{f"{prefix}{end}__gt": slice_value})
)
).order_by(f"-{prefix}{start}")
else:
# Last attempt to get only the current temporal record; order by sequence.
# does SELECT DISTINCT ON(identifier) ... ORDER BY identifier, sequence DESC
# Alternative would be something like `HAVING sequence = MAX(SELECT sequence FROM ..)`
identifier = self.table_schema.identifier[0] # from ["identificatie", "volgnummer"]
sequence_name = self.table_schema.temporal.identifier
return queryset.distinct(
*queryset.query.distinct_fields, f"{prefix}{identifier}"
).order_by(f"{prefix}{identifier}", f"-{prefix}{sequence_name}")
def get_fields_by_rain_date(date: date,
fields_qs: QuerySet = None) -> QuerySet:
if fields_qs is None:
fields_qs = get_fields()
fields_qs = fields_qs.filter(rain__date__gte=date)
return fields_qs.distinct()
def get_fields_by_rain_quantity(quantity: float,
fields_qs: QuerySet = None) -> QuerySet:
if fields_qs is None:
fields_qs = get_fields()
fields_qs = fields_qs.filter(rain__quantity__gte=quantity)
return fields_qs.distinct()
def export_raw_accesslogs_to_stream_lowlevel(self, stream: IO,
queryset: QuerySet):
text_id_to_text = {
dt['id']: dt['text']
for dt in DimensionText.objects.all().values('id', 'text')
}
rt_to_dimensions = {
rt.pk: rt.dimensions_sorted
for rt in ReportType.objects.filter(pk__in=queryset.distinct(
'report_type_id').values('report_type_id'))
}
# get all field names for the CSV
field_name_map = {(f'{dim}__{attr}' if attr else dim): dim
for dim, attr in self.implicit_dims.items()}
field_name_map.update(
{f'target__{attr}': attr
for attr in self.title_attrs})
field_names = list(field_name_map.values())
for tr, dims in rt_to_dimensions.items():
field_names += [
dim.short_name for dim in dims
if dim.short_name not in field_names
]
field_names.append('value')
# values that will be retrieved from the accesslogs
values = ['value', 'report_type_id']
values += list(field_name_map.keys())
values += [f'dim{i+1}' for i in range(7)]
# crate the writer
writer = csv.DictWriter(stream, field_names)
writer.writeheader()
# write the records
for rec_num, log in enumerate(
queryset.values(*values).iterator()): # type: int, dict
record = {
attr_out: log.get(attr_in)
for attr_in, attr_out in field_name_map.items()
}
record['value'] = log['value']
record['date'] = log['date']
for i, dim in enumerate(rt_to_dimensions[log['report_type_id']]):
value = log.get(f'dim{i+1}')
if dim.type == dim.TYPE_TEXT:
record[dim.short_name] = text_id_to_text.get(value, value)
else:
record[dim.short_name] = value
writer.writerow(record)
if rec_num % 999 == 0:
self.store_progress(rec_num + 1)
self.store_progress(rec_num + 1)
def build_queryset(self, queryset: models.QuerySet) -> models.QuerySet:
if self.query:
queryset = self.query.apply_to_queryset(queryset)
if self.order_by:
queryset = queryset.order_by(*self.order_by)
if self.distinct:
queryset = queryset.distinct(*self.distinct)
if self.prefetch_trees:
select_related = []
for prefetch_tree in self.prefetch_trees:
tree_select_related = self._flatten_prefetch_tree(
prefetch_tree=prefetch_tree)
if tree_select_related is not None:
select_related += tree_select_related
queryset = queryset.select_related(
*[sr for sr in select_related if sr is not None])
return queryset
def filter_list(queryset: QuerySet, request: Request, name_filter: str,
name_field: str) -> QuerySet:
items_selected = request.query_params.getlist(
'{name_filter}[]'.format(name_filter=name_filter))
items_selected = [name.upper() for name in items_selected]
if len(items_selected) > 0:
if json.loads(
request.query_params.get(
'{name_filter}Exclude'.format(name_filter=name_filter),
'false')):
queryset = queryset.exclude(
**{
'{name_field}__in'.format(name_field=name_field):
items_selected
})
else:
queryset = queryset.filter(
**{
'{name_field}__in'.format(name_field=name_field):
items_selected
})
return queryset.distinct()
def filter_queryset(
self, queryset: QuerySet, feature_type: FeatureType
) -> QuerySet:
"""Apply the filters and lookups to the queryset.
:param queryset: The queryset to filter.
:param feature_type: The feature type that the queryset originated from.
"""
if self.is_empty:
return queryset.none()
if self.extra_lookups:
# Each time an expression node calls add_extra_lookup(),
# the parent should have used apply_extra_lookups()
raise RuntimeError("apply_extra_lookups() was not called")
# All are applied at once.
if self.annotations:
queryset = queryset.annotate(**self.annotations)
lookups = self.lookups
try:
lookups += self.typed_lookups[feature_type.name]
except KeyError:
pass
if lookups:
queryset = queryset.filter(*lookups)
if self.ordering:
queryset = queryset.order_by(*self.ordering)
if self.distinct:
queryset = queryset.distinct()
return queryset
def build_queryset(queryset: QuerySet,
params: QueryDict) -> Tuple[QuerySet, bool]:
filtered = False
name = params.get('name', None)
if name is not None:
queryset = queryset.filter(name__icontains=name)
filtered = True
last_name = params.get('last_name', None)
if last_name is not None:
queryset = queryset.filter(last_name__icontains=last_name)
filtered = True
naming = params.get('naming', None)
if naming == SubjectSegment.NAMING_NAMED:
queryset = queryset.exclude(name='', last_name='')
filtered = True
elif naming == SubjectSegment.NAMING_UNNAMED:
queryset = queryset.filter(name='', last_name='')
filtered = True
tasks = params.getlist('tasks', None)
if tasks is not None and len(tasks):
queryset = queryset.filter(faces__task__in=tasks)
filtered = True
tasks_tags = params.getlist('tasks_tags', None)
if tasks_tags is not None and len(tasks_tags):
queryset = queryset.filter(faces__task__tags__in=tasks_tags)
filtered = True
min_time = params.get('min_time', None)
if min_time is not None:
queryset = queryset.filter(faces__created_at__time__gte=min_time)
filtered = True
max_time = params.get('max_time', None)
if max_time is not None:
queryset = queryset.filter(faces__created_at__time__lte=max_time)
filtered = True
min_date = params.get('min_date', None)
if min_date is not None:
queryset = queryset.filter(faces__created_at__date__gte=min_date)
filtered = True
max_date = params.get('max_date', None)
if max_date is not None:
queryset = queryset.filter(faces__created_at__date__lte=max_date)
filtered = True
max_age = params.get('max_age', None)
if max_age is not None:
try:
max_age = int(max_age)
except ValueError:
pass
else:
queryset = queryset.filter(
birthdate__gt=Subject.birthdate_from_age(max_age))
filtered = True
min_age = params.get('min_age', None)
if min_age is not None:
try:
min_age = int(min_age)
except ValueError:
pass
else:
queryset = queryset.filter(
birthdate__lt=Subject.birthdate_from_age(min_age))
filtered = True
min_pred_age = params.get('min_pred_age', None)
if min_pred_age is not None:
try:
min_pred_age = int(min_pred_age)
except ValueError:
pass
else:
queryset = queryset.filter(pred_age__gte=min_pred_age)
filtered = True
max_pred_age = params.get('max_pred_age', None)
if max_pred_age is not None:
try:
max_pred_age = int(max_pred_age)
except ValueError:
pass
else:
queryset = queryset.filter(pred_age__lte=max_pred_age)
filtered = True
sex = params.get('sex', None)
if sex is not None:
queryset = queryset.filter(sex=sex)
filtered = True
pred_sex = params.get('pred_sex', None)
if pred_sex is not None:
queryset = queryset.filter(pred_sex=pred_sex)
filtered = True
skin = params.get('skin', None)
if skin is not None:
queryset = queryset.filter(skin=skin)
filtered = True
order_by = params.get('order_by', None)
if order_by is not None:
queryset = queryset.order_by(order_by)
if filtered:
queryset = queryset.distinct()
return queryset, filtered
def export_raw_accesslogs_to_stream_lowlevel(self, stream: IO,
queryset: QuerySet):
start = monotonic()
text_id_to_text = {
dt['id']: dt['text']
for dt in DimensionText.objects.all().values('id', 'text')
}
logger.debug('Finished loading text remaps: %.2f s',
monotonic() - start)
rt_to_dimensions = {
rt.pk: rt.dimensions_sorted
for rt in ReportType.objects.filter(pk__in=queryset.distinct(
'report_type_id').values('report_type_id'))
}
logger.debug('Finished loading report_types and dimensions: %.2f s',
monotonic() - start)
# get all field names for the CSV
field_name_map = {(f'{dim}__{attr}' if attr else dim): dim
for dim, attr in self.implicit_dims.items()}
field_name_map.update(
{f'target__{attr}': attr
for attr in self.title_attrs})
field_names = list(field_name_map.values())
for tr, dims in rt_to_dimensions.items():
field_names += [
dim.short_name for dim in dims
if dim.short_name not in field_names
]
field_names.append('value')
logger.debug('Finished preparing field names: %.2f s',
monotonic() - start)
# values that will be retrieved from the accesslogs
values = ['value', 'report_type_id']
values += list(field_name_map.keys())
values += [f'dim{i+1}' for i in range(7)]
# crate the writer
writer = csv.DictWriter(stream, field_names)
writer.writeheader()
logger.debug('Finished preparing CSV writer: %.2f s',
monotonic() - start)
# write the records
rec_num = 0
with cachalot_disabled(True):
# disable cachalot for this query because it returns a potentially huge number of records
# and would clog the cache
for rec_num, log in enumerate(
queryset.values(*values).iterator()): # type: int, dict
record = {
attr_out: log.get(attr_in)
for attr_in, attr_out in field_name_map.items()
}
record['value'] = log['value']
record['date'] = log['date']
for i, dim in enumerate(
rt_to_dimensions[log['report_type_id']]):
value = log.get(f'dim{i+1}')
if dim.type == dim.TYPE_TEXT:
record[dim.short_name] = text_id_to_text.get(
value, value)
else:
record[dim.short_name] = value
writer.writerow(record)
if rec_num % 999 == 0:
self.store_progress(rec_num + 1)
if rec_num % 99999 == 0:
logger.debug('Stored %d records: %.2f s', rec_num,
monotonic() - start)
self.store_progress(rec_num + 1)
def bulk_updater(self,
queryset: QuerySet,
update_fields: Optional[Set[str]] = None,
return_pks: bool = False,
local_only: bool = False,
querysize: Optional[int] = None) -> Optional[Set[Any]]:
"""
Update local computed fields and descent in the dependency tree by calling
``update_dependent`` for dependent models.
This method does the local field updates on `queryset`:
- eval local `MRO` of computed fields
- expand `update_fields`
- apply optional `select_related` and `prefetch_related` rules to `queryset`
- walk all records and recalculate fields in `update_fields`
- aggregate changeset and save as batched `bulk_update` to the database
By default this method triggers the update of dependent models by calling
``update_dependent`` with `update_fields` (next level of tree traversal).
This can be suppressed by setting `local_only=True`.
If `return_pks` is set, the method returns a set of altered pks of `queryset`.
"""
model: Type[Model] = queryset.model
# distinct issue workaround
# the workaround is needed for already sliced/distinct querysets coming from outside
# TODO: distinct is a major query perf smell, and is in fact only needed on back relations
# may need some rework in _querysets_for_update
# ideally we find a way to avoid it for forward relations
# also see #101
if queryset.query.can_filter() and not queryset.query.distinct_fields:
queryset = queryset.distinct()
else:
queryset = model.objects.filter(
pk__in=subquery_pk(queryset, queryset.db))
# correct update_fields by local mro
mro = self.get_local_mro(model, update_fields)
fields: Any = set(
mro) # FIXME: narrow type once issue in django-stubs is resolved
if update_fields:
update_fields.update(fields)
select = self.get_select_related(model, fields)
prefetch = self.get_prefetch_related(model, fields)
if select:
queryset = queryset.select_related(*select)
if prefetch:
queryset = queryset.prefetch_related(*prefetch)
pks = []
if fields:
q_size = self.get_querysize(model, fields, querysize)
change: List[Model] = []
for elem in slice_iterator(queryset, q_size):
# note on the loop: while it is technically not needed to batch things here,
# we still prebatch to not cause memory issues for very big querysets
has_changed = False
for comp_field in mro:
new_value = self._compute(elem, model, comp_field)
if new_value != getattr(elem, comp_field):
has_changed = True
setattr(elem, comp_field, new_value)
if has_changed:
change.append(elem)
pks.append(elem.pk)
if len(change) >= self._batchsize:
self._update(queryset, change, fields)
change = []
if change:
self._update(queryset, change, fields)
# trigger dependent comp field updates from changed records
# other than before we exit the update tree early, if we have no changes at all
# also cuts the update tree for recursive deps (tree-like)
if not local_only and pks:
self.update_dependent(model.objects.filter(pk__in=pks),
model,
fields,
update_local=False)
return set(pks) if return_pks else None
