Ordering in get_tasks crashes
konradmalik opened this issue · 1 comments
konradmalik commented
Hi, I encountered this bug today.
This works ok:
def _get_tasks(
self,
project_client: Project,
larger_than_id: int,
only_without_predictions: bool,
) -> List[JSONDict]:
"""
Download all labeled tasks from project using the Label Studio SDK.
Read more about SDK here https://labelstud.io/sdk/
project_id: project ID
only_without_predictions: get only those tasks that do not have predictions yet
larger_than_id: filters results to tasks GREATER than this id
:return:
"""
items = [
Filters.item(
name=Column.id,
operator=Operator.GREATER,
column_type=Type.Number,
value=Filters.value(larger_than_id),
)
]
if only_without_predictions:
items.append(
Filters.item(
name=Column.total_predictions,
operator=Operator.EQUAL,
column_type=Type.Number,
value=Filters.value(0),
)
)
filters = Filters.create(
conjunction="and",
items=items,
)
return sorted(
project_client.get_tasks(filters=filters), key=lambda x: x["id"]
)While this (adding seemingly proper ordering) crashes:
def _get_tasks(
self,
project_client: Project,
larger_than_id: int,
only_without_predictions: bool,
) -> List[JSONDict]:
"""
Download all labeled tasks from project using the Label Studio SDK.
Read more about SDK here https://labelstud.io/sdk/
project_id: project ID
only_without_predictions: get only those tasks that do not have predictions yet
larger_than_id: filters results to tasks GREATER than this id
:return:
"""
items = [
Filters.item(
name=Column.id,
operator=Operator.GREATER,
column_type=Type.Number,
value=Filters.value(larger_than_id),
)
]
if only_without_predictions:
items.append(
Filters.item(
name=Column.total_predictions,
operator=Operator.EQUAL,
column_type=Type.Number,
value=Filters.value(0),
)
)
filters = Filters.create(
conjunction="and",
items=items,
)
ordering = [Column.id]
return project_client.get_tasks(filters=filters, ordering=ordering)Looks like the column from the ordering arguments gets badly translated along the road.
Versions:
label-studio-ml == 1.0.4
label-studio-sdk == 0.0.8
label-studio server == 1.4.0
This is the exception:
studio | [01/Mar/2022 14:58:21] "GET /api/projects/2 HTTP/1.1" 200 10822 [788/7125]
studio | [2022-03-01 14:58:21,851] [core.utils.common::custom_exception_handler::83] [ERROR] d24ed865-9792-4971-b8ea-9b45d8d207d2 Cannot resolve keyword 'filter:id' into field. Choices
are: annotations, created_at, data, drafts, file_upload, file_upload_id, id, io_storages_azureblobimportstoragelink, io_storages_gcsimportstoragelink, io_storages_localfilesimportstora
gelink, io_storages_redisimportstoragelink, io_storages_s3importstoragelink, is_labeled, locks, meta, overlap, predictions, project, project_id, total_predictions, updated_at
studio | Traceback (most recent call last):
studio | File "/usr/local/lib/python3.8/dist-packages/rest_framework/views.py", line 506, in dispatch
studio | response = handler(request, *args, **kwargs)
studio | File "/label-studio/label_studio/data_manager/api.py", line 230, in get
studio | page = self.paginate_queryset(queryset)
studio | File "/usr/local/lib/python3.8/dist-packages/rest_framework/generics.py", line 171, in paginate_queryset
studio | return self.paginator.paginate_queryset(queryset, self.request, view=self)
studio | File "/label-studio/label_studio/data_manager/api.py", line 138, in paginate_queryset
studio | return super().paginate_queryset(queryset, request, view)
studio | File "/usr/local/lib/python3.8/dist-packages/rest_framework/pagination.py", line 216, in paginate_queryset
studio | return list(self.page)
studio | File "/usr/local/lib/python3.8/dist-packages/django/core/paginator.py", line 143, in __len__
studio | return len(self.object_list)
studio | File "/usr/local/lib/python3.8/dist-packages/django/db/models/query.py", line 269, in __len__
studio | self._fetch_all()
studio | File "/usr/local/lib/python3.8/dist-packages/django/db/models/query.py", line 1308, in _fetch_all
studio | self._result_cache = list(self._iterable_class(self))
studio | File "/usr/local/lib/python3.8/dist-packages/django/db/models/query.py", line 53, in __iter__
studio | results = compiler.execute_sql(chunked_fetch=self.chunked_fetch, chunk_size=self.chunk_size)
studio | File "/usr/local/lib/python3.8/dist-packages/django/db/models/sql/compiler.py", line 1143, in execute_sql
studio | sql, params = self.as_sql()
studio | File "/usr/local/lib/python3.8/dist-packages/django/db/models/sql/compiler.py", line 498, in as_sql
studio | extra_select, order_by, group_by = self.pre_sql_setup()
studio | File "/usr/local/lib/python3.8/dist-packages/django/db/models/sql/compiler.py", line 56, in pre_sql_setup
studio | order_by = self.get_order_by()
studio | File "/usr/local/lib/python3.8/dist-packages/django/db/models/sql/compiler.py", line 361, in get_order_by
studio | resolved = expr.resolve_expression(self.query, allow_joins=True, reuse=None)
studio | File "/usr/local/lib/python3.8/dist-packages/django/db/models/expressions.py", line 249, in resolve_expression
studio | c.set_source_expressions([
studio | File "/usr/local/lib/python3.8/dist-packages/django/db/models/expressions.py", line 250, in <listcomp>
studio | expr.resolve_expression(query, allow_joins, reuse, summarize)
studio | File "/usr/local/lib/python3.8/dist-packages/django/db/models/expressions.py", line 539, in resolve_expression
studio | return query.resolve_ref(self.name, allow_joins, reuse, summarize)
studio | File "/usr/local/lib/python3.8/dist-packages/django/db/models/sql/query.py", line 1685, in resolve_ref
studio | join_info = self.setup_joins(field_list, self.get_meta(), self.get_initial_alias(), can_reuse=reuse)
studio | File "/usr/local/lib/python3.8/dist-packages/django/db/models/sql/query.py", line 1567, in setup_joins
studio | path, final_field, targets, rest = self.names_to_path(
studio | File "/usr/local/lib/python3.8/dist-packages/django/db/models/sql/query.py", line 1483, in names_to_path
studio | raise FieldError("Cannot resolve keyword '%s' into field. "
studio | django.core.exceptions.FieldError: Cannot resolve keyword 'filter:id' into field. Choices are: annotations, created_at, data, drafts, file_upload, file_upload_id, id, io_stora
ges_azureblobimportstoragelink, io_storages_gcsimportstoragelink, io_storages_localfilesimportstoragelink, io_storages_redisimportstoragelink, io_storages_s3importstoragelink, is_labele
d, locks, meta, overlap, predictions, project, project_id, total_predictions, updated_at
studio | [2022-03-01 14:58:21,871] [django.request::log_response::224] [ERROR] Internal Server Error: /api/dm/tasks/
studio | [01/Mar/2022 14:58:21] "GET /api/dm/tasks?project=2&page=1&page_size=-1&query=%7B%22filters%22%3A+%7B%22conjunction%22%3A+%22and%22%2C+%22items%22%3A+%5B%7B%22filter%22%3A+%22
filter%3Atasks%3Aid%22%2C+%22operator%22%3A+%22greater%22%2C+%22type%22%3A+%22Number%22%2C+%22value%22%3A+-1%7D%2C+%7B%22filter%22%3A+%22filter%3Atasks%3Atotal_predictions%22%2C+%22oper
ator%22%3A+%22equal%22%2C+%22type%22%3A+%22Number%22%2C+%22value%22%3A+0%7D%5D%7D%2C+%22ordering%22%3A+%5B%22filter%3Atasks%3Aid%22%5D%2C+%22selectedItems%22%3A+%7B%22all%22%3A+true%2C+
%22excluded%22%3A+%5B%5D%7D%7D&fields=all HTTP/1.1" 500 4091
makseq commented
Please, check the latest master branch of SDK with the latest LS develop branch. This issue is fixed there.