nucliadb 2.46.1.post382__py3-none-any.whl → 6.2.1.post2777__py3-none-any.whl

nucliadb/train/tests/test_get_entities.py

@@ -1,80 +0,0 @@
-# Copyright (C) 2021 Bosutech XXI S.L.
-#
-# nucliadb is offered under the AGPL v3.0 and as commercial software.
-# For commercial licensing, contact us at info@nuclia.com.
-#
-# AGPL:
-# This program is free software: you can redistribute it and/or modify
-# it under the terms of the GNU Affero General Public License as
-# published by the Free Software Foundation, either version 3 of the
-# License, or (at your option) any later version.
-#
-# This program is distributed in the hope that it will be useful,
-# but WITHOUT ANY WARRANTY; without even the implied warranty of
-# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
-# GNU Affero General Public License for more details.
-#
-# You should have received a copy of the GNU Affero General Public License
-# along with this program. If not, see <http://www.gnu.org/licenses/>.
-#
-import uuid
-from unittest.mock import AsyncMock, Mock
-
-import pytest
-from nucliadb_protos.train_pb2 import GetEntitiesRequest
-from nucliadb_protos.train_pb2_grpc import TrainStub
-from nucliadb_protos.writer_pb2 import GetEntitiesResponse
-
-
-@pytest.fixture(scope="function")
-async def entities_manager_mock():
-    from nucliadb.train import nodes
-
-    original = nodes.EntitiesManager
-
-    mock = Mock()
-    nodes.EntitiesManager = Mock(return_value=mock)
-
-    yield mock
-
-    nodes.EntitiesManager = original
-
-
-@pytest.mark.asyncio
-async def test_get_entities(
-    train_client: TrainStub,
-    knowledgebox_ingest: str,
-    entities_manager_mock: Mock,
-) -> None:
-    def get_entities_mock(response):
-        response.groups["group1"].entities["entity1"].value = "PERSON"
-
-    entities_manager_mock.get_entities = AsyncMock(side_effect=get_entities_mock)
-
-    req = GetEntitiesRequest()
-    req.kb.uuid = knowledgebox_ingest
-    entities: GetEntitiesResponse = await train_client.GetEntities(req)  # type: ignore
-
-    assert entities.groups["group1"].entities["entity1"].value == "PERSON"
-
-
-@pytest.mark.asyncio
-async def test_get_entities_kb_not_found(train_client: TrainStub) -> None:
-    req = GetEntitiesRequest()
-    req.kb.uuid = str(uuid.uuid4())
-    entities: GetEntitiesResponse = await train_client.GetEntities(req)  # type: ignore
-    assert entities.status == GetEntitiesResponse.Status.NOTFOUND
-
-
-@pytest.mark.asyncio
-async def test_get_entities_error(
-    train_client: TrainStub, knowledgebox_ingest: str, entities_manager_mock
-) -> None:
-    entities_manager_mock.get_entities = AsyncMock(
-        side_effect=Exception("Testing exception on ingest")
-    )
-
-    req = GetEntitiesRequest()
-    req.kb.uuid = knowledgebox_ingest
-    entities: GetEntitiesResponse = await train_client.GetEntities(req)  # type: ignore
-    assert entities.status == GetEntitiesResponse.Status.ERROR

nucliadb/train/tests/test_get_info.py

@@ -1,51 +0,0 @@
-# Copyright (C) 2021 Bosutech XXI S.L.
-#
-# nucliadb is offered under the AGPL v3.0 and as commercial software.
-# For commercial licensing, contact us at info@nuclia.com.
-#
-# AGPL:
-# This program is free software: you can redistribute it and/or modify
-# it under the terms of the GNU Affero General Public License as
-# published by the Free Software Foundation, either version 3 of the
-# License, or (at your option) any later version.
-#
-# This program is distributed in the hope that it will be useful,
-# but WITHOUT ANY WARRANTY; without even the implied warranty of
-# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
-# GNU Affero General Public License for more details.
-#
-# You should have received a copy of the GNU Affero General Public License
-# along with this program. If not, see <http://www.gnu.org/licenses/>.
-
-import sys
-
-import pytest
-from aioresponses import aioresponses
-from nucliadb_protos.train_pb2 import GetInfoRequest, TrainInfo
-from nucliadb_protos.train_pb2_grpc import TrainStub
-
-VERSION = sys.version_info
-PY_GEQ_3_11 = VERSION.major > 3 or VERSION.major == 3 and VERSION.minor >= 11
-
-
-@pytest.mark.asyncio
-@pytest.mark.skipif(
-    PY_GEQ_3_11, reason="aioresponses not compatible with python 3.11 yet"
-)
-async def test_get_info(
-    train_client: TrainStub, knowledgebox_ingest: str, test_pagination_resources
-) -> None:
-    req = GetInfoRequest()
-    req.kb.uuid = knowledgebox_ingest
-
-    with aioresponses() as m:
-        m.get(
-            f"http://search.nuclia.svc.cluster.local:8030/api/v1/kb/{knowledgebox_ingest}/counters",
-            payload={"resources": 4, "paragraphs": 89, "fields": 4, "sentences": 90},
-        )
-
-        labels: TrainInfo = await train_client.GetInfo(req)  # type: ignore
-    assert labels.fields == 4
-    assert labels.resources == 4
-    assert labels.paragraphs == 89
-    assert labels.sentences == 90

nucliadb/train/tests/test_get_ontology.py

@@ -1,34 +0,0 @@
-# Copyright (C) 2021 Bosutech XXI S.L.
-#
-# nucliadb is offered under the AGPL v3.0 and as commercial software.
-# For commercial licensing, contact us at info@nuclia.com.
-#
-# AGPL:
-# This program is free software: you can redistribute it and/or modify
-# it under the terms of the GNU Affero General Public License as
-# published by the Free Software Foundation, either version 3 of the
-# License, or (at your option) any later version.
-#
-# This program is distributed in the hope that it will be useful,
-# but WITHOUT ANY WARRANTY; without even the implied warranty of
-# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
-# GNU Affero General Public License for more details.
-#
-# You should have received a copy of the GNU Affero General Public License
-# along with this program. If not, see <http://www.gnu.org/licenses/>.
-#
-import pytest
-from nucliadb_protos.train_pb2 import GetLabelsRequest
-from nucliadb_protos.train_pb2_grpc import TrainStub
-from nucliadb_protos.writer_pb2 import GetLabelsResponse
-
-
-@pytest.mark.asyncio
-async def test_get_ontology(
-    train_client: TrainStub, knowledgebox_ingest: str, test_pagination_resources
-) -> None:
-    req = GetLabelsRequest()
-    req.kb.uuid = knowledgebox_ingest
-
-    labels: GetLabelsResponse = await train_client.GetOntology(req)  # type: ignore
-    assert labels.labels.labelset["label1"].labels[0].title == "label1"

nucliadb/train/tests/test_get_ontology_count.py

@@ -1,63 +0,0 @@
-# Copyright (C) 2021 Bosutech XXI S.L.
-#
-# nucliadb is offered under the AGPL v3.0 and as commercial software.
-# For commercial licensing, contact us at info@nuclia.com.
-#
-# AGPL:
-# This program is free software: you can redistribute it and/or modify
-# it under the terms of the GNU Affero General Public License as
-# published by the Free Software Foundation, either version 3 of the
-# License, or (at your option) any later version.
-#
-# This program is distributed in the hope that it will be useful,
-# but WITHOUT ANY WARRANTY; without even the implied warranty of
-# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
-# GNU Affero General Public License for more details.
-#
-# You should have received a copy of the GNU Affero General Public License
-# along with this program. If not, see <http://www.gnu.org/licenses/>.
-#
-import sys
-
-import pytest
-from aioresponses import aioresponses
-from nucliadb_protos.train_pb2 import GetLabelsetsCountRequest, LabelsetsCount
-from nucliadb_protos.train_pb2_grpc import TrainStub
-
-VERSION = sys.version_info
-PY_GEQ_3_11 = VERSION.major > 3 or VERSION.major == 3 and VERSION.minor >= 11
-
-
-@pytest.mark.asyncio
-@pytest.mark.skipif(
-    PY_GEQ_3_11, reason="aioresponses not compatible with python 3.11 yet"
-)
-async def test_get_ontology_count(
-    train_client: TrainStub, knowledgebox_ingest: str, test_pagination_resources
-) -> None:
-    req = GetLabelsetsCountRequest()
-    req.kb.uuid = knowledgebox_ingest
-
-    with aioresponses() as m:
-        m.get(
-            f"http://search.nuclia.svc.cluster.local:8030/api/v1/kb/{knowledgebox_ingest}/search?faceted=/l/my-labelset",  # noqa
-            payload={
-                "resources": {},
-                "sentences": {"results": [], "facets": {}},
-                "paragraphs": {
-                    "results": [],
-                    "facets": {
-                        "/l/my-labelset": {
-                            "facetresults": [
-                                {"tag": "/l/my-labelset/Label 1", "total": 1}
-                            ]
-                        }
-                    },
-                },
-                "fulltext": {"results": [], "facets": {}},
-            },
-        )
-
-        req.resource_labelsets.append("my-labelset")
-        labels: LabelsetsCount = await train_client.GetOntologyCount(req)  # type: ignore
-    assert labels.labelsets["/l/my-labelset"].paragraphs["Label 1"] == 1

nucliadb/train/tests/test_image_classification.py

@@ -1,222 +0,0 @@
-# Copyright (C) 2021 Bosutech XXI S.L.
-#
-# nucliadb is offered under the AGPL v3.0 and as commercial software.
-# For commercial licensing, contact us at info@nuclia.com.
-#
-# AGPL:
-# This program is free software: you can redistribute it and/or modify
-# it under the terms of the GNU Affero General Public License as
-# published by the Free Software Foundation, either version 3 of the
-# License, or (at your option) any later version.
-#
-# This program is distributed in the hope that it will be useful,
-# but WITHOUT ANY WARRANTY; without even the implied warranty of
-# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
-# GNU Affero General Public License for more details.
-#
-# You should have received a copy of the GNU Affero General Public License
-# along with this program. If not, see <http://www.gnu.org/licenses/>.
-
-import asyncio
-import base64
-import json
-import os
-from typing import Any
-from unittest.mock import AsyncMock, patch
-
-import aiohttp
-import pytest
-from nucliadb_protos.dataset_pb2 import ImageClassificationBatch, TaskType, TrainSet
-from nucliadb_protos.resources_pb2 import (
-    CloudFile,
-    FileExtractedData,
-    FilePages,
-    PageStructure,
-    PageStructurePage,
-    PageStructureToken,
-)
-from nucliadb_protos.writer_pb2 import BrokerMessage, OpStatusWriter
-from nucliadb_protos.writer_pb2_grpc import WriterStub
-
-from nucliadb.train import API_PREFIX
-from nucliadb.train.api.v1.router import KB_PREFIX
-from nucliadb.train.tests.utils import get_batches_from_train_response_stream
-from nucliadb_utils.utilities import Utility, get_utility, set_utility
-
-_dir = os.path.dirname(__file__)
-_testdata_dir = os.path.join(_dir, "..", "..", "tests", "testdata")
-
-INVOICE_FILENAME = os.path.join(_testdata_dir, "invoice.pdf")
-INVOICE_SELECTIONS_FILENAME = os.path.join(_testdata_dir, "invoice_selections.json")
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize("knowledgebox", ["STABLE", "EXPERIMENTAL"], indirect=True)
-async def test_generation_image_classification(
-    train_rest_api: aiohttp.ClientSession,
-    knowledgebox: str,
-    image_classification_resource,
-):
-    async with train_rest_api.get(
-        f"/{API_PREFIX}/v1/{KB_PREFIX}/{knowledgebox}/trainset"
-    ) as partitions:
-        assert partitions.status == 200
-        data = await partitions.json()
-        assert len(data["partitions"]) == 1
-        partition_id = data["partitions"][0]
-
-    trainset = TrainSet()
-    trainset.type = TaskType.IMAGE_CLASSIFICATION
-    trainset.batch_size = 10
-
-    await asyncio.sleep(0.1)
-    async with train_rest_api.post(
-        f"/{API_PREFIX}/v1/{KB_PREFIX}/{knowledgebox}/trainset/{partition_id}",
-        data=trainset.SerializeToString(),
-    ) as response:
-        assert response.status == 200
-        batches = []
-        async for batch in get_batches_from_train_response_stream(
-            response, ImageClassificationBatch
-        ):
-            batches.append(batch)
-            assert len(batch.data) == 1
-            selections = json.loads(batch.data[0].selections)
-            assert selections["width"] == 10
-            assert selections["height"] == 10
-            assert len(selections["tokens"]) == 87
-            assert len(selections["annotations"]) == 18
-            assert batch.data[0].page_uri == "DUMMY-URI"
-        assert len(batches) == 1
-
-
-@pytest.fixture
-@pytest.mark.asyncio
-async def image_classification_resource(
-    writer_rest_api: aiohttp.ClientSession, nucliadb_grpc: WriterStub, knowledgebox: str
-):
-    kbid = knowledgebox
-    field_id = "invoice"
-
-    with open(INVOICE_SELECTIONS_FILENAME) as f:
-        selections = json.load(f)
-        assert len(selections["tokens"]) == 87
-        assert len(selections["annotations"]) == 18
-
-    fieldmetadata = generate_image_classification_fieldmetadata(selections, field_id)
-
-    with open(INVOICE_FILENAME, "rb") as f:
-        invoice_content = f.read()
-
-    resp = await writer_rest_api.post(
-        f"/{API_PREFIX}/v1/{KB_PREFIX}/{knowledgebox}/resources",
-        headers={"x-synchronous": "true"},
-        json={
-            "title": "My invoice",
-            "files": {
-                field_id: {
-                    "file": {
-                        "filename": "invoice.pdf",
-                        "content_type": "application/pdf",
-                        "payload": base64.b64encode(invoice_content).decode(),
-                    }
-                }
-            },
-            "fieldmetadata": fieldmetadata,
-        },
-    )
-    assert resp.status == 201
-    body = await resp.json()
-    rid = body["uuid"]
-
-    broker_message = generate_image_classification_broker_message(
-        selections, kbid, rid, field_id
-    )
-
-    original_storage = get_utility(Utility.STORAGE)
-    set_utility(Utility.STORAGE, AsyncMock())
-    mock_set = AsyncMock(return_value=None)
-    mock_get = AsyncMock(return_value=broker_message.file_extracted_data[0])
-    with (
-        patch(
-            "nucliadb.ingest.fields.file.File.set_file_extracted_data", new=mock_set
-        ) as _,
-        patch(
-            "nucliadb.ingest.fields.file.File.get_file_extracted_data", new=mock_get
-        ) as _,
-    ):
-        resp = await nucliadb_grpc.ProcessMessage(  # type: ignore
-            iter([broker_message]), timeout=10, wait_for_ready=True
-        )
-        assert resp.status == OpStatusWriter.Status.OK
-        yield
-
-    set_utility(Utility.STORAGE, original_storage)
-
-
-def generate_image_classification_fieldmetadata(
-    selections: dict, field_id: str
-) -> list[dict[str, Any]]:
-    selections_by_page = {}  # type: ignore
-    for annotation in selections["annotations"]:
-        page_selections = selections_by_page.setdefault(annotation["page"], [])
-        page_selections.append(
-            {
-                "label": annotation["label"]["text"],
-                "top": annotation["bounds"]["top"],
-                "left": annotation["bounds"]["left"],
-                "right": annotation["bounds"]["right"],
-                "bottom": annotation["bounds"]["bottom"],
-                "token_ids": [token["tokenIndex"] for token in annotation["tokens"]],
-            }
-        )
-
-    fieldmetadata = {
-        "field": {"field": field_id, "field_type": "file"},
-        "selections": [
-            {
-                "page": page,
-                "visual": selections,
-            }
-            for page, selections in selections_by_page.items()
-        ],
-    }
-    return [fieldmetadata]
-
-
-def generate_image_classification_broker_message(
-    selections: dict, kbid: str, rid: str, field_id: str
-) -> BrokerMessage:
-    bm = BrokerMessage(
-        kbid=kbid,
-        uuid=rid,
-        source=BrokerMessage.MessageSource.PROCESSOR,
-        file_extracted_data=[
-            FileExtractedData(
-                field=field_id,
-                file_pages_previews=FilePages(
-                    pages=[
-                        CloudFile(uri="DUMMY-URI"),
-                    ],
-                    structures=[
-                        PageStructure(
-                            page=PageStructurePage(width=10, height=10),
-                            tokens=[
-                                PageStructureToken(
-                                    x=token["x"],
-                                    y=token["y"],
-                                    width=token["width"],
-                                    height=token["height"],
-                                    text=token["text"],
-                                    line=0,
-                                )
-                                for token in selections["tokens"]
-                            ],
-                        )
-                    ],
-                ),
-            )
-        ],
-    )
-
-    return bm

nucliadb/train/tests/test_list_fields.py

@@ -1,39 +0,0 @@
-# Copyright (C) 2021 Bosutech XXI S.L.
-#
-# nucliadb is offered under the AGPL v3.0 and as commercial software.
-# For commercial licensing, contact us at info@nuclia.com.
-#
-# AGPL:
-# This program is free software: you can redistribute it and/or modify
-# it under the terms of the GNU Affero General Public License as
-# published by the Free Software Foundation, either version 3 of the
-# License, or (at your option) any later version.
-#
-# This program is distributed in the hope that it will be useful,
-# but WITHOUT ANY WARRANTY; without even the implied warranty of
-# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
-# GNU Affero General Public License for more details.
-#
-# You should have received a copy of the GNU Affero General Public License
-# along with this program. If not, see <http://www.gnu.org/licenses/>.
-#
-import pytest
-from nucliadb_protos.train_pb2 import GetFieldsRequest
-from nucliadb_protos.train_pb2_grpc import TrainStub
-
-
-@pytest.mark.asyncio
-async def test_list_fields(
-    train_client: TrainStub, knowledgebox_ingest: str, test_pagination_resources
-) -> None:
-    req = GetFieldsRequest()
-    req.kb.uuid = knowledgebox_ingest
-    req.metadata.entities = True
-    req.metadata.labels = True
-    req.metadata.text = True
-    req.metadata.vector = True
-    count = 0
-    async for _ in train_client.GetParagraphs(req):  # type: ignore
-        count += 1
-
-    assert count == 30

nucliadb/train/tests/test_list_paragraphs.py

@@ -1,73 +0,0 @@
-# Copyright (C) 2021 Bosutech XXI S.L.
-#
-# nucliadb is offered under the AGPL v3.0 and as commercial software.
-# For commercial licensing, contact us at info@nuclia.com.
-#
-# AGPL:
-# This program is free software: you can redistribute it and/or modify
-# it under the terms of the GNU Affero General Public License as
-# published by the Free Software Foundation, either version 3 of the
-# License, or (at your option) any later version.
-#
-# This program is distributed in the hope that it will be useful,
-# but WITHOUT ANY WARRANTY; without even the implied warranty of
-# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
-# GNU Affero General Public License for more details.
-#
-# You should have received a copy of the GNU Affero General Public License
-# along with this program. If not, see <http://www.gnu.org/licenses/>.
-#
-import pytest
-from nucliadb_protos.train_pb2 import GetParagraphsRequest
-from nucliadb_protos.train_pb2_grpc import TrainStub
-
-
-@pytest.mark.asyncio
-async def test_list_paragraphs(
-    train_client: TrainStub, knowledgebox_ingest: str, test_pagination_resources
-) -> None:
-    req = GetParagraphsRequest()
-    req.kb.uuid = knowledgebox_ingest
-    req.metadata.entities = True
-    req.metadata.labels = True
-    req.metadata.text = True
-    req.metadata.vector = True
-    count = 0
-    async for _ in train_client.GetParagraphs(req):  # type: ignore
-        count += 1
-
-    assert count == 30
-
-
-@pytest.mark.asyncio
-async def test_list_paragraphs_shows_ners_with_positions(
-    train_client: TrainStub, knowledgebox_ingest: str, test_pagination_resources
-) -> None:
-    req = GetParagraphsRequest()
-    req.kb.uuid = knowledgebox_ingest
-    req.metadata.entities = True
-    req.metadata.labels = True
-    req.metadata.text = True
-    req.metadata.vector = True
-
-    found_barcelona = found_manresa = False
-    async for paragraph in train_client.GetParagraphs(req):  # type: ignore
-        if "Barcelona" in paragraph.metadata.text:
-            found_barcelona = True
-            assert paragraph.metadata.entities == {"Barcelona": "CITY"}
-            positions = paragraph.metadata.entity_positions["CITY/Barcelona"]
-            assert positions.entity == "Barcelona"
-            assert len(positions.positions) == 1
-            assert positions.positions[0].start == 43
-            assert positions.positions[0].end == 52
-        elif "Manresa" in paragraph.metadata.text:
-            found_manresa = True
-            assert paragraph.metadata.entities == {"Manresa": "CITY"}
-            positions = paragraph.metadata.entity_positions["CITY/Manresa"]
-            assert positions.entity == "Manresa"
-            assert len(positions.positions) == 2
-            assert positions.positions[0].start == 22
-            assert positions.positions[0].end == 29
-            assert positions.positions[1].start == 38
-            assert positions.positions[1].end == 45
-    assert found_manresa and found_barcelona

nucliadb/train/tests/test_list_resources.py

@@ -1,39 +0,0 @@
-# Copyright (C) 2021 Bosutech XXI S.L.
-#
-# nucliadb is offered under the AGPL v3.0 and as commercial software.
-# For commercial licensing, contact us at info@nuclia.com.
-#
-# AGPL:
-# This program is free software: you can redistribute it and/or modify
-# it under the terms of the GNU Affero General Public License as
-# published by the Free Software Foundation, either version 3 of the
-# License, or (at your option) any later version.
-#
-# This program is distributed in the hope that it will be useful,
-# but WITHOUT ANY WARRANTY; without even the implied warranty of
-# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
-# GNU Affero General Public License for more details.
-#
-# You should have received a copy of the GNU Affero General Public License
-# along with this program. If not, see <http://www.gnu.org/licenses/>.
-#
-import pytest
-from nucliadb_protos.train_pb2 import GetResourcesRequest
-from nucliadb_protos.train_pb2_grpc import TrainStub
-
-
-@pytest.mark.asyncio
-async def test_list_resource(
-    train_client: TrainStub, knowledgebox_ingest: str, test_pagination_resources
-) -> None:
-    req = GetResourcesRequest()
-    req.kb.uuid = knowledgebox_ingest
-    req.metadata.entities = True
-    req.metadata.labels = True
-    req.metadata.text = True
-    req.metadata.vector = True
-    count = 0
-    async for _ in train_client.GetResources(req):  # type: ignore
-        count += 1
-
-    assert count == 10

nucliadb/train/tests/test_list_sentences.py

@@ -1,71 +0,0 @@
-# Copyright (C) 2021 Bosutech XXI S.L.
-#
-# nucliadb is offered under the AGPL v3.0 and as commercial software.
-# For commercial licensing, contact us at info@nuclia.com.
-#
-# AGPL:
-# This program is free software: you can redistribute it and/or modify
-# it under the terms of the GNU Affero General Public License as
-# published by the Free Software Foundation, either version 3 of the
-# License, or (at your option) any later version.
-#
-# This program is distributed in the hope that it will be useful,
-# but WITHOUT ANY WARRANTY; without even the implied warranty of
-# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
-# GNU Affero General Public License for more details.
-#
-# You should have received a copy of the GNU Affero General Public License
-# along with this program. If not, see <http://www.gnu.org/licenses/>.
-#
-import pytest
-from nucliadb_protos.train_pb2 import GetSentencesRequest
-from nucliadb_protos.train_pb2_grpc import TrainStub
-
-
-@pytest.mark.asyncio
-async def test_list_sentences(
-    train_client: TrainStub, knowledgebox_ingest: str, test_pagination_resources
-) -> None:
-    req = GetSentencesRequest()
-    req.kb.uuid = knowledgebox_ingest
-    req.metadata.entities = True
-    req.metadata.labels = True
-    req.metadata.text = True
-    req.metadata.vector = True
-    count = 0
-
-    async for _ in train_client.GetSentences(req):  # type: ignore
-        count += 1
-
-    assert count == 40
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize("knowledgebox", ["STABLE", "EXPERIMENTAL"], indirect=True)
-async def test_list_sentences_shows_ners_with_positions(
-    train_client: TrainStub, knowledgebox: str, test_pagination_resources
-) -> None:
-    req = GetSentencesRequest()
-    req.kb.uuid = knowledgebox
-    req.metadata.entities = True
-    async for sentence in train_client.GetSentences(req):  # type: ignore
-        if "Barcelona" in sentence.metadata.text:
-            assert sentence.metadata.entities == {"Barcelona": "CITY"}
-            positions = sentence.metadata.entity_positions["CITY/Barcelona"]
-            assert positions.entity == "Barcelona"
-            assert len(positions.positions) == 1
-            assert positions.positions[0].start == 43
-            assert positions.positions[0].end == 52
-        elif "Manresa" in sentence.metadata.text:
-            assert sentence.metadata.entities == {"Manresa": "CITY"}
-            positions = sentence.metadata.entity_positions["CITY/Manresa"]
-            assert positions.entity == "Manresa"
-            assert len(positions.positions) == 2
-            assert positions.positions[0].start == 22
-            assert positions.positions[0].end == 29
-            assert positions.positions[1].start == 38
-            assert positions.positions[1].end == 45
-        else:
-            # Other sentences should not have entities nor positions
-            assert sentence.metadata.entities == {}
-            assert sentence.metadata.entity_positions == {}