nucliadb 4.0.0.post542__py3-none-any.whl → 6.2.1.post2798__py3-none-any.whl

nucliadb/search/tests/unit/api/v1/test_ask.py

@@ -1,120 +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/>.
-#
-from unittest import mock
-from unittest.mock import Mock
-
-import pytest
-from starlette.requests import Request
-
-from nucliadb.common.datamanagers.exceptions import KnowledgeBoxNotFound
-from nucliadb.models.responses import HTTPClientError
-from nucliadb.search import predict
-from nucliadb.search.api.v1.ask import ask_knowledgebox_endpoint
-from nucliadb.search.search.exceptions import (
-    IncompleteFindResultsError,
-    InvalidQueryError,
-)
-from nucliadb_utils.exceptions import LimitsExceededError
-
-pytestmark = pytest.mark.asyncio
-
-
-class DummyTestRequest(Request):
-    @property
-    def auth(self):
-        return Mock(scopes=["READER"])
-
-    @property
-    def user(self):
-        return Mock(display_name="username")
-
-
-@pytest.fixture(scope="function")
-def ask_mock():
-    with mock.patch(
-        "nucliadb.search.api.v1.ask.ask",
-    ) as mocked:
-        yield mocked
-
-
-@pytest.mark.parametrize(
-    "ask_error,http_error_response",
-    [
-        (
-            KnowledgeBoxNotFound(),
-            HTTPClientError(status_code=404, detail="Knowledge Box not found."),
-        ),
-        (
-            LimitsExceededError(402, "over the quota"),
-            HTTPClientError(status_code=402, detail="over the quota"),
-        ),
-        (
-            predict.ProxiedPredictAPIError(status=999, detail="foo"),
-            HTTPClientError(status_code=999, detail="foo"),
-        ),
-        (
-            IncompleteFindResultsError(),
-            HTTPClientError(
-                status_code=529,
-                detail="Temporary error on information retrieval. Please try again.",
-            ),
-        ),
-        (
-            predict.RephraseMissingContextError(),
-            HTTPClientError(
-                status_code=412,
-                detail="Unable to rephrase the query with the provided context.",
-            ),
-        ),
-        (
-            predict.RephraseError("foobar"),
-            HTTPClientError(
-                status_code=529,
-                detail="Temporary error while rephrasing the query. Please try again later. Error: foobar",
-            ),
-        ),
-        (
-            InvalidQueryError("foobar", "baz"),
-            HTTPClientError(
-                status_code=412, detail="Invalid query. Error in foobar: baz"
-            ),
-        ),
-    ],
-)
-async def test_ask_endpoint_handles_errors(ask_mock, ask_error, http_error_response):
-    ask_mock.side_effect = ask_error
-    request = DummyTestRequest(
-        scope={
-            "type": "http",
-            "http_version": "1.1",
-            "method": "GET",
-            "headers": [],
-        }
-    )
-    response = await ask_knowledgebox_endpoint(
-        request=request,
-        kbid="kbid",
-        item=Mock(),
-        x_ndb_client=None,
-        x_nucliadb_user="",
-        x_forwarded_for="",
-    )
-    assert response.status_code == http_error_response.status_code
-    assert response.body == http_error_response.body

nucliadb/search/tests/unit/api/v1/test_chat.py

@@ -1,96 +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/>.
-#
-from unittest import mock
-from unittest.mock import Mock
-
-import pytest
-from starlette.requests import Request
-
-from nucliadb.models.responses import HTTPClientError
-from nucliadb.search import predict
-from nucliadb.search.api.v1.chat import chat_knowledgebox_endpoint
-from nucliadb_utils.exceptions import LimitsExceededError
-
-pytestmark = pytest.mark.asyncio
-
-
-class DummyTestRequest(Request):
-    @property
-    def auth(self):
-        return Mock(scopes=["READER"])
-
-    @property
-    def user(self):
-        return Mock(display_name="username")
-
-
-@pytest.fixture(scope="function")
-def create_chat_response_mock():
-    with mock.patch(
-        "nucliadb.search.api.v1.chat.create_chat_response",
-    ) as mocked:
-        yield mocked
-
-
-@pytest.mark.parametrize(
-    "predict_error,http_error_response",
-    [
-        (
-            LimitsExceededError(402, "over the quota"),
-            HTTPClientError(status_code=402, detail="over the quota"),
-        ),
-        (
-            predict.RephraseError("foobar"),
-            HTTPClientError(
-                status_code=529,
-                detail="Temporary error while rephrasing the query. Please try again later. Error: foobar",
-            ),
-        ),
-        (
-            predict.RephraseMissingContextError(),
-            HTTPClientError(
-                status_code=412,
-                detail="Unable to rephrase the query with the provided context.",
-            ),
-        ),
-    ],
-)
-async def test_chat_endpoint_handles_errors(
-    create_chat_response_mock, predict_error, http_error_response
-):
-    create_chat_response_mock.side_effect = predict_error
-    request = DummyTestRequest(
-        scope={
-            "type": "http",
-            "http_version": "1.1",
-            "method": "GET",
-            "headers": [],
-        }
-    )
-    response = await chat_knowledgebox_endpoint(
-        request=request,
-        kbid="kbid",
-        item=Mock(),
-        x_ndb_client=None,
-        x_nucliadb_user="",
-        x_forwarded_for="",
-    )
-    assert response.status_code == http_error_response.status_code
-    assert response.body == http_error_response.body

nucliadb/search/tests/unit/api/v1/test_predict_proxy.py

@@ -1,98 +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 json
-from unittest import mock
-from unittest.mock import Mock
-
-import pytest
-from starlette.requests import Request
-
-from nucliadb.models.responses import HTTPClientError
-from nucliadb.search import predict
-from nucliadb.search.api.v1.predict_proxy import predict_proxy_endpoint
-from nucliadb.search.search.predict_proxy import PredictProxiedEndpoints
-from nucliadb_utils.exceptions import LimitsExceededError
-
-pytestmark = pytest.mark.asyncio
-
-MODULE = "nucliadb.search.api.v1.predict_proxy"
-
-
-class DummyTestRequest(Request):
-    @property
-    def auth(self):
-        return Mock(scopes=["READER"])
-
-    @property
-    def user(self):
-        return Mock(display_name="username")
-
-    async def json(self):
-        raise json.JSONDecodeError("test", "test", 0)
-
-
-@pytest.fixture(scope="function")
-def dummy_request():
-    return DummyTestRequest(
-        scope={
-            "type": "http",
-            "http_version": "1.1",
-            "method": "GET",
-            "headers": [],
-            "query_string": "",
-        }
-    )
-
-
-@pytest.fixture(scope="function")
-def predict_proxy():
-    with mock.patch(f"{MODULE}.predict_proxy") as mocked:
-        yield mocked
-
-
-@pytest.mark.parametrize(
-    "predict_error,http_error_response",
-    [
-        (
-            LimitsExceededError(402, "over the quota"),
-            HTTPClientError(status_code=402, detail="over the quota"),
-        ),
-        (
-            predict.ProxiedPredictAPIError(status=500, detail="Temporary error"),
-            HTTPClientError(
-                status_code=500,
-                detail="Temporary error",
-            ),
-        ),
-    ],
-)
-async def test_predict_proxy_endpoint_error_handling(
-    predict_proxy,
-    predict_error,
-    http_error_response,
-    dummy_request,
-):
-    predict_proxy.side_effect = predict_error
-    response = await predict_proxy_endpoint(
-        request=dummy_request,
-        kbid="kbid",
-        endpoint=PredictProxiedEndpoints.CHAT,
-    )
-    assert response.status_code == http_error_response.status_code

nucliadb/search/tests/unit/api/v1/test_summarize.py

@@ -1,99 +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/>.
-#
-from unittest import mock
-from unittest.mock import Mock
-
-import pytest
-from starlette.requests import Request
-
-from nucliadb.models.responses import HTTPClientError
-from nucliadb.search import predict
-from nucliadb.search.api.v1.summarize import summarize_endpoint
-from nucliadb.search.search.summarize import NoResourcesToSummarize
-from nucliadb_utils.exceptions import LimitsExceededError
-
-pytestmark = pytest.mark.asyncio
-
-
-class DummyTestRequest(Request):
-    @property
-    def auth(self):
-        return Mock(scopes=["READER"])
-
-    @property
-    def user(self):
-        return Mock(display_name="username")
-
-
-@pytest.fixture(scope="function")
-def dummy_request():
-    return DummyTestRequest(
-        scope={
-            "type": "http",
-            "http_version": "1.1",
-            "method": "GET",
-            "headers": [],
-        }
-    )
-
-
-@pytest.fixture(scope="function")
-def summarize():
-    with mock.patch("nucliadb.search.api.v1.summarize.summarize") as mocked:
-        yield mocked
-
-
-@pytest.mark.parametrize(
-    "error,http_error_response",
-    [
-        (
-            LimitsExceededError(402, "over the quota"),
-            HTTPClientError(status_code=402, detail="over the quota"),
-        ),
-        (
-            predict.ProxiedPredictAPIError(status=500, detail="Temporary error"),
-            HTTPClientError(
-                status_code=500,
-                detail="Temporary error",
-            ),
-        ),
-        (
-            NoResourcesToSummarize(),
-            HTTPClientError(
-                status_code=412,
-                detail="Could not summarize: No resources or extracted text found.",
-            ),
-        ),
-    ],
-)
-async def test_summarize_endpoint_handles_errors(
-    summarize,
-    error,
-    http_error_response,
-    dummy_request,
-):
-    summarize.side_effect = error
-    response = await summarize_endpoint(
-        request=dummy_request,
-        kbid="kbid",
-        item=Mock(),
-    )
-    assert response.status_code == http_error_response.status_code
-    assert response.body == http_error_response.body

nucliadb/search/tests/unit/search/__init__.py

@@ -1,18 +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/>.

nucliadb/search/tests/unit/search/requesters/__init__.py

@@ -1,18 +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/>.

nucliadb/search/tests/unit/search/requesters/test_utils.py

@@ -1,211 +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/>.
-
-from unittest.mock import AsyncMock, Mock
-
-import pytest
-from fastapi import HTTPException
-from grpc import StatusCode
-from grpc.aio import AioRpcError
-
-from nucliadb.common.cluster.base import AbstractIndexNode
-from nucliadb.search.requesters import utils
-from nucliadb_protos import nodereader_pb2, writer_pb2
-from nucliadb_utils.utilities import Utility, clean_utility, get_utility, set_utility
-
-
-@pytest.fixture
-def fake_nodes():
-    from nucliadb.common.cluster import manager
-
-    original = manager.INDEX_NODES
-    manager.INDEX_NODES.clear()
-
-    manager.add_index_node(
-        id="node-0",
-        address="nohost",
-        shard_count=0,
-        available_disk=100,
-        dummy=True,
-    )
-    manager.add_index_node(
-        id="node-replica-0",
-        address="nohost",
-        shard_count=0,
-        available_disk=100,
-        dummy=True,
-        primary_id="node-0",
-    )
-
-    yield (["node-0"], ["node-replica-0"])
-
-    manager.INDEX_NODES = original
-
-
-@pytest.fixture
-def shard_manager():
-    original = get_utility(Utility.SHARD_MANAGER)
-
-    manager = AsyncMock()
-    manager.get_shards_by_kbid = AsyncMock(
-        return_value=[
-            writer_pb2.ShardObject(
-                shard="shard-id",
-                replicas=[
-                    writer_pb2.ShardReplica(
-                        shard=writer_pb2.ShardCreated(id="shard-id"), node="node-0"
-                    )
-                ],
-            )
-        ]
-    )
-
-    set_utility(Utility.SHARD_MANAGER, manager)
-
-    yield manager
-
-    if original is None:
-        clean_utility(Utility.SHARD_MANAGER)
-    else:
-        set_utility(Utility.SHARD_MANAGER, original)
-
-
-@pytest.fixture()
-def search_methods():
-    def fake_search(
-        node: AbstractIndexNode, shard: str, query: nodereader_pb2.SearchRequest
-    ):
-        if node.is_read_replica():
-            raise Exception()
-        return nodereader_pb2.SearchResponse()
-
-    original = utils.METHODS
-    utils.METHODS = {
-        utils.Method.SEARCH: AsyncMock(side_effect=fake_search),
-        utils.Method.PARAGRAPH: AsyncMock(),
-    }
-
-    yield utils.METHODS
-
-    utils.METHODS = original
-
-
-@pytest.mark.asyncio
-async def test_node_query_retries_primary_if_secondary_fails(
-    fake_nodes,
-    shard_manager,
-    search_methods,
-):
-    """Setting up a node and a faulty replica, validate primary is queried if
-    secondary fails.
-
-    """
-    pb_query = nodereader_pb2.SearchRequest(shard="shard-id", body="question")
-    results, incomplete_results, queried_nodes = await utils.node_query(
-        kbid="my-kbid",
-        method=utils.Method.SEARCH,
-        pb_query=pb_query,
-        use_read_replica_nodes=True,
-    )
-    # secondary fails, primary is called
-    assert search_methods[utils.Method.SEARCH].await_count == 2
-    assert len(queried_nodes) == 2
-    assert queried_nodes[0][0].is_read_replica()
-    assert not queried_nodes[1][0].is_read_replica()
-
-    results, incomplete_results, queried_nodes = await utils.node_query(
-        kbid="my-kbid",
-        method=utils.Method.PARAGRAPH,
-        pb_query=Mock(),
-        use_read_replica_nodes=True,
-    )
-    # secondary succeeds, no fallback call to primary
-    assert search_methods[utils.Method.PARAGRAPH].await_count == 1
-    assert len(queried_nodes) == 1
-    assert queried_nodes[0][0].is_read_replica()
-
-
-def test_debug_nodes_info(fake_nodes: tuple[list[str], list[str]]):
-    from nucliadb.common.cluster import manager
-
-    primary = manager.get_index_node(fake_nodes[0][0])
-    assert primary is not None
-    secondary = manager.get_index_node(fake_nodes[1][0])
-    assert secondary is not None
-
-    info = utils.debug_nodes_info([(primary, "shard-a"), (secondary, "shard-b")])
-    assert len(info) == 2
-
-    primary_keys = ["id", "shard_id", "address"]
-    secondary_keys = primary_keys + ["primary_id"]
-
-    for key in primary_keys:
-        assert key in info[0]
-
-    for key in secondary_keys:
-        assert key in info[1]
-
-
-def test_validate_node_query_results():
-    assert utils.validate_node_query_results([Mock()]) is None
-
-
-def test_validate_node_query_results_no_results():
-    assert isinstance(utils.validate_node_query_results([]), HTTPException)
-    assert isinstance(utils.validate_node_query_results(None), HTTPException)
-
-
-def test_validate_node_query_results_unhandled_error():
-    error = utils.validate_node_query_results([Exception()])
-    assert isinstance(error, HTTPException)
-
-
-def test_validate_node_query_results_invalid_query():
-    result = utils.validate_node_query_results(
-        [
-            AioRpcError(
-                code=StatusCode.INTERNAL,
-                initial_metadata=Mock(),
-                trailing_metadata=Mock(),
-                details="An invalid argument was passed: 'Query is invalid. AllButQueryForbidden'",
-                debug_error_string="",
-            )
-        ]
-    )
-
-    assert isinstance(result, HTTPException)
-    assert result.status_code == 412
-    assert result.detail == "Query is invalid. AllButQueryForbidden"
-
-
-def test_validate_node_query_results_internal_unhandled():
-    result = utils.validate_node_query_results(
-        [
-            AioRpcError(
-                code=StatusCode.INTERNAL,
-                initial_metadata=Mock(),
-                trailing_metadata=Mock(),
-                details="There is something wrong with your query, my friend!",
-                debug_error_string="This query is simply wrong",
-            )
-        ]
-    )
-    assert isinstance(result, HTTPException)
-    assert result.status_code == 500
-    assert result.detail == "There is something wrong with your query, my friend!"

nucliadb/search/tests/unit/search/search/__init__.py

@@ -1,19 +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/>.
-#