From 5010d4f4b3f0cf218759df8296e6973412638e02 Mon Sep 17 00:00:00 2001
From: Valentin Rigal <rigal@teklia.com>
Date: Wed, 15 Feb 2023 09:01:48 +0000
Subject: [PATCH] Start corpus indexation
---
arkindex/documents/api/search.py | 43 ++++-
arkindex/documents/serializers/search.py | 24 +++
arkindex/documents/tasks.py | 34 ++++
.../tests/tasks/test_reindex_corpus.py | 77 +++++++++
arkindex/documents/tests/test_search_api.py | 148 +++++++++++++++++-
arkindex/project/api_v1.py | 3 +-
arkindex/project/config.py | 1 +
arkindex/project/rq_overrides.py | 13 ++
.../tests/config_samples/defaults.yaml | 1 +
.../project/tests/config_samples/errors.yaml | 1 +
.../tests/config_samples/expected_errors.yaml | 1 +
.../tests/config_samples/override.yaml | 3 +-
arkindex/project/triggers.py | 13 ++
arkindex/templates/reindex_corpus.html | 10 ++
14 files changed, 367 insertions(+), 5 deletions(-)
create mode 100644 arkindex/documents/tests/tasks/test_reindex_corpus.py
create mode 100644 arkindex/templates/reindex_corpus.html
diff --git a/arkindex/documents/api/search.py b/arkindex/documents/api/search.py
index a11358737f..4535efea25 100644
--- a/arkindex/documents/api/search.py
+++ b/arkindex/documents/api/search.py
@@ -1,18 +1,25 @@
import math
from django.conf import settings
+from django.utils.functional import cached_property
from drf_spectacular.utils import extend_schema, extend_schema_view
from rest_framework import status
from rest_framework.exceptions import NotFound, ValidationError
-from rest_framework.generics import RetrieveAPIView
+from rest_framework.generics import CreateAPIView, RetrieveAPIView
from rest_framework.response import Response
from rest_framework.utils.urls import replace_query_param
from SolrClient import SolrClient
from SolrClient.exceptions import SolrError
from arkindex.documents.models import EntityType, MetaType
-from arkindex.documents.serializers.search import CorpusSearchQuerySerializer, CorpusSearchResultSerializer
+from arkindex.documents.serializers.search import (
+ CorpusSearchQuerySerializer,
+ CorpusSearchResultSerializer,
+ ReindexCorpusSerializer,
+)
from arkindex.project.mixins import CorpusACLMixin
+from arkindex.project.permissions import IsVerified
+from arkindex.users.models import Role
solr = SolrClient(settings.SOLR_API_URL)
@@ -123,3 +130,35 @@ class CorpusSearch(CorpusACLMixin, RetrieveAPIView):
'results': results.docs if not only_facets else None,
'facets': results.get_facets()
}).data, status=status.HTTP_200_OK)
+
+
+@extend_schema_view(post=extend_schema(operation_id="BuildSearchIndex", tags=['search']))
+class SearchIndexBuild(CorpusACLMixin, CreateAPIView):
+ """
+ Starts an indexation task for a specific corpus, ran by RQ
+ """
+ permission_classes = (IsVerified, )
+ serializer_class = ReindexCorpusSerializer
+
+ @cached_property
+ def corpus(self):
+ corpus = self.get_corpus(self.kwargs['pk'], role=Role.Admin)
+ if not corpus.indexable:
+ raise ValidationError({'__all__': ['This project is not indexable.']})
+ if not corpus.types.filter(indexable=True).exists():
+ raise ValidationError({'__all__': ['There are no indexable element types for this project.']})
+ return corpus
+
+ def get_serializer_context(self):
+ return {
+ **super().get_serializer_context(),
+ 'corpus_id': self.corpus.id,
+ 'user_id': self.request.user.id,
+ }
+
+ def create(self, request, *args, **kwargs):
+ if not settings.ARKINDEX_FEATURES['search']:
+ raise ValidationError({
+ '__all__': ['Building search index is unavailable due to the search feature being disabled.']
+ })
+ return super().create(request, *args, **kwargs)
diff --git a/arkindex/documents/serializers/search.py b/arkindex/documents/serializers/search.py
index 145f76d154..39f7c78c11 100644
--- a/arkindex/documents/serializers/search.py
+++ b/arkindex/documents/serializers/search.py
@@ -1,9 +1,12 @@
from textwrap import dedent
from rest_framework import serializers
+from rest_framework.exceptions import ValidationError
from arkindex.documents.models import MetaType
+from arkindex.project.rq_overrides import get_existing_job
from arkindex.project.serializer_fields import EnumField
+from arkindex.project.triggers import reindex_corpus
class SolrDocumentSerializer(serializers.Serializer):
@@ -164,3 +167,24 @@ class CorpusSearchQuerySerializer(serializers.Serializer):
required=False,
help_text='Filter by name of the worker that created the entity.',
)
+
+
+class ReindexCorpusSerializer(serializers.Serializer):
+ drop = serializers.BooleanField(default=True, help_text='Drop existing collections for this corpus.')
+
+ def save(self, **kwargs):
+ corpus_id = self.context.get('corpus_id')
+ user_id = self.context.get('user_id')
+ assert corpus_id and user_id, 'corpus_id and user_id must be passed in the serializer context'
+
+ # Ensure the reindex job has not already been started
+ job_id = f'reindex-{corpus_id}'
+ if (job := get_existing_job(job_id)) is not None:
+ # A previous job can only be removed if finished
+ if job.ended_at is None:
+ raise ValidationError({
+ '__all__': [f'A job is already running to build search index on corpus {corpus_id}.']
+ })
+ job.delete()
+
+ reindex_corpus(**self.validated_data, corpus_id=corpus_id, user_id=user_id, job_id=job_id)
diff --git a/arkindex/documents/tasks.py b/arkindex/documents/tasks.py
index 49c624e4f0..d42a9cbfb0 100644
--- a/arkindex/documents/tasks.py
+++ b/arkindex/documents/tasks.py
@@ -3,10 +3,13 @@ from typing import Optional
from uuid import UUID
from django.conf import settings
+from django.core.mail import send_mail
from django.db.models import Q
+from django.template.loader import render_to_string
from django_rq import job
from rq import Retry, get_current_job
+from arkindex.documents.indexer import Indexer
from arkindex.documents.managers import ElementQuerySet
from arkindex.documents.models import (
Classification,
@@ -20,6 +23,7 @@ from arkindex.documents.models import (
TranscriptionEntity,
)
from arkindex.process.models import Process, ProcessElement, WorkerActivity, WorkerRun
+from arkindex.users.models import User
logger = logging.getLogger(__name__)
@@ -227,3 +231,33 @@ def add_parent_selection(corpus_id: UUID, parent: Element) -> None:
for i, item in enumerate(queryset):
rq_job.set_progress(i / total)
item.add_parent(parent)
+
+
+@job('default', timeout=settings.RQ_TIMEOUTS['reindex_corpus'])
+def reindex_corpus(corpus_id: UUID, drop: bool = True) -> None:
+ rq_job = get_current_job()
+ assert rq_job is not None, 'This task can only be run in a RQ job context.'
+ assert rq_job.user_id is not None, 'This task requires a user ID to be defined on the RQ job.'
+
+ indexer = Indexer(corpus_id)
+ if drop:
+ indexer.drop_index()
+ indexer.setup()
+ indexer.index()
+
+ # Report to the user that the index build finished
+ user = User.objects.get(id=rq_job.user_id)
+ corpus = Corpus.objects.get(id=corpus_id)
+ send_mail(
+ subject=f'Project {corpus.name} was successfully indexed',
+ message=render_to_string(
+ 'reindex_corpus.html',
+ context={
+ 'user': user,
+ 'corpus': corpus,
+ },
+ ),
+ from_email=None,
+ recipient_list=[user.email],
+ fail_silently=True,
+ )
diff --git a/arkindex/documents/tests/tasks/test_reindex_corpus.py b/arkindex/documents/tests/tasks/test_reindex_corpus.py
new file mode 100644
index 0000000000..81abde9ea0
--- /dev/null
+++ b/arkindex/documents/tests/tasks/test_reindex_corpus.py
@@ -0,0 +1,77 @@
+from unittest.mock import call, patch
+
+from arkindex.documents.tasks import reindex_corpus
+from arkindex.project.tests import FixtureTestCase
+
+EXPECTED_EMAIL_BODY = """
+Hello Test user,
+
+The search index build you started on project Unit Tests has been finished successfully.
+
+Indexed elements can be explored from the search interface on Arkindex.
+
+--
+Arkindex
+
+"""
+
+
+class TestReindexCorpus(FixtureTestCase):
+
+ def test_no_rq_job(self):
+ with self.assertRaises(AssertionError) as ctx:
+ reindex_corpus(corpus_id=self.corpus.id)
+ self.assertEqual(str(ctx.exception), 'This task can only be run in a RQ job context.')
+
+ @patch('arkindex.documents.tasks.get_current_job')
+ def test_no_user_id(self, job_mock):
+ job_mock.return_value.user_id = None
+ with self.assertRaises(AssertionError) as ctx:
+ reindex_corpus(corpus_id=self.corpus.id)
+ self.assertEqual(str(ctx.exception), 'This task requires a user ID to be defined on the RQ job.')
+
+ @patch('arkindex.documents.tasks.send_mail')
+ @patch('arkindex.documents.tasks.get_current_job')
+ @patch('arkindex.documents.tasks.Indexer')
+ def test_run_drop_false(self, indexer_mock, job_mock, send_mail_mock):
+ job_mock.return_value.user_id = self.user.id
+ reindex_corpus(corpus_id=self.corpus.id, drop=False)
+
+ self.assertEqual(indexer_mock.call_count, 1)
+ self.assertEqual(indexer_mock.call_args, call(self.corpus.id))
+ self.assertEqual(indexer_mock.return_value.drop_index.call_count, 0)
+ self.assertEqual(indexer_mock.return_value.setup.call_count, 1)
+ self.assertEqual(indexer_mock.return_value.index.call_count, 1)
+
+ self.assertListEqual(send_mail_mock.call_args_list, [
+ call(
+ subject=f'Project {self.corpus.name} was successfully indexed',
+ message=EXPECTED_EMAIL_BODY,
+ from_email=None,
+ recipient_list=[self.user.email],
+ fail_silently=True,
+ )
+ ])
+
+ @patch('arkindex.documents.tasks.send_mail')
+ @patch('arkindex.documents.tasks.get_current_job')
+ @patch('arkindex.documents.tasks.Indexer')
+ def test_run(self, indexer_mock, job_mock, send_mail_mock):
+ job_mock.return_value.user_id = self.user.id
+ reindex_corpus(corpus_id=self.corpus.id)
+
+ self.assertEqual(indexer_mock.call_count, 1)
+ self.assertEqual(indexer_mock.call_args, call(self.corpus.id))
+ self.assertEqual(indexer_mock.return_value.drop_index.call_count, 1)
+ self.assertEqual(indexer_mock.return_value.setup.call_count, 1)
+ self.assertEqual(indexer_mock.return_value.index.call_count, 1)
+
+ self.assertListEqual(send_mail_mock.call_args_list, [
+ call(
+ subject=f'Project {self.corpus.name} was successfully indexed',
+ message=EXPECTED_EMAIL_BODY,
+ from_email=None,
+ recipient_list=[self.user.email],
+ fail_silently=True,
+ )
+ ])
diff --git a/arkindex/documents/tests/test_search_api.py b/arkindex/documents/tests/test_search_api.py
index 7e2997f86b..7899c731a2 100644
--- a/arkindex/documents/tests/test_search_api.py
+++ b/arkindex/documents/tests/test_search_api.py
@@ -1,13 +1,15 @@
-from unittest.mock import patch
+from unittest.mock import call, patch
from django.test import override_settings
from django.urls import reverse
from rest_framework import status
+from rq.exceptions import NoSuchJobError
from SolrClient import SolrResponse
from SolrClient.exceptions import SolrError
from arkindex.documents.models import Corpus
from arkindex.project.tests import FixtureAPITestCase
+from arkindex.users.models import Role
class TestSearchApi(FixtureAPITestCase):
@@ -16,6 +18,7 @@ class TestSearchApi(FixtureAPITestCase):
def setUpTestData(cls):
super().setUpTestData()
cls.corpus.indexable = True
+ cls.corpus.types.filter(slug="page").update(indexable=True)
cls.corpus.save()
def build_solr_response(
@@ -313,3 +316,146 @@ class TestSearchApi(FixtureAPITestCase):
'sort': ['"element_atomic_number" is not a valid choice.']
})
self.assertFalse(mock_solr.query.called)
+
+ def test_build_index_requires_auth(self):
+ with self.assertNumQueries(0):
+ response = self.client.post(reverse('api:build-search-index', kwargs={'pk': self.corpus.id}))
+ self.assertEqual(response.status_code, status.HTTP_403_FORBIDDEN)
+
+ def test_build_index_requires_verified(self):
+ self.user.verified_email = False
+ self.user.save()
+ self.client.force_login(self.user)
+ with self.assertNumQueries(2):
+ response = self.client.post(reverse('api:build-search-index', kwargs={'pk': self.corpus.id}))
+ self.assertEqual(response.status_code, status.HTTP_403_FORBIDDEN)
+
+ @override_settings(ARKINDEX_FEATURES={'search': True})
+ def test_build_index_requires_corpus_admin(self):
+ self.corpus.memberships.update(level=Role.Contributor.value)
+ self.client.force_login(self.user)
+ with self.assertNumQueries(5):
+ response = self.client.post(reverse('api:build-search-index', kwargs={'pk': self.corpus.id}))
+ self.assertEqual(response.status_code, status.HTTP_403_FORBIDDEN)
+
+ @override_settings(ARKINDEX_FEATURES={'search': False})
+ @patch('arkindex.documents.tasks.reindex_corpus.delay')
+ def test_build_index_requires_search_feature(self, reindex_trigger_mock):
+ self.client.force_login(self.user)
+ with self.assertNumQueries(2):
+ response = self.client.post(reverse('api:build-search-index', kwargs={'pk': self.corpus.id}))
+ self.assertEqual(response.status_code, status.HTTP_400_BAD_REQUEST)
+ self.assertDictEqual(response.json(), {
+ '__all__': ['Building search index is unavailable due to the search feature being disabled.']
+ })
+ self.assertFalse(reindex_trigger_mock.called)
+
+ @override_settings(ARKINDEX_FEATURES={'search': True})
+ @patch('arkindex.documents.tasks.reindex_corpus.delay')
+ @patch('arkindex.project.rq_overrides.Job')
+ def test_build_index_existing_running_job(self, job_mock, reindex_trigger_mock):
+ existing_job = job_mock()
+ existing_job.ended_at = None
+ job_mock.fetch.return_value = existing_job
+ self.client.force_login(self.user)
+ with self.assertNumQueries(7):
+ response = self.client.post(reverse('api:build-search-index', kwargs={'pk': self.corpus.id}))
+ self.assertEqual(response.status_code, status.HTTP_400_BAD_REQUEST)
+ self.assertDictEqual(response.json(), {
+ '__all__': [f'A job is already running to build search index on corpus {self.corpus.id}.']
+ })
+ self.assertFalse(reindex_trigger_mock.called)
+
+ @override_settings(ARKINDEX_FEATURES={'search': True})
+ @patch('arkindex.documents.tasks.reindex_corpus.delay')
+ @patch('arkindex.project.rq_overrides.Job')
+ def test_build_index_non_indexable_corpus(self, job_mock, reindex_trigger_mock):
+ self.corpus.indexable = False
+ self.corpus.save()
+ self.client.force_login(self.user)
+ with self.assertNumQueries(6):
+ response = self.client.post(reverse('api:build-search-index', kwargs={'pk': self.corpus.id}))
+ self.assertEqual(response.status_code, status.HTTP_400_BAD_REQUEST)
+ self.assertDictEqual(response.json(), {
+ '__all__': ['This project is not indexable.']
+ })
+ self.assertFalse(reindex_trigger_mock.called)
+
+ @override_settings(ARKINDEX_FEATURES={'search': True})
+ @patch('arkindex.documents.tasks.reindex_corpus.delay')
+ @patch('arkindex.project.rq_overrides.Job')
+ def test_build_index_non_indexable_element_types(self, job_mock, reindex_trigger_mock):
+ self.corpus.types.update(indexable=False)
+ self.client.force_login(self.user)
+ with self.assertNumQueries(7):
+ response = self.client.post(reverse('api:build-search-index', kwargs={'pk': self.corpus.id}))
+ self.assertEqual(response.status_code, status.HTTP_400_BAD_REQUEST)
+ self.assertDictEqual(response.json(), {
+ '__all__': ['There are no indexable element types for this project.']
+ })
+ self.assertFalse(reindex_trigger_mock.called)
+
+ @override_settings(ARKINDEX_FEATURES={'search': True})
+ @patch('arkindex.documents.tasks.reindex_corpus.delay')
+ @patch('arkindex.project.rq_overrides.Job')
+ def test_build_index_existing_finished_job(self, job_mock, reindex_trigger_mock):
+ """In case a job exists but is finished, remove it and index the corpus again"""
+ existing_job = job_mock()
+ existing_job.ended_at = "2000-01-01"
+ job_mock.fetch.return_value = existing_job
+ self.client.force_login(self.user)
+ with self.assertNumQueries(7):
+ response = self.client.post(reverse('api:build-search-index', kwargs={'pk': self.corpus.id}))
+ self.assertEqual(response.status_code, status.HTTP_201_CREATED)
+ self.assertDictEqual(response.json(), {"drop": True})
+ self.assertListEqual(reindex_trigger_mock.call_args_list, [call(
+ corpus_id=self.corpus.id,
+ drop=True,
+ user_id=self.user.id,
+ description=f'Reindexing project {self.corpus.id}',
+ job_id=f'reindex-{self.corpus.id}',
+ )])
+ # The previous Job has been deleted
+ self.assertTrue(existing_job.delete.called)
+
+ @override_settings(ARKINDEX_FEATURES={'search': True})
+ @patch('arkindex.documents.tasks.reindex_corpus.delay')
+ @patch('arkindex.project.rq_overrides.Job')
+ def test_build_index(self, job_mock, reindex_trigger_mock):
+ job_mock.fetch.side_effect = NoSuchJobError
+ self.client.force_login(self.user)
+ with self.assertNumQueries(7):
+ response = self.client.post(
+ reverse('api:build-search-index', kwargs={'pk': self.corpus.id}),
+ data={"drop": True},
+ )
+ self.assertEqual(response.status_code, status.HTTP_201_CREATED)
+ self.assertDictEqual(response.json(), {"drop": True})
+ self.assertListEqual(reindex_trigger_mock.call_args_list, [call(
+ corpus_id=self.corpus.id,
+ drop=True,
+ user_id=self.user.id,
+ description=f'Reindexing project {self.corpus.id}',
+ job_id=f'reindex-{self.corpus.id}',
+ )])
+
+ @override_settings(ARKINDEX_FEATURES={'search': True})
+ @patch('arkindex.documents.tasks.reindex_corpus.delay')
+ @patch('arkindex.project.rq_overrides.Job')
+ def test_build_index_drop_false(self, job_mock, reindex_trigger_mock):
+ job_mock.fetch.side_effect = NoSuchJobError
+ self.client.force_login(self.user)
+ with self.assertNumQueries(7):
+ response = self.client.post(
+ reverse('api:build-search-index', kwargs={'pk': self.corpus.id}),
+ data={"drop": False},
+ )
+ self.assertEqual(response.status_code, status.HTTP_201_CREATED)
+ self.assertDictEqual(response.json(), {"drop": False})
+ self.assertListEqual(reindex_trigger_mock.call_args_list, [call(
+ corpus_id=self.corpus.id,
+ drop=False,
+ user_id=self.user.id,
+ description=f'Reindexing project {self.corpus.id}',
+ job_id=f'reindex-{self.corpus.id}',
+ )])
diff --git a/arkindex/project/api_v1.py b/arkindex/project/api_v1.py
index a86c98c529..7f71df176f 100644
--- a/arkindex/project/api_v1.py
+++ b/arkindex/project/api_v1.py
@@ -56,7 +56,7 @@ from arkindex.documents.api.ml import (
TranscriptionCreate,
TranscriptionEdit,
)
-from arkindex.documents.api.search import CorpusSearch
+from arkindex.documents.api.search import CorpusSearch, SearchIndexBuild
from arkindex.images.api import IIIFInformationCreate, IIIFURLCreate, ImageCreate, ImageElements, ImageRetrieve
from arkindex.ponos.api import (
AgentActions,
@@ -195,6 +195,7 @@ api = [
path('corpus/<uuid:corpus>/worker-results/', WorkerResultsDestroy.as_view(), name='worker-delete-results'),
path('corpus/<uuid:corpus>/activity-stats/', CorpusWorkersActivity.as_view(), name='corpus-activity-stats'),
path('corpus/<uuid:corpus>/activity/', WorkerActivityList.as_view(), name='corpus-activity'),
+ path('corpus/<uuid:pk>/index/', SearchIndexBuild.as_view(), name='build-search-index'),
path('export/<uuid:pk>/', DownloadExport.as_view(), name='download-export'),
# Moderation
diff --git a/arkindex/project/config.py b/arkindex/project/config.py
index 301c39d28c..57fa7ba84d 100644
--- a/arkindex/project/config.py
+++ b/arkindex/project/config.py
@@ -152,6 +152,7 @@ def get_settings_parser(base_dir):
job_timeouts_parser.add_option('move_element', type=int, default=3600)
job_timeouts_parser.add_option('initialize_activity', type=int, default=3600)
job_timeouts_parser.add_option('process_delete', type=int, default=3600)
+ job_timeouts_parser.add_option('reindex_corpus', type=int, default=7200)
csrf_parser = parser.add_subparser('csrf', default={})
csrf_parser.add_option('cookie_name', type=str, default='arkindex.csrf')
diff --git a/arkindex/project/rq_overrides.py b/arkindex/project/rq_overrides.py
index 9b45dc9ccc..f9b5c80cba 100644
--- a/arkindex/project/rq_overrides.py
+++ b/arkindex/project/rq_overrides.py
@@ -1,7 +1,9 @@
from typing import Optional
+from django_rq import get_connection
from django_rq.queues import DjangoRQ
from rq.compat import as_text, decode_redis_hash
+from rq.exceptions import NoSuchJobError
from rq.job import Job as BaseJob
from rq.registry import BaseRegistry
@@ -18,6 +20,17 @@ def as_float(value) -> Optional[float]:
return float(value)
+def get_existing_job(job_id):
+ """
+ Returns a job identified by its ID, None otherwise
+ """
+ try:
+ job = Job.fetch(job_id, connection=get_connection())
+ except NoSuchJobError:
+ return None
+ return job
+
+
class Job(BaseJob):
"""
Extension of RQ jobs to provide description updates and completion percentage
diff --git a/arkindex/project/tests/config_samples/defaults.yaml b/arkindex/project/tests/config_samples/defaults.yaml
index 54b67ee979..e4f6a03daf 100644
--- a/arkindex/project/tests/config_samples/defaults.yaml
+++ b/arkindex/project/tests/config_samples/defaults.yaml
@@ -57,6 +57,7 @@ job_timeouts:
initialize_activity: 3600
move_element: 3600
process_delete: 3600
+ reindex_corpus: 7200
worker_results_delete: 3600
jwt_signing_key: null
local_imageserver_id: 1
diff --git a/arkindex/project/tests/config_samples/errors.yaml b/arkindex/project/tests/config_samples/errors.yaml
index 95c6c6c5aa..c8f2a4cef1 100644
--- a/arkindex/project/tests/config_samples/errors.yaml
+++ b/arkindex/project/tests/config_samples/errors.yaml
@@ -40,6 +40,7 @@ job_timeouts:
export_corpus: []
move_element:
a: b
+ reindex_corpus: {}
worker_results_delete: null
jwt_signing_key: null
local_imageserver_id: 1
diff --git a/arkindex/project/tests/config_samples/expected_errors.yaml b/arkindex/project/tests/config_samples/expected_errors.yaml
index ace4d8ec7a..9f9cdfe45d 100644
--- a/arkindex/project/tests/config_samples/expected_errors.yaml
+++ b/arkindex/project/tests/config_samples/expected_errors.yaml
@@ -21,6 +21,7 @@ job_timeouts:
corpus_delete: "invalid literal for int() with base 10: 'lol'"
export_corpus: "int() argument must be a string, a bytes-like object or a number, not 'list'"
move_element: "int() argument must be a string, a bytes-like object or a number, not 'dict'"
+ reindex_corpus: "int() argument must be a string, a bytes-like object or a number, not 'dict'"
worker_results_delete: "int() argument must be a string, a bytes-like object or a number, not 'NoneType'"
ponos:
artifact_max_size: cannot convert float NaN to integer
diff --git a/arkindex/project/tests/config_samples/override.yaml b/arkindex/project/tests/config_samples/override.yaml
index 7f0cce913a..562963c8c7 100644
--- a/arkindex/project/tests/config_samples/override.yaml
+++ b/arkindex/project/tests/config_samples/override.yaml
@@ -71,7 +71,8 @@ job_timeouts:
initialize_activity: 4
move_element: 5
process_delete: 6
- worker_results_delete: 7
+ reindex_corpus: 7
+ worker_results_delete: 8
jwt_signing_key: deadbeef
local_imageserver_id: 45
ponos:
diff --git a/arkindex/project/triggers.py b/arkindex/project/triggers.py
index e569952dae..c73338e817 100644
--- a/arkindex/project/triggers.py
+++ b/arkindex/project/triggers.py
@@ -121,6 +121,19 @@ def add_parent_selection(corpus_id: UUID, parent: Element, user_id: int) -> None
)
+def reindex_corpus(corpus_id: UUID, user_id: int, drop: bool = True, job_id: Optional[str] = None) -> None:
+ """
+ Index elements of a given project
+ """
+ documents_tasks.reindex_corpus.delay(
+ corpus_id=corpus_id,
+ drop=drop,
+ user_id=user_id,
+ description=f'Reindexing project {corpus_id}',
+ job_id=job_id,
+ )
+
+
def initialize_activity(process: Process):
"""
Initialize activity on every process elements for worker versions that are part of its workflow
diff --git a/arkindex/templates/reindex_corpus.html b/arkindex/templates/reindex_corpus.html
new file mode 100644
index 0000000000..581e0a2ca5
--- /dev/null
+++ b/arkindex/templates/reindex_corpus.html
@@ -0,0 +1,10 @@
+{% autoescape off %}
+Hello {{ user.display_name }},
+
+The search index build you started on project {{ corpus.name }} has been finished successfully.
+
+Indexed elements can be explored from the search interface on Arkindex.
+
+--
+Arkindex
+{% endautoescape %}
--
GitLab