def remove_search_indexes(project_slug, version_slug=None):
"""Wrapper around ``remove_indexed_files`` to make it a task."""
remove_indexed_files(
model=HTMLFile,
project_slug=project_slug,
version_slug=version_slug,
)
def test_search_ignore(self, api_client):
project = Project.objects.get(slug='docs')
version = project.versions.all().first()
page_index = HTMLFile.objects.get(
version=version,
path='index.html',
)
page_guides = HTMLFile.objects.get(
version=version,
path='guides/index.html',
)
search_params = {
'project': project.slug,
'version': version.slug,
'q': '"content from"',
}
# Query with files not ignored.
assert page_index.ignore is None
assert page_guides.ignore is None
resp = self.get_search(api_client, search_params)
assert resp.status_code == 200
results = resp.data['results']
assert len(results) == 2
assert results[0]['path'] == '/en/latest/index.html'
assert results[1]['path'] == '/en/latest/guides/index.html'
# Query with guides/index.html ignored.
page_guides.ignore = True
page_guides.save()
remove_indexed_files(HTMLFile, project.slug, version.slug)
index_new_files(HTMLFile, version, page_index.build)
resp = self.get_search(api_client, search_params)
assert resp.status_code == 200
results = resp.data['results']
assert len(results) == 1
assert results[0]['path'] == '/en/latest/index.html'
# Query with index.html and guides/index.html ignored.
page_index.ignore = True
page_index.save()
remove_indexed_files(HTMLFile, project.slug, version.slug)
index_new_files(HTMLFile, version, page_index.build)
resp = self.get_search(api_client, search_params)
assert resp.status_code == 200
results = resp.data['results']
assert len(results) == 0
def test_remove_only_one_project_index(self, api_client, all_projects):
project = 'kuma'
assert self.has_results(api_client, project, LATEST)
assert self.has_results(api_client, project, STABLE)
utils.remove_indexed_files(
HTMLFile,
project_slug=project,
)
# Deletion of indices from ES happens async,
# so we need to wait a little before checking for results.
time.sleep(1)
assert self.has_results(api_client, project, LATEST) is False
assert self.has_results(api_client, project, STABLE) is False
# Check that other projects weren't deleted
for project in ['pipeline', 'docs']:
for version in [LATEST, STABLE]:
assert self.has_results(api_client, project, version) is True
def _sync_imported_files(version, build):
"""
Sync/Update/Delete ImportedFiles objects of this version.
:param version: Version instance
:param build: Build id
"""
# Index new HTMLFiles to ElasticSearch
index_new_files(model=HTMLFile, version=version, build=build)
# Remove old HTMLFiles from ElasticSearch
remove_indexed_files(
model=HTMLFile,
project_slug=version.project.slug,
version_slug=version.slug,
build_id=build,
)
# Delete SphinxDomain objects from previous versions
# This has to be done before deleting ImportedFiles and not with a cascade,
# because multiple Domain's can reference a specific HTMLFile.
(
SphinxDomain.objects
.filter(project=version.project, version=version)
.exclude(build=build)
.delete()
)
# Delete ImportedFiles objects (including HTMLFiles)
# from the previous build of the version.
(
ImportedFile.objects
.filter(project=version.project, version=version)
.exclude(build=build)
.delete()
)
