Search: refactor serializer's context

We were generating the context of the serializer
from outside, but we can do this from
the serializer itself,
this way we don't have to duplicate this logic.

Author: stsewd

readthedocs/search/api/v2/serializers.py

@@ -61,9 +61,11 @@ class PageSearchSerializer(serializers.Serializer):
     """
     Page serializer.
 
-    If ``projects_data`` is passed into the context, the serializer
-    will try to use that to generate the link before querying the database.
-    It's a dictionary mapping the project slug to a ProjectData object.
+    If ``projects`` is passed in the constructor, the serializer
+    will pre-generate a cache with that information,
+    this is to avoid querying the database again for each result.
+
+    :param projects: A list of tuples of project and version.
     """
 
     type = serializers.CharField(default="page", source=None, read_only=True)
@@ -76,6 +78,28 @@ class PageSearchSerializer(serializers.Serializer):
     highlights = PageHighlightSerializer(source="meta.highlight", default=dict)
     blocks = serializers.SerializerMethodField()
 
+    def __init__(self, *args, projects=None, **kwargs):
+        if projects:
+            context = kwargs.setdefault("context", {})
+            context["projects_data"] = {
+                project.slug: self._build_project_data(project, version.slug)
+                for project, version in projects
+            }
+        super().__init__(*args, **kwargs)
+
+    def _build_project_data(self, project, version_slug):
+        """Build a `ProjectData` object given a project and its version."""
+        url = project.get_docs_url(version_slug=version_slug)
+        project_alias = project.superprojects.values_list("alias", flat=True).first()
+        version_data = VersionData(
+            slug=version_slug,
+            docs_url=url,
+        )
+        return ProjectData(
+            alias=project_alias,
+            version=version_data,
+        )
+
     def _get_project_data(self, obj):
         """
         Get and cache the project data.

readthedocs/search/api/v2/views.py

@@ -14,11 +14,7 @@
 from readthedocs.projects.models import Feature, Project
 from readthedocs.search import tasks
 from readthedocs.search.api import SearchPagination
-from readthedocs.search.api.v2.serializers import (
-    PageSearchSerializer,
-    ProjectData,
-    VersionData,
-)
+from readthedocs.search.api.v2.serializers import PageSearchSerializer
 from readthedocs.search.faceted_search import PageSearch
 
 log = structlog.get_logger(__name__)
@@ -80,45 +76,15 @@ def _validate_query_params(self):
             raise ValidationError(errors)
 
     @lru_cache(maxsize=1)
-    def _get_all_projects_data(self):
-        """
-        Return a dictionary of the project itself and all its subprojects.
-
-        Example:
-
-        .. code::
-
-           {
-               "requests": ProjectData(
-                   alias='alias',
-                   version=VersionData(
-                        "latest",
-                        "https://requests.readthedocs.io/en/latest/",
-                    ),
-               ),
-               "requests-oauth": ProjectData(
-                   alias=None,
-                   version=VersionData(
-                       "latest",
-                       "https://requests-oauth.readthedocs.io/en/latest/",
-                   ),
-               ),
-           }
-
-        .. note:: The response is cached into the instance.
-
-        :rtype: A dictionary of project slugs mapped to a `VersionData` object.
-        """
+    def _get_projects_to_search(self):
+        """Get all projects to search."""
         main_version = self._get_version()
         main_project = self._get_project()
 
         if not self._has_permission(self.request.user, main_version):
-            return {}
-
-        projects_data = {
-            main_project.slug: self._get_project_data(main_project, main_version),
-        }
+            return []
 
+        projects_to_search = [(main_project, main_version)]
         subprojects = Project.objects.filter(superprojects__parent_id=main_project.id)
         for subproject in subprojects:
             version = self._get_project_version(
@@ -136,24 +102,9 @@ def _get_all_projects_data(self):
                 )
 
             if version and self._has_permission(self.request.user, version):
-                projects_data[subproject.slug] = self._get_project_data(
-                    subproject, version
-                )
+                projects_to_search.append((subproject, version))
 
-        return projects_data
-
-    def _get_project_data(self, project, version):
-        """Build a `ProjectData` object given a project and its version."""
-        url = project.get_docs_url(version_slug=version.slug)
-        project_alias = project.superprojects.values_list("alias", flat=True).first()
-        version_data = VersionData(
-            slug=version.slug,
-            docs_url=url,
-        )
-        return ProjectData(
-            alias=project_alias,
-            version=version_data,
-        )
+        return projects_to_search
 
     def _get_project_version(self, project, version_slug, include_hidden=True):
         """
@@ -219,8 +170,8 @@ def get_queryset(self):
            is compatible with DRF's paginator.
         """
         projects = {
-            project: project_data.version.slug
-            for project, project_data in self._get_all_projects_data().items()
+            project.slug: version.slug
+            for project, version in self._get_projects_to_search()
         }
         # Check to avoid searching all projects in case it's empty.
         if not projects:
@@ -236,11 +187,6 @@ def get_queryset(self):
         )
         return queryset
 
-    def get_serializer_context(self):
-        context = super().get_serializer_context()
-        context["projects_data"] = self._get_all_projects_data()
-        return context
-
     def get(self, request, *args, **kwargs):
         self._validate_query_params()
         result = self.list()
@@ -255,7 +201,9 @@ def list(self):
             self.request,
             view=self,
         )
-        serializer = self.get_serializer(page, many=True)
+        serializer = self.get_serializer(
+            page, many=True, projects=self._get_projects_to_search()
+        )
         return self.paginator.get_paginated_response(serializer.data)
 
 

readthedocs/search/tests/test_api.py

@@ -342,9 +342,9 @@ def test_doc_search_unexisting_version(self, api_client, project):
         resp = self.get_search(api_client, search_params)
         assert resp.status_code == 404
 
-    @mock.patch.object(PageSearchAPIView, '_get_all_projects_data', dict)
+    @mock.patch.object(PageSearchAPIView, "_get_projects_to_search", list)
     def test_get_all_projects_returns_empty_results(self, api_client, project):
-        """If there is a case where `_get_all_projects` returns empty, we could be querying all projects."""
+        """If there is a case where `_get_projects_to_search` returns empty, we could be querying all projects."""
 
         # `documentation` word is present both in `kuma` and `docs` files
         # and not in `pipeline`, so search with this phrase but filter through project

readthedocs/search/views.py

@@ -10,9 +10,7 @@
 from readthedocs.projects.models import Feature, Project
 from readthedocs.search.api.v2.serializers import (
     PageSearchSerializer,
-    ProjectData,
     ProjectSearchSerializer,
-    VersionData,
 )
 from readthedocs.search.faceted_search import ALL_FACETS, PageSearch, ProjectSearch
 
@@ -91,26 +89,6 @@ def _get_project(self, project_slug):
         project = get_object_or_404(queryset, slug=project_slug)
         return project
 
-    def _get_project_data(self, project, version_slug):
-        docs_url = project.get_docs_url(version_slug=version_slug)
-        version_data = VersionData(
-            slug=version_slug,
-            docs_url=docs_url,
-        )
-        project_data = {
-            project.slug: ProjectData(
-                alias=None,
-                version=version_data,
-            )
-        }
-        return project_data
-
-    def get_serializer_context(self, project, version_slug):
-        context = {
-            'projects_data': self._get_project_data(project, version_slug),
-        }
-        return context
-
     def get(self, request, project_slug):
         project_obj = self._get_project(project_slug)
         use_advanced_query = not project_obj.has_feature(
@@ -132,8 +110,7 @@ def get(self, request, project_slug):
             use_advanced_query=use_advanced_query,
         )
 
-        context = self.get_serializer_context(project_obj, user_input.version)
-        results = PageSearchSerializer(results, many=True, context=context).data
+        results = PageSearchSerializer(results, many=True).data
 
         template_context = user_input._asdict()
         template_context.update({