Use CTEs for nested groups and authorizations

This commit introduces the usage of Common Table Expressions (CTEs) to
efficiently retrieve nested group hierarchies, without having to rely on
the "routes" table (which is an _incredibly_ inefficient way of getting
the data). This requires a patch to ActiveRecord (found in the added
initializer) to work properly as ActiveRecord doesn't support WITH
statements properly out of the box.

Unfortunately MySQL provides no efficient way of getting nested groups.
For example, the old routes setup could easily take 5-10 seconds
depending on the amount of "routes" in a database. Providing vastly
different logic for both MySQL and PostgreSQL will negatively impact the
development process. Because of this the various nested groups related
methods return empty relations when used in combination with MySQL.

For project authorizations the logic is split up into two classes:

* Gitlab::ProjectAuthorizations::WithNestedGroups
* Gitlab::ProjectAuthorizations::WithoutNestedGroups

Both classes get the fresh project authorizations (= as they should be
in the "project_authorizations" table), including nested groups if
PostgreSQL is used. The logic of these two classes is quite different
apart from their public interface. This complicates development a bit,
but unfortunately there is no way around this.

This commit also introduces Gitlab::GroupHierarchy. This class can be
used to get the ancestors and descendants of a base relation, or both by
using a UNION. This in turn is used by methods such as:

* Namespace#ancestors
* Namespace#descendants
* User#all_expanded_groups

Again this class relies on CTEs and thus only works on PostgreSQL. The
Namespace methods will return an empty relation when MySQL is used,
while User#all_expanded_groups will return only the groups a user is a
direct member of.

Performance wise the impact is quite large. For example, on GitLab.com
Namespace#descendants used to take around 580 ms to retrieve data for a
particular user. Using CTEs we are able to reduce this down to roughly 1
millisecond, returning the exact same data.

== On The Fly Refreshing

Refreshing of authorizations on the fly (= when
users.authorized_projects_populated was not set) is removed with this
commit. This simplifies the code, and ensures any queries used for
authorizations are not mutated because they are executed in a Rails
scope (e.g. Project.visible_to_user).

This commit includes a migration to schedule refreshing authorizations
for all users, ensuring all of them have their authorizations in place.
Said migration schedules users in batches of 5000, with 5 minutes
between every batch to smear the load around a bit.

== Spec Changes

This commit also introduces some changes to various specs. For example,
some specs for ProjectTeam assumed that creating a personal project
would _not_ lead to the owner having access, which is incorrect. Because
we also no longer refresh authorizations on the fly for new users some
code had to be added to the "empty_project" factory. This chunk of code
ensures that the owner's permissions are refreshed after creating the
project, something that is normally done in Projects::CreateService.

2 files changed, 163 insertions, 0 deletions

diff --git a/lib/gitlab/project_authorizations/with_nested_groups.rb b/lib/gitlab/project_authorizations/with_nested_groups.rb
new file mode 100644
index 00000000000..79c082c08fd
--- /dev/null
+++ b/lib/gitlab/project_authorizations/with_nested_groups.rb
@@ -0,0 +1,128 @@
+module Gitlab
+  module ProjectAuthorizations
+    # Calculating new project authorizations when supporting nested groups.
+    #
+    # This class relies on Common Table Expressions to efficiently get all data,
+    # including data for nested groups. As a result this class can only be used
+    # on PostgreSQL.
+    class WithNestedGroups
+      attr_reader :user
+
+      # user - The User object for which to calculate the authorizations.
+      def initialize(user)
+        @user = user
+      end
+
+      def calculate
+        cte = recursive_cte
+        cte_alias = cte.table.alias(Group.table_name)
+        projects = Project.arel_table
+        links = ProjectGroupLink.arel_table
+
+        # These queries don't directly use the user object so they don't depend
+        # on the state of said object, ensuring the produced queries are always
+        # the same.
+        relations = [
+          # The project a user has direct access to.
+          user.projects.select_for_project_authorization,
+
+          # The personal projects of the user.
+          user.personal_projects.select_as_master_for_project_authorization,
+
+          # Projects that belong directly to any of the groups the user has
+          # access to.
+          Namespace.
+            unscoped.
+            select([alias_as_column(projects[:id], 'project_id'),
+                    cte_alias[:access_level]]).
+            from(cte_alias).
+            joins(:projects),
+
+          # Projects shared with any of the namespaces the user has access to.
+          Namespace.
+            unscoped.
+            select([links[:project_id],
+                    least(cte_alias[:access_level],
+                          links[:group_access],
+                          'access_level')]).
+            from(cte_alias).
+            joins('INNER JOIN project_group_links ON project_group_links.group_id = namespaces.id').
+            joins('INNER JOIN projects ON projects.id = project_group_links.project_id').
+            joins('INNER JOIN namespaces p_ns ON p_ns.id = projects.namespace_id').
+            where('p_ns.share_with_group_lock IS FALSE')
+        ]
+
+        union = Gitlab::SQL::Union.new(relations)
+
+        ProjectAuthorization.
+          unscoped.
+          with.
+          recursive(cte.to_arel).
+          select_from_union(union)
+      end
+
+      private
+
+      # Builds a recursive CTE that gets all the groups the current user has
+      # access to, including any nested groups.
+      def recursive_cte
+        cte = Gitlab::SQL::RecursiveCTE.new(:namespaces_cte)
+        members = Member.arel_table
+        namespaces = Namespace.arel_table
+
+        # Namespaces the user is a member of.
+        cte << user.groups.
+          select([namespaces[:id], members[:access_level]]).
+          except(:order)
+
+        # Sub groups of any groups the user is a member of.
+        cte << Group.select([namespaces[:id],
+                             greatest(members[:access_level],
+                                      cte.table[:access_level], 'access_level')]).
+          joins(join_cte(cte)).
+          joins(join_members).
+          except(:order)
+
+        cte
+      end
+
+      # Builds a LEFT JOIN to join optional memberships onto the CTE.
+      def join_members
+        members = Member.arel_table
+        namespaces = Namespace.arel_table
+
+        cond = members[:source_id].
+          eq(namespaces[:id]).
+          and(members[:source_type].eq('Namespace')).
+          and(members[:requested_at].eq(nil)).
+          and(members[:user_id].eq(user.id))
+
+        Arel::Nodes::OuterJoin.new(members, Arel::Nodes::On.new(cond))
+      end
+
+      # Builds an INNER JOIN to join namespaces onto the CTE.
+      def join_cte(cte)
+        namespaces = Namespace.arel_table
+        cond = cte.table[:id].eq(namespaces[:parent_id])
+
+        Arel::Nodes::InnerJoin.new(cte.table, Arel::Nodes::On.new(cond))
+      end
+
+      def greatest(left, right, column_alias)
+        sql_function('GREATEST', [left, right], column_alias)
+      end
+
+      def least(left, right, column_alias)
+        sql_function('LEAST', [left, right], column_alias)
+      end
+
+      def sql_function(name, args, column_alias)
+        alias_as_column(Arel::Nodes::NamedFunction.new(name, args), column_alias)
+      end
+
+      def alias_as_column(value, alias_to)
+        Arel::Nodes::As.new(value, Arel::Nodes::SqlLiteral.new(alias_to))
+      end
+    end
+  end
+end
diff --git a/lib/gitlab/project_authorizations/without_nested_groups.rb b/lib/gitlab/project_authorizations/without_nested_groups.rb
new file mode 100644
index 00000000000..627e8c5fba2
--- /dev/null
+++ b/lib/gitlab/project_authorizations/without_nested_groups.rb
@@ -0,0 +1,35 @@
+module Gitlab
+  module ProjectAuthorizations
+    # Calculating new project authorizations when not supporting nested groups.
+    class WithoutNestedGroups
+      attr_reader :user
+
+      # user - The User object for which to calculate the authorizations.
+      def initialize(user)
+        @user = user
+      end
+
+      def calculate
+        relations = [
+          # Projects the user is a direct member of
+          user.projects.select_for_project_authorization,
+
+          # Personal projects
+          user.personal_projects.select_as_master_for_project_authorization,
+
+          # Projects of groups the user is a member of
+          user.groups_projects.select_for_project_authorization,
+
+          # Projects shared with groups the user is a member of
+          user.groups.joins(:shared_projects).select_for_project_authorization
+        ]
+
+        union = Gitlab::SQL::Union.new(relations)
+
+        ProjectAuthorization.
+          unscoped.
+          select_from_union(union)
+      end
+    end
+  end
+end