RUBY-3353 Add docs for the Atlas search index management API

docs/reference/schema-operations.txt

@@ -15,4 +15,5 @@ including managing databases, collections, indexes and users.
   /reference/database-tasks
   /reference/collection-tasks
   /reference/indexing
+  /reference/search-indexes
   /reference/collations

docs/reference/search-indexes.txt

@@ -0,0 +1,129 @@
+********************
+Atlas Search Indexes
+********************
+
+.. default-domain:: mongodb
+
+.. contents:: On this page
+   :local:
+   :backlinks: none
+   :depth: 1
+   :class: singlecol
+
+If you are using a database hosted by MongoDB Atlas, the driver provides the
+ability to create, drop and view `Atlas search indexes <https://www.mongodb.com/docs/atlas/atlas-search/>`_
+on a collection through the ``search_indexes`` attribute:
+
+.. code-block:: ruby
+
+  client = Mongo::Client.new(your_atlas_uri, database: 'music')
+  client[:bands].search_indexes
+  # => #<Mongo::SearchIndex::View:0x000055e2822b9318 @collection=#<Mongo::Collection:0x660 namespace=music.bands> ...>
+
+
+Creating Search Indexes
+=======================
+
+Search indexes can be created one at a time, or several can be created in
+parallel in a single operation.
+
+To create a single index, use ``search_indexes#create_one``, passing the index
+definition as the first argument, and an optional name for the index as the
+second argument.
+
+.. code-block:: ruby
+
+  client[:bands].search_indexes.create_one({ dynamic: true })
+
+  client[:bands].search_indexes.create_one(
+    {
+      dynamic: false,
+      fields: {
+        name: { type: 'string', analyzer: 'lucene.simple' }
+      }
+    },
+    'band-name-index'
+  )
+
+To create multiple indexes, use ``search_indexes#create_many`` which accepts
+an array of index specifications. Unlike ``create_one``, each index
+specification is a hash with at least a ``definition`` key, which
+defines the index. Each has may also specify a ``name`` key, to name
+the index.
+
+.. code-block:: ruby
+
+  client[:bands].search_indexes.create_many([
+    { definition: { dynamic: true } },
+    { name: 'band-name-index,
+      definition: {
+        dynamic: false,
+        fields: {
+          name: { type: 'string', analyzer: 'lucene.simple' }
+        }
+      }
+    },
+  ])
+
+Note that whether you call ``create_one`` or ``create_many``, the
+method will return immediately, before the indexes are created. The
+indexes are then created in the background, asynchronously.
+
+
+Update Search Indexes
+=====================
+
+You can programmatically update an Atlas search index. For example, you
+might do this to change the analyzer used, or to provide an explicit field
+mapping, instead of a dynamic one. To do this, use the ``search_indexes#update_one``
+method:
+
+.. code-block:: ruby
+
+  client[:bands].search_indexes.update_one(new_definition, id: index_id)
+
+  client[:bands].search_indexes.update_one(new_definition, name: index_name)
+
+Indexes may be identified by either id, or name, but you must specify one
+or the other. The new index definition must be a complete definition--it will
+take precedence as specified over the existing definition.
+
+To get the id or name of an index that you wish to update, you can
+`list the search indexes <#listing-search-indexes>`_.
+
+
+Dropping Search Indexes
+=======================
+
+To drop Atlas search indexes, call ``search_indexes#drop_one`` and
+provide either the ``id`` or the ``name`` of the index you wish to
+drop.
+
+.. code-block:: ruby
+
+  client[:bands].search_indexes.drop_one(id: index_id)
+
+  client[:bands].search_indexes.drop_one(name: index_name)
+
+In either case, the method will return immediately and the index will
+be dropped in the background, asynchronously.
+
+To get the id or name of an index that you wish to drop, you can
+`list the search indexes <#listing-search-indexes>`_.
+
+
+Listing Search Indexes
+======================
+
+To list the available search indexes, iterate over the
+``search_indexes`` object:
+
+.. code-block:: ruby
+
+  client[:bands].search_indexes.each do |index_spec|
+    p index_spec['id']
+    p index_spec['name']
+    p index_spec['status']
+    p index_spec['queryable']
+    p index_spec['latestDefinition']
+  end

docs/release-notes.txt

@@ -42,10 +42,12 @@ This release includes the following new features:
   when Azure Key Vault is used for client side encryption.
 - `Queryable Encryption <https://www.mongodb.com/docs/upcoming/core/queryable-encryption/queryable-encryption/>`_ support is extended.
 - Added support for Queryable Encryption Range Indexes.
-- A `crypt_shared <https://www.mongodb.com/docs/manual/core/queryable-encryption/reference/shared-library/#download-the-automatic-encryption-shared-library>`_ 
+- A `crypt_shared <https://www.mongodb.com/docs/manual/core/queryable-encryption/reference/shared-library/#download-the-automatic-encryption-shared-library>`_
   library can be now used instead of ``mongocryptd``.
 - Added support for AWS IAM Roles for service accounts, EKS in particular.
 - AWS Credentials are now cached when possible.
+- Added support for creating and managing `Atlas search indexes <https://www.mongodb.com/docs/atlas/atlas-search/>`_ via the
+  Ruby driver.
 
 .. _release-notes-2.18: