From b774ebacd2ac00a3de6d7de72a0316dbb2ee8b2f Mon Sep 17 00:00:00 2001
From: Jeremy Prevost <JPrevost@users.noreply.github.com>
Date: Wed, 11 Feb 2026 13:32:44 -0500
Subject: [PATCH] Adds support to exclude fields from _source

Why are these changes being introduced:

* We are starting to add very large fields into our OpenSearch _source,
  which we are concerned may cause performance issues

Relevant ticket(s):

* https://mitlibraries.atlassian.net/browse/USE-406

How does this address that need:

* Introduces an ENV variable to specify fields to exclude from _source
---
 README.md                      |  2 ++
 app/models/opensearch.rb       | 10 ++++++++++
 test/models/opensearch_test.rb | 18 ++++++++++++++++++
 3 files changed, 30 insertions(+)

diff --git a/README.md b/README.md
index 5b8a259..800139c 100644
--- a/README.md
+++ b/README.md
@@ -181,6 +181,8 @@ locally.
   confused.
   ```
 
+- `OPENSEARCH_SOURCE_EXCLUDES` comma separated list of fields to exclude from the OpenSearch `_source` field. Leave unset to return all fields.
+  - recommended value: `embedding_full_record,fulltext`
 - `PLATFORM_NAME`: The value set is added to the header after the MIT Libraries logo. The logic and CSS for this comes from our theme gem.
 - `PREFERRED_DOMAIN` - set this to the domain you would like to to use. Any
   other requests that come to the app will redirect to the root of this domain.
diff --git a/app/models/opensearch.rb b/app/models/opensearch.rb
index e5c2e5e..dbac195 100644
--- a/app/models/opensearch.rb
+++ b/app/models/opensearch.rb
@@ -41,6 +41,16 @@ def build_query(from)
       sort:
     }
 
+    # If ENV OPENSEARCH_SOURCE_EXCLUDES is set, use the values in it's comma-separated list;
+    #   otherwise leave out the _source attribute entirely (which will return all fields in _source)
+    # excludes are used to prevent large fields from being returned in the search results, which can cause performance issues
+    # these fields are still searchable, just not returned in the search results
+    if ENV['OPENSEARCH_SOURCE_EXCLUDES'].present?
+      query_hash[:_source] = {
+        excludes: ENV['OPENSEARCH_SOURCE_EXCLUDES'].split(',').map(&:strip)
+      }
+    end
+
     query_hash[:highlight] = highlight if @highlight
     query_hash.to_json
   end
diff --git a/test/models/opensearch_test.rb b/test/models/opensearch_test.rb
index 491d0be..dc4ff39 100644
--- a/test/models/opensearch_test.rb
+++ b/test/models/opensearch_test.rb
@@ -387,4 +387,22 @@ class OpensearchTest < ActiveSupport::TestCase
     json = JSON.parse(os.build_query(0))
     assert_equal Opensearch::MAX_SIZE, json['size']
   end
+
+  test 'can exclude fields from _source' do
+    ClimateControl.modify(OPENSEARCH_SOURCE_EXCLUDES: 'field1,field2') do
+      os = Opensearch.new
+      os.instance_variable_set(:@params, {})
+      json = JSON.parse(os.build_query(0))
+      assert_equal %w[field1 field2], json['_source']['excludes']
+    end
+  end
+
+  test 'does not include _source if OPENSEARCH_SOURCE_EXCLUDES is not set' do
+    ClimateControl.modify(OPENSEARCH_SOURCE_EXCLUDES: nil) do
+      os = Opensearch.new
+      os.instance_variable_set(:@params, {})
+      json = JSON.parse(os.build_query(0))
+      refute json.key?('_source')
+    end
+  end
 end