From 7b065029cbd40bb57ec390ce275c7f26f600e0cc Mon Sep 17 00:00:00 2001
From: Graham Hukill <ghukill@gmail.com>
Date: Fri, 30 Jan 2026 11:37:56 -0500
Subject: [PATCH] Add --skip-embeddings flag to reindex-source

Why these changes are being introduced:

There might be times we want to re-index a source to Opensearch without indexing
embeddings as a secondary pass.

How this addresses that need:

Add --skip-embeddings flag to reindex-source CLI command.

Side effects of this change:
* None

Relevant ticket(s):
* None
---
 tests/test_cli.py | 39 +++++++++++++++++++++++++++++++++++++++
 tim/cli.py        | 44 ++++++++++++++++++++++++++++----------------
 2 files changed, 67 insertions(+), 16 deletions(-)

diff --git a/tests/test_cli.py b/tests/test_cli.py
index 9dc6ebb..25054d3 100644
--- a/tests/test_cli.py
+++ b/tests/test_cli.py
@@ -415,3 +415,42 @@ def test_reindex_source_success(
         "Reindex source complete: "
         f'{{"index": {json.dumps(mock_bulk_index())}' in caplog.text
     )
+
+
+@patch("tim.opensearch.create_index")
+@patch("tim.opensearch.promote_index")
+@patch("tim.opensearch.get_index_aliases")
+@patch("tim.opensearch.bulk_update")
+@patch("tim.opensearch.bulk_index")
+def test_reindex_source_skip_embeddings(
+    mock_bulk_index,
+    mock_bulk_update,
+    mock_get_index_aliases,
+    mock_promote_index,
+    mock_create_index,
+    caplog,
+    monkeypatch,
+    runner,
+):
+    monkeypatch.delenv("TIMDEX_OPENSEARCH_ENDPOINT", raising=False)
+    mock_get_index_aliases.return_value = ["alma", "all-current", "timdex"]
+    mock_bulk_index.return_value = {
+        "created": 1000,
+        "updated": 0,
+        "errors": 0,
+        "total": 1000,
+    }
+
+    result = runner.invoke(
+        main,
+        [
+            "reindex-source",
+            "--source",
+            "alma",
+            "--skip-embeddings",
+            "tests/fixtures/dataset",
+        ],
+    )
+    assert result.exit_code == EXIT_CODES["success"]
+    assert "Skipping embeddings update." in caplog.text
+    mock_bulk_update.assert_not_called()
diff --git a/tim/cli.py b/tim/cli.py
index 093a94d..4d6d9ff 100644
--- a/tim/cli.py
+++ b/tim/cli.py
@@ -430,6 +430,12 @@ def bulk_update_embeddings(
     help="Alias to promote the index to in addition to the primary alias. May "
     "be repeated to promote the index to multiple aliases at once.",
 )
+@click.option(
+    "--skip-embeddings",
+    is_flag=True,
+    default=False,
+    help="Skip the secondary update of documents with embeddings.",
+)
 @click.argument(
     "dataset_path",
     type=click.Path(),
@@ -441,6 +447,7 @@ def reindex_source(
     ctx: click.Context,
     source: str,
     alias: tuple[str],
+    skip_embeddings: bool,  # noqa: FBT001
     dataset_path: str,
 ) -> None:
     """Perform a full refresh for a source in Opensearch for all current records.
@@ -488,23 +495,28 @@ def reindex_source(
         logger.error(f"Bulk indexing failed: {exception}")  # noqa: TRY400
 
     # bulk index embeddings
-    logger.info("Reindexing embeddings.")
     update_results = {"updated": 0, "errors": 0, "total": 0}
-    embeddings = td.embeddings.read_dicts_iter(
-        table="current_embeddings",
-        columns=[
-            "timdex_record_id",
-            "embedding_strategy",
-            "embedding_object",
-        ],
-        source=source,
-        action="index",
-    )
-    embeddings_to_index = helpers.format_embeddings(embeddings)
-    try:
-        update_results.update(tim_os.bulk_update(client, index, embeddings_to_index))
-    except BulkOperationError as exception:
-        logger.error(f"Bulk update with embeddings failed: {exception}")  # noqa: TRY400
+    if skip_embeddings:
+        logger.info("Skipping embeddings update.")
+    else:
+        logger.info("Reindexing embeddings.")
+        embeddings = td.embeddings.read_dicts_iter(
+            table="current_embeddings",
+            columns=[
+                "timdex_record_id",
+                "embedding_strategy",
+                "embedding_object",
+            ],
+            source=source,
+            action="index",
+        )
+        embeddings_to_index = helpers.format_embeddings(embeddings)
+        try:
+            update_results.update(tim_os.bulk_update(client, index, embeddings_to_index))
+        except BulkOperationError as exception:
+            logger.error(  # noqa: TRY400
+                f"Bulk update with embeddings failed: {exception}"
+            )
 
     summary_results = {"index": index_results, "update": update_results}
     logger.info(f"Reindex source complete: {json.dumps(summary_results)}")