add searchvector to commit table and index using GIN

treeherder/config/settings.py

@@ -67,6 +67,7 @@
 INSTALLED_APPS = [
     "django.contrib.auth",
     "django.contrib.contenttypes",
+    "django.contrib.postgres.search",
     # Disable Django's own staticfiles handling in favour of WhiteNoise, for
     # greater consistency between gunicorn and `./manage.py runserver`.
     "whitenoise.runserver_nostatic",

treeherder/model/migrations/0036_commit_search_vector_idx.py

@@ -0,0 +1,24 @@
+# Generated by Django 5.1.2 on 2025-01-24 07:42
+
+import django.contrib.postgres.indexes
+import django.contrib.postgres.search
+from django.db import migrations
+
+
+class Migration(migrations.Migration):
+
+    dependencies = [
+        ("model", "0035_bugscache_optional_bugzilla_ref"),
+    ]
+
+    operations = [
+        migrations.AddIndex(
+            model_name="commit",
+            index=django.contrib.postgres.indexes.GinIndex(
+                django.contrib.postgres.search.SearchVector(
+                    "revision", "author", "comments", config="english"
+                ),
+                name="search_vector_idx",
+            ),
+        ),
+    ]

treeherder/model/models.py

@@ -7,12 +7,13 @@
 
 import newrelic.agent
 from django.contrib.auth.models import User
-from django.contrib.postgres.search import TrigramSimilarity
+from django.contrib.postgres.indexes import GinIndex
 from django.core.cache import cache
 from django.core.exceptions import ObjectDoesNotExist
 from django.core.validators import MinLengthValidator
 from django.db import models, transaction
 from django.db.models import Count, Max, Min, Q, Subquery
+from django.contrib.postgres.search import TrigramSimilarity, SearchVector
 from django.db.utils import ProgrammingError
 from django.forms import model_to_dict
 from django.utils import timezone
@@ -188,11 +189,14 @@ class Commit(models.Model):
     class Meta:
         db_table = "commit"
         unique_together = ("push", "revision")
-
+        indexes = [
+            GinIndex(
+                SearchVector("revision", "author", "comments", config="english"),
+                name="search_vector_idx",
+            ),
+        ]
     def __str__(self):
         return f"{self.push.repository.name} {self.revision}"
-
-
 class MachinePlatform(models.Model):
     id = models.AutoField(primary_key=True)
     os_name = models.CharField(max_length=25)

treeherder/webapp/api/push.py

@@ -9,7 +9,7 @@
 from rest_framework.status import HTTP_400_BAD_REQUEST, HTTP_404_NOT_FOUND
 
 from treeherder.log_parser.failureline import get_group_results
-from treeherder.model.models import Job, JobType, Push, Repository
+from treeherder.model.models import Job, JobType, Push, Repository, Commit
 from treeherder.push_health.builds import get_build_failures
 from treeherder.push_health.compare import get_commit_history
 from treeherder.push_health.linting import get_lint_failures
@@ -22,6 +22,7 @@
 from treeherder.webapp.api.serializers import PushSerializer
 from treeherder.webapp.api.utils import to_datetime, to_timestamp
 
+from django.contrib.postgres.search import SearchVector, SearchQuery
 logger = logging.getLogger(__name__)
 
 
@@ -42,7 +43,6 @@ def list(self, request, project):
 
         # This will contain some meta data about the request and results
         meta = {}
-
         # support ranges for date as well as revisions(changes) like old tbpl
         for param in [
             "fromchange",
@@ -60,7 +60,6 @@ def list(self, request, project):
         all_repos = request.query_params.get("all_repos")
 
         pushes = Push.objects.order_by("-time")
-
         if not all_repos:
             try:
                 repository = Repository.objects.get(name=project)
@@ -71,6 +70,14 @@ def list(self, request, project):
 
             pushes = pushes.filter(repository=repository)
 
+        search_param = filter_params.get("search")
+        if search_param:      
+            filtered_commits = Commit.objects.annotate(
+                search=SearchVector("revision", "author", "comments", config="english")
+            ).filter(
+                search=SearchQuery(search_param, config="english")
+            ).values_list("push_id", flat=True)
+        pushes = pushes.filter(id__in=filtered_commits)
         for param, value in meta.items():
             if param == "fromchange":
                 revision_field = "revision__startswith" if len(value) < 40 else "revision"