DEV: Also make sure locale detection skips PMs that are not group PMs when public content only (#1457)

In the earlier PR #1432, when `SiteSetting.ai_translation_backfill_limit_to_public_content = false`, we **translate** PMs but **skip translating** PMs that do not involve groups.

This commit covers the missing case on **locale detection**.

Author: nattsw

app/jobs/scheduled/posts_locale_detection_backfill.rb

@@ -20,12 +20,17 @@ def execute(args)
           .where.not(raw: [nil, ""])
 
       if SiteSetting.ai_translation_backfill_limit_to_public_content
-        public_categories = Category.where(read_restricted: false).pluck(:id)
         posts =
           posts
             .joins(:topic)
-            .where(topics: { category_id: public_categories })
-            .where(topics: { archetype: "regular" })
+            .where(topics: { category_id: Category.where(read_restricted: false).select(:id) })
+            .where("archetype != ?", Archetype.private_message)
+      else
+        posts =
+          posts.joins(:topic).where(
+            "topics.archetype != ? OR EXISTS (SELECT 1 FROM topic_allowed_groups WHERE topic_id = topics.id)",
+            Archetype.private_message,
+          )
       end
 
       if SiteSetting.ai_translation_backfill_max_age_days > 0

app/jobs/scheduled/topics_locale_detection_backfill.rb

@@ -15,7 +15,17 @@ def execute(args)
       topics = Topic.where(locale: nil, deleted_at: nil).where("topics.user_id > 0")
 
       if SiteSetting.ai_translation_backfill_limit_to_public_content
-        topics = topics.where(category_id: Category.where(read_restricted: false).select(:id))
+        topics =
+          topics.where(category_id: Category.where(read_restricted: false).select(:id)).where(
+            "archetype != ?",
+            Archetype.private_message,
+          )
+      else
+        topics =
+          topics.where(
+            "archetype != ? OR EXISTS (SELECT 1 FROM topic_allowed_groups WHERE topic_id = topics.id)",
+            Archetype.private_message,
+          )
       end
 
       if SiteSetting.ai_translation_backfill_max_age_days > 0

spec/jobs/scheduled/posts_locale_detection_backfill_spec.rb

@@ -82,23 +82,40 @@
 
   describe "with public content limitation" do
     fab!(:private_category) { Fabricate(:private_category, group: Group[:staff]) }
-    fab!(:private_topic) { Fabricate(:topic, category: private_category) }
-    fab!(:private_post) { Fabricate(:post, topic: private_topic, locale: nil) }
+    fab!(:private_cat_topic) { Fabricate(:topic, category: private_category) }
+    fab!(:private_cat_post) { Fabricate(:post, topic: private_cat_topic, locale: nil) }
+
+    fab!(:group)
+    fab!(:group_pm_topic) { Fabricate(:private_message_topic, allowed_groups: [group]) }
+    fab!(:group_pm_post) { Fabricate(:post, topic: group_pm_topic, locale: nil) }
+
+    fab!(:pm_topic) { Fabricate(:private_message_topic) }
+    fab!(:pm_post) { Fabricate(:post, topic: pm_topic, locale: nil) }
 
     before { SiteSetting.ai_translation_backfill_limit_to_public_content = true }
 
     it "only processes posts from public categories" do
       DiscourseAi::Translation::PostLocaleDetector.expects(:detect_locale).with(post).once
-      DiscourseAi::Translation::PostLocaleDetector.expects(:detect_locale).with(private_post).never
+      DiscourseAi::Translation::PostLocaleDetector
+        .expects(:detect_locale)
+        .with(private_cat_post)
+        .never
+      DiscourseAi::Translation::PostLocaleDetector.expects(:detect_locale).with(group_pm_post).never
+      DiscourseAi::Translation::PostLocaleDetector.expects(:detect_locale).with(pm_post).never
 
       job.execute({})
     end
 
-    it "processes all posts when setting is disabled" do
+    it "processes all public content and group PMs and private categories when setting is disabled" do
       SiteSetting.ai_translation_backfill_limit_to_public_content = false
 
       DiscourseAi::Translation::PostLocaleDetector.expects(:detect_locale).with(post).once
-      DiscourseAi::Translation::PostLocaleDetector.expects(:detect_locale).with(private_post).once
+      DiscourseAi::Translation::PostLocaleDetector
+        .expects(:detect_locale)
+        .with(private_cat_post)
+        .once
+      DiscourseAi::Translation::PostLocaleDetector.expects(:detect_locale).with(group_pm_post).once
+      DiscourseAi::Translation::PostLocaleDetector.expects(:detect_locale).with(pm_post).never
 
       job.execute({})
     end

spec/jobs/scheduled/topics_locale_detection_backfill_spec.rb

@@ -82,8 +82,14 @@
 
   describe "with public content limitation" do
     fab!(:private_category) { Fabricate(:private_category, group: Group[:staff]) }
+    fab!(:private_cat_topic) { Fabricate(:topic, category: private_category, locale: nil) }
+
+    fab!(:group)
+    fab!(:group_pm_topic) { Fabricate(:private_message_topic, allowed_groups: [group]) }
+
+    fab!(:pm_topic) { Fabricate(:private_message_topic) }
+
     fab!(:public_topic) { Fabricate(:topic, locale: nil) }
-    fab!(:private_topic) { Fabricate(:topic, category: private_category, locale: nil) }
 
     before do
       DiscourseAi::Translation::TopicLocaleDetector.expects(:detect_locale).at_least_once
@@ -93,19 +99,33 @@
 
     it "only processes topics from public categories" do
       DiscourseAi::Translation::TopicLocaleDetector.expects(:detect_locale).with(public_topic).once
+
+      DiscourseAi::Translation::TopicLocaleDetector
+        .expects(:detect_locale)
+        .with(private_cat_topic)
+        .never
       DiscourseAi::Translation::TopicLocaleDetector
         .expects(:detect_locale)
-        .with(private_topic)
+        .with(group_pm_topic)
         .never
+      DiscourseAi::Translation::TopicLocaleDetector.expects(:detect_locale).with(pm_topic).never
 
       job.execute({ limit: 10 })
     end
 
-    it "processes all topics when setting is disabled" do
+    it "processes public category topics, group PMs, and private category topics when setting is disabled" do
       SiteSetting.ai_translation_backfill_limit_to_public_content = false
 
       DiscourseAi::Translation::TopicLocaleDetector.expects(:detect_locale).with(public_topic).once
-      DiscourseAi::Translation::TopicLocaleDetector.expects(:detect_locale).with(private_topic).once
+      DiscourseAi::Translation::TopicLocaleDetector
+        .expects(:detect_locale)
+        .with(group_pm_topic)
+        .once
+      DiscourseAi::Translation::TopicLocaleDetector
+        .expects(:detect_locale)
+        .with(private_cat_topic)
+        .once
+      DiscourseAi::Translation::TopicLocaleDetector.expects(:detect_locale).with(pm_topic).never
 
       job.execute({ limit: 10 })
     end