improve filter implementation

Author: SamSaffron

lib/personas/tools/researcher.rb

@@ -34,9 +34,12 @@ def filter_description
               Filter string to target specific content.
               - Supports user (@username)
               - date ranges (after:YYYY-MM-DD, before:YYYY-MM-DD for posts; topic_after:YYYY-MM-DD, topic_before:YYYY-MM-DD for topics)
-              - categories (category:name)
-              - tags (tag:name)
-              - groups (group:name).
+              - categories (category:category1,category2)
+              - tags (tag:tag1,tag2)
+              - groups (group:group1,group2).
+              - status (status:open, status:closed, status:archived, status:noreplies, status:single_user)
+
+              If multiple tags or categories are specified, they are treated as OR conditions.
 
               Multiple filters can be combined with spaces. Example: '@sam after:2023-01-01 tag:feature'
             TEXT

lib/utils/research/filter.rb

@@ -73,33 +73,43 @@ def self.word_to_date(str)
           end
         end
 
-        # Category filter
-        register_filter(/\Acategory:([a-zA-Z0-9_\-]+)\z/i) do |relation, slug, _|
-          category = Category.find_by("LOWER(slug) = LOWER(?)", slug)
-          if category
-            category_ids = [category.id]
-            category_ids +=
-              Category.subcategory_ids(category.id) if category.subcategory_ids.present?
-            relation.where("topics.category_id IN (?)", category_ids)
+        register_filter(/\A(?:tags?|tag):(.*)\z/i) do |relation, tag_param, _|
+          if tag_param.include?(",")
+            tag_names = tag_param.split(",").map(&:strip)
+            tag_ids = Tag.where(name: tag_names).pluck(:id)
+            return relation.where("1 = 0") if tag_ids.empty?
+            relation.where(topic_id: TopicTag.where(tag_id: tag_ids).select(:topic_id))
           else
-            relation.where("1 = 0") # No results if category doesn't exist
+            if tag = Tag.find_by(name: tag_param)
+              relation.where(topic_id: TopicTag.where(tag_id: tag.id).select(:topic_id))
+            else
+              relation.where("1 = 0")
+            end
           end
         end
 
-        # Tag filter
-        register_filter(/\Atag:([a-zA-Z0-9_\-]+)\z/i) do |relation, name, _|
-          tag = Tag.find_by_name(name)
-          if tag
-            relation.joins("INNER JOIN topic_tags ON topic_tags.topic_id = topics.id").where(
-              "topic_tags.tag_id = ?",
-              tag.id,
-            )
+        register_filter(/\A(?:categories?|category):(.*)\z/i) do |relation, category_param, _|
+          if category_param.include?(",")
+            category_names = category_param.split(",").map(&:strip)
+
+            found_category_ids = []
+            category_names.each do |name|
+              category = Category.find_by(slug: name) || Category.find_by(name: name)
+              found_category_ids << category.id if category
+            end
+
+            return relation.where("1 = 0") if found_category_ids.empty?
+            relation.where(topic_id: Topic.where(category_id: found_category_ids).select(:id))
           else
-            relation.where("1 = 0") # No results if tag doesn't exist
+            if category =
+                 Category.find_by(slug: category_param) || Category.find_by(name: category_param)
+              relation.where(topic_id: Topic.where(category_id: category.id).select(:id))
+            else
+              relation.where("1 = 0")
+            end
           end
         end
 
-        # User filter
         register_filter(/\A\@(\w+)\z/i) do |relation, username, filter|
           user = User.find_by(username_lower: username.downcase)
           if user
@@ -109,7 +119,6 @@ def self.word_to_date(str)
           end
         end
 
-        # Posted by current user
         register_filter(/\Ain:posted\z/i) do |relation, _, filter|
           if filter.guardian.user
             relation.where("posts.user_id = ?", filter.guardian.user.id)

spec/lib/utils/research/filter_spec.rb

@@ -0,0 +1,95 @@
+# frozen_string_literal: true
+
+describe DiscourseAi::Utils::Research::Filter do
+  describe "integration tests" do
+    before_all { SiteSetting.min_topic_title_length = 3 }
+
+    fab!(:user)
+
+    fab!(:feature_tag) { Fabricate(:tag, name: "feature") }
+    fab!(:bug_tag) { Fabricate(:tag, name: "bug") }
+
+    fab!(:announcement_category) { Fabricate(:category, name: "Announcements") }
+    fab!(:feedback_category) { Fabricate(:category, name: "Feedback") }
+
+    fab!(:feature_topic) do
+      Fabricate(
+        :topic,
+        user: user,
+        tags: [feature_tag],
+        category: announcement_category,
+        title: "New Feature Discussion",
+      )
+    end
+
+    fab!(:bug_topic) do
+      Fabricate(
+        :topic,
+        tags: [bug_tag],
+        user: user,
+        category: announcement_category,
+        title: "Bug Report",
+      )
+    end
+
+    fab!(:feature_bug_topic) do
+      Fabricate(
+        :topic,
+        tags: [feature_tag, bug_tag],
+        user: user,
+        category: feedback_category,
+        title: "Feature with Bug",
+      )
+    end
+
+    fab!(:no_tag_topic) do
+      Fabricate(:topic, user: user, category: feedback_category, title: "General Discussion")
+    end
+
+    fab!(:feature_post) { Fabricate(:post, topic: feature_topic, user: user) }
+    fab!(:bug_post) { Fabricate(:post, topic: bug_topic, user: user) }
+    fab!(:feature_bug_post) { Fabricate(:post, topic: feature_bug_topic, user: user) }
+    fab!(:no_tag_post) { Fabricate(:post, topic: no_tag_topic, user: user) }
+
+    describe "tag filtering" do
+      it "correctly filters posts by tags" do
+        filter = described_class.new("tag:feature")
+        expect(filter.search.pluck(:id)).to contain_exactly(feature_post.id, feature_bug_post.id)
+
+        filter = described_class.new("tag:feature,bug")
+        expect(filter.search.pluck(:id)).to contain_exactly(
+          feature_bug_post.id,
+          bug_post.id,
+          feature_post.id,
+        )
+
+        filter = described_class.new("tags:bug")
+        expect(filter.search.pluck(:id)).to contain_exactly(bug_post.id, feature_bug_post.id)
+
+        filter = described_class.new("tag:nonexistent")
+        expect(filter.search.count).to eq(0)
+      end
+    end
+
+    describe "category filtering" do
+      it "correctly filters posts by categories" do
+        filter = described_class.new("category:Announcements")
+        expect(filter.search.pluck(:id)).to contain_exactly(feature_post.id, bug_post.id)
+
+        filter = described_class.new("category:Announcements,Feedback")
+        expect(filter.search.pluck(:id)).to contain_exactly(
+          feature_post.id,
+          bug_post.id,
+          feature_bug_post.id,
+          no_tag_post.id,
+        )
+
+        filter = described_class.new("categories:Feedback")
+        expect(filter.search.pluck(:id)).to contain_exactly(feature_bug_post.id, no_tag_post.id)
+
+        filter = described_class.new("category:Feedback tag:feature")
+        expect(filter.search.pluck(:id)).to contain_exactly(feature_bug_post.id)
+      end
+    end
+  end
+end