HELIO-3888 Log and Measure epub searches

app/controllers/e_pubs_controller.rb

@@ -97,10 +97,13 @@ def search
     query = params[:q] || ''
     # due to performance issues, must have 3 or more characters to search
     return render json: { q: query, search_results: [] } if query.length < 3
+    start = (Time.now.to_f * 1000.0).to_i
     results = Rails.cache.fetch(search_cache_key(@noid, query), expires_in: 30.days) do
       epub = EPub::Publication.from_directory(UnpackService.root_path_from_noid(@noid, 'epub'))
       epub.search(query)
     end
+    finish = (Time.now.to_f * 1000.0).to_i
+    EpubSearchLog.create(noid: @noid, query: query, time: finish - start, hits: results[:search_results].count, search_results: results)
     render json: results
   rescue StandardError => e
     Rails.logger.error "EPubsController.search raised #{e}"

app/models/epub_search_log.rb

@@ -0,0 +1,6 @@
+# frozen_string_literal: true
+
+class EpubSearchLog < ApplicationRecord
+  include Filterable
+  scope :query_like, ->(like) { where("query like ?", "%#{like}%") }
+end

db/migrate/20210507185550_create_epub_search_logs.rb

@@ -0,0 +1,13 @@
+class CreateEpubSearchLogs < ActiveRecord::Migration[5.2]
+  def change
+    create_table :epub_search_logs do |t|
+      t.string :noid, index: true
+      t.text :query
+      t.integer :time
+      t.integer :hits
+      t.mediumtext :search_results
+
+      t.timestamps
+    end
+  end
+end

db/schema.rb

@@ -10,7 +10,7 @@
 #
 # It's strongly recommended that you check this file into your version control system.
 
-ActiveRecord::Schema.define(version: 2021_04_08_212028) do
+ActiveRecord::Schema.define(version: 2021_05_07_185550) do
 
   create_table "api_requests", options: "ENGINE=InnoDB DEFAULT CHARSET=utf8", force: :cascade do |t|
     t.integer "user_id"
@@ -161,6 +161,17 @@
     t.text "toc", limit: 16777215
   end
 
+  create_table "epub_search_log", options: "ENGINE=InnoDB DEFAULT CHARSET=utf8", force: :cascade do |t|
+    t.string "noid"
+    t.text "query"
+    t.integer "time"
+    t.integer "hits"
+    t.text "search_results", limit: 16777215
+    t.datetime "created_at", null: false
+    t.datetime "updated_at", null: false
+    t.index ["noid"], name: "index_epub_search_log_on_noid"
+  end
+
   create_table "featured_representatives", options: "ENGINE=InnoDB DEFAULT CHARSET=utf8", force: :cascade do |t|
     t.string "work_id"
     t.string "file_set_id"
@@ -428,14 +439,6 @@
     t.index ["grantor_id"], name: "index_proxy_deposit_rights_on_grantor_id"
   end
 
-  create_table "robotrons", options: "ENGINE=InnoDB DEFAULT CHARSET=utf8", force: :cascade do |t|
-    t.string "ip", null: false
-    t.integer "hits", default: 0
-    t.datetime "created_at", null: false
-    t.datetime "updated_at", null: false
-    t.index ["ip"], name: "index_robotrons_on_ip"
-  end
-
   create_table "roles", id: :integer, options: "ENGINE=InnoDB DEFAULT CHARSET=utf8", force: :cascade do |t|
     t.string "resource_type"
     t.integer "resource_id"

lib/e_pub/search.rb

@@ -1,5 +1,7 @@
 # frozen_string_literal: true
 
+require "skylight"
+
 module EPub
   class Search
     def initialize(publication)
@@ -65,29 +67,31 @@ def results_from_chapters(db_results, query)
         results[:search_results] = []
 
         db_results.each do |chapter|
-          file = File.join(@publication.root_path, File.dirname(@publication.content_file), chapter[:href])
-          doc = Nokogiri::XML(File.open(file))
-          doc.remove_namespaces!
-
-          matches = []
-          body = doc.xpath("//body")
-          body.children.each do |node|
-            matches << find_targets(node, query)
-          end
-
-          matches = matches.flatten.compact
-
-          matches.each_index do |index|
-            match = matches[index]
-
-            # De-duplicate identical snippets with slightly different CFIs that are neighbors.
-            # Since we need the CFIs in the reader for syntax highlighting, still send those,
-            # just not snippets
-            empty_snippet = "" if match.snippet == matches[index - 1].snippet && matches.length > 1
-
-            results[:search_results].push(cfi: "#{chapter[:basecfi]}#{match.cfi}",
-                                          title: chapter[:title],
-                                          snippet: empty_snippet || match.snippet)
+          Skylight.instrument title: "EpubSearch Text In Each Chapter" do
+            file = File.join(@publication.root_path, File.dirname(@publication.content_file), chapter[:href])
+            doc = Nokogiri::XML(File.open(file))
+            doc.remove_namespaces!
+
+            matches = []
+            body = doc.xpath("//body")
+            body.children.each do |node|
+              matches << find_targets(node, query)
+            end
+
+            matches = matches.flatten.compact
+
+            matches.each_index do |index|
+              match = matches[index]
+
+              # De-duplicate identical snippets with slightly different CFIs that are neighbors.
+              # Since we need the CFIs in the reader for syntax highlighting, still send those,
+              # just not snippets
+              empty_snippet = "" if match.snippet == matches[index - 1].snippet && matches.length > 1
+
+              results[:search_results].push(cfi: "#{chapter[:basecfi]}#{match.cfi}",
+                                            title: chapter[:title],
+                                            snippet: empty_snippet || match.snippet)
+            end
           end
         end
         results

spec/controllers/e_pubs_controller_spec.rb

@@ -253,6 +253,9 @@
           # to get search highlighting
           @snippets = JSON.parse(response.body)["search_results"].map { |result| result["snippet"].presence }.compact
           expect(@snippets.length).to eq 102
+          expect(EpubSearchLog.first.noid).to eq file_set.id
+          expect(EpubSearchLog.first.query).to eq "White Whale"
+          expect(EpubSearchLog.first.hits).to eq 105
         end
       end
 
@@ -312,6 +315,9 @@
           expect(response).to have_http_status(:success)
           expect(JSON.parse(response.body)["q"]).to eq "glubmerschmup"
           expect(JSON.parse(response.body)["search_results"]).to eq []
+          expect(EpubSearchLog.first.noid).to eq file_set.id
+          expect(EpubSearchLog.first.query).to eq "glubmerschmup"
+          expect(EpubSearchLog.first.hits).to eq 0
         end
       end
     end