Add option to not consider word boundaries when processing keyword filtering (#7975)

* Add option to not consider word boundaries when filtering phrases

* Add a few tests for keyword/phrase filtering

app/controllers/api/v1/filters_controller.rb

@@ -43,6 +43,6 @@ class Api::V1::FiltersController < Api::BaseController
   end
 
   def resource_params
-    params.permit(:phrase, :expires_in, :irreversible, context: [])
+    params.permit(:phrase, :expires_in, :irreversible, :whole_word, context: [])
   end
 end

app/javascript/mastodon/selectors/index.js

@@ -45,7 +45,10 @@ export const regexFromFilters = filters => {
     return null;
   }
 
-  return new RegExp(filters.map(filter => escapeRegExp(filter.get('phrase'))).map(expr => `\\b${expr}\\b`).join('|'), 'i');
+  return new RegExp(filters.map(filter => {
+    let expr = escapeRegExp(filter.get('phrase'));
+    return filter.get('whole_word') ? `\\b${expr}\\b` : expr;
+  }).join('|'), 'i');
 };
 
 export const makeGetStatus = () => {

app/lib/feed_manager.rb

@@ -200,7 +200,16 @@ class FeedManager
     active_filters = Rails.cache.fetch("filters:#{receiver_id}") { CustomFilter.where(account_id: receiver_id).active_irreversible.to_a }.to_a
 
     active_filters.select! { |filter| filter.context.include?(context.to_s) && !filter.expired? }
-    active_filters.map! { |filter| Regexp.new("\\b#{Regexp.escape(filter.phrase)}\\b", true) }
+    active_filters.map! do |filter|
+      if filter.whole_word
+        sb = filter.phrase =~ /\A[[:word:]]/ ? '\b' : ''
+        eb = filter.phrase =~ /[[:word:]]\Z/ ? '\b' : ''
+
+        /(?mix:#{sb}#{Regexp.escape(filter.phrase)}#{eb})/
+      else
+        /#{Regexp.escape(filter.phrase)}/i
+      end
+    end
 
     return false if active_filters.empty?
 

app/models/custom_filter.rb

@@ -8,6 +8,7 @@
 #  expires_at   :datetime
 #  phrase       :text             default(""), not null
 #  context      :string           default([]), not null, is an Array
+#  whole_word   :boolean          default(TRUE), not null
 #  irreversible :boolean          default(FALSE), not null
 #  created_at   :datetime         not null
 #  updated_at   :datetime         not null

app/serializers/rest/filter_serializer.rb

@@ -1,6 +1,6 @@
 # frozen_string_literal: true
 
 class REST::FilterSerializer < ActiveModel::Serializer
-  attributes :id, :phrase, :context, :expires_at,
+  attributes :id, :phrase, :context, :whole_word, :expires_at,
              :irreversible
 end

app/views/filters/_fields.html.haml

@@ -7,5 +7,8 @@
 .fields-group
   = f.input :irreversible, wrapper: :with_label
 
+.fields-group
+  = f.input :whole_word, wrapper: :with_label
+
 .fields-group
   = f.input :expires_in, wrapper: :with_label, collection: [30.minutes, 1.hour, 6.hours, 12.hours, 1.day, 1.week].map(&:to_i), label_method: lambda { |i| I18n.t("invites.expires_in.#{i}") }, prompt: I18n.t('invites.expires_in_prompt')

db/migrate/20180707154237_add_whole_word_to_custom_filter.rb

@@ -0,0 +1,17 @@
+require Rails.root.join('lib', 'mastodon', 'migration_helpers')
+
+class AddWholeWordToCustomFilter < ActiveRecord::Migration[5.2]
+  include Mastodon::MigrationHelpers
+
+  disable_ddl_transaction!
+
+  def change
+    safety_assured do
+      add_column_with_default :custom_filters, :whole_word, :boolean, default: true, allow_null: false
+    end
+  end
+
+  def down
+    remove_column :custom_filters, :whole_word
+  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: 2018_06_28_181026) do
+ActiveRecord::Schema.define(version: 2018_07_07_154237) do
 
   # These are extensions that must be enabled in order to support this database
   enable_extension "plpgsql"
@@ -149,6 +149,7 @@ ActiveRecord::Schema.define(version: 2018_06_28_181026) do
     t.text "phrase", default: "", null: false
     t.string "context", default: [], null: false, array: true
     t.boolean "irreversible", default: false, null: false
+    t.boolean "whole_word", default: true, null: false
     t.datetime "created_at", null: false
     t.datetime "updated_at", null: false
     t.index ["account_id"], name: "index_custom_filters_on_account_id"

spec/lib/feed_manager_spec.rb

@@ -127,12 +127,28 @@ RSpec.describe FeedManager do
         expect(FeedManager.instance.filter?(:home, reblog, alice.id)).to be true
       end
 
-      it 'returns true if status contains irreversibly muted phrase' do
-        alice.custom_filters.create!(phrase: 'farts', context: %w(home public), irreversible: true)
-        alice.custom_filters.create!(phrase: 'pop tarts', context: %w(home), irreversible: true)
-        alice.follow!(jeff)
-        status = Fabricate(:status, text: 'i sure like POP TARts', account: jeff)
-        expect(FeedManager.instance.filter?(:home, status, alice.id)).to be true
+      context 'for irreversibly muted phrases' do
+        it 'considers word boundaries when matching' do
+          alice.custom_filters.create!(phrase: 'bob', context: %w(home), irreversible: true)
+          alice.follow!(jeff)
+          status = Fabricate(:status, text: 'bobcats', account: jeff)
+          expect(FeedManager.instance.filter?(:home, status, alice.id)).to be_falsy
+        end
+
+        it 'returns true if phrase is contained' do
+          alice.custom_filters.create!(phrase: 'farts', context: %w(home public), irreversible: true)
+          alice.custom_filters.create!(phrase: 'pop tarts', context: %w(home), irreversible: true)
+          alice.follow!(jeff)
+          status = Fabricate(:status, text: 'i sure like POP TARts', account: jeff)
+          expect(FeedManager.instance.filter?(:home, status, alice.id)).to be true
+        end
+
+        it 'matches substrings if whole_word is false' do
+          alice.custom_filters.create!(phrase: 'take', context: %w(home), whole_word: false, irreversible: true)
+          alice.follow!(jeff)
+          status = Fabricate(:status, text: 'shiitake', account: jeff)
+          expect(FeedManager.instance.filter?(:home, status, alice.id)).to be true
+        end
       end
     end