From 78a6b871fe3dae308380ea88132ddadc86a1431e Mon Sep 17 00:00:00 2001 From: James Tucker Date: Wed, 9 Nov 2022 20:49:30 -0800 Subject: [PATCH] Improve performance by avoiding regex construction (#20215) ```ruby 10.times { p /#{FOO}/.object_id } 10.times { p FOO_RE.object_id } ``` --- app/controllers/api/v1/tags_controller.rb | 2 +- app/lib/hashtag_normalizer.rb | 2 +- app/models/account.rb | 3 ++- app/models/custom_emoji.rb | 3 ++- app/models/featured_tag.rb | 2 +- app/models/tag.rb | 13 ++++++++----- app/services/account_search_service.rb | 4 +++- app/services/resolve_url_service.rb | 4 +++- 8 files changed, 21 insertions(+), 12 deletions(-) diff --git a/app/controllers/api/v1/tags_controller.rb b/app/controllers/api/v1/tags_controller.rb index 9e5c53330..32f71bdce 100644 --- a/app/controllers/api/v1/tags_controller.rb +++ b/app/controllers/api/v1/tags_controller.rb @@ -24,7 +24,7 @@ class Api::V1::TagsController < Api::BaseController private def set_or_create_tag - return not_found unless /\A(#{Tag::HASHTAG_NAME_RE})\z/.match?(params[:id]) + return not_found unless Tag::HASHTAG_NAME_RE.match?(params[:id]) @tag = Tag.find_normalized(params[:id]) || Tag.new(name: Tag.normalize(params[:id]), display_name: params[:id]) end end diff --git a/app/lib/hashtag_normalizer.rb b/app/lib/hashtag_normalizer.rb index c1f99e163..49fa6101d 100644 --- a/app/lib/hashtag_normalizer.rb +++ b/app/lib/hashtag_normalizer.rb @@ -8,7 +8,7 @@ class HashtagNormalizer private def remove_invalid_characters(str) - str.gsub(/[^[:alnum:]#{Tag::HASHTAG_SEPARATORS}]/, '') + str.gsub(Tag::HASHTAG_INVALID_CHARS_RE, '') end def ascii_folding(str) diff --git a/app/models/account.rb b/app/models/account.rb index be1968fa6..cc3a8f3df 100644 --- a/app/models/account.rb +++ b/app/models/account.rb @@ -64,6 +64,7 @@ class Account < ApplicationRecord USERNAME_RE = /[a-z0-9_]+([a-z0-9_\.-]+[a-z0-9_]+)?/i MENTION_RE = /(?<=^|[^\/[:word:]])@((#{USERNAME_RE})(?:@[[:word:]\.\-]+[[:word:]]+)?)/i URL_PREFIX_RE = /\Ahttp(s?):\/\/[^\/]+/ + USERNAME_ONLY_RE = /\A#{USERNAME_RE}\z/i include Attachmentable include AccountAssociations @@ -84,7 +85,7 @@ class Account < ApplicationRecord validates_with UniqueUsernameValidator, if: -> { will_save_change_to_username? } # Remote user validations - validates :username, format: { with: /\A#{USERNAME_RE}\z/i }, if: -> { !local? && will_save_change_to_username? } + validates :username, format: { with: USERNAME_ONLY_RE }, if: -> { !local? && will_save_change_to_username? } # Local user validations validates :username, format: { with: /\A[a-z0-9_]+\z/i }, length: { maximum: 30 }, if: -> { local? && will_save_change_to_username? && actor_type != 'Application' } diff --git a/app/models/custom_emoji.rb b/app/models/custom_emoji.rb index 077ce559a..7b19cd2ac 100644 --- a/app/models/custom_emoji.rb +++ b/app/models/custom_emoji.rb @@ -30,6 +30,7 @@ class CustomEmoji < ApplicationRecord SCAN_RE = /(?<=[^[:alnum:]:]|\n|^) :(#{SHORTCODE_RE_FRAGMENT}): (?=[^[:alnum:]:]|$)/x + SHORTCODE_ONLY_RE = /\A#{SHORTCODE_RE_FRAGMENT}\z/ IMAGE_MIME_TYPES = %w(image/png image/gif image/webp).freeze @@ -41,7 +42,7 @@ class CustomEmoji < ApplicationRecord before_validation :downcase_domain validates_attachment :image, content_type: { content_type: IMAGE_MIME_TYPES }, presence: true, size: { less_than: LIMIT } - validates :shortcode, uniqueness: { scope: :domain }, format: { with: /\A#{SHORTCODE_RE_FRAGMENT}\z/ }, length: { minimum: 2 } + validates :shortcode, uniqueness: { scope: :domain }, format: { with: SHORTCODE_ONLY_RE }, length: { minimum: 2 } scope :local, -> { where(domain: nil) } scope :remote, -> { where.not(domain: nil) } diff --git a/app/models/featured_tag.rb b/app/models/featured_tag.rb index debae2212..70f949b6a 100644 --- a/app/models/featured_tag.rb +++ b/app/models/featured_tag.rb @@ -17,7 +17,7 @@ class FeaturedTag < ApplicationRecord belongs_to :account, inverse_of: :featured_tags belongs_to :tag, inverse_of: :featured_tags, optional: true # Set after validation - validates :name, presence: true, format: { with: /\A(#{Tag::HASHTAG_NAME_RE})\z/i }, on: :create + validates :name, presence: true, format: { with: Tag::HASHTAG_NAME_RE }, on: :create validate :validate_tag_uniqueness, on: :create validate :validate_featured_tags_limit, on: :create diff --git a/app/models/tag.rb b/app/models/tag.rb index 8929baf66..b66f85423 100644 --- a/app/models/tag.rb +++ b/app/models/tag.rb @@ -27,11 +27,14 @@ class Tag < ApplicationRecord has_many :followers, through: :passive_relationships, source: :account HASHTAG_SEPARATORS = "_\u00B7\u200c" - HASHTAG_NAME_RE = "([[:word:]_][[:word:]#{HASHTAG_SEPARATORS}]*[[:alpha:]#{HASHTAG_SEPARATORS}][[:word:]#{HASHTAG_SEPARATORS}]*[[:word:]_])|([[:word:]_]*[[:alpha:]][[:word:]_]*)" - HASHTAG_RE = /(?:^|[^\/\)\w])#(#{HASHTAG_NAME_RE})/i + HASHTAG_NAME_PAT = "([[:word:]_][[:word:]#{HASHTAG_SEPARATORS}]*[[:alpha:]#{HASHTAG_SEPARATORS}][[:word:]#{HASHTAG_SEPARATORS}]*[[:word:]_])|([[:word:]_]*[[:alpha:]][[:word:]_]*)" - validates :name, presence: true, format: { with: /\A(#{HASHTAG_NAME_RE})\z/i } - validates :display_name, format: { with: /\A(#{HASHTAG_NAME_RE})\z/i } + HASHTAG_RE = /(?:^|[^\/\)\w])#(#{HASHTAG_NAME_PAT})/i + HASHTAG_NAME_RE = /\A(#{HASHTAG_NAME_PAT})\z/i + HASHTAG_INVALID_CHARS_RE = /[^[:alnum:]#{HASHTAG_SEPARATORS}]/ + + validates :name, presence: true, format: { with: HASHTAG_NAME_RE } + validates :display_name, format: { with: HASHTAG_NAME_RE } validate :validate_name_change, if: -> { !new_record? && name_changed? } validate :validate_display_name_change, if: -> { !new_record? && display_name_changed? } @@ -102,7 +105,7 @@ class Tag < ApplicationRecord names = Array(name_or_names).map { |str| [normalize(str), str] }.uniq(&:first) names.map do |(normalized_name, display_name)| - tag = matching_name(normalized_name).first || create(name: normalized_name, display_name: display_name.gsub(/[^[:alnum:]#{HASHTAG_SEPARATORS}]/, '')) + tag = matching_name(normalized_name).first || create(name: normalized_name, display_name: display_name.gsub(HASHTAG_INVALID_CHARS_RE, '')) yield tag if block_given? diff --git a/app/services/account_search_service.rb b/app/services/account_search_service.rb index 9f2330a94..85538870b 100644 --- a/app/services/account_search_service.rb +++ b/app/services/account_search_service.rb @@ -3,6 +3,8 @@ class AccountSearchService < BaseService attr_reader :query, :limit, :offset, :options, :account + MENTION_ONLY_RE = /\A#{Account::MENTION_RE}\z/i + # Min. number of characters to look for non-exact matches MIN_QUERY_LENGTH = 5 @@ -180,7 +182,7 @@ class AccountSearchService < BaseService end def username_complete? - query.include?('@') && "@#{query}".match?(/\A#{Account::MENTION_RE}\Z/) + query.include?('@') && "@#{query}".match?(MENTION_ONLY_RE) end def likely_acct? diff --git a/app/services/resolve_url_service.rb b/app/services/resolve_url_service.rb index 37c856cf8..52f35daf3 100644 --- a/app/services/resolve_url_service.rb +++ b/app/services/resolve_url_service.rb @@ -4,6 +4,8 @@ class ResolveURLService < BaseService include JsonLdHelper include Authorization + USERNAME_STATUS_RE = %r{/@(?#{Account::USERNAME_RE})/(?[0-9]+)\Z} + def call(url, on_behalf_of: nil) @url = url @on_behalf_of = on_behalf_of @@ -43,7 +45,7 @@ class ResolveURLService < BaseService # We don't have an index on `url`, so try guessing the `uri` from `url` parsed_url = Addressable::URI.parse(@url) - parsed_url.path.match(%r{/@(?#{Account::USERNAME_RE})/(?[0-9]+)\Z}) do |matched| + parsed_url.path.match(USERNAME_STATUS_RE) do |matched| parsed_url.path = "/users/#{matched[:username]}/statuses/#{matched[:status_id]}" scope = scope.or(Status.where(uri: parsed_url.to_s, url: @url)) end