Improve performance by avoiding regex construction (#20215)

```ruby
10.times { p /#{FOO}/.object_id }
10.times { p FOO_RE.object_id }
```
This commit is contained in:
James Tucker 2022-11-09 20:49:30 -08:00 committed by GitHub
parent 0cd0786aef
commit 78a6b871fe
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
8 changed files with 21 additions and 12 deletions

View file

@ -24,7 +24,7 @@ class Api::V1::TagsController < Api::BaseController
private private
def set_or_create_tag 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]) @tag = Tag.find_normalized(params[:id]) || Tag.new(name: Tag.normalize(params[:id]), display_name: params[:id])
end end
end end

View file

@ -8,7 +8,7 @@ class HashtagNormalizer
private private
def remove_invalid_characters(str) def remove_invalid_characters(str)
str.gsub(/[^[:alnum:]#{Tag::HASHTAG_SEPARATORS}]/, '') str.gsub(Tag::HASHTAG_INVALID_CHARS_RE, '')
end end
def ascii_folding(str) def ascii_folding(str)

View file

@ -64,6 +64,7 @@ class Account < ApplicationRecord
USERNAME_RE = /[a-z0-9_]+([a-z0-9_\.-]+[a-z0-9_]+)?/i USERNAME_RE = /[a-z0-9_]+([a-z0-9_\.-]+[a-z0-9_]+)?/i
MENTION_RE = /(?<=^|[^\/[:word:]])@((#{USERNAME_RE})(?:@[[:word:]\.\-]+[[:word:]]+)?)/i MENTION_RE = /(?<=^|[^\/[:word:]])@((#{USERNAME_RE})(?:@[[:word:]\.\-]+[[:word:]]+)?)/i
URL_PREFIX_RE = /\Ahttp(s?):\/\/[^\/]+/ URL_PREFIX_RE = /\Ahttp(s?):\/\/[^\/]+/
USERNAME_ONLY_RE = /\A#{USERNAME_RE}\z/i
include Attachmentable include Attachmentable
include AccountAssociations include AccountAssociations
@ -84,7 +85,7 @@ class Account < ApplicationRecord
validates_with UniqueUsernameValidator, if: -> { will_save_change_to_username? } validates_with UniqueUsernameValidator, if: -> { will_save_change_to_username? }
# Remote user validations # 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 # 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' } validates :username, format: { with: /\A[a-z0-9_]+\z/i }, length: { maximum: 30 }, if: -> { local? && will_save_change_to_username? && actor_type != 'Application' }

View file

@ -30,6 +30,7 @@ class CustomEmoji < ApplicationRecord
SCAN_RE = /(?<=[^[:alnum:]:]|\n|^) SCAN_RE = /(?<=[^[:alnum:]:]|\n|^)
:(#{SHORTCODE_RE_FRAGMENT}): :(#{SHORTCODE_RE_FRAGMENT}):
(?=[^[:alnum:]:]|$)/x (?=[^[:alnum:]:]|$)/x
SHORTCODE_ONLY_RE = /\A#{SHORTCODE_RE_FRAGMENT}\z/
IMAGE_MIME_TYPES = %w(image/png image/gif image/webp).freeze IMAGE_MIME_TYPES = %w(image/png image/gif image/webp).freeze
@ -41,7 +42,7 @@ class CustomEmoji < ApplicationRecord
before_validation :downcase_domain before_validation :downcase_domain
validates_attachment :image, content_type: { content_type: IMAGE_MIME_TYPES }, presence: true, size: { less_than: LIMIT } 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 :local, -> { where(domain: nil) }
scope :remote, -> { where.not(domain: nil) } scope :remote, -> { where.not(domain: nil) }

View file

@ -17,7 +17,7 @@ class FeaturedTag < ApplicationRecord
belongs_to :account, inverse_of: :featured_tags belongs_to :account, inverse_of: :featured_tags
belongs_to :tag, inverse_of: :featured_tags, optional: true # Set after validation 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_tag_uniqueness, on: :create
validate :validate_featured_tags_limit, on: :create validate :validate_featured_tags_limit, on: :create

View file

@ -27,11 +27,14 @@ class Tag < ApplicationRecord
has_many :followers, through: :passive_relationships, source: :account has_many :followers, through: :passive_relationships, source: :account
HASHTAG_SEPARATORS = "_\u00B7\u200c" HASHTAG_SEPARATORS = "_\u00B7\u200c"
HASHTAG_NAME_RE = "([[:word:]_][[:word:]#{HASHTAG_SEPARATORS}]*[[:alpha:]#{HASHTAG_SEPARATORS}][[:word:]#{HASHTAG_SEPARATORS}]*[[:word:]_])|([[:word:]_]*[[:alpha:]][[:word:]_]*)" HASHTAG_NAME_PAT = "([[:word:]_][[:word:]#{HASHTAG_SEPARATORS}]*[[:alpha:]#{HASHTAG_SEPARATORS}][[:word:]#{HASHTAG_SEPARATORS}]*[[:word:]_])|([[:word:]_]*[[:alpha:]][[:word:]_]*)"
HASHTAG_RE = /(?:^|[^\/\)\w])#(#{HASHTAG_NAME_RE})/i
validates :name, presence: true, format: { with: /\A(#{HASHTAG_NAME_RE})\z/i } HASHTAG_RE = /(?:^|[^\/\)\w])#(#{HASHTAG_NAME_PAT})/i
validates :display_name, format: { with: /\A(#{HASHTAG_NAME_RE})\z/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_name_change, if: -> { !new_record? && name_changed? }
validate :validate_display_name_change, if: -> { !new_record? && display_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 = Array(name_or_names).map { |str| [normalize(str), str] }.uniq(&:first)
names.map do |(normalized_name, display_name)| 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? yield tag if block_given?

View file

@ -3,6 +3,8 @@
class AccountSearchService < BaseService class AccountSearchService < BaseService
attr_reader :query, :limit, :offset, :options, :account 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. number of characters to look for non-exact matches
MIN_QUERY_LENGTH = 5 MIN_QUERY_LENGTH = 5
@ -180,7 +182,7 @@ class AccountSearchService < BaseService
end end
def username_complete? def username_complete?
query.include?('@') && "@#{query}".match?(/\A#{Account::MENTION_RE}\Z/) query.include?('@') && "@#{query}".match?(MENTION_ONLY_RE)
end end
def likely_acct? def likely_acct?

View file

@ -4,6 +4,8 @@ class ResolveURLService < BaseService
include JsonLdHelper include JsonLdHelper
include Authorization include Authorization
USERNAME_STATUS_RE = %r{/@(?<username>#{Account::USERNAME_RE})/(?<status_id>[0-9]+)\Z}
def call(url, on_behalf_of: nil) def call(url, on_behalf_of: nil)
@url = url @url = url
@on_behalf_of = on_behalf_of @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` # We don't have an index on `url`, so try guessing the `uri` from `url`
parsed_url = Addressable::URI.parse(@url) parsed_url = Addressable::URI.parse(@url)
parsed_url.path.match(%r{/@(?<username>#{Account::USERNAME_RE})/(?<status_id>[0-9]+)\Z}) do |matched| parsed_url.path.match(USERNAME_STATUS_RE) do |matched|
parsed_url.path = "/users/#{matched[:username]}/statuses/#{matched[:status_id]}" parsed_url.path = "/users/#{matched[:username]}/statuses/#{matched[:status_id]}"
scope = scope.or(Status.where(uri: parsed_url.to_s, url: @url)) scope = scope.or(Status.where(uri: parsed_url.to_s, url: @url))
end end