Change hashtags to preserve first-used casing (#11416)

main
Eugen Rochko 2019-07-28 05:59:51 +02:00 committed by GitHub
parent 4cc29eb5ad
commit f371b32137
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
6 changed files with 54 additions and 19 deletions

View File

@ -148,12 +148,9 @@ class ActivityPub::Activity::Create < ActivityPub::Activity
def process_hashtag(tag) def process_hashtag(tag)
return if tag['name'].blank? return if tag['name'].blank?
hashtag = tag['name'].gsub(/\A#/, '').mb_chars.downcase Tag.find_or_create_by_names(tag['name']) do |hashtag|
hashtag = Tag.where(name: hashtag).first_or_create!(name: hashtag) @tags << hashtag unless @tags.include?(hashtag)
end
return if @tags.include?(hashtag)
@tags << hashtag
rescue ActiveRecord::RecordInvalid rescue ActiveRecord::RecordInvalid
nil nil
end end

View File

@ -20,7 +20,7 @@ class Tag < ApplicationRecord
HASHTAG_NAME_RE = '([[:word:]_][[:word:]_·]*[[:alpha:]_·][[:word:]_·]*[[:word:]_])|([[:word:]_]*[[:alpha:]][[:word:]_]*)' HASHTAG_NAME_RE = '([[:word:]_][[:word:]_·]*[[:alpha:]_·][[:word:]_·]*[[:word:]_])|([[:word:]_]*[[:alpha:]][[:word:]_]*)'
HASHTAG_RE = /(?:^|[^\/\)\w])#(#{HASHTAG_NAME_RE})/i HASHTAG_RE = /(?:^|[^\/\)\w])#(#{HASHTAG_NAME_RE})/i
validates :name, presence: true, uniqueness: true, format: { with: /\A(#{HASHTAG_NAME_RE})\z/i } validates :name, presence: true, format: { with: /\A(#{HASHTAG_NAME_RE})\z/i }
scope :discoverable, -> { joins(:account_tag_stat).where(AccountTagStat.arel_table[:accounts_count].gt(0)).where(account_tag_stats: { hidden: false }).order(Arel.sql('account_tag_stats.accounts_count desc')) } scope :discoverable, -> { joins(:account_tag_stat).where(AccountTagStat.arel_table[:accounts_count].gt(0)).where(account_tag_stats: { hidden: false }).order(Arel.sql('account_tag_stats.accounts_count desc')) }
scope :hidden, -> { where(account_tag_stats: { hidden: true }) } scope :hidden, -> { where(account_tag_stats: { hidden: true }) }
@ -64,22 +64,48 @@ class Tag < ApplicationRecord
end end
class << self class << self
def search_for(term, limit = 5, offset = 0) def find_or_create_by_names(name_or_names)
pattern = sanitize_sql_like(term.strip) + '%' Array(name_or_names).map(&method(:normalize)).uniq.map do |normalized_name|
tag = matching_name(normalized_name).first || create(name: normalized_name)
Tag.where('lower(name) like lower(?)', pattern) yield tag if block_given?
tag
end
end
def search_for(term, limit = 5, offset = 0)
pattern = sanitize_sql_like(normalize(term.strip)) + '%'
Tag.where(arel_table[:name].lower.matches(pattern.downcase))
.order(:name) .order(:name)
.limit(limit) .limit(limit)
.offset(offset) .offset(offset)
end end
def find_normalized(name) def find_normalized(name)
find_by(name: name.mb_chars.downcase.to_s) matching_name(name).first
end end
def find_normalized!(name) def find_normalized!(name)
find_normalized(name) || raise(ActiveRecord::RecordNotFound) find_normalized(name) || raise(ActiveRecord::RecordNotFound)
end end
def matching_name(name_or_names)
names = Array(name_or_names).map { |name| normalize(name).downcase }
if names.size == 1
where(arel_table[:name].lower.eq(names.first))
else
where(arel_table[:name].lower.in(names))
end
end
private
def normalize(str)
str.gsub(/\A#/, '').mb_chars.to_s
end
end end
private private

View File

@ -14,7 +14,7 @@ class HashtagQueryService < BaseService
private private
def tags_for(tags) def tags_for(names)
Tag.where(name: tags.map(&:downcase)) if tags.presence Tag.matching_name(names) if names.presence
end end
end end

View File

@ -5,9 +5,7 @@ class ProcessHashtagsService < BaseService
tags = Extractor.extract_hashtags(status.text) if status.local? tags = Extractor.extract_hashtags(status.text) if status.local?
records = [] records = []
tags.map { |str| str.mb_chars.downcase }.uniq(&:to_s).each do |name| Tag.find_or_create_by_names(tags) do |tag|
tag = Tag.where(name: name).first_or_create(name: name)
status.tags << tag status.tags << tag
records << tag records << tag

View File

@ -0,0 +1,15 @@
class AddCaseInsensitiveIndexToTags < ActiveRecord::Migration[5.2]
disable_ddl_transaction!
def up
safety_assured { execute 'CREATE UNIQUE INDEX CONCURRENTLY index_tags_on_name_lower ON tags (lower(name))' }
remove_index :tags, name: 'index_tags_on_name'
remove_index :tags, name: 'hashtag_search_index'
end
def down
add_index :tags, :name, unique: true, algorithm: :concurrently
safety_assured { execute 'CREATE INDEX CONCURRENTLY hashtag_search_index ON tags (name text_pattern_ops)' }
remove_index :tags, name: 'index_tags_on_name_lower'
end
end

View File

@ -10,7 +10,7 @@
# #
# It's strongly recommended that you check this file into your version control system. # It's strongly recommended that you check this file into your version control system.
ActiveRecord::Schema.define(version: 2019_07_15_164535) do ActiveRecord::Schema.define(version: 2019_07_26_175042) do
# These are extensions that must be enabled in order to support this database # These are extensions that must be enabled in order to support this database
enable_extension "plpgsql" enable_extension "plpgsql"
@ -652,8 +652,7 @@ ActiveRecord::Schema.define(version: 2019_07_15_164535) do
t.string "name", default: "", null: false t.string "name", default: "", null: false
t.datetime "created_at", null: false t.datetime "created_at", null: false
t.datetime "updated_at", null: false t.datetime "updated_at", null: false
t.index "lower((name)::text) text_pattern_ops", name: "hashtag_search_index" t.index "lower((name)::text)", name: "index_tags_on_name_lower", unique: true
t.index ["name"], name: "index_tags_on_name", unique: true
end end
create_table "tombstones", force: :cascade do |t| create_table "tombstones", force: :cascade do |t|