Account archive download (#6460)
* Fix #201: Account archive download * Export actor and private key in the archive * Optimize BackupService - Add conversation to cached associations of status, because somehow it was forgotten and is source of N+1 queries - Explicitly call GC between batches of records being fetched (Model class allocations are the worst offender) - Stream media files into the tar in 1MB chunks (Do not allocate media file (up to 8MB) as string into memory) - Use #bytesize instead of #size to calculate file size for JSON (Fix FileOverflow error) - Segment media into subfolders by status ID because apparently GIF-to-MP4 media are all named "media.mp4" for some reason * Keep uniquely generated filename in Paperclip::GifTranscoder * Ensure dumped files do not overwrite each other by maintaing directory partitions * Give tar archives a good name * Add scheduler to remove week-old backups * Fix code style issuemaster
Eugen Rochko
7 years ago
committed by
GitHub
parent
c1e77b56a9
commit
61ed133fea
@ -1,11 +1,23 @@ |
||||
# frozen_string_literal: true |
||||
|
||||
class Settings::ExportsController < ApplicationController |
||||
include Authorization |
||||
|
||||
layout 'admin' |
||||
|
||||
before_action :authenticate_user! |
||||
|
||||
def show |
||||
@export = Export.new(current_account) |
||||
@export = Export.new(current_account) |
||||
@backups = current_user.backups |
||||
end |
||||
|
||||
def create |
||||
authorize :backup, :create? |
||||
|
||||
backup = current_user.backups.create! |
||||
BackupWorker.perform_async(backup.id) |
||||
|
||||
redirect_to settings_export_path |
||||
end |
||||
end |
||||
|
||||
|
After Width: | Height: | Size: 205 B |
|
After Width: | Height: | Size: 271 B |
@ -0,0 +1,22 @@ |
||||
# frozen_string_literal: true |
||||
# == Schema Information |
||||
# |
||||
# Table name: backups |
||||
# |
||||
# id :integer not null, primary key |
||||
# user_id :integer |
||||
# dump_file_name :string |
||||
# dump_content_type :string |
||||
# dump_file_size :integer |
||||
# dump_updated_at :datetime |
||||
# processed :boolean default(FALSE), not null |
||||
# created_at :datetime not null |
||||
# updated_at :datetime not null |
||||
# |
||||
|
||||
class Backup < ApplicationRecord |
||||
belongs_to :user, inverse_of: :backups |
||||
|
||||
has_attached_file :dump |
||||
do_not_validate_attachment_file_type :dump |
||||
end |
||||
@ -0,0 +1,9 @@ |
||||
# frozen_string_literal: true |
||||
|
||||
class BackupPolicy < ApplicationPolicy |
||||
MIN_AGE = 1.week |
||||
|
||||
def create? |
||||
user_signed_in? && current_user.backups.where('created_at >= ?', MIN_AGE.ago).count.zero? |
||||
end |
||||
end |
||||
@ -0,0 +1,128 @@ |
||||
# frozen_string_literal: true |
||||
|
||||
require 'rubygems/package' |
||||
|
||||
class BackupService < BaseService |
||||
attr_reader :account, :backup, :collection |
||||
|
||||
def call(backup) |
||||
@backup = backup |
||||
@account = backup.user.account |
||||
|
||||
build_json! |
||||
build_archive! |
||||
end |
||||
|
||||
private |
||||
|
||||
def build_json! |
||||
@collection = serialize(collection_presenter, ActivityPub::CollectionSerializer) |
||||
|
||||
account.statuses.with_includes.find_in_batches do |statuses| |
||||
statuses.each do |status| |
||||
item = serialize(status, ActivityPub::ActivitySerializer) |
||||
item.delete(:'@context') |
||||
|
||||
unless item[:type] == 'Announce' || item[:object][:attachment].blank? |
||||
item[:object][:attachment].each do |attachment| |
||||
attachment[:url] = Addressable::URI.parse(attachment[:url]).path.gsub(/\A\/system\//, '') |
||||
end |
||||
end |
||||
|
||||
@collection[:orderedItems] << item |
||||
end |
||||
|
||||
GC.start |
||||
end |
||||
end |
||||
|
||||
def build_archive! |
||||
tmp_file = Tempfile.new(%w(archive .tar.gz)) |
||||
|
||||
File.open(tmp_file, 'wb') do |file| |
||||
Zlib::GzipWriter.wrap(file) do |gz| |
||||
Gem::Package::TarWriter.new(gz) do |tar| |
||||
dump_media_attachments!(tar) |
||||
dump_outbox!(tar) |
||||
dump_actor!(tar) |
||||
end |
||||
end |
||||
end |
||||
|
||||
archive_filename = ['archive', Time.now.utc.strftime('%Y%m%d%H%M%S'), SecureRandom.hex(2)].join('-') + '.tar.gz' |
||||
|
||||
@backup.dump = ActionDispatch::Http::UploadedFile.new(tempfile: tmp_file, filename: archive_filename) |
||||
@backup.processed = true |
||||
@backup.save! |
||||
ensure |
||||
tmp_file.close |
||||
tmp_file.unlink |
||||
end |
||||
|
||||
def dump_media_attachments!(tar) |
||||
MediaAttachment.attached.where(account: account).find_in_batches do |media_attachments| |
||||
media_attachments.each do |m| |
||||
download_to_tar(tar, m.file, m.file.path) |
||||
end |
||||
|
||||
GC.start |
||||
end |
||||
end |
||||
|
||||
def dump_outbox!(tar) |
||||
json = Oj.dump(collection) |
||||
|
||||
tar.add_file_simple('outbox.json', 0o444, json.bytesize) do |io| |
||||
io.write(json) |
||||
end |
||||
end |
||||
|
||||
def dump_actor!(tar) |
||||
actor = serialize(account, ActivityPub::ActorSerializer) |
||||
|
||||
actor[:icon][:url] = 'avatar' + File.extname(actor[:icon][:url]) if actor[:icon] |
||||
actor[:image][:url] = 'header' + File.extname(actor[:image][:url]) if actor[:image] |
||||
|
||||
download_to_tar(tar, account.avatar, 'avatar' + File.extname(account.avatar.path)) if account.avatar.exists? |
||||
download_to_tar(tar, account.header, 'header' + File.extname(account.header.path)) if account.header.exists? |
||||
|
||||
json = Oj.dump(actor) |
||||
|
||||
tar.add_file_simple('actor.json', 0o444, json.bytesize) do |io| |
||||
io.write(json) |
||||
end |
||||
|
||||
tar.add_file_simple('key.pem', 0o444, account.private_key.bytesize) do |io| |
||||
io.write(account.private_key) |
||||
end |
||||
end |
||||
|
||||
def collection_presenter |
||||
ActivityPub::CollectionPresenter.new( |
||||
id: account_outbox_url(account), |
||||
type: :ordered, |
||||
size: account.statuses_count, |
||||
items: [] |
||||
) |
||||
end |
||||
|
||||
def serialize(object, serializer) |
||||
ActiveModelSerializers::SerializableResource.new( |
||||
object, |
||||
serializer: serializer, |
||||
adapter: ActivityPub::Adapter |
||||
).as_json |
||||
end |
||||
|
||||
CHUNK_SIZE = 1.megabyte |
||||
|
||||
def download_to_tar(tar, attachment, filename) |
||||
adapter = Paperclip.io_adapters.for(attachment) |
||||
|
||||
tar.add_file_simple(filename, 0o444, adapter.size) do |io| |
||||
while (buffer = adapter.read(CHUNK_SIZE)) |
||||
io.write(buffer) |
||||
end |
||||
end |
||||
end |
||||
end |
||||
@ -0,0 +1,59 @@ |
||||
%table.email-table{ cellspacing: 0, cellpadding: 0 } |
||||
%tbody |
||||
%tr |
||||
%td.email-body |
||||
.email-container |
||||
%table.content-section{ cellspacing: 0, cellpadding: 0 } |
||||
%tbody |
||||
%tr |
||||
%td.content-cell.hero |
||||
.email-row |
||||
.col-6 |
||||
%table.column{ cellspacing: 0, cellpadding: 0 } |
||||
%tbody |
||||
%tr |
||||
%td.column-cell.text-center.padded |
||||
%table.hero-icon{ align: 'center', cellspacing: 0, cellpadding: 0 } |
||||
%tbody |
||||
%tr |
||||
%td |
||||
= image_tag full_pack_url('icon_file_download.png'), alt: '' |
||||
|
||||
%h1= t 'user_mailer.backup_ready.title' |
||||
|
||||
%table.email-table{ cellspacing: 0, cellpadding: 0 } |
||||
%tbody |
||||
%tr |
||||
%td.email-body |
||||
.email-container |
||||
%table.content-section{ cellspacing: 0, cellpadding: 0 } |
||||
%tbody |
||||
%tr |
||||
%td.content-cell.content-start |
||||
.email-row |
||||
.col-6 |
||||
%table.column{ cellspacing: 0, cellpadding: 0 } |
||||
%tbody |
||||
%tr |
||||
%td.column-cell.text-center |
||||
%p= t 'user_mailer.backup_ready.explanation' |
||||
|
||||
%table.email-table{ cellspacing: 0, cellpadding: 0 } |
||||
%tbody |
||||
%tr |
||||
%td.email-body |
||||
.email-container |
||||
%table.content-section{ cellspacing: 0, cellpadding: 0 } |
||||
%tbody |
||||
%tr |
||||
%td.content-cell |
||||
%table.column{ cellspacing: 0, cellpadding: 0 } |
||||
%tbody |
||||
%tr |
||||
%td.column-cell.button-cell |
||||
%table.button{ align: 'center', cellspacing: 0, cellpadding: 0 } |
||||
%tbody |
||||
%tr |
||||
%td.button-primary |
||||
= link_to full_asset_url(@backup.dump.url) do |
||||
%span= t 'exports.archive_takeout.download' |
||||
@ -0,0 +1,7 @@ |
||||
<%= t 'user_mailer.backup_ready.title' %> |
||||
|
||||
=== |
||||
|
||||
<%= t 'user_mailer.backup_ready.explanation' %> |
||||
|
||||
=> <%= full_asset_url(@backup.dump.url) %> |
||||
@ -0,0 +1,17 @@ |
||||
# frozen_string_literal: true |
||||
|
||||
class BackupWorker |
||||
include Sidekiq::Worker |
||||
|
||||
sidekiq_options queue: 'pull' |
||||
|
||||
def perform(backup_id) |
||||
backup = Backup.find(backup_id) |
||||
user = backup.user |
||||
|
||||
BackupService.new.call(backup) |
||||
|
||||
user.backups.where.not(id: backup.id).destroy_all |
||||
UserMailer.backup_ready(user, backup).deliver_later |
||||
end |
||||
end |
||||
@ -0,0 +1,16 @@ |
||||
# frozen_string_literal: true |
||||
require 'sidekiq-scheduler' |
||||
|
||||
class Scheduler::BackupCleanupScheduler |
||||
include Sidekiq::Worker |
||||
|
||||
def perform |
||||
old_backups.find_each(&:destroy!) |
||||
end |
||||
|
||||
private |
||||
|
||||
def old_backups |
||||
Backup.where('created_at < ?', 7.days.ago) |
||||
end |
||||
end |
||||
@ -0,0 +1,11 @@ |
||||
class CreateBackups < ActiveRecord::Migration[5.1] |
||||
def change |
||||
create_table :backups do |t| |
||||
t.references :user, foreign_key: { on_delete: :nullify } |
||||
t.attachment :dump |
||||
t.boolean :processed, null: false, default: false |
||||
|
||||
t.timestamps |
||||
end |
||||
end |
||||
end |
||||
@ -0,0 +1,3 @@ |
||||
Fabricator(:backup) do |
||||
user |
||||
end |
||||
@ -0,0 +1,5 @@ |
||||
require 'rails_helper' |
||||
|
||||
RSpec.describe Backup, type: :model do |
||||
|
||||
end |
||||
Loading…
Reference in new issue