demarches-normaliennes/app/models/export.rb
mfo 98d483ec32
Merge pull request #9552 from demarches-simplifiees/US/fix-export-with-file-bigger-than-4Go
correctif(export): lorsqu'un export .zip fait plus de 4Go, le fait del'assigner via file.attach(blob) declenche soit un identify soit un virus scanner qui par la suite le vide sur le bucket
2023-10-03 13:40:12 +00:00

160 lines
4.1 KiB
Ruby

class Export < ApplicationRecord
include TransientModelsWithPurgeableJobConcern
MAX_DUREE_CONSERVATION_EXPORT = 32.hours
MAX_DUREE_GENERATION = 16.hours
enum format: {
csv: 'csv',
ods: 'ods',
xlsx: 'xlsx',
zip: 'zip',
json: 'json'
}, _prefix: true
enum time_span_type: {
everything: 'everything',
monthly: 'monthly'
}
enum statut: {
'a-suivre': 'a-suivre',
suivis: 'suivis',
traites: 'traites',
tous: 'tous',
supprimes_recemment: 'supprimes_recemment',
archives: 'archives',
expirant: 'expirant'
}
has_and_belongs_to_many :groupe_instructeurs
belongs_to :procedure_presentation, optional: true
has_one_attached :file
validates :format, :groupe_instructeurs, :key, presence: true
scope :ante_chronological, -> { order(updated_at: :desc) }
after_create_commit :compute_async
FORMATS_WITH_TIME_SPAN = [:xlsx, :ods, :csv].flat_map do |format|
[{ format: format, time_span_type: 'everything' }]
end
FORMATS = [:xlsx, :ods, :csv, :zip, :json].map do |format|
{ format: format }
end
def compute_async
ExportJob.perform_later(self)
end
def compute
load_snapshot!
file.attach(blob.signed_id) # attaching a blob directly might run identify/virus scanner and wipe it
end
def since
time_span_type == Export.time_span_types.fetch(:monthly) ? 30.days.ago : nil
end
def filtered?
procedure_presentation_id.present?
end
def self.find_or_create_fresh_export(format, groupe_instructeurs, time_span_type: time_span_types.fetch(:everything), statut: statuts.fetch(:tous), procedure_presentation: nil)
attributes = {
format:,
time_span_type:,
statut:,
key: generate_cache_key(groupe_instructeurs.map(&:id), procedure_presentation)
}
recent_export = pending
.or(generated.where(updated_at: (5.minutes.ago)..))
.includes(:procedure_presentation)
.find_by(attributes)
return recent_export if recent_export.present?
create!(**attributes, groupe_instructeurs:,
procedure_presentation:,
procedure_presentation_snapshot: procedure_presentation&.snapshot)
end
def self.for_groupe_instructeurs(groupe_instructeurs_ids)
joins(:groupe_instructeurs).where(groupe_instructeurs: groupe_instructeurs_ids).distinct(:id)
end
def self.by_key(groupe_instructeurs_ids, procedure_presentation)
where(key: [
generate_cache_key(groupe_instructeurs_ids),
generate_cache_key(groupe_instructeurs_ids, procedure_presentation)
])
end
def self.generate_cache_key(groupe_instructeurs_ids, procedure_presentation = nil)
if procedure_presentation.present?
[
groupe_instructeurs_ids.sort.join('-'),
procedure_presentation.id,
Digest::MD5.hexdigest(procedure_presentation.snapshot.slice(:filters, :sort).to_s)
].join('--')
else
groupe_instructeurs_ids.sort.join('-')
end
end
def count
if procedure_presentation_id.present?
dossiers_for_export.count
end
end
def procedure
groupe_instructeurs.first.procedure
end
private
def load_snapshot!
if procedure_presentation_snapshot.present?
procedure_presentation.attributes = procedure_presentation_snapshot
end
end
def dossiers_for_export
@dossiers_for_export ||= begin
dossiers = Dossier.where(groupe_instructeur: groupe_instructeurs)
if since.present?
dossiers.visible_by_administration.where('dossiers.depose_at > ?', since)
elsif procedure_presentation.present?
filtered_sorted_ids = procedure_presentation
.filtered_sorted_ids(dossiers, statut)
dossiers.where(id: filtered_sorted_ids)
else
dossiers.visible_by_administration
end
end
end
def blob
service = ProcedureExportService.new(procedure, dossiers_for_export)
case format.to_sym
when :csv
service.to_csv
when :xlsx
service.to_xlsx
when :ods
service.to_ods
when :zip
service.to_zip
when :json
service.to_geo_json
end
end
end