From cfe1b59aa9d252246375e84b153a3edaa1330df3 Mon Sep 17 00:00:00 2001 From: MuK IT GmbH Date: Sat, 13 Jul 2019 10:05:52 +0000 Subject: [PATCH] publish muk_attachment_lobject - 11.0 --- .../models/ir_attachment.py | 86 ++++++++++++++++--- 1 file changed, 74 insertions(+), 12 deletions(-) diff --git a/muk_attachment_lobject/models/ir_attachment.py b/muk_attachment_lobject/models/ir_attachment.py index 9cf2157..b5d9ea9 100644 --- a/muk_attachment_lobject/models/ir_attachment.py +++ b/muk_attachment_lobject/models/ir_attachment.py @@ -1,5 +1,5 @@ ################################################################################### -# +# # MuK Document Management System # # Copyright (C) 2018 MuK IT GmbH @@ -23,6 +23,7 @@ import base64 import logging import mimetypes +import odoo from odoo import api, models, _ from odoo.exceptions import AccessError @@ -31,22 +32,78 @@ from odoo.addons.muk_fields_lobject.fields.lobject import LargeObject _logger = logging.getLogger(__name__) class LObjectIrAttachment(models.Model): - + _inherit = 'ir.attachment' store_lobject = LargeObject( string="Data") - + + def _force_storage_prepare_chunks(self): + """ Technical method to select attachments that need to be migrated + This method automaticaly splits attachment by chunks, + to speed up migration. + + :return list: list of chunks where each chunk is list of attachment ids + [[1,2,3],[40, 42, 12,33], ...] + """ + CHUNK_SIZE = 100 + attachments = self.search(['|', ['res_field', '=', False], ['res_field', '!=', False]]) + storage = self._storage() + chunks = [] + current_chunk = [] + for attach in attachments: + # Detect storage_type of attachment + if attach.db_datas: + current = 'db' + elif attach.store_lobject: + current = 'lobject' + elif attach.store_fname: + current = 'file' + else: + current = None + + if storage != current: + # This attachment needs migration, thus adding it to result + current_chunk += [attach.id] + if len(current_chunk) >= CHUNK_SIZE: + chunks += [current_chunk] + current_chunk = [] + + if current_chunk: + chunks += [current_chunk] + return chunks + @api.model def force_storage(self): if not self.env.user._is_admin(): raise AccessError(_('Only administrators can execute this action.')) - attachments = self.search(['|', ['res_field', '=', False], ['res_field', '!=', False]]) - for index, attach in enumerate(attachments): - _logger.info(_("Migrate Attachment %s of %s") % (index, len(attachments))) - attach.write({'datas': attach.datas}) - return True - + + # Do migration by chunks to make it faster. + chunks_to_migrate = self._force_storage_prepare_chunks() + for chunk_index, chunk in enumerate(self._force_storage_prepare_chunks()): + # Here we need to precess each chunk in new transaction. + # When all attachments in chunk processed, then commit. + # In case of any errors - rollback + with api.Environment.manage(): + with odoo.registry(self.env.cr.dbname).cursor() as new_cr: + new_env = api.Environment(new_cr, self.env.uid, + self.env.context.copy()) + attachments = new_env['ir.attachment'].browse(chunk) + try: + for index, attach in enumerate(attachments): + _logger.info( + "Migrate Attachment %s of %s [chunk %s of %s]", + index, len(attachments), + chunk_index, len(chunks_to_migrate)) + attach.write({'datas': attach.datas}) + except Exception: + _logger.error( + "Cannot migrate attachments.", exc_info=True) + new_cr.rollback() + raise + else: + new_cr.commit() + @api.depends('store_fname', 'db_datas', 'store_lobject') def _compute_datas(self): bin_size = self._context.get('bin_size') @@ -58,7 +115,7 @@ class LObjectIrAttachment(models.Model): attach.datas = attach.with_context({'base64': True}).store_lobject else: super(LObjectIrAttachment, attach)._compute_datas() - + def _inverse_datas(self): location = self._storage() for attach in self: @@ -79,7 +136,12 @@ class LObjectIrAttachment(models.Model): self._file_delete(fname) else: super(LObjectIrAttachment, attach)._inverse_datas() - + # It is required to set 'store_lobject' to false, because it is + # used in muk_dms_attachment to detect storage type of + # attachment, thus it is impossible to detect attachments that + # need migration 'LObject -> File' or 'LObject -> smthng' + attach.write({'store_lobject': False}) + def _compute_mimetype(self, values): mimetype = super(LObjectIrAttachment, self)._compute_mimetype(values) if not mimetype or mimetype == 'application/octet-stream': @@ -89,4 +151,4 @@ class LObjectIrAttachment(models.Model): mimetype = attach.mimetype if not mimetype and attach.datas_fname: mimetype = mimetypes.guess_type(attach.datas_fname)[0] - return mimetype or 'application/octet-stream' + return mimetype or 'application/octet-stream' \ No newline at end of file