[16.0][MIG] procurement_mto_analytic: Migration to 16.0

[AungKoKoLin1997]

@qrtl QT4346

[yostashiro]

Suggested change

	# In case of using stock_analytic and mrp_stock_analytic
	if not analytic_distribution and hasattr(self, "analytic_distribution"):
	analytic_distribution = self.analytic_distribution

Let's remove this as stock_analytic already takes care of this:

account-analytic/stock_analytic/models/stock_move.py

Lines 34 to 46 in 14ecb3f

	def _prepare_procurement_values(self):
	"""
	Allows to transmit analytic account from moves to new
	moves through procurement.
	"""
	res = super()._prepare_procurement_values()
	if self.analytic_distribution:
	res.update(
	{
	"analytic_distribution": self.analytic_distribution,
	}
	)
	return res

[yostashiro]

Can you please check if this works.

Suggested change

	def _find_matching_analytic_distribution_record(self, analytic_distribution):
	# Prepare the JSONB structure as a string for the SQL query
	analytic_distribution_str = json.dumps(analytic_distribution)
	# Use a parameterized query for safety
	query = """
	SELECT id FROM purchase_order_line
	WHERE analytic_distribution::jsonb = %s::jsonb;
	"""
	self.env.cr.execute(query, (analytic_distribution_str,))
	result = self.env.cr.fetchall()
	# Extract IDs from the result
	matching_po_line = [res[0] for res in result]
	return matching_po_line
	def _make_po_get_domain(self, company_id, values, partner):
	domain = super()._make_po_get_domain(company_id, values, partner)
	if values.get("analytic_distribution"):
	# Fetch matching record IDs based on dynamic analytic_distribution
	matching_po_line = self._find_matching_analytic_distribution_record(
	values["analytic_distribution"]
	)
	if matching_po_line:
	domain += (("order_line", "in", tuple(matching_po_line)),)
	else:
	# To create new PO
	domain += (("order_line", "=", 0000000),)
	else:
	domain += (("order_line.analytic_distribution", "=", False),)
	def _make_po_get_domain(self, company_id, values, partner):
	domain = super()._make_po_get_domain(company_id, values, partner)
	matching_lines = self.env["purchase.order"].sudo().search([dom for dom in domain]).order_line.filtered(lambda x: x.analytic_distribution == values.get("analytic_distribution"))
	if matching_lines:
	domain += (("order_line", "in", matching_lines.ids),)
	else:
	# To create new PO
	domain += (("order_line", "=", 0000000),)

[yostashiro]

Reviewing the design, I now doubt that this should be the way to go, as there can be cases where one purchase order holds mutplie analytic distributions. We should not be extending this method.

We should instead create a PR in odoo/odoo to split https://github.com/OCA/OCB/blob/1e9b8db/addons/purchase_stock/models/stock_rule.py#L125 into a method, and extend that method in this module to add analytic distribution as a key.

[aisopuro]

In Python, dict entries are (basically) random-ordered in versions 3.6 and below, and insertion-ordered after that. So we have the following:

Python 3.10.12 (main, Nov 20 2023, 15:14:05) [GCC 11.4.0]
Type 'copyright', 'credits' or 'license' for more information
IPython 8.16.1 -- An enhanced Interactive Python. Type '?' for help.

In [1]: a = {"1": 100, "2": 100}
In [2]: b = {"2": 100, "1": 100}
In [3]: tuple(a.items()) == tuple(b.items())

Out[3]: False

In [4]: a.items()
Out[4]: dict_items([('1', 100), ('2', 100)])

In [5]: b.items()
Out[5]: dict_items([('2', 100), ('1', 100)])

To me, this intuitively doesn't make sense: if I have a distribution with the same accounts and the same percentages, those should be the same regardless of the order, right? I suggest adding a sorted call into the key generation:

Python 3.10.12 (main, Nov 20 2023, 15:14:05) [GCC 11.4.0]
Type 'copyright', 'credits' or 'license' for more information
IPython 8.16.1 -- An enhanced Interactive Python. Type '?' for help.

In [1]: a = {"1": 100, "2": 100}
In [2]: b = {"2": 100, "1": 100}
In [3]: tuple(sorted(a.items())) == tuple(sorted(b.items()))

Out[3]: True

[AungKoKoLin1997]

if I have a distribution with the same accounts and the same percentages, those should be the same regardless of the order, right? I suggest adding a sorted call into the key generation

@aisopuro items of analytic_distribution are always sorted by default. So, we don't need to worry about it.

[aisopuro]

Could you link to the place in Odoo's source code that ensures they are always sorted, regardless of whether:

• The distribution was added by a user through the UI
• The distribution was added by business logic in Python
• The distribution was imported via excel
• etc

Or does the JSONB column type in Postgresql guarantee keys are ordered?

[AungKoKoLin1997]

Depends on this source(https://www.postgresql.org/docs/10/datatype-json.html#JSON-TYPE-MAPPING-TABLE:~:text=There%20are%20two,value%20is%20kept.), I conclude analytic_distribution are sorted depends on the efficiency.
Whenever we insert the key in any order, it will be sorted if they have same data.
Eg, when we insert {"1":60, "2":40} and {"2":40, "1":60}, they will be sorted. So, the result will be the same.

[aisopuro]

Nice 👍, my suggestion is then not needed.