trjhtr

I created this method that adds a new situation to the situation list. The situation has a many-to-one relation with project.task. This method seems a little longer than it should be and I don't know how to optimize it:

@api.multi
def create_situation(self):
 active_id = self.env.context.get('active_ids', ) or 
 proj_obj = self.env['project.task'].browse(active_id)
 situation_orders = self.env['project.task.situation'].browse() 
 task_lines = self.env['task.situation.line'].browse()
 situation_ids = self.env['project.task.situation'].search([('task_id', '=', proj_obj.id)], order='id desc', limit=1)
 size_list_situation_ids = len(situation_ids)
 last_id = situation_ids and max(situation_ids)
 s_sequence = "Situation N° " + str(proj_obj.sequence)

 for record in proj_obj: 
 task_lines = [(0, 0, 'ref': line.ref,
 'product_id': line.product_id.id,
 'quantity': line.quantity,
 'um_id': line.um_id,
 'price_unit': line.price_unit,
 'price_subtotal': line.price_subtotal,
 'progress': line.progress,
 'cumulative_amount_ht': line.cumulative_amount_ht,
 'previous_amount_ht':line1.cumulative_amount_ht if size_list_situation_ids != 0 else 0,
 'amount_net_ht': line.cumulative_amount_ht if size_list_situation_ids == 0 else (line.cumulative_amount_ht - line1.cumulative_amount_ht)) for line, line1 in izip_longest(record.task_line_ids, last_id.task_line)]
 situation_order = 'partner_id': proj_obj.project_id.partner_id.id,
 'task_id':proj_obj.id,
 'task_line':task_lines,
 'user_id': proj_obj.user_id.id,
 'name':s_sequence,
 'tyoe': 'situation' if self._context.get('type', 'situation') else 'final_count',
 'project_id':proj_obj.project_id.id,
 'due_date': self.date_id.id
 situation_orders+=situation_orders.with_context(self._context).create(situation_order) 
 proj_obj.write('sequence':proj_obj.sequence+1)

First off, make it more maintainable: break out that long list comprehension into a for loop so it's clearer where line is coming from.

As for optimization, since size_list_situation_ids doesn't change through the loop, you can refactor those trinary ops into functions and then set them just once at the beginning:

if size_list_situation_ids:
 prev_amount_ht = lambda l: l.cumulative_amount_ht
 amount_net_ht = lambda l, l1: l.cumulative_amount_ht - l1.cumulative_amount_ht
else:
 prev_amount_ht = lambda l: 0
 amount_net_ht = lambda l, l1: l.cumulative_amount_ht

and then do:

...
'previous_amount_ht': prev_amount_ht(line1),
'amount_net_ht': amount_net_ht(line, line1),
...

which saves a couple of comparisons every time through the loop.

Much more than that is difficult to suss out, since the code looks non-functional overall: where does situation_orders come from? Why is the initial task_lines even being loaded, as it doesn't seem to be used?