Define aggregated model

Contents

Define aggregated model#

Aggregated data are useful to analyze business. Tryton can provide such data using ModelSQL class which are not based on an existing table in the database but using a SQL query. This is done by defining a table_query() method which returns a SQL FromItem.

Let’s create a ModelSQL which aggregate the number of opportunity converted or lost per month.

First we create a ModelSQL class which defines a table_query() in opportunity.py:

from sql import Literal
from sql.aggregate import Count, Min
from sql.functions import CurrentTimestamp, DateTrunc
...
class OpportunityMonthly(ModelSQL, ModelView):
    "Opportunity Monthly"
    __name__ = 'training.opportunity.monthly'

    month = fields.Date("Month")
    converted = fields.Integer("Converted")
    lost = fields.Integer("Lost")

    @classmethod
    def table_query(cls):
        pool = Pool()
        Opportunity = pool.get('training.opportunity')
        opportunity = Opportunity.__table__()

        month = cls.month.sql_cast(
            DateTrunc('month', opportunity.end_date))
        query = opportunity.select(
            Literal(0).as_('create_uid'),
            CurrentTimestamp().as_('create_date'),
            Literal(None).as_('write_uid'),
            Literal(None).as_('write_date'),
            Min(opportunity.id).as_('id'),
            month.as_('month'),
            Count(
                Literal('*'),
                filter_=opportunity.state == 'converted').as_('converted'),
            Count(
                Literal('*'),
                filter_=opportunity.state == 'lost').as_('lost'),
            where=opportunity.state.in_(['converted', 'lost']),
            group_by=[month])
        return query

Note

The table query must return a value for all the fields of the model but also a unique id and a value for the create and write fields.

Note

We get the SQL table from the __table__() method.

Note

We use sql_cast() to convert the timestamp returned by date_trunc into a date.

Then as usual we register the ModelSQL class in the in the Pool as type model in __init__.py:

def register():
    ...
    Pool.register(
        ...,
        opportunity.OpportunityMonthly,
        module='opportunity', type_='model')

And to display we create a list view and the menu entry in opportunity.xml:

<tryton>
   <data>
      ...
      <record model="ir.ui.view" id="opportunity_monthly_view_list">
         <field name="model">training.opportunity.monthly</field>
         <field name="type">tree</field>
         <field name="name">opportunity_monthly_list</field>
      </record>

      <record model="ir.action.act_window" id="act_opportunity_monthly_form">
         <field name="name">Monthly Opportunities</field>
         <field name="res_model">training.opportunity.monthly</field>
      </record>
      <record model="ir.action.act_window.view" id="act_opportunity_monthly_form_view">
         <field name="sequence" eval="10"/>
         <field name="view" ref="opportunity_monthly_view_list"/>
         <field name="act_window" ref="act_opportunity_monthly_form"/>
      </record>

      <menuitem
         parent="menu_opportunity"
         action="act_opportunity_monthly_form"
         sequence="50"
         id="menu_opportunity_monthly_form"/>
   </data>
</tryton>

And now the view in view/opportunity_monthly_list.xml:

<tree>
   <field name="month"/>
   <field name="converted"/>
   <field name="lost"/>
</tree>

Update database#

As we have registered new model and XML records, we need to update the database with:

$ trytond-admin -d test --all

And restart the server and reconnect with the client to test computing aggregate:

$ trytond

Note

As you can see the model behaves like the other models, except that you can not create, delete nor write on them.

This is all for your first module. If you want to learn more about Tryton, you can continue on specific topics.