oca-ocb-core/odoo-bringout-oca-ocb-rating/rating/models/rating_mixin.py

# Part of Odoo. See LICENSE file for full copyright and licensing details.

from collections import defaultdict

from odoo import api, fields, models
from odoo.addons.rating.models import rating_data
from odoo.fields import Domain
from odoo.tools.float_utils import float_compare, float_round


class RatingMixin(models.AbstractModel):
    """This mixin adds rating statistics to mail.thread that already support ratings."""
    _name = 'rating.mixin'
    _description = "Rating Mixin"
    _inherit = ['mail.thread']

    rating_last_value = fields.Float('Rating Last Value', groups='base.group_user', compute='_compute_rating_last_value', compute_sudo=True, store=True, aggregator="avg")
    rating_last_feedback = fields.Text('Rating Last Feedback', groups='base.group_user', related='rating_ids.feedback')
    rating_last_image = fields.Binary('Rating Last Image', groups='base.group_user', related='rating_ids.rating_image')
    rating_count = fields.Integer('Rating count', compute="_compute_rating_stats", compute_sudo=True)
    rating_avg = fields.Float("Average Rating", groups='base.group_user',
        compute='_compute_rating_stats', compute_sudo=True, search='_search_rating_avg')
    rating_avg_text = fields.Selection(rating_data.RATING_TEXT, groups='base.group_user',
        compute='_compute_rating_avg_text', compute_sudo=True)
    rating_percentage_satisfaction = fields.Float("Rating Satisfaction", compute='_compute_rating_satisfaction', compute_sudo=True)
    rating_last_text = fields.Selection(string="Rating Text", groups='base.group_user', related="rating_ids.rating_text")

    @api.depends('rating_ids', 'rating_ids.rating', 'rating_ids.consumed')
    def _compute_rating_last_value(self):
        # Pure SQL instead of calling read_group to allow ordering array_agg
        self.flush_model(['rating_ids'])
        self.env['rating.rating'].flush_model(['consumed', 'rating'])
        if not self.ids:
            self.rating_last_value = 0
            return
        self.env.cr.execute("""
            SELECT
                array_agg(rating ORDER BY write_date DESC, id DESC) AS "ratings",
                res_id as res_id
            FROM "rating_rating"
            WHERE
                res_model = %s
            AND res_id in %s
            AND consumed = true
            GROUP BY res_id""", [self._name, tuple(self.ids)])
        read_group_raw = self.env.cr.dictfetchall()
        rating_by_res_id = {e['res_id']: e['ratings'][0] for e in read_group_raw}
        for record in self:
            record.rating_last_value = rating_by_res_id.get(record.id, 0)

    @api.depends('rating_ids.res_id', 'rating_ids.rating')
    def _compute_rating_stats(self):
        """ Compute avg and count in one query, as thoses fields will be used together most of the time. """
        domain = self._rating_domain() & Domain('rating', '>=', rating_data.RATING_LIMIT_MIN)
        read_group_res = self.env['rating.rating']._read_group(domain, ['res_id'], aggregates=['__count', 'rating:avg'])  # force average on rating column
        mapping = {res_id: {'rating_count': count, 'rating_avg': rating_avg} for res_id, count, rating_avg in read_group_res}
        for record in self:
            record.rating_count = mapping.get(record.id, {}).get('rating_count', 0)
            record.rating_avg = mapping.get(record.id, {}).get('rating_avg', 0)

    def _search_rating_avg(self, operator, value):
        op = rating_data.OPERATOR_MAPPING.get(operator)
        if not op:
            return NotImplemented
        rating_read_group = self.env['rating.rating'].sudo()._read_group(
            [('res_model', '=', self._name), ('consumed', '=', True), ('rating', '>=', rating_data.RATING_LIMIT_MIN)],
            ['res_id'], ['rating:avg'])
        res_ids = [
            res_id
            for res_id, rating_avg in rating_read_group
            if op(float_compare(rating_avg, value, 2), 0)
        ]
        return [('id', 'in', res_ids)]

    @api.depends('rating_avg')
    def _compute_rating_avg_text(self):
        for record in self:
            record.rating_avg_text = rating_data._rating_avg_to_text(record.rating_avg)

    @api.depends('rating_ids.res_id', 'rating_ids.rating')
    def _compute_rating_satisfaction(self):
        """ Compute the rating satisfaction percentage, this is done separately from rating_count and rating_avg
            since the query is different, to avoid computing if it is not necessary"""
        domain = self._rating_domain() & Domain('rating', '>=', rating_data.RATING_LIMIT_MIN)
        # See `_compute_rating_percentage_satisfaction` above
        read_group_res = self.env['rating.rating']._read_group(domain, ['res_id', 'rating'], aggregates=['__count'])
        default_grades = {'great': 0, 'okay': 0, 'bad': 0}
        grades_per_record = {record_id: default_grades.copy() for record_id in self.ids}

        for record_id, rating, count in read_group_res:
            grade = rating_data._rating_to_grade(rating)
            grades_per_record[record_id][grade] += count

        for record in self:
            grade_repartition = grades_per_record.get(record.id, default_grades)
            grade_count = sum(grade_repartition.values())
            record.rating_percentage_satisfaction = grade_repartition['great'] * 100 / grade_count if grade_count else -1

    def write(self, vals):
        """ If the rated ressource name is modified, we should update the rating res_name too.
            If the rated ressource parent is changed we should update the parent_res_id too"""
        result = super().write(vals)
        for record in self.sudo():  # ratings may be inaccessible
            if record._rec_name in vals:  # set the res_name of ratings to be recomputed
                res_name_field = self.env['rating.rating']._fields['res_name']
                self.env.add_to_compute(res_name_field, record.rating_ids)
            if record._rating_get_parent_field_name() in vals:
                record.rating_ids.write({'parent_res_id': record[record._rating_get_parent_field_name()].id})

        return result

    def _rating_get_parent_field_name(self):
        """Return the parent relation field name. Should return a Many2One"""
        return None

    def _rating_domain(self):
        """ Returns a normalized domain on rating.rating to select the records to
            include in count, avg, ... computation of current model.
        """
        return Domain([('res_model', '=', self._name), ('res_id', 'in', self.ids), ('consumed', '=', True)])

    def _rating_get_repartition(self, add_stats=False, domain=None):
        """ get the repatition of rating grade for the given res_ids.
            :param add_stats : flag to add stat to the result
            :type add_stats : boolean
            :param domain : optional extra domain of the rating to include/exclude in repartition
            :return dictionnary
                if not add_stats, the dict is like
                    - key is the rating value (integer)
                    - value is the number of object (res_model, res_id) having the value
                otherwise, key is the value of the information (string) : either stat name (avg, total, ...) or 'repartition'
                containing the same dict if add_stats was False.
        """
        base_domain = self._rating_domain() & Domain('rating', '>=', 1)
        if domain:
            base_domain &= Domain(domain)
        rg_data = self.env['rating.rating']._read_group(base_domain, ['rating'], ['__count'])
        # init dict with all possible rate value, except 0 (no value for the rating)
        values = dict.fromkeys(range(1, 6), 0)
        for rating, count in rg_data:
            rating_val_round = float_round(rating, precision_digits=1)
            values[rating_val_round] = values.get(rating_val_round, 0) + count
        # add other stats
        if add_stats:
            rating_number = sum(values.values())
            return {
                'repartition': values,
                'avg': sum(float(key * values[key]) for key in values) / rating_number if rating_number > 0 else 0,
                'total': sum(count for __, count in rg_data),
            }
        return values

    def rating_get_grades(self, domain=None):
        """ Get the repartitions of rating grade for the given res_ids.
        :param domain: Optional domain of the rating to include/exclude
            in the grades computation.
        :returns: A dictionary where the key is the rating and the value
            is the count of unique ``(res_model, res_id)`` pairs whose
            grades are associated with that rating.

            The rates are:

            * ``"great"``, graded between 70 and 100
            * ``"okay"``, graded between 31 and 69
            * ``"bad"``, graded between 0 and 30
        :rtype: dict[typing.Literal["great", "okay", "bad"], int]
        """
        data = self._rating_get_repartition(domain=domain)
        res = dict.fromkeys(['great', 'okay', 'bad'], 0)
        for key in data:
            grade = rating_data._rating_to_grade(key)
            res[grade] += data[key]
        return res

    def rating_get_stats(self, domain=None):
        """Get the statistics of the rating repartitions

        :param domain : optional domain of the rating to include/exclude in statistic computation
        :returns: A dictionnary where:

            - key is the name of the information (stat name)
            - value is statistic value : 'percent' contains the repartition in percentage, 'avg' is the average rate
              and 'total' is the number of rating
        """
        data = self._rating_get_repartition(domain=domain, add_stats=True)
        result = {
            'avg': data['avg'],
            'total': data['total'],
            'percent': dict.fromkeys(range(1, 6), 0),
        }
        for rate in data['repartition']:
            result['percent'][rate] = (data['repartition'][rate] * 100) / data['total'] if data['total'] > 0 else 0
        return result

    def _rating_get_stats_per_record(self, domain=None):
        """
        Computes rating statistics for each record individually.

        :param domain: Optional domain to apply on the ratings.
        :return: A dictionary mapping each record ID to its statistics dictionary.
        :rtype: dict
        """
        base_domain = self._rating_domain() & Domain("rating", ">=", 1)
        if domain:
            base_domain &= Domain(domain)
        rg_data = self.env["rating.rating"]._read_group(
            base_domain,
            groupby=["res_id", "rating"],
            aggregates=["__count"],
        )
        stats_per_record = defaultdict(
            lambda: {"total": 0, "weighted_sum": 0.0, "counts": defaultdict(int), "percent": {}}
        )
        for res_id, rating, count in rg_data:
            stats = stats_per_record[res_id]
            stats["total"] += count
            stats["weighted_sum"] += rating * count
            stats["counts"][int(rating)] = count
        for stats in stats_per_record.values():
            total = stats["total"]
            if total > 0:
                stats["avg"] = stats["weighted_sum"] / total
                stats["percent"] = {
                    rate: (stats["counts"].get(rate, 0) * 100) / total for rate in range(1, 6)
                }
            else:
                stats["avg"] = 0
                stats["percent"] = dict.fromkeys(range(1, 6), 0.0)
            del stats["weighted_sum"]
            del stats["counts"]
        return stats_per_record

    @api.model
    def _allow_publish_rating_stats(self):
        """Override to allow the rating stats to be demonstrated."""
        return False