# -*- coding: utf-8 -*-
# Part of Odoo. See LICENSE file for full copyright and licensing details.

import logging
import textwrap
import uuid

from dateutil.relativedelta import relativedelta

from odoo import api, fields, models, _
from odoo.exceptions import ValidationError, UserError
from odoo.tools import float_is_zero

_logger = logging.getLogger(__name__)


class SurveyUserInput(models.Model):
    """ Metadata for a set of one user's answers to a particular survey """
    _name = "survey.user_input"
    _description = "Survey User Input"
    _rec_name = "survey_id"
    _order = "create_date desc"
    _inherit = ['mail.thread', 'mail.activity.mixin']

    # answer description
    survey_id = fields.Many2one('survey.survey', string='Survey', required=True, readonly=True, index=True, ondelete='cascade')
    scoring_type = fields.Selection(string="Scoring", related="survey_id.scoring_type")
    start_datetime = fields.Datetime('Start date and time', readonly=True)
    end_datetime = fields.Datetime('End date and time', readonly=True)
    deadline = fields.Datetime('Deadline', help="Datetime until customer can open the survey and submit answers")
    state = fields.Selection([
        ('new', 'New'),
        ('in_progress', 'In Progress'),
        ('done', 'Completed')], string='Status', default='new', readonly=True)
    test_entry = fields.Boolean(readonly=True)
    last_displayed_page_id = fields.Many2one('survey.question', string='Last displayed question/page')
    # attempts management
    is_attempts_limited = fields.Boolean("Limited number of attempts", related='survey_id.is_attempts_limited')
    attempts_limit = fields.Integer("Number of attempts", related='survey_id.attempts_limit')
    attempts_count = fields.Integer("Attempts Count", compute='_compute_attempts_info')
    attempts_number = fields.Integer("Attempt n°", compute='_compute_attempts_info')
    survey_time_limit_reached = fields.Boolean("Survey Time Limit Reached", compute='_compute_survey_time_limit_reached')
    # identification / access
    access_token = fields.Char('Identification token', default=lambda self: str(uuid.uuid4()), readonly=True, required=True, copy=False)
    invite_token = fields.Char('Invite token', readonly=True, copy=False)  # no unique constraint, as it identifies a pool of attempts
    partner_id = fields.Many2one('res.partner', string='Contact', readonly=True, index='btree_not_null')
    email = fields.Char('Email', readonly=True)
    nickname = fields.Char('Nickname', help="Attendee nickname, mainly used to identify them in the survey session leaderboard.")
    # questions / answers
    user_input_line_ids = fields.One2many('survey.user_input.line', 'user_input_id', string='Answers', copy=True)
    predefined_question_ids = fields.Many2many('survey.question', string='Predefined Questions', readonly=True)
    scoring_percentage = fields.Float("Score (%)", compute="_compute_scoring_values", store=True, compute_sudo=True)  # stored for perf reasons
    scoring_total = fields.Float("Total Score", compute="_compute_scoring_values", store=True, compute_sudo=True, digits=(10, 2))  # stored for perf reasons
    scoring_success = fields.Boolean('Quizz Passed', compute='_compute_scoring_success', store=True, compute_sudo=True)  # stored for perf reasons
    survey_first_submitted = fields.Boolean(string='Survey First Submitted')
    # live sessions
    is_session_answer = fields.Boolean('Is in a Session', help="Is that user input part of a survey session or not.")
    question_time_limit_reached = fields.Boolean("Question Time Limit Reached", compute='_compute_question_time_limit_reached')

    _sql_constraints = [
        ('unique_token', 'UNIQUE (access_token)', 'An access token must be unique!'),
    ]

    @api.depends('user_input_line_ids.answer_score', 'user_input_line_ids.question_id', 'predefined_question_ids.answer_score')
    def _compute_scoring_values(self):
        for user_input in self:
            # sum(multi-choice question scores) + sum(simple answer_type scores)
            total_possible_score = 0
            for question in user_input.predefined_question_ids:
                if question.question_type == 'simple_choice':
                    total_possible_score += max([score for score in question.mapped('suggested_answer_ids.answer_score') if score > 0], default=0)
                elif question.question_type == 'multiple_choice':
                    total_possible_score += sum(score for score in question.mapped('suggested_answer_ids.answer_score') if score > 0)
                elif question.is_scored_question:
                    total_possible_score += question.answer_score

            if total_possible_score == 0:
                user_input.scoring_percentage = 0
                user_input.scoring_total = 0
            else:
                score_total = sum(user_input.user_input_line_ids.mapped('answer_score'))
                user_input.scoring_total = score_total
                score_percentage = (score_total / total_possible_score) * 100
                user_input.scoring_percentage = round(score_percentage, 2) if score_percentage > 0 else 0

    @api.depends('scoring_percentage', 'survey_id')
    def _compute_scoring_success(self):
        for user_input in self:
            user_input.scoring_success = user_input.scoring_percentage >= user_input.survey_id.scoring_success_min

    @api.depends(
        'start_datetime',
        'survey_id.is_time_limited',
        'survey_id.time_limit')
    def _compute_survey_time_limit_reached(self):
        """ Checks that the user_input is not exceeding the survey's time limit. """
        for user_input in self:
            if not user_input.is_session_answer and user_input.start_datetime:
                start_time = user_input.start_datetime
                time_limit = user_input.survey_id.time_limit
                user_input.survey_time_limit_reached = user_input.survey_id.is_time_limited and \
                    fields.Datetime.now() >= start_time + relativedelta(minutes=time_limit)
            else:
                user_input.survey_time_limit_reached = False

    @api.depends(
        'survey_id.session_question_id.time_limit',
        'survey_id.session_question_id.is_time_limited',
        'survey_id.session_question_start_time')
    def _compute_question_time_limit_reached(self):
        """ Checks that the user_input is not exceeding the question's time limit.
        Only used in the context of survey sessions. """
        for user_input in self:
            if user_input.is_session_answer and user_input.survey_id.session_question_start_time:
                start_time = user_input.survey_id.session_question_start_time
                time_limit = user_input.survey_id.session_question_id.time_limit
                user_input.question_time_limit_reached = user_input.survey_id.session_question_id.is_time_limited and \
                    fields.Datetime.now() >= start_time + relativedelta(seconds=time_limit)
            else:
                user_input.question_time_limit_reached = False

    @api.depends('state', 'test_entry', 'survey_id.is_attempts_limited', 'partner_id', 'email', 'invite_token')
    def _compute_attempts_info(self):
        attempts_to_compute = self.filtered(
            lambda user_input: user_input.state == 'done' and not user_input.test_entry and user_input.survey_id.is_attempts_limited
        )

        for user_input in (self - attempts_to_compute):
            user_input.attempts_count = 1
            user_input.attempts_number = 1

        if attempts_to_compute:
            self.flush_model(['email', 'invite_token', 'partner_id', 'state', 'survey_id', 'test_entry'])

            self.env.cr.execute("""
                SELECT user_input.id,
                       COUNT(all_attempts_user_input.id) AS attempts_count,
                       COUNT(CASE WHEN all_attempts_user_input.id < user_input.id THEN all_attempts_user_input.id END) + 1 AS attempts_number
                FROM survey_user_input user_input
                LEFT OUTER JOIN survey_user_input all_attempts_user_input
                ON user_input.survey_id = all_attempts_user_input.survey_id
                AND all_attempts_user_input.state = 'done'
                AND all_attempts_user_input.test_entry IS NOT TRUE
                AND (user_input.invite_token IS NULL OR user_input.invite_token = all_attempts_user_input.invite_token)
                AND (user_input.partner_id = all_attempts_user_input.partner_id OR user_input.email = all_attempts_user_input.email)
                WHERE user_input.id IN %s
                GROUP BY user_input.id;
            """, (tuple(attempts_to_compute.ids),))

            attempts_number_results = self.env.cr.dictfetchall()

            attempts_number_results = {
                attempts_number_result['id']: {
                    'attempts_number': attempts_number_result['attempts_number'],
                    'attempts_count': attempts_number_result['attempts_count'],
                }
                for attempts_number_result in attempts_number_results
            }

            for user_input in attempts_to_compute:
                attempts_number_result = attempts_number_results.get(user_input.id, {})
                user_input.attempts_number = attempts_number_result.get('attempts_number', 1)
                user_input.attempts_count = attempts_number_result.get('attempts_count', 1)

    @api.model_create_multi
    def create(self, vals_list):
        for vals in vals_list:
            if 'predefined_question_ids' not in vals:
                suvey_id = vals.get('survey_id', self.env.context.get('default_survey_id'))
                survey = self.env['survey.survey'].browse(suvey_id)
                vals['predefined_question_ids'] = [(6, 0, survey._prepare_user_input_predefined_questions().ids)]
        return super(SurveyUserInput, self).create(vals_list)

    # ------------------------------------------------------------
    # ACTIONS / BUSINESS
    # ------------------------------------------------------------

    def action_resend(self):
        partners = self.env['res.partner']
        emails = []
        for user_answer in self:
            if user_answer.partner_id:
                partners |= user_answer.partner_id
            elif user_answer.email:
                emails.append(user_answer.email)

        return self.survey_id.with_context(
            default_existing_mode='resend',
            default_partner_ids=partners.ids,
            default_emails=','.join(emails)
        ).action_send_survey()

    def action_print_answers(self):
        """ Open the website page with the survey form """
        self.ensure_one()
        return {
            'type': 'ir.actions.act_url',
            'name': "View Answers",
            'target': 'self',
            'url': '/survey/print/%s?answer_token=%s' % (self.survey_id.access_token, self.access_token)
        }

    def action_redirect_to_attempts(self):
        self.ensure_one()

        action = self.env['ir.actions.act_window']._for_xml_id('survey.action_survey_user_input')
        context = dict(self.env.context or {})

        context['create'] = False
        context['search_default_survey_id'] = self.survey_id.id
        context['search_default_group_by_survey'] = False
        if self.partner_id:
            context['search_default_partner_id'] = self.partner_id.id
        elif self.email:
            context['search_default_email'] = self.email

        action['context'] = context
        return action

    @api.model
    def _generate_invite_token(self):
        return str(uuid.uuid4())

    def _mark_in_progress(self):
        """ marks the state as 'in_progress' and updates the start_datetime accordingly. """
        self.write({
            'start_datetime': fields.Datetime.now(),
            'state': 'in_progress'
        })

    def _mark_done(self):
        """ This method will:
        1. mark the state as 'done'
        2. send the certification email with attached document if
        - The survey is a certification
        - It has a certification_mail_template_id set
        - The user succeeded the test
        3. Notify survey subtype subscribers of the newly completed input
        Will also run challenge Cron to give the certification badge if any."""
        self.write({
            'end_datetime': fields.Datetime.now(),
            'state': 'done',
        })

        Challenge_sudo = self.env['gamification.challenge'].sudo()
        badge_ids = []
        self._notify_new_participation_subscribers()
        for user_input in self:
            if user_input.survey_id.certification and user_input.scoring_success:
                if user_input.survey_id.certification_mail_template_id and not user_input.test_entry:
                    user_input.survey_id.certification_mail_template_id.send_mail(user_input.id, email_layout_xmlid="mail.mail_notification_light")
                if user_input.survey_id.certification_give_badge:
                    badge_ids.append(user_input.survey_id.certification_badge_id.id)

            # Update predefined_question_id to remove inactive questions
            user_input.predefined_question_ids -= user_input._get_inactive_conditional_questions()

        if badge_ids:
            challenges = Challenge_sudo.search([('reward_id', 'in', badge_ids)])
            if challenges:
                Challenge_sudo._cron_update(ids=challenges.ids, commit=False)

    def get_start_url(self):
        self.ensure_one()
        return '%s?answer_token=%s' % (self.survey_id.get_start_url(), self.access_token)

    def get_print_url(self):
        self.ensure_one()
        return '%s?answer_token=%s' % (self.survey_id.get_print_url(), self.access_token)

    # ------------------------------------------------------------
    # CREATE / UPDATE LINES FROM SURVEY FRONTEND INPUT
    # ------------------------------------------------------------

    def _save_lines(self, question, answer, comment=None, overwrite_existing=True):
        """ Save answers to questions, depending on question type.

        :param bool overwrite_existing: if an answer already exists for question and user_input_id
        it will be overwritten (or deleted for 'choice' questions) in order to maintain data consistency.
        :raises UserError: if line exists and overwrite_existing is False
        """
        old_answers = self.env['survey.user_input.line'].search([
            ('user_input_id', '=', self.id),
            ('question_id', '=', question.id)
        ])
        if old_answers and not overwrite_existing:
            raise UserError(_("This answer cannot be overwritten."))

        if question.question_type in ['char_box', 'text_box', 'scale', 'numerical_box', 'date', 'datetime']:
            self._save_line_simple_answer(question, old_answers, answer)
            if question.save_as_email and answer:
                self.write({'email': answer})
            if question.save_as_nickname and answer:
                self.write({'nickname': answer})

        elif question.question_type in ['simple_choice', 'multiple_choice']:
            self._save_line_choice(question, old_answers, answer, comment)
        elif question.question_type == 'matrix':
            self._save_line_matrix(question, old_answers, answer, comment)
        else:
            raise AttributeError(question.question_type + ": This type of question has no saving function")

    def _save_line_simple_answer(self, question, old_answers, answer):
        vals = self._get_line_answer_values(question, answer, question.question_type)
        if old_answers:
            old_answers.write(vals)
            return old_answers
        else:
            return self.env['survey.user_input.line'].create(vals)

    def _save_line_choice(self, question, old_answers, answers, comment):
        if not (isinstance(answers, list)):
            answers = [answers]

        if not answers:
            # add a False answer to force saving a skipped line
            # this will make this question correctly considered as skipped in statistics
            answers = [False]

        vals_list = []

        if question.question_type == 'simple_choice':
            if not question.comment_count_as_answer or not question.comments_allowed or not comment:
                vals_list = [self._get_line_answer_values(question, answer, 'suggestion') for answer in answers]
        elif question.question_type == 'multiple_choice':
            vals_list = [self._get_line_answer_values(question, answer, 'suggestion') for answer in answers]

        if comment:
            vals_list.append(self._get_line_comment_values(question, comment))

        old_answers.sudo().unlink()
        return self.env['survey.user_input.line'].create(vals_list)

    def _save_line_matrix(self, question, old_answers, answers, comment):
        vals_list = []

        if not answers and question.matrix_row_ids:
            # add a False answer to force saving a skipped line
            # this will make this question correctly considered as skipped in statistics
            answers = {question.matrix_row_ids[0].id: [False]}

        if answers:
            for row_key, row_answer in answers.items():
                for answer in row_answer:
                    vals = self._get_line_answer_values(question, answer, 'suggestion')
                    vals['matrix_row_id'] = int(row_key)
                    vals_list.append(vals.copy())

        if comment:
            vals_list.append(self._get_line_comment_values(question, comment))

        old_answers.sudo().unlink()
        return self.env['survey.user_input.line'].create(vals_list)

    def _get_line_answer_values(self, question, answer, answer_type):
        vals = {
            'user_input_id': self.id,
            'question_id': question.id,
            'skipped': False,
            'answer_type': answer_type,
        }
        if not answer or (isinstance(answer, str) and not answer.strip()):
            vals.update(answer_type=None, skipped=True)
            return vals

        if answer_type == 'suggestion':
            vals['suggested_answer_id'] = int(answer)
        elif answer_type == 'numerical_box':
            vals['value_numerical_box'] = float(answer)
        elif answer_type == 'scale':
            vals['value_scale'] = int(answer)
        else:
            vals['value_%s' % answer_type] = answer
        return vals

    def _get_line_comment_values(self, question, comment):
        return {
            'user_input_id': self.id,
            'question_id': question.id,
            'skipped': False,
            'answer_type': 'char_box',
            'value_char_box': comment,
        }

    # ------------------------------------------------------------
    # STATISTICS / RESULTS
    # ------------------------------------------------------------

    def _prepare_statistics(self):
        """ Prepares survey.user_input's statistics to display various charts on the frontend.
        Returns a structure containing answers statistics "by section" and "totals" for every input in self.

        e.g returned structure:
        {
            survey.user_input(1,): {
                'by_section': {
                    'Uncategorized': {
                        'question_count': 2,
                        'correct': 2,
                        'partial': 0,
                        'incorrect': 0,
                        'skipped': 0,
                    },
                    'Mathematics': {
                        'question_count': 3,
                        'correct': 1,
                        'partial': 1,
                        'incorrect': 0,
                        'skipped': 1,
                    },
                    'Geography': {
                        'question_count': 4,
                        'correct': 2,
                        'partial': 0,
                        'incorrect': 2,
                        'skipped': 0,
                    }
                },
                'totals' [{
                    'text': 'Correct',
                    'count': 5,
                }, {
                    'text': 'Partially',
                    'count': 1,
                }, {
                    'text': 'Incorrect',
                    'count': 2,
                }, {
                    'text': 'Unanswered',
                    'count': 1,
                }]
            }
        }"""
        res = dict((user_input, {
            'by_section': {}
        }) for user_input in self)

        scored_questions = self.mapped('predefined_question_ids').filtered(lambda question: question.is_scored_question)

        for question in scored_questions:
            if question.question_type == 'simple_choice':
                question_incorrect_scored_answers = question.suggested_answer_ids.filtered(lambda answer: not answer.is_correct and answer.answer_score > 0)

            if question.question_type in ['simple_choice', 'multiple_choice']:
                question_correct_suggested_answers = question.suggested_answer_ids.filtered(lambda answer: answer.is_correct)

            question_section = question.page_id.title or _('Uncategorized')
            for user_input in self:
                user_input_lines = user_input.user_input_line_ids.filtered(lambda line:
                    line.question_id == question and (line.answer_type != 'char_box' or question.comment_count_as_answer))
                if question.question_type == 'simple_choice':
                    answer_result_key = self._simple_choice_question_answer_result(user_input_lines, question_correct_suggested_answers, question_incorrect_scored_answers)
                elif question.question_type == 'multiple_choice':
                    answer_result_key = self._multiple_choice_question_answer_result(user_input_lines, question_correct_suggested_answers)
                else:
                    answer_result_key = self._simple_question_answer_result(user_input_lines)

                if question_section not in res[user_input]['by_section']:
                    res[user_input]['by_section'][question_section] = {
                        'question_count': 0,
                        'correct': 0,
                        'partial': 0,
                        'incorrect': 0,
                        'skipped': 0,
                    }

                res[user_input]['by_section'][question_section]['question_count'] += 1
                res[user_input]['by_section'][question_section][answer_result_key] += 1

        for user_input in self:
            correct_count = 0
            partial_count = 0
            incorrect_count = 0
            skipped_count = 0

            for section_counts in res[user_input]['by_section'].values():
                correct_count += section_counts.get('correct', 0)
                partial_count += section_counts.get('partial', 0)
                incorrect_count += section_counts.get('incorrect', 0)
                skipped_count += section_counts.get('skipped', 0)

            res[user_input]['totals'] = [
                {'text': _("Correct"), 'count': correct_count},
                {'text': _("Partially"), 'count': partial_count},
                {'text': _("Incorrect"), 'count': incorrect_count},
                {'text': _("Unanswered"), 'count': skipped_count}
            ]

        return res

    def _multiple_choice_question_answer_result(self, user_input_lines, question_correct_suggested_answers):
        correct_user_input_lines = user_input_lines.filtered(lambda line: line.answer_is_correct and not line.skipped).mapped('suggested_answer_id')
        incorrect_user_input_lines = user_input_lines.filtered(lambda line: not line.answer_is_correct and not line.skipped)
        if question_correct_suggested_answers and correct_user_input_lines == question_correct_suggested_answers:
            return 'correct'
        elif correct_user_input_lines and correct_user_input_lines < question_correct_suggested_answers:
            return 'partial'
        elif not correct_user_input_lines and incorrect_user_input_lines:
            return 'incorrect'
        else:
            return 'skipped'

    def _simple_choice_question_answer_result(self, user_input_line, question_correct_suggested_answers, question_incorrect_scored_answers):
        user_answer = user_input_line.suggested_answer_id if not user_input_line.skipped else self.env['survey.question.answer']
        if user_answer in question_correct_suggested_answers:
            return 'correct'
        elif user_answer in question_incorrect_scored_answers:
            return 'partial'
        elif user_answer:
            return 'incorrect'
        else:
            return 'skipped'

    def _simple_question_answer_result(self, user_input_line):
        if user_input_line.skipped:
            return 'skipped'
        elif user_input_line.answer_is_correct:
            return 'correct'
        else:
            return 'incorrect'

    # ------------------------------------------------------------
    # Conditional Questions Management
    # ------------------------------------------------------------

    def _get_conditional_values(self):
        """ For survey containing conditional questions, we need a triggered_questions_by_answer map that contains
                {key: answer, value: the question that the answer triggers, if selected},
         The idea is to be able to verify, on every answer check, if this answer is triggering the display
         of another question.
         If answer is not in the conditional map:
            - nothing happens.
         If the answer is in the conditional map:
            - If we are in ONE PAGE survey : (handled at CLIENT side)
                -> display immediately the depending question
            - If we are in PAGE PER SECTION : (handled at CLIENT side)
                - If related question is on the same page :
                    -> display immediately the depending question
                - If the related question is not on the same page :
                    -> keep the answers in memory and check at next page load if the depending question is in there and
                       display it, if so.
            - If we are in PAGE PER QUESTION : (handled at SERVER side)
                -> During submit, determine which is the next question to display getting the next question
                   that is the next in sequence and that is either not triggered by another question's answer, or that
                   is triggered by an already selected answer.
         To do all this, we need to return:
            - triggering_answers_by_question: dict -> for a given question, the answers that triggers it
                Used mainly to ease template rendering
            - triggered_questions_by_answer: dict -> for a given answer, list of questions triggered by this answer;
                Used mainly for dynamic show/hide behaviour at client side
            - list of all selected answers: [answer_id1, answer_id2, ...] (for survey reloading, otherwise, this list is
              updated at client side)
        """
        triggering_answers_by_question = {}
        triggered_questions_by_answer = {}
        # Ignore conditional configuration if randomised questions selection
        if self.survey_id.questions_selection != 'random':
            triggering_answers_by_question, triggered_questions_by_answer = self.survey_id._get_conditional_maps()
        selected_answers = self._get_selected_suggested_answers()

        return triggering_answers_by_question, triggered_questions_by_answer, selected_answers

    def _get_selected_suggested_answers(self):
        """
        For now, only simple and multiple choices question type are handled by the conditional questions feature.
        Mapping all the suggested answers selected by the user will also include answers from matrix question type,
        Those ones won't be used.
        Maybe someday, conditional questions feature will be extended to work with matrix question.
        :return: all the suggested answer selected by the user.
        """
        return self.mapped('user_input_line_ids.suggested_answer_id')

    def _clear_inactive_conditional_answers(self):
        """
        Clean eventual answers on conditional questions that should not have been displayed to user.
        This method is used mainly for page per question survey, a similar method does the same treatment
        at client side for the other survey layouts.
        E.g.: if depending answer was uncheck after answering conditional question, we need to clear answers
              of that conditional question, for two reasons:
              - ensure correct scoring
              - if the selected answer triggers another question later in the survey, if the answer is not cleared,
                a question that should not be displayed to the user will be.

        TODO DBE: Maybe this can be the only cleaning method, even for section_per_page or one_page where
        conditional questions are, for now, cleared in JS directly. But this can be annoying if user typed a long
        answer, changed their mind unchecking depending answer and changed again their mind by rechecking the depending
        answer -> For now, the long answer will be lost. If we use this as the master cleaning method,
        long answer will be cleared only during submit.
        """
        inactive_questions = self._get_inactive_conditional_questions()

        # delete user.input.line on question that should not be answered.
        answers_to_delete = self.user_input_line_ids.filtered(lambda answer: answer.question_id in inactive_questions)
        answers_to_delete.unlink()

    def _get_inactive_conditional_questions(self):
        triggering_answers_by_question, _, selected_answers = self._get_conditional_values()

        # get questions that should not be answered
        inactive_questions = self.env['survey.question']
        for question, triggering_answers in triggering_answers_by_question.items():
            if triggering_answers and not triggering_answers & selected_answers:
                inactive_questions |= question
        return inactive_questions

    def _get_print_questions(self):
        """ Get the questions to display : the ones that should have been answered = active questions
            In case of session, active questions are based on most voted answers
        :return: active survey.question browse records
        """
        survey = self.survey_id
        if self.is_session_answer:
            most_voted_answers = survey._get_session_most_voted_answers()
            inactive_questions = most_voted_answers._get_inactive_conditional_questions()
        else:
            inactive_questions = self._get_inactive_conditional_questions()
        return survey.question_ids - inactive_questions

    def _get_next_skipped_page_or_question(self):
        """Get next skipped question or page in case the option 'can_go_back' is set on the survey
        It loops to the first skipped question or page if 'last_displayed_page_id' is the last
        skipped question or page."""
        self.ensure_one()
        skipped_mandatory_answer_ids = self.user_input_line_ids.filtered(
            lambda answer: answer.skipped and answer.question_id.constr_mandatory)

        if not skipped_mandatory_answer_ids:
            return self.env['survey.question']

        page_or_question_key = 'page_id' if self.survey_id.questions_layout == 'page_per_section' else 'question_id'
        page_or_question_ids = skipped_mandatory_answer_ids.mapped(page_or_question_key).sorted()

        if self.last_displayed_page_id not in page_or_question_ids\
            or self.last_displayed_page_id == page_or_question_ids[-1]:
            return page_or_question_ids[0]

        current_page_index = page_or_question_ids.ids.index(self.last_displayed_page_id.id)
        return page_or_question_ids[current_page_index + 1]

    def _get_skipped_questions(self):
        self.ensure_one()

        return self.user_input_line_ids.filtered(
            lambda answer: answer.skipped and answer.question_id.constr_mandatory).question_id

    def _is_last_skipped_page_or_question(self, page_or_question):
        """In case of a submitted survey tells if the question or page is the last
        skipped page or question.

        This is used to :

        - Display a Submit button if the actual question is the last skipped question.
        - Avoid displaying a Submit button on the last survey question if there are
          still skipped questions before.
        - Avoid displaying the next page if submitting the latest skipped question.

        :param page_or_question: page if survey's layout is page_per_section, question if page_per_question.
        """
        if self.survey_id.questions_layout == 'one_page':
            return True
        skipped = self._get_skipped_questions()
        if not skipped:
            return True
        if self.survey_id.questions_layout == 'page_per_section':
            skipped = skipped.page_id
        return skipped == page_or_question

    # ------------------------------------------------------------
    # MESSAGING
    # ------------------------------------------------------------

    def _message_get_suggested_recipients(self):
        recipients = super()._message_get_suggested_recipients()
        if self.partner_id:
            self._message_add_suggested_recipient(
                recipients,
                partner=self.partner_id,
                reason=_('Survey Participant')
            )
        return recipients

    def _notify_new_participation_subscribers(self):
        subtype_id = self.env.ref('survey.mt_survey_survey_user_input_completed', raise_if_not_found=False)
        if not self.ids or not subtype_id:
            return
        author_id = self.env.ref('base.partner_root').id if self.env.user.is_public else self.env.user.partner_id.id
        # Only post if there are any followers
        recipients_data = self.env['mail.followers']._get_recipient_data(self.survey_id, 'notification', subtype_id.id)
        followed_survey_ids = [survey_id for survey_id, followers in recipients_data.items() if followers]
        for user_input in self.filtered(lambda user_input_: user_input_.survey_id.id in followed_survey_ids):
            survey_title = user_input.survey_id.title
            if user_input.partner_id:
                body = _(
                    '%(participant)s just participated in "%(survey_title)s".',
                    participant=user_input.partner_id.display_name,
                    survey_title=survey_title,
                )
            else:
                body = _('Someone just participated in "%(survey_title)s".', survey_title=survey_title)

            user_input.message_post(author_id=author_id, body=body, subtype_xmlid='survey.mt_survey_user_input_completed')


class SurveyUserInputLine(models.Model):
    _name = 'survey.user_input.line'
    _description = 'Survey User Input Line'
    _rec_name = 'user_input_id'
    _order = 'question_sequence, id'

    # survey data
    user_input_id = fields.Many2one('survey.user_input', string='User Input', ondelete='cascade', required=True, index=True)
    survey_id = fields.Many2one(related='user_input_id.survey_id', string='Survey', store=True, readonly=False)
    question_id = fields.Many2one('survey.question', string='Question', ondelete='cascade', required=True, index=True)
    page_id = fields.Many2one(related='question_id.page_id', string="Section", readonly=False)
    question_sequence = fields.Integer('Sequence', related='question_id.sequence', store=True)
    # answer
    skipped = fields.Boolean('Skipped')
    answer_type = fields.Selection([
        ('text_box', 'Free Text'),
        ('char_box', 'Text'),
        ('numerical_box', 'Number'),
        ('scale', 'Number'),
        ('date', 'Date'),
        ('datetime', 'Datetime'),
        ('suggestion', 'Suggestion')], string='Answer Type')
    value_char_box = fields.Char('Text answer')
    value_numerical_box = fields.Float('Numerical answer')
    value_scale = fields.Integer('Scale value')
    value_date = fields.Date('Date answer')
    value_datetime = fields.Datetime('Datetime answer')
    value_text_box = fields.Text('Free Text answer')
    suggested_answer_id = fields.Many2one('survey.question.answer', string="Suggested answer")
    matrix_row_id = fields.Many2one('survey.question.answer', string="Row answer")
    # scoring
    answer_score = fields.Float('Score')
    answer_is_correct = fields.Boolean('Correct')

    @api.depends(
        'answer_type', 'value_text_box', 'value_numerical_box',
        'value_char_box', 'value_date', 'value_datetime',
        'suggested_answer_id.value', 'matrix_row_id.value',
    )
    def _compute_display_name(self):
        for line in self:
            if line.answer_type == 'char_box':
                line.display_name = line.value_char_box
            elif line.answer_type == 'text_box' and line.value_text_box:
                line.display_name = textwrap.shorten(line.value_text_box, width=50, placeholder=" [...]")
            elif line.answer_type == 'numerical_box':
                line.display_name = line.value_numerical_box
            elif line.answer_type == 'date':
                line.display_name = fields.Date.to_string(line.value_date)
            elif line.answer_type == 'datetime':
                line.display_name = fields.Datetime.to_string(line.value_datetime)
            elif line.answer_type == 'scale':
                line.display_name = line.value_scale
            elif line.answer_type == 'suggestion':
                if line.matrix_row_id:
                    line.display_name = f'{line.suggested_answer_id.value}: {line.matrix_row_id.value}'
                else:
                    line.display_name = line.suggested_answer_id.value

            if not line.display_name:
                line.display_name = _('Skipped')

    @api.constrains('skipped', 'answer_type')
    def _check_answer_type_skipped(self):
        for line in self:
            if (line.skipped == bool(line.answer_type)):
                raise ValidationError(_('A question can either be skipped or answered, not both.'))

            # allow 0 for numerical box and scale
            if line.answer_type == 'numerical_box' and float_is_zero(line['value_numerical_box'], precision_digits=6):
                continue
            if line.answer_type == 'scale' and line['value_scale'] == 0:
                continue

            if line.answer_type == 'suggestion':
                field_name = 'suggested_answer_id'
            elif line.answer_type:
                field_name = 'value_%s' % line.answer_type
            else:  # skipped
                field_name = False

            if field_name and not line[field_name]:
                raise ValidationError(_('The answer must be in the right type'))

    @api.model_create_multi
    def create(self, vals_list):
        for vals in vals_list:
            if not vals.get('answer_score'):
                score_vals = self._get_answer_score_values(vals)
                vals.update(score_vals)
        return super(SurveyUserInputLine, self).create(vals_list)

    def write(self, vals):
        res = True
        for line in self:
            vals_copy = {**vals}
            getter_params = {
                'user_input_id': line.user_input_id.id,
                'answer_type': line.answer_type,
                'question_id': line.question_id.id,
                **vals_copy
            }
            if not vals_copy.get('answer_score'):
                score_vals = self._get_answer_score_values(getter_params, compute_speed_score=False)
                vals_copy.update(score_vals)
            res = super(SurveyUserInputLine, line).write(vals_copy) and res
        return res

    def _get_answer_matching_domain(self):
        self.ensure_one()
        if self.answer_type in ('char_box', 'text_box', 'numerical_box', 'scale', 'date', 'datetime'):
            value_field = {
                'char_box': 'value_char_box',
                'text_box': 'value_text_box',
                'numerical_box': 'value_numerical_box',
                'scale': 'value_scale',
                'date': 'value_date',
                'datetime': 'value_datetime',
            }
            operators = {
                'char_box': 'ilike',
                'text_box': 'ilike',
                'numerical_box': '=',
                'scale': '=',
                'date': '=',
                'datetime': '=',
            }
            return ['&', ('question_id', '=', self.question_id.id), (value_field[self.answer_type], operators[self.answer_type], self._get_answer_value())]
        elif self.answer_type == 'suggestion':
            return self.suggested_answer_id._get_answer_matching_domain(self.matrix_row_id.id if self.matrix_row_id else False)

    @api.model
    def _get_answer_score_values(self, vals, compute_speed_score=True):
        """ Get values for: answer_is_correct and associated answer_score.

        Requires vals to contain 'answer_type', 'question_id', and 'user_input_id'.
        Depending on 'answer_type' additional value of 'suggested_answer_id' may also be
        required.

        Calculates whether an answer_is_correct and its score based on 'answer_type' and
        corresponding question. Handles choice (answer_type == 'suggestion') questions
        separately from other question types. Each selected choice answer is handled as an
        individual answer.

        If score depends on the speed of the answer, it is adjusted as follows:
         - If the user answers in less than 2 seconds, they receive 100% of the possible points.
         - If user answers after that, they receive 50% of the possible points + the remaining
            50% scaled by the time limit and time taken to answer [i.e. a minimum of 50% of the
            possible points is given to all correct answers]

        Example of returned values:
            * {'answer_is_correct': False, 'answer_score': 0} (default)
            * {'answer_is_correct': True, 'answer_score': 2.0}
        """
        user_input_id = vals.get('user_input_id')
        answer_type = vals.get('answer_type')
        question_id = vals.get('question_id')
        if not question_id:
            raise ValueError(_('Computing score requires a question in arguments.'))
        question = self.env['survey.question'].browse(int(question_id))

        # default and non-scored questions
        answer_is_correct = False
        answer_score = 0

        # record selected suggested choice answer_score (can be: pos, neg, or 0)
        if question.question_type in ['simple_choice', 'multiple_choice']:
            if answer_type == 'suggestion':
                suggested_answer_id = vals.get('suggested_answer_id')
                if suggested_answer_id:
                    question_answer = self.env['survey.question.answer'].browse(int(suggested_answer_id))
                    answer_score = question_answer.answer_score
                    answer_is_correct = question_answer.is_correct
        # for all other scored question cases, record question answer_score (can be: pos or 0)
        elif question.question_type in ['date', 'datetime', 'numerical_box']:
            answer = vals.get('value_%s' % answer_type)
            if answer_type == 'numerical_box':
                answer = float(answer)
            elif answer_type == 'date':
                answer = fields.Date.from_string(answer)
            elif answer_type == 'datetime':
                answer = fields.Datetime.from_string(answer)
            if answer and answer == question['answer_%s' % answer_type]:
                answer_is_correct = True
                answer_score = question.answer_score

        if compute_speed_score and answer_score > 0:
            user_input = self.env['survey.user_input'].browse(user_input_id)
            session_speed_rating = user_input.exists() and user_input.is_session_answer and user_input.survey_id.session_speed_rating
            if session_speed_rating and question.is_time_limited:
                max_score_delay = 2
                time_limit = question.time_limit
                now = fields.Datetime.now()
                seconds_to_answer = (now - user_input.survey_id.session_question_start_time).total_seconds()
                question_remaining_time = time_limit - seconds_to_answer
                # if answered within the max_score_delay => leave score as is
                if question_remaining_time < 0:  # if no time left
                    answer_score /= 2
                elif seconds_to_answer > max_score_delay:  # linear decrease in score after 2 sec
                    score_proportion = (time_limit - seconds_to_answer) / (time_limit - max_score_delay)
                    answer_score = (answer_score / 2) * (1 + score_proportion)

        return {
            'answer_is_correct': answer_is_correct,
            'answer_score': answer_score
        }

    def _get_answer_value(self):
        self.ensure_one()
        if self.answer_type == 'char_box':
            return self.value_char_box
        elif self.answer_type == 'text_box':
            return self.value_text_box
        elif self.answer_type == 'numerical_box':
            return self.value_numerical_box
        elif self.answer_type == 'scale':
            return self.value_scale
        elif self.answer_type == 'date':
            return self.value_date
        elif self.answer_type == 'datetime':
            return self.value_datetime
        elif self.answer_type == 'suggestion':
            return self.suggested_answer_id.value