views.py

#
# Licensed to the Apache Software Foundation (ASF) under one
# or more contributor license agreements.  See the NOTICE file
# distributed with this work for additional information
# regarding copyright ownership.  The ASF licenses this file
# to you under the Apache License, Version 2.0 (the
# "License"); you may not use this file except in compliance
# with the License.  You may obtain a copy of the License at
#
#   http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing,
# software distributed under the License is distributed on an
# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
# KIND, either express or implied.  See the License for the
# specific language governing permissions and limitations
# under the License.
#
import collections
import copy
import itertools
import json
import logging
import math
import re
import sys
import traceback
import warnings
from bisect import insort_left
from collections import defaultdict
from datetime import datetime, timedelta
from functools import wraps
from json import JSONDecodeError
from operator import itemgetter
from typing import Any, Callable, Dict, List, Optional, Set, Tuple, Union
from urllib.parse import parse_qsl, unquote, urlencode, urlparse

import lazy_object_proxy
import markupsafe
import nvd3
import sqlalchemy as sqla
from croniter import croniter
from flask import (
    Response,
    abort,
    before_render_template,
    flash,
    g,
    jsonify,
    make_response,
    redirect,
    render_template,
    request,
    send_from_directory,
    session as flask_session,
    url_for,
)
from flask_appbuilder import BaseView, ModelView, expose
from flask_appbuilder.actions import action
from flask_appbuilder.fieldwidgets import Select2Widget
from flask_appbuilder.models.sqla.filters import BaseFilter
from flask_appbuilder.security.decorators import has_access
from flask_appbuilder.security.views import (
    PermissionModelView,
    PermissionViewModelView,
    ResetMyPasswordView,
    ResetPasswordView,
    RoleModelView,
    UserDBModelView,
    UserInfoEditView,
    UserLDAPModelView,
    UserOAuthModelView,
    UserOIDModelView,
    UserRemoteUserModelView,
    UserStatsChartView,
    ViewMenuModelView,
)
from flask_appbuilder.urltools import get_order_args, get_page_args, get_page_size_args
from flask_appbuilder.widgets import FormWidget
from flask_babel import lazy_gettext
from jinja2.utils import htmlsafe_json_dumps, pformat  # type: ignore
from markupsafe import Markup, escape
from pendulum.datetime import DateTime
from pendulum.parsing.exceptions import ParserError
from pygments import highlight, lexers
from pygments.formatters import HtmlFormatter
from sqlalchemy import Date, and_, desc, func, inspect, union_all
from sqlalchemy.exc import IntegrityError
from sqlalchemy.orm import Session, joinedload
from wtforms import SelectField, validators
from wtforms.validators import InputRequired

import airflow
from airflow import models, plugins_manager, settings
from airflow.api.common.mark_tasks import (
    set_dag_run_state_to_failed,
    set_dag_run_state_to_queued,
    set_dag_run_state_to_success,
    set_state,
)
from airflow.compat.functools import cached_property
from airflow.configuration import AIRFLOW_CONFIG, conf
from airflow.exceptions import AirflowException, ParamValidationError
from airflow.executors.executor_loader import ExecutorLoader
from airflow.jobs.base_job import BaseJob
from airflow.jobs.scheduler_job import SchedulerJob
from airflow.jobs.triggerer_job import TriggererJob
from airflow.models import DAG, Connection, DagModel, DagTag, Log, SlaMiss, TaskFail, XCom, errors
from airflow.models.abstractoperator import AbstractOperator
from airflow.models.dagcode import DagCode
from airflow.models.dagrun import DagRun, DagRunType
from airflow.models.operator import Operator
from airflow.models.serialized_dag import SerializedDagModel
from airflow.models.taskinstance import TaskInstance
from airflow.providers_manager import ProvidersManager
from airflow.security import permissions
from airflow.ti_deps.dep_context import DepContext
from airflow.ti_deps.dependencies_deps import RUNNING_DEPS, SCHEDULER_QUEUED_DEPS
from airflow.timetables.base import DataInterval, TimeRestriction
from airflow.timetables.interval import CronDataIntervalTimetable
from airflow.utils import json as utils_json, timezone, yaml
from airflow.utils.airflow_flask_app import get_airflow_app
from airflow.utils.dates import infer_time_unit, scale_time_units
from airflow.utils.docs import get_doc_url_for_provider, get_docs_url
from airflow.utils.helpers import alchemy_to_dict
from airflow.utils.log import secrets_masker
from airflow.utils.log.log_reader import TaskLogReader
from airflow.utils.net import get_hostname
from airflow.utils.session import NEW_SESSION, create_session, provide_session
from airflow.utils.state import State, TaskInstanceState
from airflow.utils.strings import to_boolean
from airflow.utils.timezone import td_format, utcnow
from airflow.version import version
from airflow.www import auth, utils as wwwutils
from airflow.www.decorators import action_logging, gzipped
from airflow.www.forms import (
    ConnectionForm,
    DagRunEditForm,
    DateTimeForm,
    DateTimeWithNumRunsForm,
    DateTimeWithNumRunsWithDagRunsForm,
    TaskInstanceEditForm,
)
from airflow.www.widgets import AirflowModelListWidget, AirflowVariableShowWidget

PAGE_SIZE = conf.getint('webserver', 'page_size')
FILTER_TAGS_COOKIE = 'tags_filter'
FILTER_STATUS_COOKIE = 'dag_status_filter'
LINECHART_X_AXIS_TICKFORMAT = (
    "function (d, i) { let xLabel;"
    "if (i === undefined) {xLabel = d3.time.format('%H:%M, %d %b %Y')(new Date(parseInt(d)));"
    "} else {xLabel = d3.time.format('%H:%M, %d %b')(new Date(parseInt(d)));} return xLabel;}"
)


def truncate_task_duration(task_duration):
    """
    Cast the task_duration to an int was for optimization for large/huge dags if task_duration > 10s
    otherwise we keep it as a float with 3dp
    """
    return int(task_duration) if task_duration > 10.0 else round(task_duration, 3)


def get_safe_url(url):
    """Given a user-supplied URL, ensure it points to our web server"""
    valid_schemes = ['http', 'https', '']
    valid_netlocs = [request.host, '']

    if not url:
        return url_for('Airflow.index')

    parsed = urlparse(url)

    # If the url contains semicolon, redirect it to homepage to avoid
    # potential XSS. (Similar to https://github.com/python/cpython/pull/24297/files (bpo-42967))
    if ';' in unquote(url):
        return url_for('Airflow.index')

    query = parse_qsl(parsed.query, keep_blank_values=True)

    url = parsed._replace(query=urlencode(query)).geturl()

    if parsed.scheme in valid_schemes and parsed.netloc in valid_netlocs:
        return url

    return url_for('Airflow.index')


def get_date_time_num_runs_dag_runs_form_data(www_request, session, dag):
    """Get Execution Data, Base Date & Number of runs from a Request"""
    date_time = www_request.args.get('execution_date')
    if date_time:
        date_time = _safe_parse_datetime(date_time)
    else:
        date_time = dag.get_latest_execution_date(session=session) or timezone.utcnow()

    base_date = www_request.args.get('base_date')
    if base_date:
        base_date = _safe_parse_datetime(base_date)
    else:
        # The DateTimeField widget truncates milliseconds and would loose
        # the first dag run. Round to next second.
        base_date = (date_time + timedelta(seconds=1)).replace(microsecond=0)

    default_dag_run = conf.getint('webserver', 'default_dag_run_display_number')
    num_runs = www_request.args.get('num_runs', default=default_dag_run, type=int)

    # When base_date has been rounded up because of the DateTimeField widget, we want
    # to use the execution_date as the starting point for our query just to ensure a
    # link targeting a specific dag run actually loads that dag run.  If there are
    # more than num_runs dag runs in the "rounded period" then those dagruns would get
    # loaded and the actual requested run would be excluded by the limit().  Once
    # the user has changed base date to be anything else we want to use that instead.
    query_date = base_date
    if date_time < base_date and date_time + timedelta(seconds=1) >= base_date:
        query_date = date_time

    drs = (
        session.query(DagRun)
        .filter(DagRun.dag_id == dag.dag_id, DagRun.execution_date <= query_date)
        .order_by(desc(DagRun.execution_date))
        .limit(num_runs)
        .all()
    )
    dr_choices = []
    dr_state = None
    for dr in drs:
        dr_choices.append((dr.execution_date.isoformat(), dr.run_id))
        if date_time == dr.execution_date:
            dr_state = dr.state

    # Happens if base_date was changed and the selected dag run is not in result
    if not dr_state and drs:
        dr = drs[0]
        date_time = dr.execution_date
        dr_state = dr.state

    return {
        'dttm': date_time,
        'base_date': base_date,
        'num_runs': num_runs,
        'execution_date': date_time.isoformat(),
        'dr_choices': dr_choices,
        'dr_state': dr_state,
    }


def _safe_parse_datetime(v):
    """Parse datetime and return error message for invalid dates"""
    try:
        return timezone.parse(v)
    except (TypeError, ParserError):
        abort(400, f"Invalid datetime: {v!r}")


def dag_to_grid(dag, dag_runs, session):
    """
    Create a nested dict representation of the DAG's TaskGroup and its children
    used to construct the Graph and Grid views.
    """
    query = (
        session.query(
            TaskInstance.task_id,
            TaskInstance.run_id,
            TaskInstance.state,
            sqla.func.count(sqla.func.coalesce(TaskInstance.state, sqla.literal('no_status'))).label(
                'state_count'
            ),
            sqla.func.min(TaskInstance.start_date).label('start_date'),
            sqla.func.max(TaskInstance.end_date).label('end_date'),
            sqla.func.max(TaskInstance._try_number).label('_try_number'),
        )
        .filter(
            TaskInstance.dag_id == dag.dag_id,
            TaskInstance.run_id.in_([dag_run.run_id for dag_run in dag_runs]),
        )
        .group_by(TaskInstance.task_id, TaskInstance.run_id, TaskInstance.state)
        .order_by(TaskInstance.task_id, TaskInstance.run_id)
    )

    grouped_tis = {task_id: list(tis) for task_id, tis in itertools.groupby(query, key=lambda ti: ti.task_id)}

    def task_group_to_grid(item, dag_runs, grouped_tis):
        if isinstance(item, AbstractOperator):

            def _get_summary(task_instance):
                try_count = (
                    task_instance._try_number
                    if task_instance._try_number != 0 or task_instance.state in State.running
                    else task_instance._try_number + 1
                )

                return {
                    'task_id': task_instance.task_id,
                    'run_id': task_instance.run_id,
                    'state': task_instance.state,
                    'start_date': task_instance.start_date,
                    'end_date': task_instance.end_date,
                    'try_number': try_count,
                }

            def _mapped_summary(ti_summaries):
                run_id = None
                record = None

                def set_overall_state(record):
                    for state in wwwutils.priority:
                        if state in record['mapped_states']:
                            record['state'] = state
                            break
                    if None in record['mapped_states']:
                        # When turnong the dict into JSON we can't have None as a key, so use the string that
                        # the UI does
                        record['mapped_states']['no_status'] = record['mapped_states'].pop(None)

                for ti_summary in ti_summaries:
                    if ti_summary.state is None:
                        ti_summary.state == 'no_status'
                    if run_id != ti_summary.run_id:
                        run_id = ti_summary.run_id
                        if record:
                            set_overall_state(record)
                            yield record
                        record = {
                            'task_id': ti_summary.task_id,
                            'run_id': run_id,
                            'start_date': ti_summary.start_date,
                            'end_date': ti_summary.end_date,
                            'mapped_states': {ti_summary.state: ti_summary.state_count},
                            'state': None,  # We change this before yielding
                        }
                        continue
                    record['start_date'] = min(
                        filter(None, [record['start_date'], ti_summary.start_date]), default=None
                    )
                    record['end_date'] = max(
                        filter(None, [record['end_date'], ti_summary.end_date]), default=None
                    )
                    record['mapped_states'][ti_summary.state] = ti_summary.state_count
                if record:
                    set_overall_state(record)
                    yield record

            if item.is_mapped:
                instances = list(_mapped_summary(grouped_tis.get(item.task_id, [])))
            else:
                instances = list(map(_get_summary, grouped_tis.get(item.task_id, [])))

            return {
                'id': item.task_id,
                'instances': instances,
                'label': item.label,
                'extra_links': item.extra_links,
                'is_mapped': item.is_mapped,
            }

        # Task Group
        task_group = item

        children = [
            task_group_to_grid(child, dag_runs, grouped_tis) for child in task_group.topological_sort()
        ]

        def get_summary(dag_run, children):
            child_instances = [child['instances'] for child in children if 'instances' in child]
            child_instances = [
                item for sublist in child_instances for item in sublist if item['run_id'] == dag_run.run_id
            ]

            children_start_dates = (item['start_date'] for item in child_instances if item)
            children_end_dates = (item['end_date'] for item in child_instances if item)
            children_states = {item['state'] for item in child_instances if item}

            group_state = None
            for state in wwwutils.priority:
                if state in children_states:
                    group_state = state
                    break
            group_start_date = min(filter(None, children_start_dates), default=None)
            group_end_date = max(filter(None, children_end_dates), default=None)

            return {
                'task_id': task_group.group_id,
                'run_id': dag_run.run_id,
                'state': group_state,
                'start_date': group_start_date,
                'end_date': group_end_date,
            }

        # We don't need to calculate summaries for the root
        if task_group.group_id is None:
            return {
                'id': task_group.group_id,
                'label': task_group.label,
                'children': children,
                'instances': [],
            }

        group_summaries = [get_summary(dr, children) for dr in dag_runs]

        return {
            'id': task_group.group_id,
            'label': task_group.label,
            'children': children,
            'tooltip': task_group.tooltip,
            'instances': group_summaries,
        }

    return task_group_to_grid(dag.task_group, dag_runs, grouped_tis)


def task_group_to_dict(task_item_or_group):
    """
    Create a nested dict representation of this TaskGroup and its children used to construct
    the Graph.
    """
    if isinstance(task_item_or_group, AbstractOperator):
        return {
            'id': task_item_or_group.task_id,
            'value': {
                'label': task_item_or_group.label,
                'labelStyle': f"fill:{task_item_or_group.ui_fgcolor};",
                'style': f"fill:{task_item_or_group.ui_color};",
                'rx': 5,
                'ry': 5,
            },
        }
    task_group = task_item_or_group
    children = [
        task_group_to_dict(child) for child in sorted(task_group.children.values(), key=lambda t: t.label)
    ]

    if task_group.upstream_group_ids or task_group.upstream_task_ids:
        children.append(
            {
                'id': task_group.upstream_join_id,
                'value': {
                    'label': '',
                    'labelStyle': f"fill:{task_group.ui_fgcolor};",
                    'style': f"fill:{task_group.ui_color};",
                    'shape': 'circle',
                },
            }
        )

    if task_group.downstream_group_ids or task_group.downstream_task_ids:
        # This is the join node used to reduce the number of edges between two TaskGroup.
        children.append(
            {
                'id': task_group.downstream_join_id,
                'value': {
                    'label': '',
                    'labelStyle': f"fill:{task_group.ui_fgcolor};",
                    'style': f"fill:{task_group.ui_color};",
                    'shape': 'circle',
                },
            }
        )

    return {
        "id": task_group.group_id,
        'value': {
            'label': task_group.label,
            'labelStyle': f"fill:{task_group.ui_fgcolor};",
            'style': f"fill:{task_group.ui_color}",
            'rx': 5,
            'ry': 5,
            'clusterLabelPos': 'top',
            'tooltip': task_group.tooltip,
        },
        'children': children,
    }


def get_key_paths(input_dict):
    """Return a list of dot-separated dictionary paths"""
    for key, value in input_dict.items():
        if isinstance(value, dict):
            for sub_key in get_key_paths(value):
                yield '.'.join((key, sub_key))
        else:
            yield key


def get_value_from_path(key_path, content):
    """Return the value from a dictionary based on dot-separated path of keys"""
    elem = content
    for x in key_path.strip(".").split("."):
        try:
            x = int(x)
            elem = elem[x]
        except ValueError:
            elem = elem.get(x)

    return elem


def dag_edges(dag):
    """
    Create the list of edges needed to construct the Graph view.

    A special case is made if a TaskGroup is immediately upstream/downstream of another
    TaskGroup or task. Two dummy nodes named upstream_join_id and downstream_join_id are
    created for the TaskGroup. Instead of drawing an edge onto every task in the TaskGroup,
    all edges are directed onto the dummy nodes. This is to cut down the number of edges on
    the graph.

    For example: A DAG with TaskGroups group1 and group2:
        group1: task1, task2, task3
        group2: task4, task5, task6

    group2 is downstream of group1:
        group1 >> group2

    Edges to add (This avoids having to create edges between every task in group1 and group2):
        task1 >> downstream_join_id
        task2 >> downstream_join_id
        task3 >> downstream_join_id
        downstream_join_id >> upstream_join_id
        upstream_join_id >> task4
        upstream_join_id >> task5
        upstream_join_id >> task6
    """
    # Edges to add between TaskGroup
    edges_to_add = set()
    # Edges to remove between individual tasks that are replaced by edges_to_add.
    edges_to_skip = set()

    task_group_map = dag.task_group.get_task_group_dict()

    def collect_edges(task_group):
        """Update edges_to_add and edges_to_skip according to TaskGroups."""
        if isinstance(task_group, AbstractOperator):
            return

        for target_id in task_group.downstream_group_ids:
            # For every TaskGroup immediately downstream, add edges between downstream_join_id
            # and upstream_join_id. Skip edges between individual tasks of the TaskGroups.
            target_group = task_group_map[target_id]
            edges_to_add.add((task_group.downstream_join_id, target_group.upstream_join_id))

            for child in task_group.get_leaves():
                edges_to_add.add((child.task_id, task_group.downstream_join_id))
                for target in target_group.get_roots():
                    edges_to_skip.add((child.task_id, target.task_id))
                edges_to_skip.add((child.task_id, target_group.upstream_join_id))

            for child in target_group.get_roots():
                edges_to_add.add((target_group.upstream_join_id, child.task_id))
                edges_to_skip.add((task_group.downstream_join_id, child.task_id))

        # For every individual task immediately downstream, add edges between downstream_join_id and
        # the downstream task. Skip edges between individual tasks of the TaskGroup and the
        # downstream task.
        for target_id in task_group.downstream_task_ids:
            edges_to_add.add((task_group.downstream_join_id, target_id))

            for child in task_group.get_leaves():
                edges_to_add.add((child.task_id, task_group.downstream_join_id))
                edges_to_skip.add((child.task_id, target_id))

        # For every individual task immediately upstream, add edges between the upstream task
        # and upstream_join_id. Skip edges between the upstream task and individual tasks
        # of the TaskGroup.
        for source_id in task_group.upstream_task_ids:
            edges_to_add.add((source_id, task_group.upstream_join_id))
            for child in task_group.get_roots():
                edges_to_add.add((task_group.upstream_join_id, child.task_id))
                edges_to_skip.add((source_id, child.task_id))

        for child in task_group.children.values():
            collect_edges(child)

    collect_edges(dag.task_group)

    # Collect all the edges between individual tasks
    edges = set()

    def get_downstream(task):
        for child in task.downstream_list:
            edge = (task.task_id, child.task_id)
            if edge not in edges:
                edges.add(edge)
                get_downstream(child)

    for root in dag.roots:
        get_downstream(root)

    result = []
    # Build result dicts with the two ends of the edge, plus any extra metadata
    # if we have it.
    for source_id, target_id in sorted(edges.union(edges_to_add) - edges_to_skip):
        record = {"source_id": source_id, "target_id": target_id}
        label = dag.get_edge_info(source_id, target_id).get("label")
        if label:
            record["label"] = label
        result.append(record)
    return result


def get_task_stats_from_query(qry):
    """
    Return a dict of the task quantity, grouped by dag id and task status.

    :param qry: The data in the format (<dag id>, <task state>, <is dag running>, <task count>),
        ordered by <dag id> and <is dag running>
    """
    data = {}
    last_dag_id = None
    has_running_dags = False
    for dag_id, state, is_dag_running, count in qry:
        if last_dag_id != dag_id:
            last_dag_id = dag_id
            has_running_dags = False
        elif not is_dag_running and has_running_dags:
            continue

        if is_dag_running:
            has_running_dags = True
        if dag_id not in data:
            data[dag_id] = {}
        data[dag_id][state] = count
    return data


def redirect_or_json(origin, msg, status="", status_code=200):
    """
    Some endpoints are called by javascript,
    returning json will allow us to more elegantly handle side-effects in-page
    """
    if request.headers.get('Accept') == 'application/json':
        if status == 'error' and status_code == 200:
            status_code = 500
        return Response(response=msg, status=status_code, mimetype="application/json")
    else:
        if status:
            flash(msg, status)
        else:
            flash(msg)
        return redirect(origin)


######################################################################################
#                                    Error handlers
######################################################################################


def not_found(error):
    """Show Not Found on screen for any error in the Webserver"""
    return (
        render_template(
            'airflow/not_found.html',
            hostname=get_hostname()
            if conf.getboolean('webserver', 'EXPOSE_HOSTNAME', fallback=True)
            else 'redact',
        ),
        404,
    )


def show_traceback(error):
    """Show Traceback for a given error"""
    return (
        render_template(
            'airflow/traceback.html',
            python_version=sys.version.split(" ")[0],
            airflow_version=version,
            hostname=get_hostname()
            if conf.getboolean('webserver', 'EXPOSE_HOSTNAME', fallback=True)
            else 'redact',
            info=traceback.format_exc()
            if conf.getboolean('webserver', 'EXPOSE_STACKTRACE', fallback=True)
            else 'Error! Please contact server admin.',
        ),
        500,
    )


######################################################################################
#                                    BaseViews
######################################################################################


class AirflowBaseView(BaseView):
    """Base View to set Airflow related properties"""

    from airflow import macros

    route_base = ''

    extra_args = {
        # Make our macros available to our UI templates too.
        'macros': macros,
        'get_docs_url': get_docs_url,
    }

    if not conf.getboolean('core', 'unit_test_mode'):
        extra_args['sqlite_warning'] = settings.engine.dialect.name == 'sqlite'
        extra_args['sequential_executor_warning'] = conf.get('core', 'executor') == 'SequentialExecutor'

    line_chart_attr = {
        'legend.maxKeyLength': 200,
    }

    def render_template(self, *args, **kwargs):
        # Add triggerer_job only if we need it
        if TriggererJob.is_needed():
            kwargs["triggerer_job"] = lazy_object_proxy.Proxy(TriggererJob.most_recent_job)
        return super().render_template(
            *args,
            # Cache this at most once per request, not for the lifetime of the view instance
            scheduler_job=lazy_object_proxy.Proxy(SchedulerJob.most_recent_job),
            **kwargs,
        )


def add_user_permissions_to_dag(sender, template, context, **extra):
    """
    Adds `.can_edit`, `.can_trigger`, and `.can_delete` properties
    to DAG based on current user's permissions.
    Located in `views.py` rather than the DAG model to keep
    permissions logic out of the Airflow core.
    """
    if 'dag' in context:
        dag = context['dag']
        can_create_dag_run = get_airflow_app().appbuilder.sm.has_access(
            permissions.ACTION_CAN_CREATE, permissions.RESOURCE_DAG_RUN
        )

        dag.can_edit = get_airflow_app().appbuilder.sm.can_edit_dag(dag.dag_id)
        dag.can_trigger = dag.can_edit and can_create_dag_run
        dag.can_delete = get_airflow_app().appbuilder.sm.can_delete_dag(dag.dag_id)
        context['dag'] = dag


before_render_template.connect(add_user_permissions_to_dag)


class Airflow(AirflowBaseView):
    """Main Airflow application."""

    @expose('/health')
    def health(self):
        """
        An endpoint helping check the health status of the Airflow instance,
        including metadatabase and scheduler.
        """
        payload = {'metadatabase': {'status': 'unhealthy'}}

        latest_scheduler_heartbeat = None
        scheduler_status = 'unhealthy'
        payload['metadatabase'] = {'status': 'healthy'}
        try:
            scheduler_job = SchedulerJob.most_recent_job()

            if scheduler_job:
                latest_scheduler_heartbeat = scheduler_job.latest_heartbeat.isoformat()
                if scheduler_job.is_alive():
                    scheduler_status = 'healthy'
        except Exception:
            payload['metadatabase']['status'] = 'unhealthy'

        payload['scheduler'] = {
            'status': scheduler_status,
            'latest_scheduler_heartbeat': latest_scheduler_heartbeat,
        }

        return wwwutils.json_response(payload)

    @expose('/home')
    @auth.has_access(
        [
            (permissions.ACTION_CAN_READ, permissions.RESOURCE_WEBSITE),
        ]
    )
    def index(self):
        """Home view."""
        hide_paused_dags_by_default = conf.getboolean('webserver', 'hide_paused_dags_by_default')
        default_dag_run = conf.getint('webserver', 'default_dag_run_display_number')

        num_runs = request.args.get('num_runs', default=default_dag_run, type=int)
        current_page = request.args.get('page', default=0, type=int)
        arg_search_query = request.args.get('search')
        arg_tags_filter = request.args.getlist('tags')
        arg_status_filter = request.args.get('status')
        arg_sorting_key = request.args.get('sorting_key', 'dag_id')
        arg_sorting_direction = request.args.get('sorting_direction', default='asc')

        if request.args.get('reset_tags') is not None:
            flask_session[FILTER_TAGS_COOKIE] = None
            # Remove the reset_tags=reset from the URL
            return redirect(url_for('Airflow.index'))

        cookie_val = flask_session.get(FILTER_TAGS_COOKIE)
        if arg_tags_filter:
            flask_session[FILTER_TAGS_COOKIE] = ','.join(arg_tags_filter)
        elif cookie_val:
            # If tags exist in cookie, but not URL, add them to the URL
            return redirect(url_for('Airflow.index', tags=cookie_val.split(',')))

        if arg_status_filter is None:
            cookie_val = flask_session.get(FILTER_STATUS_COOKIE)
            if cookie_val:
                arg_status_filter = cookie_val
            else:
                arg_status_filter = 'active' if hide_paused_dags_by_default else 'all'
                flask_session[FILTER_STATUS_COOKIE] = arg_status_filter
        else:
            status = arg_status_filter.strip().lower()
            flask_session[FILTER_STATUS_COOKIE] = status
            arg_status_filter = status

        dags_per_page = PAGE_SIZE

        start = current_page * dags_per_page
        end = start + dags_per_page

        # Get all the dag id the user could access
        filter_dag_ids = get_airflow_app().appbuilder.sm.get_accessible_dag_ids(g.user)

        with create_session() as session:
            # read orm_dags from the db
            dags_query = session.query(DagModel).filter(~DagModel.is_subdag, DagModel.is_active)

            if arg_search_query:
                dags_query = dags_query.filter(
                    DagModel.dag_id.ilike('%' + arg_search_query + '%')
                    | DagModel.owners.ilike('%' + arg_search_query + '%')
                )

            if arg_tags_filter:
                dags_query = dags_query.filter(DagModel.tags.any(DagTag.name.in_(arg_tags_filter)))

            dags_query = dags_query.filter(DagModel.dag_id.in_(filter_dag_ids))

            all_dags = dags_query
            active_dags = dags_query.filter(~DagModel.is_paused)
            paused_dags = dags_query.filter(DagModel.is_paused)

            is_paused_count = dict(
                all_dags.with_entities(DagModel.is_paused, func.count(DagModel.dag_id))
                .group_by(DagModel.is_paused)
                .all()
            )
            status_count_active = is_paused_count.get(False, 0)
            status_count_paused = is_paused_count.get(True, 0)
            all_dags_count = status_count_active + status_count_paused
            if arg_status_filter == 'active':
                current_dags = active_dags
                num_of_all_dags = status_count_active
            elif arg_status_filter == 'paused':
                current_dags = paused_dags
                num_of_all_dags = status_count_paused
            else:
                current_dags = all_dags
                num_of_all_dags = all_dags_count

            sort_column = DagModel.__table__.c.get(arg_sorting_key)
            if sort_column is not None:
                if arg_sorting_direction == 'desc':
                    sort_column = sort_column.desc()
                current_dags = current_dags.order_by(sort_column)

            dags = current_dags.options(joinedload(DagModel.tags)).offset(start).limit(dags_per_page).all()
            user_permissions = g.user.perms
            all_dags_editable = (permissions.ACTION_CAN_EDIT, permissions.RESOURCE_DAG) in user_permissions
            can_create_dag_run = (
                permissions.ACTION_CAN_CREATE,
                permissions.RESOURCE_DAG_RUN,
            ) in user_permissions

            all_dags_deletable = (
                permissions.ACTION_CAN_DELETE,
                permissions.RESOURCE_DAG,
            ) in user_permissions

            for dag in dags:
                dag_resource_name = permissions.RESOURCE_DAG_PREFIX + dag.dag_id
                if all_dags_editable:
                    dag.can_edit = True
                else:
                    dag.can_edit = (permissions.ACTION_CAN_EDIT, dag_resource_name) in user_permissions
                dag.can_trigger = dag.can_edit and can_create_dag_run
                if all_dags_deletable:
                    dag.can_delete = True
                else:
                    dag.can_delete = (permissions.ACTION_CAN_DELETE, dag_resource_name) in user_permissions

            dagtags = session.query(DagTag.name).distinct(DagTag.name).all()
            tags = [
                {"name": name, "selected": bool(arg_tags_filter and name in arg_tags_filter)}
                for name, in dagtags
            ]

            import_errors = session.query(errors.ImportError).order_by(errors.ImportError.id)

            if (permissions.ACTION_CAN_READ, permissions.RESOURCE_DAG) not in user_permissions:
                # if the user doesn't have access to all DAGs, only display errors from visible DAGs
                import_errors = import_errors.join(
                    DagModel, DagModel.fileloc == errors.ImportError.filename
                ).filter(DagModel.dag_id.in_(filter_dag_ids))

            for import_error in import_errors:
                flash(
                    "Broken DAG: [{ie.filename}] {ie.stacktrace}".format(ie=import_error),
                    "dag_import_error",
                )

        from airflow.plugins_manager import import_errors as plugin_import_errors

        for filename, stacktrace in plugin_import_errors.items():
            flash(
                f"Broken plugin: [{filename}] {stacktrace}",
                "error",
            )

        num_of_pages = int(math.ceil(num_of_all_dags / float(dags_per_page)))

        state_color_mapping = State.state_color.copy()
        state_color_mapping["null"] = state_color_mapping.pop(None)

        page_title = conf.get(section="webserver", key="instance_name", fallback="DAGs")
        page_title_has_markup = conf.getboolean(
            section="webserver", key="instance_name_has_markup", fallback=False
        )

        dashboard_alerts = [
            fm for fm in settings.DASHBOARD_UIALERTS if fm.should_show(get_airflow_app().appbuilder.sm)
        ]

        def _iter_parsed_moved_data_table_names():
            for table_name in inspect(session.get_bind()).get_table_names():
                segments = table_name.split("__", 3)
                if len(segments) < 3:
                    continue
                if segments[0] != settings.AIRFLOW_MOVED_TABLE_PREFIX:
                    continue
                # Second segment is a version marker that we don't need to show.
                yield segments[-1], table_name

        if (
            permissions.ACTION_CAN_ACCESS_MENU,
            permissions.RESOURCE_ADMIN_MENU,
        ) in user_permissions and conf.getboolean("webserver", "warn_deployment_exposure"):
            robots_file_access_count = (
                session.query(Log)
                .filter(Log.event == "robots")
                .filter(Log.dttm > (utcnow() - timedelta(days=7)))
                .count()
            )
            if robots_file_access_count > 0:
                flash(
                    Markup(
                        'Recent requests have been made to /robots.txt. '
                        'This indicates that this deployment may be accessible to the public internet. '
                        'This warning can be disabled by setting webserver.warn_deployment_exposure=False in '
                        'airflow.cfg. Read more about web deployment security <a href='
                        f'"{get_docs_url("security/webserver.html")}">'
                        'here</a>'
                    ),
                    "warning",
                )

        return self.render_template(
            'airflow/dags.html',
            dags=dags,
            dashboard_alerts=dashboard_alerts,
            migration_moved_data_alerts=sorted(set(_iter_parsed_moved_data_table_names())),
            current_page=current_page,
            search_query=arg_search_query if arg_search_query else '',
            page_title=Markup(page_title) if page_title_has_markup else page_title,
            page_size=dags_per_page,
            num_of_pages=num_of_pages,
            num_dag_from=min(start + 1, num_of_all_dags),
            num_dag_to=min(end, num_of_all_dags),
            num_of_all_dags=num_of_all_dags,
            paging=wwwutils.generate_pages(
                current_page,
                num_of_pages,
                search=escape(arg_search_query) if arg_search_query else None,
                status=arg_status_filter if arg_status_filter else None,
                tags=arg_tags_filter if arg_tags_filter else None,
            ),
            num_runs=num_runs,
            tags=tags,
            state_color=state_color_mapping,
            status_filter=arg_status_filter,
            status_count_all=all_dags_count,
            status_count_active=status_count_active,
            status_count_paused=status_count_paused,
            tags_filter=arg_tags_filter,
            sorting_key=arg_sorting_key,
            sorting_direction=arg_sorting_direction,
            auto_refresh_interval=conf.getint('webserver', 'auto_refresh_interval'),
        )

    @expose('/dag_stats', methods=['POST'])
    @auth.has_access(
        [
            (permissions.ACTION_CAN_READ, permissions.RESOURCE_DAG),
            (permissions.ACTION_CAN_READ, permissions.RESOURCE_DAG_RUN),
        ]
    )
    @provide_session
    def dag_stats(self, session=None):
        """Dag statistics."""
        dr = models.DagRun

        allowed_dag_ids = get_airflow_app().appbuilder.sm.get_accessible_dag_ids(g.user)

        dag_state_stats = session.query(dr.dag_id, dr.state, sqla.func.count(dr.state)).group_by(
            dr.dag_id, dr.state
        )

        # Filter by post parameters
        selected_dag_ids = {unquote(dag_id) for dag_id in request.form.getlist('dag_ids') if dag_id}

        if selected_dag_ids:
            filter_dag_ids = selected_dag_ids.intersection(allowed_dag_ids)
        else:
            filter_dag_ids = allowed_dag_ids

        if not filter_dag_ids:
            return wwwutils.json_response({})

        payload = {}
        dag_state_stats = dag_state_stats.filter(dr.dag_id.in_(filter_dag_ids))
        data = {}

        for dag_id, state, count in dag_state_stats:
            if dag_id not in data:
                data[dag_id] = {}
            data[dag_id][state] = count

        for dag_id in filter_dag_ids:
            payload[dag_id] = []
            for state in State.dag_states:
                count = data.get(dag_id, {}).get(state, 0)
                payload[dag_id].append({'state': state, 'count': count})

        return wwwutils.json_response(payload)

    @expose('/task_stats', methods=['POST'])
    @auth.has_access(
        [
            (permissions.ACTION_CAN_READ, permissions.RESOURCE_DAG),
            (permissions.ACTION_CAN_READ, permissions.RESOURCE_DAG_RUN),
            (permissions.ACTION_CAN_READ, permissions.RESOURCE_TASK_INSTANCE),
        ]
    )
    @provide_session
    def task_stats(self, session=None):
        """Task Statistics"""
        allowed_dag_ids = get_airflow_app().appbuilder.sm.get_accessible_dag_ids(g.user)

        if not allowed_dag_ids:
            return wwwutils.json_response({})

        # Filter by post parameters
        selected_dag_ids = {unquote(dag_id) for dag_id in request.form.getlist('dag_ids') if dag_id}

        if selected_dag_ids:
            filter_dag_ids = selected_dag_ids.intersection(allowed_dag_ids)
        else:
            filter_dag_ids = allowed_dag_ids

        running_dag_run_query_result = (
            session.query(DagRun.dag_id, DagRun.run_id)
            .join(DagModel, DagModel.dag_id == DagRun.dag_id)
            .filter(DagRun.state == State.RUNNING, DagModel.is_active)
        )

        running_dag_run_query_result = running_dag_run_query_result.filter(DagRun.dag_id.in_(filter_dag_ids))

        running_dag_run_query_result = running_dag_run_query_result.subquery('running_dag_run')

        # Select all task_instances from active dag_runs.
        running_task_instance_query_result = session.query(
            TaskInstance.dag_id.label('dag_id'),
            TaskInstance.state.label('state'),
            sqla.literal(True).label('is_dag_running'),
        ).join(
            running_dag_run_query_result,
            and_(
                running_dag_run_query_result.c.dag_id == TaskInstance.dag_id,
                running_dag_run_query_result.c.run_id == TaskInstance.run_id,
            ),
        )

        if conf.getboolean('webserver', 'SHOW_RECENT_STATS_FOR_COMPLETED_RUNS', fallback=True):

            last_dag_run = (
                session.query(DagRun.dag_id, sqla.func.max(DagRun.execution_date).label('execution_date'))
                .join(DagModel, DagModel.dag_id == DagRun.dag_id)
                .filter(DagRun.state != State.RUNNING, DagModel.is_active)
                .group_by(DagRun.dag_id)
            )

            last_dag_run = last_dag_run.filter(DagRun.dag_id.in_(filter_dag_ids))
            last_dag_run = last_dag_run.subquery('last_dag_run')

            # Select all task_instances from active dag_runs.
            # If no dag_run is active, return task instances from most recent dag_run.
            last_task_instance_query_result = (
                session.query(
                    TaskInstance.dag_id.label('dag_id'),
                    TaskInstance.state.label('state'),
                    sqla.literal(False).label('is_dag_running'),
                )
                .join(TaskInstance.dag_run)
                .join(
                    last_dag_run,
                    and_(
                        last_dag_run.c.dag_id == TaskInstance.dag_id,
                        last_dag_run.c.execution_date == DagRun.execution_date,
                    ),
                )
            )

            final_task_instance_query_result = union_all(
                last_task_instance_query_result, running_task_instance_query_result
            ).alias('final_ti')
        else:
            final_task_instance_query_result = running_task_instance_query_result.subquery('final_ti')

        qry = (
            session.query(
                final_task_instance_query_result.c.dag_id,
                final_task_instance_query_result.c.state,
                final_task_instance_query_result.c.is_dag_running,
                sqla.func.count(),
            )
            .group_by(
                final_task_instance_query_result.c.dag_id,
                final_task_instance_query_result.c.state,
                final_task_instance_query_result.c.is_dag_running,
            )
            .order_by(
                final_task_instance_query_result.c.dag_id,
                final_task_instance_query_result.c.is_dag_running.desc(),
            )
        )

        data = get_task_stats_from_query(qry)
        payload = {}
        for dag_id in filter_dag_ids:
            payload[dag_id] = []
            for state in State.task_states:
                count = data.get(dag_id, {}).get(state, 0)
                payload[dag_id].append({'state': state, 'count': count})
        return wwwutils.json_response(payload)

    @expose('/last_dagruns', methods=['POST'])
    @auth.has_access(
        [
            (permissions.ACTION_CAN_READ, permissions.RESOURCE_DAG),
            (permissions.ACTION_CAN_READ, permissions.RESOURCE_DAG_RUN),
        ]
    )
    @provide_session
    def last_dagruns(self, session=None):
        """Last DAG runs"""
        allowed_dag_ids = get_airflow_app().appbuilder.sm.get_accessible_dag_ids(g.user)

        # Filter by post parameters
        selected_dag_ids = {unquote(dag_id) for dag_id in request.form.getlist('dag_ids') if dag_id}

        if selected_dag_ids:
            filter_dag_ids = selected_dag_ids.intersection(allowed_dag_ids)
        else:
            filter_dag_ids = allowed_dag_ids

        if not filter_dag_ids:
            return wwwutils.json_response({})

        last_runs_subquery = (
            session.query(
                DagRun.dag_id,
                sqla.func.max(DagRun.execution_date).label("max_execution_date"),
            )
            .group_by(DagRun.dag_id)
            .filter(DagRun.dag_id.in_(filter_dag_ids))  # Only include accessible/selected DAGs.
            .subquery("last_runs")
        )

        query = session.query(
            DagRun.dag_id,
            DagRun.start_date,
            DagRun.end_date,
            DagRun.state,
            DagRun.execution_date,
            DagRun.data_interval_start,
            DagRun.data_interval_end,
        ).join(
            last_runs_subquery,
            and_(
                last_runs_subquery.c.dag_id == DagRun.dag_id,
                last_runs_subquery.c.max_execution_date == DagRun.execution_date,
            ),
        )

        resp = {
            r.dag_id.replace('.', '__dot__'): {
                "dag_id": r.dag_id,
                "state": r.state,
                "execution_date": wwwutils.datetime_to_string(r.execution_date),
                "start_date": wwwutils.datetime_to_string(r.start_date),
                "end_date": wwwutils.datetime_to_string(r.end_date),
                "data_interval_start": wwwutils.datetime_to_string(r.data_interval_start),
                "data_interval_end": wwwutils.datetime_to_string(r.data_interval_end),
            }
            for r in query
        }
        return wwwutils.json_response(resp)

    @expose('/code')
    @auth.has_access(
        [
            (permissions.ACTION_CAN_READ, permissions.RESOURCE_DAG),
            (permissions.ACTION_CAN_READ, permissions.RESOURCE_DAG_CODE),
        ]
    )
    def legacy_code(self):
        """Redirect from url param."""
        return redirect(url_for('Airflow.code', **request.args))

    @expose('/dags/<string:dag_id>/code')
    @auth.has_access(
        [
            (permissions.ACTION_CAN_READ, permissions.RESOURCE_DAG),
            (permissions.ACTION_CAN_READ, permissions.RESOURCE_DAG_CODE),
        ]
    )
    @provide_session
    def code(self, dag_id, session=None):
        """Dag Code."""
        all_errors = ""
        dag_orm = None

        try:
            dag_orm = DagModel.get_dagmodel(dag_id, session=session)
            code = DagCode.get_code_by_fileloc(dag_orm.fileloc)
            html_code = Markup(highlight(code, lexers.PythonLexer(), HtmlFormatter(linenos=True)))

        except Exception as e:
            all_errors += (
                "Exception encountered during "
                f"dag_id retrieval/dag retrieval fallback/code highlighting:\n\n{e}\n"
            )
            html_code = Markup('<p>Failed to load DAG file Code.</p><p>Details: {}</p>').format(
                escape(all_errors)
            )

        wwwutils.check_import_errors(dag_orm.fileloc, session)
        wwwutils.check_dag_warnings(dag_orm.dag_id, session)

        return self.render_template(
            'airflow/dag_code.html',
            html_code=html_code,
            dag=dag_orm,
            dag_model=dag_orm,
            title=dag_id,
            root=request.args.get('root'),
            wrapped=conf.getboolean('webserver', 'default_wrap'),
        )

    @expose('/dag_details')
    @auth.has_access(
        [
            (permissions.ACTION_CAN_READ, permissions.RESOURCE_DAG),
            (permissions.ACTION_CAN_READ, permissions.RESOURCE_DAG_RUN),
        ]
    )
    def legacy_dag_details(self):
        """Redirect from url param."""
        return redirect(url_for('Airflow.dag_details', **request.args))

    @expose('/dags/<string:dag_id>/details')
    @auth.has_access(
        [
            (permissions.ACTION_CAN_READ, permissions.RESOURCE_DAG),
            (permissions.ACTION_CAN_READ, permissions.RESOURCE_DAG_RUN),
        ]
    )
    @provide_session
    def dag_details(self, dag_id, session=None):
        """Get Dag details."""
        dag = get_airflow_app().dag_bag.get_dag(dag_id)
        dag_model = DagModel.get_dagmodel(dag_id, session=session)

        title = "DAG Details"
        root = request.args.get('root', '')

        wwwutils.check_import_errors(dag.fileloc, session)
        wwwutils.check_dag_warnings(dag.dag_id, session)

        states = (
            session.query(TaskInstance.state, sqla.func.count(TaskInstance.dag_id))
            .filter(TaskInstance.dag_id == dag_id)
            .group_by(TaskInstance.state)
            .all()
        )

        active_runs = models.DagRun.find(dag_id=dag_id, state=State.RUNNING, external_trigger=False)

        tags = session.query(models.DagTag).filter(models.DagTag.dag_id == dag_id).all()

        attrs_to_avoid = [
            "NUM_DAGS_PER_DAGRUN_QUERY",
            "serialized_dag",
            "tags",
            "default_view",
            "relative_fileloc",
            "dag_id",
            "description",
            "max_active_runs",
            "max_active_tasks",
            "schedule_interval",
            "owners",
            "is_paused",
        ]
        attrs_to_avoid.extend(wwwutils.get_attr_renderer().keys())
        dag_model_attrs: List[Tuple[str, Any]] = [
            (attr_name, attr)
            for attr_name, attr in (
                (attr_name, getattr(dag_model, attr_name))
                for attr_name in dir(dag_model)
                if not attr_name.startswith("_") and attr_name not in attrs_to_avoid
            )
            if not callable(attr)
        ]

        return self.render_template(
            'airflow/dag_details.html',
            dag=dag,
            title=title,
            root=root,
            states=states,
            State=State,
            active_runs=active_runs,
            tags=tags,
            dag_model_attrs=dag_model_attrs,
        )

    @expose('/rendered-templates')
    @auth.has_access(
        [
            (permissions.ACTION_CAN_READ, permissions.RESOURCE_DAG),
            (permissions.ACTION_CAN_READ, permissions.RESOURCE_TASK_INSTANCE),
        ]
    )
    @action_logging
    @provide_session
    def rendered_templates(self, session):
        """Get rendered Dag."""
        dag_id = request.args.get('dag_id')
        task_id = request.args.get('task_id')
        map_index = request.args.get('map_index', -1, type=int)
        execution_date = request.args.get('execution_date')
        dttm = _safe_parse_datetime(execution_date)
        form = DateTimeForm(data={'execution_date': dttm})
        root = request.args.get('root', '')

        logging.info("Retrieving rendered templates.")
        dag: DAG = get_airflow_app().dag_bag.get_dag(dag_id)
        dag_run = dag.get_dagrun(execution_date=dttm, session=session)
        raw_task = dag.get_task(task_id).prepare_for_execution()

        ti: TaskInstance
        if dag_run is None:
            # No DAG run matching given logical date. This usually means this
            # DAG has never been run. Task instance rendering does not really
            # make sense in this situation, but "works" prior to AIP-39. This
            # "fakes" a temporary DagRun-TaskInstance association (not saved to
            # database) for presentation only.
            ti = TaskInstance(raw_task, map_index=map_index)
            ti.dag_run = DagRun(dag_id=dag_id, execution_date=dttm)
        else:
            ti = dag_run.get_task_instance(task_id=task_id, map_index=map_index, session=session)
            ti.refresh_from_task(raw_task)

        try:
            ti.get_rendered_template_fields(session=session)
        except AirflowException as e:
            msg = "Error rendering template: " + escape(e)
            if e.__cause__:
                msg += Markup("<br><br>OriginalError: ") + escape(e.__cause__)
            flash(msg, "error")
        except Exception as e:
            flash("Error rendering template: " + str(e), "error")

        # Ensure we are rendering the unmapped operator. Unmapping should be
        # done automatically if template fields are rendered successfully; this
        # only matters if get_rendered_template_fields() raised an exception.
        # The following rendering won't show useful values in this case anyway,
        # but we'll display some quasi-meaingful field names.
        task = ti.task.unmap()

        title = "Rendered Template"
        html_dict = {}
        renderers = wwwutils.get_attr_renderer()

        for template_field in task.template_fields:
            content = getattr(task, template_field)
            renderer = task.template_fields_renderers.get(template_field, template_field)
            if renderer in renderers:
                if isinstance(content, (dict, list)):
                    json_content = json.dumps(content, sort_keys=True, indent=4)
                    html_dict[template_field] = renderers[renderer](json_content)
                else:
                    html_dict[template_field] = renderers[renderer](content)
            else:
                html_dict[template_field] = Markup("<pre><code>{}</pre></code>").format(pformat(content))

            if isinstance(content, dict):
                if template_field == 'op_kwargs':
                    for key, value in content.items():
                        renderer = task.template_fields_renderers.get(key, key)
                        if renderer in renderers:
                            html_dict['.'.join([template_field, key])] = renderers[renderer](value)
                        else:
                            html_dict['.'.join([template_field, key])] = Markup(
                                "<pre><code>{}</pre></code>"
                            ).format(pformat(value))
                else:
                    for dict_keys in get_key_paths(content):
                        template_path = '.'.join((template_field, dict_keys))
                        renderer = task.template_fields_renderers.get(template_path, template_path)
                        if renderer in renderers:
                            content_value = get_value_from_path(dict_keys, content)
                            html_dict[template_path] = renderers[renderer](content_value)

        return self.render_template(
            'airflow/ti_code.html',
            html_dict=html_dict,
            dag=dag,
            task_id=task_id,
            execution_date=execution_date,
            map_index=map_index,
            form=form,
            root=root,
            title=title,
        )

    @expose('/rendered-k8s')
    @auth.has_access(
        [
            (permissions.ACTION_CAN_READ, permissions.RESOURCE_DAG),
            (permissions.ACTION_CAN_READ, permissions.RESOURCE_TASK_INSTANCE),
        ]
    )
    @action_logging
    @provide_session
    def rendered_k8s(self, session: Session = NEW_SESSION):
        """Get rendered k8s yaml."""
        if not settings.IS_K8S_OR_K8SCELERY_EXECUTOR:
            abort(404)
        dag_id = request.args.get('dag_id')
        task_id = request.args.get('task_id')
        if task_id is None:
            logging.warning("Task id not passed in the request")
            abort(400)
        execution_date = request.args.get('execution_date')
        dttm = _safe_parse_datetime(execution_date)
        form = DateTimeForm(data={'execution_date': dttm})
        root = request.args.get('root', '')
        map_index = request.args.get('map_index', -1, type=int)
        logging.info("Retrieving rendered templates.")

        dag: DAG = get_airflow_app().dag_bag.get_dag(dag_id)
        task = dag.get_task(task_id)
        dag_run = dag.get_dagrun(execution_date=dttm, session=session)
        ti = dag_run.get_task_instance(task_id=task.task_id, map_index=map_index, session=session)

        pod_spec = None
        try:
            pod_spec = ti.get_rendered_k8s_spec(session=session)
        except AirflowException as e:
            msg = "Error rendering Kubernetes POD Spec: " + escape(e)
            if e.__cause__:
                msg += Markup("<br><br>OriginalError: ") + escape(e.__cause__)
            flash(msg, "error")
        except Exception as e:
            flash("Error rendering Kubernetes Pod Spec: " + str(e), "error")
        title = "Rendered K8s Pod Spec"
        html_dict = {}
        renderers = wwwutils.get_attr_renderer()
        if pod_spec:
            content = yaml.dump(pod_spec)
            content = renderers["yaml"](content)
        else:
            content = Markup("<pre><code>Error rendering Kubernetes POD Spec</pre></code>")
        html_dict['k8s'] = content

        return self.render_template(
            'airflow/ti_code.html',
            html_dict=html_dict,
            dag=dag,
            task_id=task_id,
            execution_date=execution_date,
            map_index=map_index,
            form=form,
            root=root,
            title=title,
        )

    @expose('/get_logs_with_metadata')
    @auth.has_access(
        [
            (permissions.ACTION_CAN_READ, permissions.RESOURCE_DAG),
            (permissions.ACTION_CAN_READ, permissions.RESOURCE_TASK_INSTANCE),
            (permissions.ACTION_CAN_READ, permissions.RESOURCE_TASK_LOG),
        ]
    )
    @action_logging
    @provide_session
    def get_logs_with_metadata(self, session=None):
        """Retrieve logs including metadata."""
        dag_id = request.args.get('dag_id')
        task_id = request.args.get('task_id')
        execution_date = request.args.get('execution_date')
        map_index = request.args.get('map_index', -1, type=int)
        try_number = request.args.get('try_number', type=int)
        metadata = request.args.get('metadata', '{}')
        response_format = request.args.get('format', 'json')

        # Validate JSON metadata
        try:
            metadata = json.loads(metadata)
            # metadata may be null
            if not metadata:
                metadata = {}
        except json.decoder.JSONDecodeError:
            error_message = "Invalid JSON metadata"
            response = jsonify({"error": error_message})
            response.status_code = 400

            return response

        # Convert string datetime into actual datetime
        try:
            execution_date = timezone.parse(execution_date)
        except ValueError:
            error_message = (
                f'Given execution date, {execution_date}, could not be identified as a date. '
                'Example date format: 2015-11-16T14:34:15+00:00'
            )
            response = jsonify({'error': error_message})
            response.status_code = 400

            return response

        task_log_reader = TaskLogReader()
        if not task_log_reader.supports_read:
            return jsonify(
                message="Task log handler does not support read logs.",
                error=True,
                metadata={"end_of_log": True},
            )

        ti = (
            session.query(models.TaskInstance)
            .filter_by(dag_id=dag_id, task_id=task_id, execution_date=execution_date, map_index=map_index)
            .first()
        )

        if ti is None:
            return jsonify(
                message="*** Task instance did not exist in the DB\n",
                error=True,
                metadata={"end_of_log": True},
            )

        try:
            dag = get_airflow_app().dag_bag.get_dag(dag_id)
            if dag:
                ti.task = dag.get_task(ti.task_id)

            if response_format == 'json':
                logs, metadata = task_log_reader.read_log_chunks(ti, try_number, metadata)
                message = logs[0] if try_number is not None else logs
                return jsonify(message=message, metadata=metadata)

            metadata['download_logs'] = True
            attachment_filename = task_log_reader.render_log_filename(ti, try_number, session=session)
            log_stream = task_log_reader.read_log_stream(ti, try_number, metadata)
            return Response(
                response=log_stream,
                mimetype="text/plain",
                headers={"Content-Disposition": f"attachment; filename={attachment_filename}"},
            )
        except AttributeError as e:
            error_message = [f"Task log handler does not support read logs.\n{str(e)}\n"]
            metadata['end_of_log'] = True
            return jsonify(message=error_message, error=True, metadata=metadata)

    @expose('/log')
    @auth.has_access(
        [
            (permissions.ACTION_CAN_READ, permissions.RESOURCE_DAG),
            (permissions.ACTION_CAN_READ, permissions.RESOURCE_TASK_INSTANCE),
            (permissions.ACTION_CAN_READ, permissions.RESOURCE_TASK_LOG),
        ]
    )
    @action_logging
    @provide_session
    def log(self, session=None):
        """Retrieve log."""
        dag_id = request.args.get('dag_id')
        task_id = request.args.get('task_id')
        map_index = request.args.get('map_index', -1, type=int)
        execution_date = request.args.get('execution_date')

        if execution_date:
            dttm = _safe_parse_datetime(execution_date)
        else:
            dttm = None

        form = DateTimeForm(data={'execution_date': dttm})
        dag_model = DagModel.get_dagmodel(dag_id)

        ti = (
            session.query(models.TaskInstance)
            .filter_by(dag_id=dag_id, task_id=task_id, execution_date=dttm, map_index=map_index)
            .first()
        )

        num_logs = 0
        if ti is not None:
            num_logs = ti.next_try_number - 1
            if ti.state in (State.UP_FOR_RESCHEDULE, State.DEFERRED):
                # Tasks in reschedule state decremented the try number
                num_logs += 1
        logs = [''] * num_logs
        root = request.args.get('root', '')
        return self.render_template(
            'airflow/ti_log.html',
            logs=logs,
            dag=dag_model,
            title="Log by attempts",
            dag_id=dag_id,
            task_id=task_id,
            execution_date=execution_date,
            map_index=map_index,
            form=form,
            root=root,
            wrapped=conf.getboolean('webserver', 'default_wrap'),
        )

    @expose('/redirect_to_external_log')
    @auth.has_access(
        [
            (permissions.ACTION_CAN_READ, permissions.RESOURCE_DAG),
            (permissions.ACTION_CAN_READ, permissions.RESOURCE_TASK_INSTANCE),
            (permissions.ACTION_CAN_READ, permissions.RESOURCE_TASK_LOG),
        ]
    )
    @action_logging
    @provide_session
    def redirect_to_external_log(self, session=None):
        """Redirects to external log."""
        dag_id = request.args.get('dag_id')
        task_id = request.args.get('task_id')
        execution_date = request.args.get('execution_date')
        dttm = _safe_parse_datetime(execution_date)
        map_index = request.args.get('map_index', -1, type=int)
        try_number = request.args.get('try_number', 1)

        ti = (
            session.query(models.TaskInstance)
            .filter_by(dag_id=dag_id, task_id=task_id, execution_date=dttm, map_index=map_index)
            .first()
        )

        if not ti:
            flash(f"Task [{dag_id}.{task_id}] does not exist", "error")
            return redirect(url_for('Airflow.index'))

        task_log_reader = TaskLogReader()
        if not task_log_reader.supports_external_link:
            flash("Task log handler does not support external links", "error")
            return redirect(url_for('Airflow.index'))

        handler = task_log_reader.log_handler
        url = handler.get_external_log_url(ti, try_number)
        return redirect(url)

    @expose('/task')
    @auth.has_access(
        [
            (permissions.ACTION_CAN_READ, permissions.RESOURCE_DAG),
            (permissions.ACTION_CAN_READ, permissions.RESOURCE_TASK_INSTANCE),
        ]
    )
    @action_logging
    @provide_session
    def task(self, session):
        """Retrieve task."""
        dag_id = request.args.get('dag_id')
        task_id = request.args.get('task_id')
        execution_date = request.args.get('execution_date')
        dttm = _safe_parse_datetime(execution_date)
        map_index = request.args.get('map_index', -1, type=int)
        form = DateTimeForm(data={'execution_date': dttm})
        root = request.args.get('root', '')
        dag = get_airflow_app().dag_bag.get_dag(dag_id)

        if not dag or task_id not in dag.task_ids:
            flash(f"Task [{dag_id}.{task_id}] doesn't seem to exist at the moment", "error")
            return redirect(url_for('Airflow.index'))
        task = copy.copy(dag.get_task(task_id))
        task.resolve_template_files()

        ti: Optional[TaskInstance] = (
            session.query(TaskInstance)
            .options(
                # HACK: Eager-load relationships. This is needed because
                # multiple properties mis-use provide_session() that destroys
                # the session object ti is bounded to.
                joinedload(TaskInstance.queued_by_job, innerjoin=False),
                joinedload(TaskInstance.trigger, innerjoin=False),
            )
            .filter_by(execution_date=dttm, dag_id=dag_id, task_id=task_id, map_index=map_index)
            .one_or_none()
        )
        if ti is None:
            ti_attrs: Optional[List[Tuple[str, Any]]] = None
        else:
            ti.refresh_from_task(task)
            ti_attrs_to_skip = [
                'dag_id',
                'key',
                'mark_success_url',
                'log',
                'log_url',
                'task',
            ]
            # Some fields on TI are deprecated, but we don't want those warnings here.
            with warnings.catch_warnings():
                warnings.simplefilter("ignore", DeprecationWarning)
                all_ti_attrs = (
                    (name, getattr(ti, name))
                    for name in dir(ti)
                    if not name.startswith("_") and name not in ti_attrs_to_skip
                )
            ti_attrs = sorted((name, attr) for name, attr in all_ti_attrs if not callable(attr))

        attr_renderers = wwwutils.get_attr_renderer()

        attrs_to_skip = getattr(task, 'HIDE_ATTRS_FROM_UI', set())

        def include_task_attrs(attr_name):
            return not (
                attr_name == 'HIDE_ATTRS_FROM_UI'
                or attr_name.startswith("_")
                or attr_name in attr_renderers
                or attr_name in attrs_to_skip
            )

        task_attrs = [
            (attr_name, attr)
            for attr_name, attr in (
                (attr_name, getattr(task, attr_name)) for attr_name in filter(include_task_attrs, dir(task))
            )
            if not callable(attr)
        ]

        # Color coding the special attributes that are code
        special_attrs_rendered = {
            attr_name: renderer(getattr(task, attr_name))
            for attr_name, renderer in attr_renderers.items()
            if hasattr(task, attr_name)
        }

        no_failed_deps_result = [
            (
                "Unknown",
                "All dependencies are met but the task instance is not running. In most "
                "cases this just means that the task will probably be scheduled soon "
                "unless:<br>\n- The scheduler is down or under heavy load<br>\n{}\n"
                "<br>\nIf this task instance does not start soon please contact your "
                "Airflow administrator for assistance.".format(
                    "- This task instance already ran and had it's state changed manually "
                    "(e.g. cleared in the UI)<br>"
                    if ti and ti.state == State.NONE
                    else ""
                ),
            )
        ]

        # Use the scheduler's context to figure out which dependencies are not met
        if ti is None:
            failed_dep_reasons: List[Tuple[str, str]] = []
        else:
            dep_context = DepContext(SCHEDULER_QUEUED_DEPS)
            failed_dep_reasons = [
                (dep.dep_name, dep.reason) for dep in ti.get_failed_dep_statuses(dep_context=dep_context)
            ]

        title = "Task Instance Details"
        return self.render_template(
            'airflow/task.html',
            task_attrs=task_attrs,
            ti_attrs=ti_attrs,
            failed_dep_reasons=failed_dep_reasons or no_failed_deps_result,
            task_id=task_id,
            execution_date=execution_date,
            map_index=map_index,
            special_attrs_rendered=special_attrs_rendered,
            form=form,
            root=root,
            dag=dag,
            title=title,
        )

    @expose('/xcom')
    @auth.has_access(
        [
            (permissions.ACTION_CAN_READ, permissions.RESOURCE_DAG),
            (permissions.ACTION_CAN_READ, permissions.RESOURCE_TASK_INSTANCE),
            (permissions.ACTION_CAN_READ, permissions.RESOURCE_XCOM),
        ]
    )
    @action_logging
    @provide_session
    def xcom(self, session=None):
        """Retrieve XCOM."""
        dag_id = request.args.get('dag_id')
        task_id = request.args.get('task_id')
        map_index = request.args.get('map_index', -1, type=int)
        # Carrying execution_date through, even though it's irrelevant for
        # this context
        execution_date = request.args.get('execution_date')
        dttm = _safe_parse_datetime(execution_date)

        form = DateTimeForm(data={'execution_date': dttm})
        root = request.args.get('root', '')
        dag = DagModel.get_dagmodel(dag_id)
        ti = session.query(TaskInstance).filter_by(dag_id=dag_id, task_id=task_id).first()

        if not ti:
            flash(f"Task [{dag_id}.{task_id}] doesn't seem to exist at the moment", "error")
            return redirect(url_for('Airflow.index'))

        xcomlist = (
            session.query(XCom)
            .filter_by(dag_id=dag_id, task_id=task_id, execution_date=dttm, map_index=map_index)
            .all()
        )

        attributes = []
        for xcom in xcomlist:
            if not xcom.key.startswith('_'):
                attributes.append((xcom.key, xcom.value))

        title = "XCom"
        return self.render_template(
            'airflow/xcom.html',
            attributes=attributes,
            task_id=task_id,
            execution_date=execution_date,
            map_index=map_index,
            form=form,
            root=root,
            dag=dag,
            title=title,
        )

    @expose('/run', methods=['POST'])
    @auth.has_access(
        [
            (permissions.ACTION_CAN_EDIT, permissions.RESOURCE_DAG),
            (permissions.ACTION_CAN_CREATE, permissions.RESOURCE_TASK_INSTANCE),
        ]
    )
    @action_logging
    @provide_session
    def run(self, session=None):
        """Runs Task Instance."""
        dag_id = request.form.get('dag_id')
        task_id = request.form.get('task_id')
        dag_run_id = request.form.get('dag_run_id')
        map_index = request.args.get('map_index', -1, type=int)
        origin = get_safe_url(request.form.get('origin'))
        dag: DAG = get_airflow_app().dag_bag.get_dag(dag_id)
        task = dag.get_task(task_id)

        ignore_all_deps = request.form.get('ignore_all_deps') == "true"
        ignore_task_deps = request.form.get('ignore_task_deps') == "true"
        ignore_ti_state = request.form.get('ignore_ti_state') == "true"

        executor = ExecutorLoader.get_default_executor()

        if not getattr(executor, "supports_ad_hoc_ti_run", False):
            msg = "Only works with the Celery, CeleryKubernetes or Kubernetes executors"
            return redirect_or_json(origin, msg, "error", 400)

        dag_run = dag.get_dagrun(run_id=dag_run_id)
        ti = dag_run.get_task_instance(task_id=task.task_id, map_index=map_index)
        if not ti:
            msg = "Could not queue task instance for execution, task instance is missing"
            return redirect_or_json(origin, msg, "error", 400)

        ti.refresh_from_task(task)

        # Make sure the task instance can be run
        dep_context = DepContext(
            deps=RUNNING_DEPS,
            ignore_all_deps=ignore_all_deps,
            ignore_task_deps=ignore_task_deps,
            ignore_ti_state=ignore_ti_state,
        )
        failed_deps = list(ti.get_failed_dep_statuses(dep_context=dep_context))
        if failed_deps:
            failed_deps_str = ", ".join(f"{dep.dep_name}: {dep.reason}" for dep in</