rhodecode/subscribers.py

# Copyright (C) 2010-2024 RhodeCode GmbH
#
# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU Affero General Public License, version 3
# (only), as published by the Free Software Foundation.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU Affero General Public License
# along with this program.  If not, see <http://www.gnu.org/licenses/>.
#
# This program is dual-licensed. If you wish to learn more about the
# RhodeCode Enterprise Edition, including its added features, Support services,
# and proprietary license terms, please see https://rhodecode.com/licenses/
import io
import shlex

import math
import re
import os
import datetime
import logging
import queue
import subprocess


from dateutil.parser import parse
from pyramid.interfaces import IRoutesMapper
from pyramid.settings import asbool
from pyramid.path import AssetResolver
from threading import Thread

from rhodecode.config.jsroutes import generate_jsroutes_content
from rhodecode.lib.base import get_auth_user
from rhodecode.lib.celerylib.loader import set_celery_conf

import rhodecode


log = logging.getLogger(__name__)


def add_renderer_globals(event):
    from rhodecode.lib import helpers

    # TODO: When executed in pyramid view context the request is not available
    # in the event. Find a better solution to get the request.
    from pyramid.threadlocal import get_current_request
    request = event['request'] or get_current_request()

    # Add Pyramid translation as '_' to context
    event['_'] = request.translate
    event['_ungettext'] = request.plularize
    event['h'] = helpers


def auto_merge_pr_if_needed(event):
    from rhodecode.model.db import PullRequest
    from rhodecode.model.pull_request import (
        PullRequestModel, ChangesetStatus, MergeCheck
    )

    pr_event_data = event.as_dict()['pullrequest']
    pull_request = PullRequest.get(pr_event_data['pull_request_id'])
    calculated_status = pr_event_data['status']
    if (calculated_status == ChangesetStatus.STATUS_APPROVED
            and PullRequestModel().is_automatic_merge_enabled(pull_request)):
        user = pull_request.author.AuthUser()

        merge_check = MergeCheck.validate(
            pull_request, user, translator=lambda x: x, fail_early=True
        )
        if merge_check.merge_possible:
            from rhodecode.lib.base import vcs_operation_context
            extras = vcs_operation_context(
                event.request.environ, repo_name=pull_request.target_repo.repo_name,
                username=user.username, action='push',
                scm=pull_request.target_repo.repo_type)
            from rc_ee.lib.celerylib.tasks import auto_merge_repo
            auto_merge_repo.apply_async(
                args=(pull_request.pull_request_id, extras)
            )


def set_user_lang(event):
    request = event.request
    cur_user = getattr(request, 'user', None)

    if cur_user:
        user_lang = cur_user.get_instance().user_data.get('language')
        if user_lang:
            log.debug('lang: setting current user:%s language to: %s', cur_user, user_lang)
            event.request._LOCALE_ = user_lang


def update_celery_conf(event):
    log.debug('Setting celery config from new request')
    set_celery_conf(request=event.request, registry=event.request.registry)


def add_request_user_context(event):
    """
    Adds auth user into request context
    """

    request = event.request
    # access req_id as soon as possible
    req_id = request.req_id

    if hasattr(request, 'vcs_call'):
        # skip vcs calls
        return

    if hasattr(request, 'rpc_method'):
        # skip api calls
        return

    auth_user, auth_token = get_auth_user(request)
    request.user = auth_user
    request.user_auth_token = auth_token
    request.environ['rc_auth_user'] = auth_user
    request.environ['rc_auth_user_id'] = str(auth_user.user_id)
    request.environ['rc_req_id'] = req_id


def reset_log_bucket(event):
    """
    reset the log bucket on new request
    """
    request = event.request
    request.req_id_records_init()


def scan_repositories_if_enabled(event):
    """
    This is subscribed to the `pyramid.events.ApplicationCreated` event. It
    does a repository scan if enabled in the settings.
    """

    settings = event.app.registry.settings
    vcs_server_enabled = settings['vcs.server.enable']
    import_on_startup = settings['startup.import_repos']

    if vcs_server_enabled and import_on_startup:
        from rhodecode.model.scm import ScmModel
        from rhodecode.lib.utils import repo2db_mapper
        scm = ScmModel()
        repositories = scm.repo_scan(scm.repos_path)
        repo2db_mapper(repositories)


def write_metadata_if_needed(event):
    """
    Writes upgrade metadata
    """
    import rhodecode
    from rhodecode.lib import system_info
    from rhodecode.lib import ext_json

    fname = '.rcmetadata.json'
    ini_loc = os.path.dirname(rhodecode.CONFIG.get('__file__'))
    metadata_destination = os.path.join(ini_loc, fname)

    def get_update_age():
        now = datetime.datetime.utcnow()

        with open(metadata_destination, 'rb') as f:
            data = ext_json.json.loads(f.read())
            if 'created_on' in data:
                update_date = parse(data['created_on'])
                diff = now - update_date
                return diff.total_seconds() / 60.0

        return 0

    def write():
        configuration = system_info.SysInfo(
            system_info.rhodecode_config)()['value']
        license_token = configuration['config']['license_token']

        setup = dict(
            workers=configuration['config']['server:main'].get(
                'workers', '?'),
            worker_type=configuration['config']['server:main'].get(
                'worker_class', 'sync'),
        )
        dbinfo = system_info.SysInfo(system_info.database_info)()['value']
        del dbinfo['url']

        metadata = dict(
            desc='upgrade metadata info',
            license_token=license_token,
            created_on=datetime.datetime.utcnow().isoformat(),
            usage=system_info.SysInfo(system_info.usage_info)()['value'],
            platform=system_info.SysInfo(system_info.platform_type)()['value'],
            database=dbinfo,
            cpu=system_info.SysInfo(system_info.cpu)()['value'],
            memory=system_info.SysInfo(system_info.memory)()['value'],
            setup=setup
        )

        with open(metadata_destination, 'wb') as f:
            f.write(ext_json.json.dumps(metadata))

    settings = event.app.registry.settings
    if settings.get('metadata.skip'):
        return

    # only write this every 24h, workers restart caused unwanted delays
    try:
        age_in_min = get_update_age()
    except Exception:
        age_in_min = 0

    if age_in_min > 60 * 60 * 24:
        return

    try:
        write()
    except Exception:
        pass


def write_usage_data(event):
    import rhodecode
    from rhodecode.lib import system_info
    from rhodecode.lib import ext_json

    settings = event.app.registry.settings
    instance_tag = settings.get('metadata.write_usage_tag')
    if not settings.get('metadata.write_usage'):
        return

    def get_update_age(dest_file):
        now = datetime.datetime.now(datetime.UTC)

        with open(dest_file, 'rb') as f:
            data = ext_json.json.loads(f.read())
            if 'created_on' in data:
                update_date = parse(data['created_on'])
                diff = now - update_date
                return math.ceil(diff.total_seconds() / 60.0)

        return 0

    utc_date = datetime.datetime.now(datetime.UTC)
    hour_quarter = int(math.ceil((utc_date.hour + utc_date.minute/60.0) / 6.))
    fname = f'.rc_usage_{utc_date.year}{utc_date.month:02d}{utc_date.day:02d}_{hour_quarter}.json'
    ini_loc = os.path.dirname(rhodecode.CONFIG.get('__file__'))

    usage_dir = os.path.join(ini_loc, '.rcusage')
    if not os.path.isdir(usage_dir):
        os.makedirs(usage_dir)
    usage_metadata_destination = os.path.join(usage_dir, fname)

    try:
        age_in_min = get_update_age(usage_metadata_destination)
    except Exception:
        age_in_min = 0

    # write every 6th hour
    if age_in_min and age_in_min < 60 * 6:
        log.debug('Usage file created %s minutes ago, skipping (threshold: %s minutes)...',
                  age_in_min, 60 * 6)
        return

    def write(dest_file):
        configuration = system_info.SysInfo(system_info.rhodecode_config)()['value']
        license_token = configuration['config']['license_token']

        metadata = dict(
            desc='Usage data',
            instance_tag=instance_tag,
            license_token=license_token,
            created_on=datetime.datetime.utcnow().isoformat(),
            usage=system_info.SysInfo(system_info.usage_info)()['value'],
        )

        with open(dest_file, 'wb') as f:
            f.write(ext_json.formatted_json(metadata))

    try:
        log.debug('Writing usage file at: %s', usage_metadata_destination)
        write(usage_metadata_destination)
    except Exception:
        pass


def write_js_routes_if_enabled(event):
    registry = event.app.registry

    mapper = registry.queryUtility(IRoutesMapper)
    _argument_prog = re.compile(r'\{(.*?)\}|:\((.*)\)')

    def _extract_route_information(route):
        """
        Convert a route into tuple(name, path, args), eg:
            ('show_user', '/profile/%(username)s', ['username'])
        """

        route_path = route.pattern
        pattern = route.pattern

        def replace(matchobj):
            if matchobj.group(1):
                return "%%(%s)s" % matchobj.group(1).split(':')[0]
            else:
                return "%%(%s)s" % matchobj.group(2)

        route_path = _argument_prog.sub(replace, route_path)

        if not route_path.startswith('/'):
            route_path = f'/{route_path}'

        return (
            route.name,
            route_path,
            [(arg[0].split(':')[0] if arg[0] != '' else arg[1])
             for arg in _argument_prog.findall(pattern)]
        )

    def get_routes():
        # pyramid routes
        for route in mapper.get_routes():
            if not route.name.startswith('__'):
                yield _extract_route_information(route)

    if asbool(registry.settings.get('generate_js_files', 'false')):
        static_path = AssetResolver().resolve('rhodecode:public').abspath()
        jsroutes = get_routes()
        jsroutes_file_content = generate_jsroutes_content(jsroutes)
        jsroutes_file_path = os.path.join(
            static_path, 'js', 'rhodecode', 'routes.js')

        try:
            with open(jsroutes_file_path, 'w', encoding='utf-8') as f:
                f.write(jsroutes_file_content)
            log.debug('generated JS files in %s', jsroutes_file_path)
        except Exception:
            log.exception('Failed to write routes.js into %s', jsroutes_file_path)


def import_license_if_present(event):
    """
    This is subscribed to the `pyramid.events.ApplicationCreated` event. It
    does a import license key based on a presence of the file.
    """
    settings = event.app.registry.settings

    rhodecode_edition_id = settings.get('rhodecode.edition_id')
    license_file_path = settings.get('license.import_path')
    force = settings.get('license.import_path_mode') == 'force'

    if license_file_path and rhodecode_edition_id == 'EE':
        log.debug('license.import_path= is set importing license from %s', license_file_path)
        from rhodecode.model.meta import Session
        from rhodecode.model.license import apply_license_from_file
        try:
            apply_license_from_file(license_file_path, force=force)
            Session().commit()
        except OSError:
            log.exception('Failed to import license from %s, make sure this file exists', license_file_path)


class Subscriber(object):
    """
    Base class for subscribers to the pyramid event system.
    """
    def __call__(self, event):
        self.run(event)

    def run(self, event):
        raise NotImplementedError('Subclass has to implement this.')


class AsyncSubscriber(Subscriber):
    """
    Subscriber that handles the execution of events in a separate task to not
    block the execution of the code which triggers the event. It puts the
    received events into a queue from which the worker process takes them in
    order.
    """
    def __init__(self):
        self._stop = False
        self._eventq = queue.Queue()
        self._worker = self.create_worker()
        self._worker.start()

    def __call__(self, event):
        self._eventq.put(event)

    def create_worker(self):
        worker = Thread(target=self.do_work)
        worker.daemon = True
        return worker

    def stop_worker(self):
        self._stop = False
        self._eventq.put(None)
        self._worker.join()

    def do_work(self):
        while not self._stop:
            event = self._eventq.get()
            if event is not None:
                self.run(event)


class AsyncSubprocessSubscriber(AsyncSubscriber):
    """
    Subscriber that uses the subprocess module to execute a command if an
    event is received. Events are handled asynchronously::

        subscriber = AsyncSubprocessSubscriber('ls -la', timeout=10)
        subscriber(dummyEvent) # running __call__(event)

    """

    def __init__(self, cmd, timeout=None):
        if not isinstance(cmd, (list, tuple)):
            cmd = shlex.split(cmd)
        super().__init__()
        self._cmd = cmd
        self._timeout = timeout

    def run(self, event):
        cmd = self._cmd
        timeout = self._timeout
        log.debug('Executing command %s.', cmd)

        try:
            output = subprocess.check_output(
                cmd, timeout=timeout, stderr=subprocess.STDOUT)
            log.debug('Command finished %s', cmd)
            if output:
                log.debug('Command output: %s', output)
        except subprocess.TimeoutExpired as e:
            log.exception('Timeout while executing command.')
            if e.output:
                log.error('Command output: %s', e.output)
        except subprocess.CalledProcessError as e:
            log.exception('Error while executing command.')
            if e.output:
                log.error('Command output: %s', e.output)
        except Exception:
            log.exception(
                'Exception while executing command %s.', cmd)