modernize: python3 updates

rhodecode/lib/rc_cache/cache_key_meta.py

@@ -1,71 +1,69 b''
-
-
 # Copyright (C) 2015-2023 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 atexit
 import logging
 import os
 import signal
 import sys
 
 import rhodecode
 
 log = logging.getLogger(__name__)
 
 cache_keys_by_pid = set()
 
 
 def sigHandler(signo, frame):
     """
     signals trigger sys.exit so there's a single handler to cleanup the code.
     """
     if rhodecode.is_test:
         return
 
     sys.exit(0)
 
 
 def free_cache_keys(*args):
     from rhodecode.model.db import CacheKey, Session
 
     if rhodecode.is_test:
         return
 
     ssh_cmd = os.environ.get('RC_CMD_SSH_WRAPPER')
     if ssh_cmd:
         return
 
     if cache_keys_by_pid:
         try:
             for cache_proc in set(cache_keys_by_pid):
-                like_expression = '{}%'.format(cache_proc)
+                like_expression = f'{cache_proc}%'
                 qry = CacheKey.query().filter(CacheKey.cache_key.like(like_expression))
                 count = qry.count()
                 log.info('Clearing %s: %s cache keys, total: %s', cache_proc, len(cache_keys_by_pid), count)
                 qry.delete(synchronize_session='fetch')
                 cache_keys_by_pid.remove(cache_proc)
             Session().commit()
         except Exception:
             log.exception('Failed to clear keys, exiting gracefully')
 
 atexit.register(free_cache_keys)
 
 signal.signal(signal.SIGTERM, sigHandler)
 signal.signal(signal.SIGINT, sigHandler)
 

rhodecode/lib/rc_cache/utils.py

@@ -1,405 +1,405 b''
 # Copyright (C) 2015-2023 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 functools
 import logging
 import os
 import threading
 import time
 
 import decorator
 from dogpile.cache import CacheRegion
 
 import rhodecode
 from rhodecode.lib.hash_utils import sha1
 from rhodecode.lib.str_utils import safe_bytes
 from rhodecode.lib.type_utils import str2bool
 
 from . import region_meta, cache_key_meta
 
 log = logging.getLogger(__name__)
 
 
 def isCython(func):
     """
     Private helper that checks if a function is a cython function.
     """
     return func.__class__.__name__ == 'cython_function_or_method'
 
 
 class RhodeCodeCacheRegion(CacheRegion):
 
     def __repr__(self):
         return f'{self.__class__}(name={self.name})'
 
     def conditional_cache_on_arguments(
             self, namespace=None,
             expiration_time=None,
             should_cache_fn=None,
             to_str=str,
             function_key_generator=None,
             condition=True):
         """
         Custom conditional decorator, that will not touch any dogpile internals if
         condition isn't meet. This works a bit different from should_cache_fn
         And it's faster in cases we don't ever want to compute cached values
         """
         expiration_time_is_callable = callable(expiration_time)
         if not namespace:
             namespace = getattr(self, '_default_namespace', None)
 
         if function_key_generator is None:
             function_key_generator = self.function_key_generator
 
         def get_or_create_for_user_func(func_key_generator, user_func, *arg, **kw):
 
             if not condition:
                 log.debug('Calling un-cached method:%s', user_func.__name__)
                 start = time.time()
                 result = user_func(*arg, **kw)
                 total = time.time() - start
                 log.debug('un-cached method:%s took %.4fs', user_func.__name__, total)
                 return result
 
             key = func_key_generator(*arg, **kw)
 
             timeout = expiration_time() if expiration_time_is_callable \
                 else expiration_time
 
             log.debug('Calling cached method:`%s`', user_func.__name__)
             return self.get_or_create(key, user_func, timeout, should_cache_fn, (arg, kw))
 
         def cache_decorator(user_func):
             if to_str is str:
                 # backwards compatible
                 key_generator = function_key_generator(namespace, user_func)
             else:
                 key_generator = function_key_generator(namespace, user_func, to_str=to_str)
 
             def refresh(*arg, **kw):
                 """
                 Like invalidate, but regenerates the value instead
                 """
                 key = key_generator(*arg, **kw)
                 value = user_func(*arg, **kw)
                 self.set(key, value)
                 return value
 
             def invalidate(*arg, **kw):
                 key = key_generator(*arg, **kw)
                 self.delete(key)
 
             def set_(value, *arg, **kw):
                 key = key_generator(*arg, **kw)
                 self.set(key, value)
 
             def get(*arg, **kw):
                 key = key_generator(*arg, **kw)
                 return self.get(key)
 
             user_func.set = set_
             user_func.invalidate = invalidate
             user_func.get = get
             user_func.refresh = refresh
             user_func.key_generator = key_generator
             user_func.original = user_func
 
             # Use `decorate` to preserve the signature of :param:`user_func`.
             return decorator.decorate(user_func, functools.partial(
                 get_or_create_for_user_func, key_generator))
 
         return cache_decorator
 
 
 def make_region(*arg, **kw):
     return RhodeCodeCacheRegion(*arg, **kw)
 
 
 def get_default_cache_settings(settings, prefixes=None):
     prefixes = prefixes or []
     cache_settings = {}
     for key in settings.keys():
         for prefix in prefixes:
             if key.startswith(prefix):
                 name = key.split(prefix)[1].strip()
                 val = settings[key]
                 if isinstance(val, str):
                     val = val.strip()
                 cache_settings[name] = val
     return cache_settings
 
 
 def compute_key_from_params(*args):
     """
     Helper to compute key from given params to be used in cache manager
     """
     return sha1(safe_bytes("_".join(map(str, args))))
 
 
 def custom_key_generator(backend, namespace, fn):
     func_name = fn.__name__
 
     def generate_key(*args):
         backend_pref = getattr(backend, 'key_prefix', None) or 'backend_prefix'
         namespace_pref = namespace or 'default_namespace'
         arg_key = compute_key_from_params(*args)
         final_key = f"{backend_pref}:{namespace_pref}:{func_name}_{arg_key}"
 
         return final_key
 
     return generate_key
 
 
 def backend_key_generator(backend):
     """
     Special wrapper that also sends over the backend to the key generator
     """
     def wrapper(namespace, fn):
         return custom_key_generator(backend, namespace, fn)
     return wrapper
 
 
 def get_or_create_region(region_name, region_namespace: str = None, use_async_runner=False):
     from .backends import FileNamespaceBackend
     from . import async_creation_runner
 
     region_obj = region_meta.dogpile_cache_regions.get(region_name)
     if not region_obj:
         reg_keys = list(region_meta.dogpile_cache_regions.keys())
-        raise EnvironmentError(f'Region `{region_name}` not in configured: {reg_keys}.')
+        raise OSError(f'Region `{region_name}` not in configured: {reg_keys}.')
 
     region_uid_name = f'{region_name}:{region_namespace}'
 
     if isinstance(region_obj.actual_backend, FileNamespaceBackend):
         if not region_namespace:
             raise ValueError(f'{FileNamespaceBackend} used requires to specify region_namespace param')
 
         region_exist = region_meta.dogpile_cache_regions.get(region_namespace)
         if region_exist:
             log.debug('Using already configured region: %s', region_namespace)
             return region_exist
 
         expiration_time = region_obj.expiration_time
 
         cache_dir = region_meta.dogpile_config_defaults['cache_dir']
         namespace_cache_dir = cache_dir
 
         # we default the namespace_cache_dir to our default cache dir.
         # however if this backend is configured with filename= param, we prioritize that
         # so all caches within that particular region, even those namespaced end up in the same path
         if region_obj.actual_backend.filename:
             namespace_cache_dir = os.path.dirname(region_obj.actual_backend.filename)
 
         if not os.path.isdir(namespace_cache_dir):
             os.makedirs(namespace_cache_dir)
         new_region = make_region(
             name=region_uid_name,
             function_key_generator=backend_key_generator(region_obj.actual_backend)
         )
 
         namespace_filename = os.path.join(
             namespace_cache_dir, f"{region_name}_{region_namespace}.cache_db")
         # special type that allows 1db per namespace
         new_region.configure(
             backend='dogpile.cache.rc.file_namespace',
             expiration_time=expiration_time,
             arguments={"filename": namespace_filename}
         )
 
         # create and save in region caches
         log.debug('configuring new region: %s', region_uid_name)
         region_obj = region_meta.dogpile_cache_regions[region_namespace] = new_region
 
     region_obj._default_namespace = region_namespace
     if use_async_runner:
         region_obj.async_creation_runner = async_creation_runner
     return region_obj
 
 
 def clear_cache_namespace(cache_region: str | RhodeCodeCacheRegion, cache_namespace_uid: str, method: str):
     from . import CLEAR_DELETE, CLEAR_INVALIDATE
 
     if not isinstance(cache_region, RhodeCodeCacheRegion):
         cache_region = get_or_create_region(cache_region, cache_namespace_uid)
     log.debug('clearing cache region: %s with method=%s', cache_region, method)
 
     num_affected_keys = None
 
     if method == CLEAR_INVALIDATE:
         # NOTE: The CacheRegion.invalidate() method’s default mode of
         # operation is to set a timestamp local to this CacheRegion in this Python process only.
         # It does not impact other Python processes or regions as the timestamp is only stored locally in memory.
         cache_region.invalidate(hard=True)
 
     if method == CLEAR_DELETE:
         cache_keys = cache_region.backend.list_keys(prefix=cache_namespace_uid)
         num_affected_keys = len(cache_keys)
         if num_affected_keys:
             cache_region.delete_multi(cache_keys)
 
     return num_affected_keys
 
 
 class ActiveRegionCache(object):
     def __init__(self, context, cache_data):
         self.context = context
         self.cache_data = cache_data
 
     def should_invalidate(self):
         return False
 
 
 class FreshRegionCache(object):
     def __init__(self, context, cache_data):
         self.context = context
         self.cache_data = cache_data
 
     def should_invalidate(self):
         return True
 
 
 class InvalidationContext(object):
     """
     usage::
 
         from rhodecode.lib import rc_cache
 
         cache_namespace_uid = CacheKey.SOME_NAMESPACE.format(1)
         region = rc_cache.get_or_create_region('cache_perms', cache_namespace_uid)
 
         @region.conditional_cache_on_arguments(namespace=cache_namespace_uid, condition=True)
         def heavy_compute(cache_name, param1, param2):
             print('COMPUTE {}, {}, {}'.format(cache_name, param1, param2))
 
         # invalidation namespace is shared namespace key for all process caches
         # we use it to send a global signal
         invalidation_namespace = 'repo_cache:1'
 
         inv_context_manager = rc_cache.InvalidationContext(
             uid=cache_namespace_uid, invalidation_namespace=invalidation_namespace)
         with inv_context_manager as invalidation_context:
             args = ('one', 'two')
             # re-compute and store cache if we get invalidate signal
             if invalidation_context.should_invalidate():
                 result = heavy_compute.refresh(*args)
             else:
                 result = heavy_compute(*args)
 
             compute_time = inv_context_manager.compute_time
             log.debug('result computed in %.4fs', compute_time)
 
         # To send global invalidation signal, simply run
         CacheKey.set_invalidate(invalidation_namespace)
 
     """
 
     def __repr__(self):
         return f'<InvalidationContext:{self.cache_key}[{self.uid}]>'
 
     def __init__(self, uid, invalidation_namespace='',
                  raise_exception=False, thread_scoped=None):
         self.uid = uid
         self.invalidation_namespace = invalidation_namespace
         self.raise_exception = raise_exception
         self.proc_id = rhodecode.CONFIG.get('instance_id') or 'DEFAULT'
         self.thread_id = 'global'
 
         if thread_scoped is None:
             # if we set "default" we can override this via .ini settings
             thread_scoped = rhodecode.ConfigGet().get_bool('cache_thread_scoped')
 
         # Append the thread id to the cache key if this invalidation context
         # should be scoped to the current thread.
         if thread_scoped is True:
             self.thread_id = threading.current_thread().ident
 
         self.cache_key = compute_key_from_params(uid)
         self.cache_key = 'proc:{}|thread:{}|params:{}'.format(
             self.proc_id, self.thread_id, self.cache_key)
-        self.proc_key = 'proc:{}'.format(self.proc_id)
+        self.proc_key = f'proc:{self.proc_id}'
         self.compute_time = 0
 
     def get_or_create_cache_obj(self, cache_type, invalidation_namespace=''):
         from rhodecode.model.db import CacheKey
 
         invalidation_namespace = invalidation_namespace or self.invalidation_namespace
         # fetch all cache keys for this namespace and convert them to a map to find if we
         # have specific cache_key object registered. We do this because we want to have
         # all consistent cache_state_uid for newly registered objects
         cache_obj_map = CacheKey.get_namespace_map(invalidation_namespace)
         cache_obj = cache_obj_map.get(self.cache_key)
         log.debug('Fetched cache obj %s using %s cache key.', cache_obj, self.cache_key)
 
         if not cache_obj:
             new_cache_args = invalidation_namespace
             first_cache_obj = next(iter(cache_obj_map.values())) if cache_obj_map else None
             cache_state_uid = None
             if first_cache_obj:
                 cache_state_uid = first_cache_obj.cache_state_uid
             cache_obj = CacheKey(self.cache_key, cache_args=new_cache_args,
                                  cache_state_uid=cache_state_uid)
             cache_key_meta.cache_keys_by_pid.add(self.proc_key)
 
         return cache_obj
 
     def __enter__(self):
         """
         Test if current object is valid, and return CacheRegion function
         that does invalidation and calculation
         """
         log.debug('Entering cache invalidation check context: %s', self.invalidation_namespace)
         # register or get a new key based on uid
         self.cache_obj = self.get_or_create_cache_obj(cache_type=self.uid)
         cache_data = self.cache_obj.get_dict()
         self._start_time = time.time()
         if self.cache_obj.cache_active:
             # means our cache obj is existing and marked as it's
             # cache is not outdated, we return ActiveRegionCache
             self.skip_cache_active_change = True
 
             return ActiveRegionCache(context=self, cache_data=cache_data)
 
         # the key is either not existing or set to False, we return
         # the real invalidator which re-computes value. We additionally set
         # the flag to actually update the Database objects
         self.skip_cache_active_change = False
         return FreshRegionCache(context=self, cache_data=cache_data)
 
     def __exit__(self, exc_type, exc_val, exc_tb):
         from rhodecode.model.db import IntegrityError, Session
 
         # save compute time
         self.compute_time = time.time() - self._start_time
 
         if self.skip_cache_active_change:
             return
 
         try:
             self.cache_obj.cache_active = True
             Session().add(self.cache_obj)
             Session().commit()
         except IntegrityError:
             # if we catch integrity error, it means we inserted this object
             # assumption is that's really an edge race-condition case and
             # it's safe is to skip it
             Session().rollback()
         except Exception:
             log.exception('Failed to commit on cache key update')
             Session().rollback()
             if self.raise_exception:
                 raise

rhodecode/lib/rc_commands/add_artifact.py

@@ -1,107 +1,105 b''
-
-
 # Copyright (C) 2016-2023 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 sys
 import logging
 
 import click
 
 from rhodecode.lib.pyramid_utils import bootstrap
 from rhodecode.model.db import Session, User, Repository
 from rhodecode.model.user import UserModel
 from rhodecode.apps.file_store import utils as store_utils
 
 log = logging.getLogger(__name__)
 
 
 @click.command()
 @click.argument('ini_path', type=click.Path(exists=True))
 @click.option(
     '--filename',
     required=True,
     help='Filename for artifact.')
 @click.option(
     '--file-path',
     required=True,
     type=click.Path(exists=True, dir_okay=False, readable=True),
     help='Path to a file to be added as artifact')
 @click.option(
     '--repo-id',
     required=True,
     type=int,
     help='ID of repository to add this artifact to.')
 @click.option(
     '--user-id',
     default=None,
     type=int,
     help='User ID for creator of artifact. '
          'Default would be first super admin.')
 @click.option(
     '--description',
     default=None,
     type=str,
     help='Add description to this artifact')
 def main(ini_path, filename, file_path, repo_id, user_id, description):
     return command(ini_path, filename, file_path, repo_id, user_id, description)
 
 
 def command(ini_path, filename, file_path, repo_id, user_id, description):
     with bootstrap(ini_path, env={'RC_CMD_SETUP_RC': '1'}) as env:
         try:
             from rc_ee.api.views.store_api import _store_file
         except ImportError:
             click.secho('ERROR: Unable to import store_api. '
                         'store_api is only available in EE edition of RhodeCode',
                         fg='red')
             sys.exit(-1)
 
         request = env['request']
 
         repo = Repository.get(repo_id)
         if not repo:
-            click.secho('ERROR: Unable to find repository with id `{}`'.format(repo_id),
+            click.secho(f'ERROR: Unable to find repository with id `{repo_id}`',
                         fg='red')
             sys.exit(-1)
 
         # if we don't give user, or it's "DEFAULT" user we pick super-admin
         if user_id is not None or user_id == 1:
             db_user = User.get(user_id)
         else:
             db_user = User.get_first_super_admin()
 
         if not db_user:
-            click.secho('ERROR: Unable to find user with id/username `{}`'.format(user_id),
+            click.secho(f'ERROR: Unable to find user with id/username `{user_id}`',
                         fg='red')
             sys.exit(-1)
 
         auth_user = db_user.AuthUser(ip_addr='127.0.0.1')
 
         storage = store_utils.get_file_storage(request.registry.settings)
 
         with open(file_path, 'rb') as f:
-            click.secho('Adding new artifact from path: `{}`'.format(file_path),
+            click.secho(f'Adding new artifact from path: `{file_path}`',
                         fg='green')
 
             file_data = _store_file(
                 storage, auth_user, filename, content=None, check_acl=True,
                 file_obj=f, description=description,
                 scope_repo_id=repo.repo_id)
-            click.secho('File Data: {}'.format(file_data),
+            click.secho(f'File Data: {file_data}',
                         fg='green')

rhodecode/lib/rc_commands/ishell.py

@@ -1,59 +1,57 b''
-
-
 # Copyright (C) 2016-2023 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 click
 
 from rhodecode.lib.pyramid_utils import bootstrap
 import pyramid.paster
 
 # imports, used in ipython shell
 import os
 import sys
 import time
 import shutil
 import datetime
 from rhodecode.model.db import *
 
 welcome_banner = """Welcome to RhodeCode iShell.
 Type `exit` to exit the shell.
 iShell is interactive shell to interact directly with the
 internal RhodeCode APIs. You can rescue your lost password,
 or reset some user/system settings.
 """
 
 
 @click.command()
 @click.argument('ini_path', type=click.Path(exists=True))
 def main(ini_path):
     pyramid.paster.setup_logging(ini_path)
 
     with bootstrap(ini_path) as env:
 
         try:
             from IPython import embed
             from traitlets.config import Config
             cfg = Config()
             cfg.InteractiveShellEmbed.confirm_exit = False
             embed(config=cfg, banner1=welcome_banner)
         except ImportError:
             print('ipython installation required for ishell')
             sys.exit(-1)
 
 

rhodecode/lib/rc_commands/setup_rc.py

@@ -1,125 +1,123 b''
-
-
 # Copyright (C) 2016-2023 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 logging
 
 import click
 import pyramid.paster
 
 from rhodecode.lib.pyramid_utils import bootstrap, get_app_config
 from rhodecode.lib.db_manage import DbManage
 from rhodecode.lib.utils2 import get_encryption_key
 from rhodecode.model.db import Session
 
 
 log = logging.getLogger(__name__)
 
 
 @click.command()
 @click.argument('ini_path', type=click.Path(exists=True))
 @click.option(
     '--force-yes/--force-no', default=None,
     help="Force yes/no to every question")
 @click.option(
     '--user',
     default=None,
     help='Initial super-admin username')
 @click.option(
     '--email',
     default=None,
     help='Initial super-admin email address.')
 @click.option(
     '--password',
     default=None,
     help='Initial super-admin password. Minimum 6 chars.')
 @click.option(
     '--api-key',
     help='Initial API key for the admin user')
 @click.option(
     '--repos',
     default=None,
     help='Absolute path to storage location. This is storage for all '
          'existing and future repositories, and repository groups.')
 @click.option(
     '--public-access/--no-public-access',
     default=None,
     help='Enable public access on this installation. '
          'Default is public access enabled.')
 @click.option(
     '--skip-existing-db',
     default=False,
     is_flag=True,
     help='Do not destroy and re-initialize the database if it already exist.')
 @click.option(
     '--apply-license-key',
     default=False,
     is_flag=True,
     help='Get the license key from a license file or ENV and apply during DB creation.')
 def main(ini_path, force_yes, user, email, password, api_key, repos,
          public_access, skip_existing_db, apply_license_key):
     return command(ini_path, force_yes, user, email, password, api_key,
                    repos, public_access, skip_existing_db, apply_license_key)
 
 
 def command(ini_path, force_yes, user, email, password, api_key, repos,
             public_access, skip_existing_db, apply_license_key):
     # mapping of old parameters to new CLI from click
     options = dict(
         username=user,
         email=email,
         password=password,
         api_key=api_key,
         repos_location=repos,
         force_ask=force_yes,
         public_access=public_access
     )
     pyramid.paster.setup_logging(ini_path)
 
     config = get_app_config(ini_path)
 
     db_uri = config['sqlalchemy.db1.url']
     enc_key = get_encryption_key(config)
     dbmanage = DbManage(log_sql=True, dbconf=db_uri, root='.',
                         tests=False, cli_args=options, enc_key=enc_key)
     if skip_existing_db and dbmanage.db_exists():
         return
 
     dbmanage.create_tables(override=True)
     dbmanage.set_db_version()
     opts = dbmanage.config_prompt(None)
     dbmanage.create_settings(opts)
     dbmanage.create_default_user()
     dbmanage.create_admin_and_prompt()
     dbmanage.create_permissions()
     dbmanage.populate_default_permissions()
     if apply_license_key:
         try:
             from rc_license.models import apply_trial_license_if_missing
             apply_trial_license_if_missing(force=True)
         except ImportError:
             pass
 
     Session().commit()
 
     with bootstrap(ini_path, env={'RC_CMD_SETUP_RC': '1'}) as env:
         msg = 'Successfully initialized database, schema and default data.'
         print()
         print('*' * len(msg))
         print(msg.upper())
         print('*' * len(msg))

rhodecode/lib/rc_commands/upgrade_db.py

@@ -1,56 +1,54 b''
-
-
 # Copyright (C) 2016-2023 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 logging
 
 import click
 import pyramid.paster
 
 from rhodecode.lib.pyramid_utils import bootstrap
 from rhodecode.lib.db_manage import DbManage
 from rhodecode.lib.utils2 import safe_int, get_encryption_key
 
 log = logging.getLogger(__name__)
 
 
 @click.command()
 @click.argument('ini_path', type=click.Path(exists=True))
 @click.option('--force-yes/--force-no', default=None,
               help="Force yes/no to every question")
 @click.option('--force-version', default=None,
               help="Force upgrade from version")
 def main(ini_path, force_yes, force_version):
     return command(ini_path, force_yes, force_version)
 
 
 def command(ini_path, force_yes, force_version):
     pyramid.paster.setup_logging(ini_path)
 
     with bootstrap(ini_path, env={'RC_CMD_UPGRADE_DB': '1'}) as env:
         config = env['registry'].settings
         db_uri = config['sqlalchemy.db1.url']
         enc_key = get_encryption_key(config)
         options = {}
         if force_yes is not None:
             options['force_ask'] = force_yes
         dbmanage = DbManage(
             log_sql=True, dbconf=db_uri, root='.', tests=False,
             cli_args=options, enc_key=enc_key)
         dbmanage.upgrade(version=safe_int(force_version))

rhodecode/lib/repo_maintenance.py

@@ -1,193 +1,191 b''
-
-
 # Copyright (C) 2017-2023 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 logging
 
 log = logging.getLogger(__name__)
 
 
 class MaintenanceTask(object):
     human_name = 'undefined'
 
     def __init__(self, db_repo):
         self.db_repo = db_repo
 
     def run(self):
         """Execute task and return task human value"""
         raise NotImplementedError()
 
 
 class GitGC(MaintenanceTask):
     human_name = 'GIT Garbage collect'
 
     def _count_objects(self, repo):
         stdout, stderr = repo.run_git_command(
             ['count-objects', '-v'], fail_on_stderr=False)
 
         errors = ' '
         objects = ' '.join(stdout.splitlines())
 
         if stderr:
             errors = '\nSTD ERR:' + '\n'.join(stderr.splitlines())
         return objects + errors
 
     def run(self):
         output = []
         instance = self.db_repo.scm_instance()
 
         objects_before = self._count_objects(instance)
 
         log.debug('GIT objects:%s', objects_before)
         cmd = ['gc', '--aggressive']
         stdout, stderr = instance.run_git_command(cmd, fail_on_stderr=False)
 
         out = 'executed {}'.format(' '.join(cmd))
         output.append(out)
 
         out = ''
         if stderr:
             out += ''.join(stderr.splitlines())
 
         if stdout:
             out += ''.join(stdout.splitlines())
 
         if out:
             output.append(out)
 
         objects_after = self._count_objects(instance)
         log.debug('GIT objects:%s', objects_after)
         output.append('objects before :' + objects_before)
         output.append('objects after  :' + objects_after)
 
         return '\n'.join(output)
 
 
 class GitFSCK(MaintenanceTask):
     human_name = 'GIT FSCK'
 
     def run(self):
         output = []
         instance = self.db_repo.scm_instance()
 
         cmd = ['fsck', '--full']
         stdout, stderr = instance.run_git_command(cmd, fail_on_stderr=False)
 
         out = 'executed {}'.format(' '.join(cmd))
         output.append(out)
 
         out = ''
         if stderr:
             out += ''.join(stderr.splitlines())
 
         if stdout:
             out += ''.join(stdout.splitlines())
 
         if out:
             output.append(out)
 
         return '\n'.join(output)
 
 
 class GitRepack(MaintenanceTask):
     human_name = 'GIT Repack'
 
     def run(self):
         output = []
         instance = self.db_repo.scm_instance()
         cmd = ['repack', '-a', '-d',
                '--window-memory', '10m', '--max-pack-size', '100m']
         stdout, stderr = instance.run_git_command(cmd, fail_on_stderr=False)
 
         out = 'executed {}'.format(' '.join(cmd))
         output.append(out)
         out = ''
 
         if stderr:
             out += ''.join(stderr.splitlines())
 
         if stdout:
             out += ''.join(stdout.splitlines())
 
         if out:
             output.append(out)
 
         return '\n'.join(output)
 
 
 class HGVerify(MaintenanceTask):
     human_name = 'HG Verify repo'
 
     def run(self):
         instance = self.db_repo.scm_instance()
         res = instance.verify()
         return res
 
 
 class HGUpdateCaches(MaintenanceTask):
     human_name = 'HG update caches'
 
     def run(self):
         instance = self.db_repo.scm_instance()
         res = instance.hg_update_cache()
         return res
 
 
 class HGRebuildFnCaches(MaintenanceTask):
     human_name = 'HG rebuild fn caches'
 
     def run(self):
         instance = self.db_repo.scm_instance()
         res = instance.hg_rebuild_fn_cache()
         return res
 
 
 class SVNVerify(MaintenanceTask):
     human_name = 'SVN Verify repo'
 
     def run(self):
         instance = self.db_repo.scm_instance()
         res = instance.verify()
         return res
 
 
 class RepoMaintenance(object):
     """
     Performs maintenance of repository based on it's type
     """
     tasks = {
         'hg': [HGVerify, HGUpdateCaches, HGRebuildFnCaches],
         'git': [GitFSCK, GitGC, GitRepack],
         'svn': [SVNVerify],
     }
 
     def get_tasks_for_repo(self, db_repo):
         """
         fetches human names of tasks pending for execution for given type of repo
         """
         tasks = []
         for task in self.tasks[db_repo.repo_type]:
             tasks.append(task.human_name)
         return tasks
 
     def execute(self, db_repo):
         executed_tasks = []
         for task in self.tasks[db_repo.repo_type]:
             output = task.human_name + ':\n' + task(db_repo).run() + '\n--\n'
             executed_tasks.append(output)
         return executed_tasks

rhodecode/lib/request.py

@@ -1,109 +1,107 b''
-
-
 # Copyright (C) 2017-2023 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/
 
 from uuid import uuid4
 import pyramid.testing
 from pyramid.decorator import reify
 from pyramid.request import Request as _Request
 from rhodecode.lib.type_utils import StrictAttributeDict
 
 
 class TemplateArgs(StrictAttributeDict):
     pass
 
 
 # Base Class with DummyMethods, testing / CLI scripts
 class RequestBase(object):
     _req_id_bucket = list()
     _call_context = TemplateArgs()
     _call_context.visual = TemplateArgs()
     _call_context.visual.show_sha_length = 12
     _call_context.visual.show_revision_number = True
 
     @reify
     def req_id(self):
         return str(uuid4())
 
     @property
     def req_id_bucket(self):
         return self._req_id_bucket
 
     def req_id_records_init(self):
         self._req_id_bucket = list()
 
     def translate(self, *args, **kwargs):
         raise NotImplementedError()
 
     def plularize(self, *args, **kwargs):
         raise NotImplementedError()
 
     def get_partial_renderer(self, tmpl_name):
         raise NotImplementedError()
 
     @property
     def call_context(self):
         return self._call_context
 
     def set_call_context(self, new_context):
         self._call_context = new_context
 
 
 # for thin non-web/cli etc
 class ThinRequest(RequestBase, pyramid.testing.DummyRequest):
 
     def translate(self, msg):
         return msg
 
     def plularize(self, singular, plural, n):
         return singular
 
     def get_partial_renderer(self, tmpl_name):
         from rhodecode.lib.partial_renderer import get_partial_renderer
         return get_partial_renderer(request=self, tmpl_name=tmpl_name)
 
 
 # for real-web-based
 class RealRequest(RequestBase, _Request):
     def get_partial_renderer(self, tmpl_name):
         from rhodecode.lib.partial_renderer import get_partial_renderer
         return get_partial_renderer(request=self, tmpl_name=tmpl_name)
 
     def request_count(self):
         from rhodecode.lib.request_counter import get_request_counter
         return get_request_counter()
 
     def plularize(self, *args, **kwargs):
         return self.localizer.pluralize(*args, **kwargs)
 
     def translate(self, *args, **kwargs):
         localizer = self.localizer
         from rhodecode.translation import _ as tsf
 
         def auto_translate(*_args, **_kwargs):
             return localizer.translate(tsf(*_args, **_kwargs))
 
         return auto_translate(*args, **kwargs)
 
 
 class Request(RealRequest):
     """
     This is the main request object used in web-context
     """
     pass

rhodecode/lib/request_counter.py

@@ -1,27 +1,25 b''
-
-
 # Copyright (C) 2017-2023 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/
 
 counter = 0
 
 
 def get_request_counter():
     global counter
     counter += 1
     return counter

rhodecode/lib/statsd_client.py

@@ -1,49 +1,49 b''
 from rhodecode.lib._vendor.statsd import client_from_config
 
 
 class StatsdClientNotInitialised(Exception):
     pass
 
 
 class _Singleton(type):
     """A metaclass that creates a Singleton base class when called."""
 
     _instances = {}
 
     def __call__(cls, *args, **kwargs):
         if cls not in cls._instances:
-            cls._instances[cls] = super(_Singleton, cls).__call__(*args, **kwargs)
+            cls._instances[cls] = super().__call__(*args, **kwargs)
         return cls._instances[cls]
 
 
 class Singleton(_Singleton("SingletonMeta", (object,), {})):
     pass
 
 
 class StatsdClientClass(Singleton):
     setup_run = False
     statsd_client = None
     statsd = None
 
     def __getattribute__(self, name):
 
         if name.startswith("statsd"):
             if self.setup_run:
-                return super(StatsdClientClass, self).__getattribute__(name)
+                return super().__getattribute__(name)
             else:
                 return None
                 #raise StatsdClientNotInitialised("requested key was %s" % name)
 
-        return super(StatsdClientClass, self).__getattribute__(name)
+        return super().__getattribute__(name)
 
     def setup(self, settings):
         """
         Initialize the client
         """
         statsd = client_from_config(settings)
         self.statsd = statsd
         self.statsd_client = statsd
         self.setup_run = True
 
 
 StatsdClient = StatsdClientClass()

rhodecode/lib/str_utils.py

@@ -1,170 +1,169 b''
-
 # Copyright (C) 2011-2023 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 typing
 import base64
 import logging
 from unidecode import unidecode
 
 import rhodecode
 from rhodecode.lib.type_utils import aslist
 
 
 log = logging.getLogger(__name__)
 
 
 def safe_int(val, default=None) -> int:
     """
     Returns int() of val if val is not convertable to int use default
     instead
 
     :param val:
     :param default:
     """
 
     try:
         val = int(val)
     except (ValueError, TypeError):
         val = default
 
     return val
 
 
 def safe_float(val, default=None) -> float:
     """
     Returns float() of val if val is not convertable to float use default
     instead
 
     :param val:
     :param default:
     """
 
     try:
         val = float(val)
     except (ValueError, TypeError):
         val = default
 
     return val
 
 
 def base64_to_str(text) -> str:
     return safe_str(base64.encodebytes(safe_bytes(text))).strip()
 
 
-def get_default_encodings() -> typing.List[str]:
+def get_default_encodings() -> list[str]:
     return aslist(rhodecode.CONFIG.get('default_encoding', 'utf8'), sep=',')
 
 
 DEFAULT_ENCODINGS = get_default_encodings()
 
 
 def safe_str(str_, to_encoding=None) -> str:
     """
     safe str function. Does few trick to turn unicode_ into string
 
     :param str_: str to encode
     :param to_encoding: encode to this type UTF8 default
     """
     if isinstance(str_, str):
         return str_
 
     # if it's bytes cast to str
     if not isinstance(str_, bytes):
         return str(str_)
 
     to_encoding = to_encoding or DEFAULT_ENCODINGS
     if not isinstance(to_encoding, (list, tuple)):
         to_encoding = [to_encoding]
 
     for enc in to_encoding:
         try:
             return str(str_, enc)
         except UnicodeDecodeError:
             pass
 
     return str(str_, to_encoding[0], 'replace')
 
 
 def safe_bytes(str_, from_encoding=None) -> bytes:
     """
     safe bytes function. Does few trick to turn str_ into bytes string:
 
     :param str_: string to decode
     :param from_encoding: encode from this type UTF8 default
     """
     if isinstance(str_, bytes):
         return str_
 
     if not isinstance(str_, str):
         raise ValueError(f'safe_bytes cannot convert other types than str: got: {type(str_)}')
 
     from_encoding = from_encoding or get_default_encodings()
     if not isinstance(from_encoding, (list, tuple)):
         from_encoding = [from_encoding]
 
     for enc in from_encoding:
         try:
             return str_.encode(enc)
         except UnicodeDecodeError:
             pass
 
     return str_.encode(from_encoding[0], 'replace')
 
 
 def ascii_bytes(str_, allow_bytes=False) -> bytes:
     """
     Simple conversion from str to bytes, with assumption that str_ is pure ASCII.
     Fails with UnicodeError on invalid input.
     This should be used where encoding and "safe" ambiguity should be avoided.
     Where strings already have been encoded in other ways but still are unicode
     string - for example to hex, base64, json, urlencoding, or are known to be
     identifiers.
     """
     if allow_bytes and isinstance(str_, bytes):
         return str_
 
     if not isinstance(str_, str):
         raise ValueError(f'ascii_bytes cannot convert other types than str: got: {type(str_)}')
     return str_.encode('ascii')
 
 
 def ascii_str(str_) -> str:
     """
     Simple conversion from bytes to str, with assumption that str_ is pure ASCII.
     Fails with UnicodeError on invalid input.
     This should be used where encoding and "safe" ambiguity should be avoided.
     Where strings are encoded but also in other ways are known to be ASCII, and
     where a unicode string is wanted without caring about encoding. For example
     to hex, base64, urlencoding, or are known to be identifiers.
     """
 
     if not isinstance(str_, bytes):
         raise ValueError(f'ascii_str cannot convert other types than bytes: got: {type(str_)}')
     return str_.decode('ascii')
 
 
 def convert_special_chars(str_) -> str:
     """
         trie to replace non-ascii letters to their ascii representation eg::
 
             `żołw` converts into `zolw`
     """
     value = safe_str(str_)
     converted_value = unidecode(value)
     return converted_value

rhodecode/lib/string_renderer.py

@@ -1,34 +1,33 b''
-
 # Copyright (C) 2010-2023 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/
 
 
 def html(info):
     """
     Custom string as html content_type renderer for pyramid
     """
     def _render(value, system):
         request = system.get('request')
         if request is not None:
             response = request.response
             ct = response.content_type
             if ct == response.default_content_type:
                 response.content_type = 'text/html'
         return value
 
     return _render

rhodecode/lib/system_info.py

@@ -1,848 +1,846 b''
-
-
 # Copyright (C) 2017-2023 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 os
 import sys
 import time
 import platform
 import collections
 import psutil
 from functools import wraps
 
 import pkg_resources
 import logging
 import resource
 
 import configparser
 
 from rhodecode.lib.str_utils import safe_str
 
 log = logging.getLogger(__name__)
 
 
 _NA = 'NOT AVAILABLE'
 _NA_FLOAT = 0.0
 
 STATE_OK = 'ok'
 STATE_ERR = 'error'
 STATE_WARN = 'warning'
 
 STATE_OK_DEFAULT = {'message': '', 'type': STATE_OK}
 
 
 registered_helpers = {}
 
 
 def register_sysinfo(func):
     """
     @register_helper
     def db_check():
         pass
 
     db_check == registered_helpers['db_check']
     """
     global registered_helpers
     registered_helpers[func.__name__] = func
 
     @wraps(func)
     def _wrapper(*args, **kwargs):
         return func(*args, **kwargs)
     return _wrapper
 
 
 # HELPERS
 def percentage(part: (int, float), whole: (int, float)):
     whole = float(whole)
     if whole > 0:
         return round(100 * float(part) / whole, 1)
     return 0.0
 
 
 def get_storage_size(storage_path):
     sizes = []
     for file_ in os.listdir(storage_path):
         storage_file = os.path.join(storage_path, file_)
         if os.path.isfile(storage_file):
             try:
                 sizes.append(os.path.getsize(storage_file))
             except OSError:
                 log.exception('Failed to get size of storage file %s', storage_file)
                 pass
 
     return sum(sizes)
 
 
 def get_resource(resource_type):
     try:
         return resource.getrlimit(resource_type)
     except Exception:
         return 'NOT_SUPPORTED'
 
 
 def get_cert_path(ini_path):
     default = '/etc/ssl/certs/ca-certificates.crt'
     control_ca_bundle = os.path.join(
         os.path.dirname(os.path.dirname(os.path.dirname(os.path.abspath(ini_path)))),
         '.rccontrol-profile/etc/ca-bundle.crt')
     if os.path.isfile(control_ca_bundle):
         default = control_ca_bundle
 
     return default
 
 
 class SysInfoRes(object):
     def __init__(self, value, state=None, human_value=None):
         self.value = value
         self.state = state or STATE_OK_DEFAULT
         self.human_value = human_value or value
 
     def __json__(self):
         return {
             'value': self.value,
             'state': self.state,
             'human_value': self.human_value,
         }
 
     def get_value(self):
         return self.__json__()
 
     def __str__(self):
-        return '<SysInfoRes({})>'.format(self.__json__())
+        return f'<SysInfoRes({self.__json__()})>'
 
 
 class SysInfo(object):
 
     def __init__(self, func_name, **kwargs):
         self.function_name = func_name
         self.value = _NA
         self.state = None
         self.kwargs = kwargs or {}
 
     def __call__(self):
         computed = self.compute(**self.kwargs)
         if not isinstance(computed, SysInfoRes):
             raise ValueError(
                 'computed value for {} is not instance of '
                 '{}, got {} instead'.format(
                     self.function_name, SysInfoRes, type(computed)))
         return computed.__json__()
 
     def __str__(self):
-        return '<SysInfo({})>'.format(self.function_name)
+        return f'<SysInfo({self.function_name})>'
 
     def compute(self, **kwargs):
         return self.function_name(**kwargs)
 
 
 # SysInfo functions
 @register_sysinfo
 def python_info():
     value = dict(version=f'{platform.python_version()}:{platform.python_implementation()}',
                  executable=sys.executable)
     return SysInfoRes(value=value)
 
 
 @register_sysinfo
 def py_modules():
     mods = dict([(p.project_name, {'version': p.version, 'location': p.location})
                  for p in pkg_resources.working_set])
 
     value = sorted(mods.items(), key=lambda k: k[0].lower())
     return SysInfoRes(value=value)
 
 
 @register_sysinfo
 def platform_type():
     from rhodecode.lib.utils import generate_platform_uuid
 
     value = dict(
         name=safe_str(platform.platform()),
         uuid=generate_platform_uuid()
     )
     return SysInfoRes(value=value)
 
 
 @register_sysinfo
 def locale_info():
     import locale
 
     def safe_get_locale(locale_name):
         try:
             locale.getlocale(locale_name)
         except TypeError:
             return f'FAILED_LOCALE_GET:{locale_name}'
 
     value = dict(
         locale_default=locale.getlocale(),
         locale_lc_all=safe_get_locale(locale.LC_ALL),
         locale_lc_ctype=safe_get_locale(locale.LC_CTYPE),
         lang_env=os.environ.get('LANG'),
         lc_all_env=os.environ.get('LC_ALL'),
         local_archive_env=os.environ.get('LOCALE_ARCHIVE'),
     )
     human_value = \
         f"LANG: {value['lang_env']}, \
           locale LC_ALL: {value['locale_lc_all']},  \
           locale LC_CTYPE: {value['locale_lc_ctype']},  \
           Default locales: {value['locale_default']}"
 
     return SysInfoRes(value=value, human_value=human_value)
 
 
 @register_sysinfo
 def ulimit_info():
     data = collections.OrderedDict([
         ('cpu time (seconds)', get_resource(resource.RLIMIT_CPU)),
         ('file size', get_resource(resource.RLIMIT_FSIZE)),
         ('stack size', get_resource(resource.RLIMIT_STACK)),
         ('core file size', get_resource(resource.RLIMIT_CORE)),
         ('address space size', get_resource(resource.RLIMIT_AS)),
         ('locked in mem size', get_resource(resource.RLIMIT_MEMLOCK)),
         ('heap size', get_resource(resource.RLIMIT_DATA)),
         ('rss size', get_resource(resource.RLIMIT_RSS)),
         ('number of processes', get_resource(resource.RLIMIT_NPROC)),
         ('open files', get_resource(resource.RLIMIT_NOFILE)),
     ])
 
     text = ', '.join(f'{k}:{v}' for k, v in data.items())
 
     value = {
         'limits': data,
         'text': text,
     }
     return SysInfoRes(value=value)
 
 
 @register_sysinfo
 def uptime():
     from rhodecode.lib.helpers import age, time_to_datetime
     from rhodecode.translation import TranslationString
 
     value = dict(boot_time=0, uptime=0, text='')
     state = STATE_OK_DEFAULT
 
     boot_time = psutil.boot_time()
     value['boot_time'] = boot_time
     value['uptime'] = time.time() - boot_time
 
     date_or_age = age(time_to_datetime(boot_time))
     if isinstance(date_or_age, TranslationString):
         date_or_age = date_or_age.interpolate()
 
     human_value = value.copy()
     human_value['boot_time'] = time_to_datetime(boot_time)
     human_value['uptime'] = age(time_to_datetime(boot_time), show_suffix=False)
 
-    human_value['text'] = 'Server started {}'.format(date_or_age)
+    human_value['text'] = f'Server started {date_or_age}'
     return SysInfoRes(value=value, human_value=human_value)
 
 
 @register_sysinfo
 def memory():
     from rhodecode.lib.helpers import format_byte_size_binary
     value = dict(available=0, used=0, used_real=0, cached=0, percent=0,
                  percent_used=0, free=0, inactive=0, active=0, shared=0,
                  total=0, buffers=0, text='')
 
     state = STATE_OK_DEFAULT
 
     value.update(dict(psutil.virtual_memory()._asdict()))
     value['used_real'] = value['total'] - value['available']
     value['percent_used'] = psutil._common.usage_percent(
         value['used_real'], value['total'], 1)
 
     human_value = value.copy()
-    human_value['text'] = '%s/%s, %s%% used' % (
+    human_value['text'] = '{}/{}, {}% used'.format(
         format_byte_size_binary(value['used_real']),
         format_byte_size_binary(value['total']),
-        value['percent_used'],)
+        value['percent_used'])
 
     keys = list(value.keys())[::]
     keys.pop(keys.index('percent'))
     keys.pop(keys.index('percent_used'))
     keys.pop(keys.index('text'))
     for k in keys:
         human_value[k] = format_byte_size_binary(value[k])
 
     if state['type'] == STATE_OK and value['percent_used'] > 90:
         msg = 'Critical: your available RAM memory is very low.'
         state = {'message': msg, 'type': STATE_ERR}
 
     elif state['type'] == STATE_OK and value['percent_used'] > 70:
         msg = 'Warning: your available RAM memory is running low.'
         state = {'message': msg, 'type': STATE_WARN}
 
     return SysInfoRes(value=value, state=state, human_value=human_value)
 
 
 @register_sysinfo
 def machine_load():
     value = {'1_min': _NA_FLOAT, '5_min': _NA_FLOAT, '15_min': _NA_FLOAT, 'text': ''}
     state = STATE_OK_DEFAULT
 
     # load averages
     if hasattr(psutil.os, 'getloadavg'):
         value.update(dict(
             list(zip(['1_min', '5_min', '15_min'], psutil.os.getloadavg()))
         ))
 
     human_value = value.copy()
     human_value['text'] = '1min: {}, 5min: {}, 15min: {}'.format(
         value['1_min'], value['5_min'], value['15_min'])
 
     if state['type'] == STATE_OK and value['15_min'] > 5.0:
         msg = 'Warning: your machine load is very high.'
         state = {'message': msg, 'type': STATE_WARN}
 
     return SysInfoRes(value=value, state=state, human_value=human_value)
 
 
 @register_sysinfo
 def cpu():
     value = {'cpu': 0, 'cpu_count': 0, 'cpu_usage': []}
     state = STATE_OK_DEFAULT
 
     value['cpu'] = psutil.cpu_percent(0.5)
     value['cpu_usage'] = psutil.cpu_percent(0.5, percpu=True)
     value['cpu_count'] = psutil.cpu_count()
 
     human_value = value.copy()
     human_value['text'] = '{} cores at {} %'.format(
         value['cpu_count'], value['cpu'])
 
     return SysInfoRes(value=value, state=state, human_value=human_value)
 
 
 @register_sysinfo
 def storage():
     from rhodecode.lib.helpers import format_byte_size_binary
     from rhodecode.model.settings import VcsSettingsModel
     path = VcsSettingsModel().get_repos_location()
 
     value = dict(percent=0, used=0, total=0, path=path, text='')
     state = STATE_OK_DEFAULT
 
     try:
         value.update(dict(psutil.disk_usage(path)._asdict()))
     except Exception as e:
         log.exception('Failed to fetch disk info')
         state = {'message': str(e), 'type': STATE_ERR}
 
     human_value = value.copy()
     human_value['used'] = format_byte_size_binary(value['used'])
     human_value['total'] = format_byte_size_binary(value['total'])
     human_value['text'] = "{}/{}, {}% used".format(
         format_byte_size_binary(value['used']),
         format_byte_size_binary(value['total']),
         value['percent'])
 
     if state['type'] == STATE_OK and value['percent'] > 90:
         msg = 'Critical: your disk space is very low.'
         state = {'message': msg, 'type': STATE_ERR}
 
     elif state['type'] == STATE_OK and value['percent'] > 70:
         msg = 'Warning: your disk space is running low.'
         state = {'message': msg, 'type': STATE_WARN}
 
     return SysInfoRes(value=value, state=state, human_value=human_value)
 
 
 @register_sysinfo
 def storage_inodes():
     from rhodecode.model.settings import VcsSettingsModel
     path = VcsSettingsModel().get_repos_location()
 
     value = dict(percent=0.0, free=0, used=0, total=0, path=path, text='')
     state = STATE_OK_DEFAULT
 
     try:
         i_stat = os.statvfs(path)
         value['free'] = i_stat.f_ffree
         value['used'] = i_stat.f_files-i_stat.f_favail
         value['total'] = i_stat.f_files
         value['percent'] = percentage(value['used'], value['total'])
     except Exception as e:
         log.exception('Failed to fetch disk inodes info')
         state = {'message': str(e), 'type': STATE_ERR}
 
     human_value = value.copy()
     human_value['text'] = "{}/{}, {}% used".format(
         value['used'], value['total'], value['percent'])
 
     if state['type'] == STATE_OK and value['percent'] > 90:
         msg = 'Critical: your disk free inodes are very low.'
         state = {'message': msg, 'type': STATE_ERR}
 
     elif state['type'] == STATE_OK and value['percent'] > 70:
         msg = 'Warning: your disk free inodes are running low.'
         state = {'message': msg, 'type': STATE_WARN}
 
     return SysInfoRes(value=value, state=state, human_value=human_value)
 
 
 @register_sysinfo
 def storage_archives():
     import rhodecode
     from rhodecode.lib.utils import safe_str
     from rhodecode.lib.helpers import format_byte_size_binary
 
     msg = 'Archive cache storage is controlled by ' \
           'archive_cache.store_dir=/path/to/cache option in the .ini file'
     path = safe_str(rhodecode.CONFIG.get('archive_cache.store_dir', msg))
 
     value = dict(percent=0, used=0, total=0, items=0, path=path, text='')
     state = STATE_OK_DEFAULT
     try:
         items_count = 0
         used = 0
         for root, dirs, files in os.walk(path):
             if root == path:
                 items_count = len(dirs)
 
             for f in files:
                 try:
                     used += os.path.getsize(os.path.join(root, f))
                 except OSError:
                     pass
         value.update({
             'percent': 100,
             'used': used,
             'total': used,
             'items': items_count
         })
 
     except Exception as e:
         log.exception('failed to fetch archive cache storage')
         state = {'message': str(e), 'type': STATE_ERR}
 
     human_value = value.copy()
     human_value['used'] = format_byte_size_binary(value['used'])
     human_value['total'] = format_byte_size_binary(value['total'])
     human_value['text'] = "{} ({} items)".format(
         human_value['used'], value['items'])
 
     return SysInfoRes(value=value, state=state, human_value=human_value)
 
 
 @register_sysinfo
 def storage_gist():
     from rhodecode.model.gist import GIST_STORE_LOC
     from rhodecode.model.settings import VcsSettingsModel
     from rhodecode.lib.utils import safe_str
     from rhodecode.lib.helpers import format_byte_size_binary
     path = safe_str(os.path.join(
         VcsSettingsModel().get_repos_location(), GIST_STORE_LOC))
 
     # gist storage
     value = dict(percent=0, used=0, total=0, items=0, path=path, text='')
     state = STATE_OK_DEFAULT
 
     try:
         items_count = 0
         used = 0
         for root, dirs, files in os.walk(path):
             if root == path:
                 items_count = len(dirs)
 
             for f in files:
                 try:
                     used += os.path.getsize(os.path.join(root, f))
                 except OSError:
                     pass
         value.update({
             'percent': 100,
             'used': used,
             'total': used,
             'items': items_count
         })
     except Exception as e:
         log.exception('failed to fetch gist storage items')
         state = {'message': str(e), 'type': STATE_ERR}
 
     human_value = value.copy()
     human_value['used'] = format_byte_size_binary(value['used'])
     human_value['total'] = format_byte_size_binary(value['total'])
     human_value['text'] = "{} ({} items)".format(
         human_value['used'], value['items'])
 
     return SysInfoRes(value=value, state=state, human_value=human_value)
 
 
 @register_sysinfo
 def storage_temp():
     import tempfile
     from rhodecode.lib.helpers import format_byte_size_binary
 
     path = tempfile.gettempdir()
     value = dict(percent=0, used=0, total=0, items=0, path=path, text='')
     state = STATE_OK_DEFAULT
 
     if not psutil:
         return SysInfoRes(value=value, state=state)
 
     try:
         value.update(dict(psutil.disk_usage(path)._asdict()))
     except Exception as e:
         log.exception('Failed to fetch temp dir info')
         state = {'message': str(e), 'type': STATE_ERR}
 
     human_value = value.copy()
     human_value['used'] = format_byte_size_binary(value['used'])
     human_value['total'] = format_byte_size_binary(value['total'])
     human_value['text'] = "{}/{}, {}% used".format(
         format_byte_size_binary(value['used']),
         format_byte_size_binary(value['total']),
         value['percent'])
 
     return SysInfoRes(value=value, state=state, human_value=human_value)
 
 
 @register_sysinfo
 def search_info():
     import rhodecode
     from rhodecode.lib.index import searcher_from_config
 
     backend = rhodecode.CONFIG.get('search.module', '')
     location = rhodecode.CONFIG.get('search.location', '')
 
     try:
         searcher = searcher_from_config(rhodecode.CONFIG)
         searcher = searcher.__class__.__name__
     except Exception:
         searcher = None
 
     value = dict(
         backend=backend, searcher=searcher, location=location, text='')
     state = STATE_OK_DEFAULT
 
     human_value = value.copy()
     human_value['text'] = "backend:`{}`".format(human_value['backend'])
 
     return SysInfoRes(value=value, state=state, human_value=human_value)
 
 
 @register_sysinfo
 def git_info():
     from rhodecode.lib.vcs.backends import git
     state = STATE_OK_DEFAULT
     value = human_value = ''
     try:
         value = git.discover_git_version(raise_on_exc=True)
-        human_value = 'version reported from VCSServer: {}'.format(value)
+        human_value = f'version reported from VCSServer: {value}'
     except Exception as e:
         state = {'message': str(e), 'type': STATE_ERR}
 
     return SysInfoRes(value=value, state=state, human_value=human_value)
 
 
 @register_sysinfo
 def hg_info():
     from rhodecode.lib.vcs.backends import hg
     state = STATE_OK_DEFAULT
     value = human_value = ''
     try:
         value = hg.discover_hg_version(raise_on_exc=True)
-        human_value = 'version reported from VCSServer: {}'.format(value)
+        human_value = f'version reported from VCSServer: {value}'
     except Exception as e:
         state = {'message': str(e), 'type': STATE_ERR}
     return SysInfoRes(value=value, state=state, human_value=human_value)
 
 
 @register_sysinfo
 def svn_info():
     from rhodecode.lib.vcs.backends import svn
     state = STATE_OK_DEFAULT
     value = human_value = ''
     try:
         value = svn.discover_svn_version(raise_on_exc=True)
-        human_value = 'version reported from VCSServer: {}'.format(value)
+        human_value = f'version reported from VCSServer: {value}'
     except Exception as e:
         state = {'message': str(e), 'type': STATE_ERR}
     return SysInfoRes(value=value, state=state, human_value=human_value)
 
 
 @register_sysinfo
 def vcs_backends():
     import rhodecode
     value = rhodecode.CONFIG.get('vcs.backends')
     human_value = 'Enabled backends in order: {}'.format(','.join(value))
     return SysInfoRes(value=value, human_value=human_value)
 
 
 @register_sysinfo
 def vcs_server():
     import rhodecode
     from rhodecode.lib.vcs.backends import get_vcsserver_service_data
 
     server_url = rhodecode.CONFIG.get('vcs.server')
     enabled = rhodecode.CONFIG.get('vcs.server.enable')
     protocol = rhodecode.CONFIG.get('vcs.server.protocol') or 'http'
     state = STATE_OK_DEFAULT
     version = None
     workers = 0
 
     try:
         data = get_vcsserver_service_data()
         if data and 'version' in data:
             version = data['version']
 
         if data and 'config' in data:
             conf = data['config']
             workers = conf.get('workers', 'NOT AVAILABLE')
 
         connection = 'connected'
     except Exception as e:
         connection = 'failed'
         state = {'message': str(e), 'type': STATE_ERR}
 
     value = dict(
         url=server_url,
         enabled=enabled,
         protocol=protocol,
         connection=connection,
         version=version,
         text='',
     )
 
     human_value = value.copy()
     human_value['text'] = \
         '{url}@ver:{ver} via {mode} mode[workers:{workers}], connection:{conn}'.format(
             url=server_url, ver=version, workers=workers, mode=protocol,
             conn=connection)
 
     return SysInfoRes(value=value, state=state, human_value=human_value)
 
 
 @register_sysinfo
 def vcs_server_config():
     from rhodecode.lib.vcs.backends import get_vcsserver_service_data
     state = STATE_OK_DEFAULT
 
     value = {}
     try:
         data = get_vcsserver_service_data()
         value = data['app_config']
     except Exception as e:
         state = {'message': str(e), 'type': STATE_ERR}
 
     human_value = value.copy()
     human_value['text'] = 'VCS Server config'
 
     return SysInfoRes(value=value, state=state, human_value=human_value)
 
 
 @register_sysinfo
 def rhodecode_app_info():
     import rhodecode
     edition = rhodecode.CONFIG.get('rhodecode.edition')
 
     value = dict(
         rhodecode_version=rhodecode.__version__,
         rhodecode_lib_path=os.path.abspath(rhodecode.__file__),
         text=''
     )
     human_value = value.copy()
     human_value['text'] = 'RhodeCode {edition}, version {ver}'.format(
         edition=edition, ver=value['rhodecode_version']
     )
     return SysInfoRes(value=value, human_value=human_value)
 
 
 @register_sysinfo
 def rhodecode_config():
     import rhodecode
     path = rhodecode.CONFIG.get('__file__')
     rhodecode_ini_safe = rhodecode.CONFIG.copy()
     cert_path = get_cert_path(path)
 
     try:
         config = configparser.ConfigParser()
         config.read(path)
         parsed_ini = config
         if parsed_ini.has_section('server:main'):
             parsed_ini = dict(parsed_ini.items('server:main'))
     except Exception:
         log.exception('Failed to read .ini file for display')
         parsed_ini = {}
 
     rhodecode_ini_safe['server:main'] = parsed_ini
 
     blacklist = [
         'rhodecode_license_key',
         'routes.map',
         'sqlalchemy.db1.url',
         'channelstream.secret',
         'beaker.session.secret',
         'rhodecode.encrypted_values.secret',
         'rhodecode_auth_github_consumer_key',
         'rhodecode_auth_github_consumer_secret',
         'rhodecode_auth_google_consumer_key',
         'rhodecode_auth_google_consumer_secret',
         'rhodecode_auth_bitbucket_consumer_secret',
         'rhodecode_auth_bitbucket_consumer_key',
         'rhodecode_auth_twitter_consumer_secret',
         'rhodecode_auth_twitter_consumer_key',
 
         'rhodecode_auth_twitter_secret',
         'rhodecode_auth_github_secret',
         'rhodecode_auth_google_secret',
         'rhodecode_auth_bitbucket_secret',
 
         'appenlight.api_key',
         ('app_conf', 'sqlalchemy.db1.url')
     ]
     for k in blacklist:
         if isinstance(k, tuple):
             section, key = k
             if section in rhodecode_ini_safe:
                 rhodecode_ini_safe[section] = '**OBFUSCATED**'
         else:
             rhodecode_ini_safe.pop(k, None)
 
     # TODO: maybe put some CONFIG checks here ?
     return SysInfoRes(value={'config': rhodecode_ini_safe,
                              'path': path, 'cert_path': cert_path})
 
 
 @register_sysinfo
 def database_info():
     import rhodecode
     from sqlalchemy.engine import url as engine_url
     from rhodecode.model import meta
     from rhodecode.model.meta import Session
     from rhodecode.model.db import DbMigrateVersion
 
     state = STATE_OK_DEFAULT
 
     db_migrate = DbMigrateVersion.query().filter(
         DbMigrateVersion.repository_id == 'rhodecode_db_migrations').one()
 
     db_url_obj = engine_url.make_url(rhodecode.CONFIG['sqlalchemy.db1.url'])
 
     try:
         engine = meta.get_engine()
         db_server_info = engine.dialect._get_server_version_info(
             Session.connection(bind=engine))
         db_version = '.'.join(map(str, db_server_info))
     except Exception:
         log.exception('failed to fetch db version')
         db_version = 'UNKNOWN'
 
     db_info = dict(
         migrate_version=db_migrate.version,
         type=db_url_obj.get_backend_name(),
         version=db_version,
         url=repr(db_url_obj)
     )
     current_version = db_migrate.version
     expected_version = rhodecode.__dbversion__
     if state['type'] == STATE_OK and current_version != expected_version:
         msg = 'Critical: database schema mismatch, ' \
               'expected version {}, got {}. ' \
               'Please run migrations on your database.'.format(
             expected_version, current_version)
         state = {'message': msg, 'type': STATE_ERR}
 
     human_value = db_info.copy()
     human_value['url'] = "{} @ migration version: {}".format(
         db_info['url'], db_info['migrate_version'])
     human_value['version'] = "{} {}".format(db_info['type'], db_info['version'])
     return SysInfoRes(value=db_info, state=state, human_value=human_value)
 
 
 @register_sysinfo
 def server_info(environ):
     import rhodecode
     from rhodecode.lib.base import get_server_ip_addr, get_server_port
 
     value = {
-        'server_ip': '%s:%s' % (
+        'server_ip': '{}:{}'.format(
             get_server_ip_addr(environ, log_errors=False),
             get_server_port(environ)
         ),
         'server_id': rhodecode.CONFIG.get('instance_id'),
     }
     return SysInfoRes(value=value)
 
 
 @register_sysinfo
 def usage_info():
     from rhodecode.model.db import User, Repository
     value = {
         'users': User.query().count(),
         'users_active': User.query().filter(User.active == True).count(),
         'repositories': Repository.query().count(),
         'repository_types': {
             'hg': Repository.query().filter(
                 Repository.repo_type == 'hg').count(),
             'git': Repository.query().filter(
                 Repository.repo_type == 'git').count(),
             'svn': Repository.query().filter(
                 Repository.repo_type == 'svn').count(),
         },
     }
     return SysInfoRes(value=value)
 
 
 def get_system_info(environ):
     environ = environ or {}
     return {
         'rhodecode_app': SysInfo(rhodecode_app_info)(),
         'rhodecode_config': SysInfo(rhodecode_config)(),
         'rhodecode_usage': SysInfo(usage_info)(),
         'python': SysInfo(python_info)(),
         'py_modules': SysInfo(py_modules)(),
 
         'platform': SysInfo(platform_type)(),
         'locale': SysInfo(locale_info)(),
         'server': SysInfo(server_info, environ=environ)(),
         'database': SysInfo(database_info)(),
         'ulimit': SysInfo(ulimit_info)(),
         'storage': SysInfo(storage)(),
         'storage_inodes': SysInfo(storage_inodes)(),
         'storage_archive': SysInfo(storage_archives)(),
         'storage_gist': SysInfo(storage_gist)(),
         'storage_temp': SysInfo(storage_temp)(),
 
         'search': SysInfo(search_info)(),
 
         'uptime': SysInfo(uptime)(),
         'load': SysInfo(machine_load)(),
         'cpu': SysInfo(cpu)(),
         'memory': SysInfo(memory)(),
 
         'vcs_backends': SysInfo(vcs_backends)(),
         'vcs_server': SysInfo(vcs_server)(),
 
         'vcs_server_config': SysInfo(vcs_server_config)(),
 
         'git': SysInfo(git_info)(),
         'hg': SysInfo(hg_info)(),
         'svn': SysInfo(svn_info)(),
     }
 
 
 def load_system_info(key):
     """
     get_sys_info('vcs_server')
     get_sys_info('database')
     """
     return SysInfo(registered_helpers[key])()

rhodecode/lib/type_utils.py

@@ -1,93 +1,91 b''
-
-
 # Copyright (C) 2011-2023 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 logging
 
 log = logging.getLogger(__name__)
 
 
 def str2bool(str_):
     """
     returns True/False value from given string, it tries to translate the
     string into boolean
 
     :param str_: string value to translate into boolean
     :rtype: boolean
     :returns: boolean from given string
     """
     if str_ is None:
         return False
     if str_ in (True, False):
         return str_
     str_ = str(str_).strip().lower()
     return str_ in ('t', 'true', 'y', 'yes', 'on', '1')
 
 
 def aslist(obj, sep=None, strip=True) -> list:
     """
     Returns given string separated by sep as list
 
     :param obj:
     :param sep:
     :param strip:
     """
     if isinstance(obj, str):
         if obj in ['', ""]:
             return []
 
         lst = obj.split(sep)
         if strip:
             lst = [v.strip() for v in lst]
         return lst
     elif isinstance(obj, (list, tuple)):
         return obj
     elif obj is None:
         return []
     else:
         return [obj]
 
 
 class AttributeDictBase(dict):
     def __getstate__(self):
         odict = self.__dict__  # get attribute dictionary
         return odict
 
     def __setstate__(self, dict):
         self.__dict__ = dict
 
     __setattr__ = dict.__setitem__
     __delattr__ = dict.__delitem__
 
 
 class StrictAttributeDict(AttributeDictBase):
     """
     Strict Version of Attribute dict which raises an Attribute error when
     requested attribute is not set
     """
     def __getattr__(self, attr):
         try:
             return self[attr]
         except KeyError:
             raise AttributeError(f'{self.__class__} object has no attribute {attr}')
 
 
 class AttributeDict(AttributeDictBase):
     def __getattr__(self, attr):
         return self.get(attr, None)

rhodecode/lib/user_log_filter.py

@@ -1,128 +1,127 b''
-
 # Copyright (C) 2010-2023 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 logging
 
 from whoosh.qparser.default import QueryParser, query
 from whoosh.qparser.dateparse import DateParserPlugin
 from whoosh.fields import (TEXT, Schema, DATETIME, KEYWORD)
 from sqlalchemy.sql.expression import or_, and_, not_, func
 
 from rhodecode.model.db import UserLog
 from rhodecode.lib.utils2 import remove_prefix, remove_suffix
 from rhodecode.lib.str_utils import safe_str
 
 
 # JOURNAL SCHEMA used only to generate queries in journal. We use whoosh
 # querylang to build sql queries and filter journals
 AUDIT_LOG_SCHEMA = Schema(
     username=KEYWORD(),
     repository=KEYWORD(),
 
     date=DATETIME(),
     action=TEXT(),
     ip=TEXT(),
 )
 
 log = logging.getLogger(__name__)
 
 
 def user_log_filter(user_log, search_term):
     """
     Filters sqlalchemy user_log based on search_term with whoosh Query language
     http://packages.python.org/Whoosh/querylang.html
 
     :param user_log:
     :param search_term:
     """
     log.debug('Initial search term: %r', search_term)
     qry = None
     if search_term:
         qp = QueryParser('repository', schema=AUDIT_LOG_SCHEMA)
         qp.add_plugin(DateParserPlugin())
         qry = qp.parse(safe_str(search_term))
         log.debug('Filtering using parsed query %r', qry)
 
     def wildcard_handler(col, wc_term):
         if wc_term.startswith('*') and not wc_term.endswith('*'):
             # postfix == endswith
             wc_term = remove_prefix(wc_term, prefix='*')
             return func.lower(col).endswith(wc_term)
         elif wc_term.startswith('*') and wc_term.endswith('*'):
             # wildcard == ilike
             wc_term = remove_prefix(wc_term, prefix='*')
             wc_term = remove_suffix(wc_term, suffix='*')
             return func.lower(col).contains(wc_term)
 
     def get_filterion(field, val, term):
 
         if field == 'repository':
             field = getattr(UserLog, 'repository_name')
         elif field == 'ip':
             field = getattr(UserLog, 'user_ip')
         elif field == 'date':
             field = getattr(UserLog, 'action_date')
         elif field == 'username':
             field = getattr(UserLog, 'username')
         else:
             field = getattr(UserLog, field)
         log.debug('filter field: %s val=>%s', field, val)
 
         # sql filtering
         if isinstance(term, query.Wildcard):
             return wildcard_handler(field, val)
         elif isinstance(term, query.Prefix):
             return func.lower(field).startswith(func.lower(val))
         elif isinstance(term, query.DateRange):
             return and_(field >= val[0], field <= val[1])
         elif isinstance(term, query.Not):
             return not_(field == val)
         return func.lower(field) == func.lower(val)
 
     if isinstance(qry, (query.And, query.Not, query.Term, query.Prefix,
                         query.Wildcard, query.DateRange)):
         if not isinstance(qry, query.And):
             qry = [qry]
 
         for term in qry:
             if isinstance(term, query.Not):
                 not_term = [z for z in term.leaves()][0]
                 field = not_term.fieldname
                 val = not_term.text
             elif isinstance(term, query.DateRange):
                 field = term.fieldname
                 val = [term.startdate, term.enddate]
             elif isinstance(term, query.NullQuery.__class__):
                 field = ''
                 val = ''
             else:
                 field = term.fieldname
                 val = term.text
             if field:
                 user_log = user_log.filter(get_filterion(field, val, term))
     elif isinstance(qry, query.Or):
         filters = []
         for term in qry:
             field = term.fieldname
             val = (term.text if not isinstance(term, query.DateRange)
                    else [term.startdate, term.enddate])
             filters.append(get_filterion(field, val, term))
         user_log = user_log.filter(or_(*filters))
 
     return user_log

rhodecode/lib/user_sessions.py

@@ -1,264 +1,261 b''
-
-
 # Copyright (C) 2017-2023 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 os
 import re
 import time
 import datetime
 import dateutil
 import pickle
 
 from rhodecode.model.db import DbSession, Session
 
 
 class CleanupCommand(Exception):
     pass
 
 
 class BaseAuthSessions(object):
     SESSION_TYPE = None
     NOT_AVAILABLE = 'NOT AVAILABLE'
 
     def __init__(self, config):
         session_conf = {}
         for k, v in list(config.items()):
             if k.startswith('beaker.session'):
                 session_conf[k] = v
         self.config = session_conf
 
     def get_count(self):
         raise NotImplementedError
 
     def get_expired_count(self, older_than_seconds=None):
         raise NotImplementedError
 
     def clean_sessions(self, older_than_seconds=None):
         raise NotImplementedError
 
     def _seconds_to_date(self, seconds):
         return datetime.datetime.utcnow() - dateutil.relativedelta.relativedelta(
             seconds=seconds)
 
 
 class DbAuthSessions(BaseAuthSessions):
     SESSION_TYPE = 'ext:database'
 
     def get_count(self):
         return DbSession.query().count()
 
     def get_expired_count(self, older_than_seconds=None):
         expiry_date = self._seconds_to_date(older_than_seconds)
         return DbSession.query().filter(DbSession.accessed < expiry_date).count()
 
     def clean_sessions(self, older_than_seconds=None):
         expiry_date = self._seconds_to_date(older_than_seconds)
         to_remove = DbSession.query().filter(DbSession.accessed < expiry_date).count()
         DbSession.query().filter(DbSession.accessed < expiry_date).delete()
         Session().commit()
         return to_remove
 
 
 class FileAuthSessions(BaseAuthSessions):
     SESSION_TYPE = 'file sessions'
 
     def _get_sessions_dir(self):
         data_dir = self.config.get('beaker.session.data_dir')
         return data_dir
 
     def _count_on_filesystem(self, path, older_than=0, callback=None):
         value = dict(percent=0, used=0, total=0, items=0, callbacks=0,
                      path=path, text='')
         items_count = 0
         used = 0
         callbacks = 0
         cur_time = time.time()
         for root, dirs, files in os.walk(path):
             for f in files:
                 final_path = os.path.join(root, f)
                 try:
                     mtime = os.stat(final_path).st_mtime
                     if (cur_time - mtime) > older_than:
                         items_count += 1
                         if callback:
                             callback_res = callback(final_path)
                             callbacks += 1
                         else:
                             used += os.path.getsize(final_path)
                 except OSError:
                     pass
         value.update({
             'percent': 100,
             'used': used,
             'total': used,
             'items': items_count,
             'callbacks': callbacks
         })
         return value
 
     def get_count(self):
         try:
             sessions_dir = self._get_sessions_dir()
             items_count = self._count_on_filesystem(sessions_dir)['items']
         except Exception:
             items_count = self.NOT_AVAILABLE
         return items_count
 
     def get_expired_count(self, older_than_seconds=0):
         try:
             sessions_dir = self._get_sessions_dir()
             items_count = self._count_on_filesystem(
                 sessions_dir, older_than=older_than_seconds)['items']
         except Exception:
             items_count = self.NOT_AVAILABLE
         return items_count
 
     def clean_sessions(self, older_than_seconds=0):
         # find . -mtime +60 -exec rm {} \;
 
         sessions_dir = self._get_sessions_dir()
 
         def remove_item(path):
             os.remove(path)
 
         stats = self._count_on_filesystem(
             sessions_dir, older_than=older_than_seconds,
             callback=remove_item)
         return stats['callbacks']
 
 
 class MemcachedAuthSessions(BaseAuthSessions):
     SESSION_TYPE = 'ext:memcached'
     _key_regex = re.compile(r'ITEM (.*_session) \[(.*); (.*)\]')
 
     def _get_client(self):
         import memcache
         client = memcache.Client([self.config.get('beaker.session.url')])
         return client
 
     def _get_telnet_client(self, host, port):
         import telnetlib
         client = telnetlib.Telnet(host, port, None)
         return client
 
     def _run_telnet_cmd(self, client, cmd):
         client.write("%s\n" % cmd)
         return client.read_until('END')
 
     def key_details(self, client, slab_ids, limit=100):
         """  Return a list of tuples containing keys and details """
         cmd = 'stats cachedump %s %s'
         for slab_id in slab_ids:
-            for key in self._key_regex.finditer(
-                    self._run_telnet_cmd(client, cmd % (slab_id, limit))):
-                yield key
+            yield from self._key_regex.finditer(
+                    self._run_telnet_cmd(client, cmd % (slab_id, limit)))
 
     def get_count(self):
         client = self._get_client()
         count = self.NOT_AVAILABLE
         try:
             slabs = []
             for server, slabs_data in client.get_slabs():
                 slabs.extend(list(slabs_data.keys()))
 
             host, port = client.servers[0].address
             telnet_client = self._get_telnet_client(host, port)
             keys = self.key_details(telnet_client, slabs)
             count = 0
             for _k in keys:
                 count += 1
         except Exception:
             return count
 
         return count
 
     def get_expired_count(self, older_than_seconds=None):
         return self.NOT_AVAILABLE
 
     def clean_sessions(self, older_than_seconds=None):
         raise CleanupCommand('Cleanup for this session type not yet available')
 
 
 class RedisAuthSessions(BaseAuthSessions):
     SESSION_TYPE = 'ext:redis'
 
     def _get_client(self):
         import redis
         args = {
             'socket_timeout': 60,
             'url': self.config.get('beaker.session.url')
         }
 
         client = redis.StrictRedis.from_url(**args)
         return client
 
     def get_count(self):
         client = self._get_client()
         return len(client.keys('beaker_cache:*'))
 
     def get_expired_count(self, older_than_seconds=None):
         expiry_date = self._seconds_to_date(older_than_seconds)
         return self.NOT_AVAILABLE
 
     def clean_sessions(self, older_than_seconds=None):
         client = self._get_client()
         expiry_time = time.time() - older_than_seconds
         deleted_keys = 0
         for key in client.keys('beaker_cache:*'):
             data = client.get(key)
             if data:
                 json_data = pickle.loads(data)
                 try:
                     accessed_time = json_data['_accessed_time']
                 except KeyError:
                     accessed_time = 0
                 if accessed_time < expiry_time:
                     client.delete(key)
                     deleted_keys += 1
 
         return deleted_keys
 
 
 class MemoryAuthSessions(BaseAuthSessions):
     SESSION_TYPE = 'memory'
 
     def get_count(self):
         return self.NOT_AVAILABLE
 
     def get_expired_count(self, older_than_seconds=None):
         return self.NOT_AVAILABLE
 
     def clean_sessions(self, older_than_seconds=None):
         raise CleanupCommand('Cleanup for this session type not yet available')
 
 
 def get_session_handler(session_type):
     types = {
         'file': FileAuthSessions,
         'ext:memcached': MemcachedAuthSessions,
         'ext:redis': RedisAuthSessions,
         'ext:database': DbAuthSessions,
         'memory': MemoryAuthSessions
     }
 
     try:
         return types[session_type]
     except KeyError:
         raise ValueError(
-            'This type {} is not supported'.format(session_type))
+            f'This type {session_type} is not supported')

rhodecode/lib/utils.py

@@ -1,809 +1,807 b''
-
 # Copyright (C) 2010-2023 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/
 
 """
 Utilities library for RhodeCode
 """
 
 import datetime
 import decorator
 import logging
 import os
 import re
 import sys
 import shutil
 import socket
 import tempfile
 import traceback
 import tarfile
 import warnings
 from os.path import join as jn
 
 import paste
 import pkg_resources
 from webhelpers2.text import collapse, strip_tags, convert_accented_entities, convert_misc_entities
 
 from mako import exceptions
 
 from rhodecode.lib.hash_utils import sha256_safe, md5, sha1
 from rhodecode.lib.str_utils import safe_bytes, safe_str
 from rhodecode.lib.vcs.backends.base import Config
 from rhodecode.lib.vcs.exceptions import VCSError
 from rhodecode.lib.vcs.utils.helpers import get_scm, get_scm_backend
 from rhodecode.lib.ext_json import sjson as json
 from rhodecode.model import meta
 from rhodecode.model.db import (
     Repository, User, RhodeCodeUi, UserLog, RepoGroup, UserGroup)
 from rhodecode.model.meta import Session
 
 
 log = logging.getLogger(__name__)
 
 REMOVED_REPO_PAT = re.compile(r'rm__\d{8}_\d{6}_\d{6}__.*')
 
 # String which contains characters that are not allowed in slug names for
 # repositories or repository groups. It is properly escaped to use it in
 # regular expressions.
 SLUG_BAD_CHARS = re.escape(r'`?=[]\;\'"<>,/~!@#$%^&*()+{}|:')
 
 # Regex that matches forbidden characters in repo/group slugs.
 SLUG_BAD_CHAR_RE = re.compile(r'[{}\x00-\x08\x0b-\x0c\x0e-\x1f]'.format(SLUG_BAD_CHARS))
 
 # Regex that matches allowed characters in repo/group slugs.
 SLUG_GOOD_CHAR_RE = re.compile(r'[^{}]'.format(SLUG_BAD_CHARS))
 
 # Regex that matches whole repo/group slugs.
 SLUG_RE = re.compile(r'[^{}]+'.format(SLUG_BAD_CHARS))
 
 _license_cache = None
 
 
 def repo_name_slug(value):
     """
     Return slug of name of repository
     This function is called on each creation/modification
     of repository to prevent bad names in repo
     """
 
     replacement_char = '-'
 
     slug = strip_tags(value)
     slug = convert_accented_entities(slug)
     slug = convert_misc_entities(slug)
 
     slug = SLUG_BAD_CHAR_RE.sub('', slug)
     slug = re.sub(r'[\s]+', '-', slug)
     slug = collapse(slug, replacement_char)
 
     return slug
 
 
 #==============================================================================
 # PERM DECORATOR HELPERS FOR EXTRACTING NAMES FOR PERM CHECKS
 #==============================================================================
 def get_repo_slug(request):
     _repo = ''
 
     if hasattr(request, 'db_repo_name'):
         # if our requests has set db reference use it for name, this
         # translates the example.com/_<id> into proper repo names
         _repo = request.db_repo_name
     elif getattr(request, 'matchdict', None):
         # pyramid
         _repo = request.matchdict.get('repo_name')
 
     if _repo:
         _repo = _repo.rstrip('/')
     return _repo
 
 
 def get_repo_group_slug(request):
     _group = ''
     if hasattr(request, 'db_repo_group'):
         # if our requests has set db reference use it for name, this
         # translates the example.com/_<id> into proper repo group names
         _group = request.db_repo_group.group_name
     elif getattr(request, 'matchdict', None):
         # pyramid
         _group = request.matchdict.get('repo_group_name')
 
     if _group:
         _group = _group.rstrip('/')
     return _group
 
 
 def get_user_group_slug(request):
     _user_group = ''
 
     if hasattr(request, 'db_user_group'):
         _user_group = request.db_user_group.users_group_name
     elif getattr(request, 'matchdict', None):
         # pyramid
         _user_group = request.matchdict.get('user_group_id')
         _user_group_name = request.matchdict.get('user_group_name')
         try:
             if _user_group:
                 _user_group = UserGroup.get(_user_group)
             elif _user_group_name:
                 _user_group = UserGroup.get_by_group_name(_user_group_name)
 
             if _user_group:
                 _user_group = _user_group.users_group_name
         except Exception:
             log.exception('Failed to get user group by id and name')
             # catch all failures here
             return None
 
     return _user_group
 
 
 def get_filesystem_repos(path, recursive=False, skip_removed_repos=True):
     """
     Scans given path for repos and return (name,(type,path)) tuple
 
     :param path: path to scan for repositories
     :param recursive: recursive search and return names with subdirs in front
     """
 
     # remove ending slash for better results
     path = path.rstrip(os.sep)
     log.debug('now scanning in %s location recursive:%s...', path, recursive)
 
     def _get_repos(p):
         dirpaths = get_dirpaths(p)
         if not _is_dir_writable(p):
             log.warning('repo path without write access: %s', p)
 
         for dirpath in dirpaths:
             if os.path.isfile(os.path.join(p, dirpath)):
                 continue
             cur_path = os.path.join(p, dirpath)
 
             # skip removed repos
             if skip_removed_repos and REMOVED_REPO_PAT.match(dirpath):
                 continue
 
             #skip .<somethin> dirs
             if dirpath.startswith('.'):
                 continue
 
             try:
                 scm_info = get_scm(cur_path)
                 yield scm_info[1].split(path, 1)[-1].lstrip(os.sep), scm_info
             except VCSError:
                 if not recursive:
                     continue
                 #check if this dir containts other repos for recursive scan
                 rec_path = os.path.join(p, dirpath)
                 if os.path.isdir(rec_path):
-                    for inner_scm in _get_repos(rec_path):
-                        yield inner_scm
+                    yield from _get_repos(rec_path)
 
     return _get_repos(path)
 
 
 def get_dirpaths(p: str) -> list:
     try:
         # OS-independable way of checking if we have at least read-only
         # access or not.
         dirpaths = os.listdir(p)
     except OSError:
         log.warning('ignoring repo path without read access: %s', p)
         return []
 
     # os.listpath has a tweak: If a unicode is passed into it, then it tries to
     # decode paths and suddenly returns unicode objects itself. The items it
     # cannot decode are returned as strings and cause issues.
     #
     # Those paths are ignored here until a solid solution for path handling has
     # been built.
     expected_type = type(p)
 
     def _has_correct_type(item):
         if type(item) is not expected_type:
             log.error(
                 "Ignoring path %s since it cannot be decoded into str.",
                 # Using "repr" to make sure that we see the byte value in case
                 # of support.
                 repr(item))
             return False
         return True
 
     dirpaths = [item for item in dirpaths if _has_correct_type(item)]
 
     return dirpaths
 
 
 def _is_dir_writable(path):
     """
     Probe if `path` is writable.
 
     Due to trouble on Cygwin / Windows, this is actually probing if it is
     possible to create a file inside of `path`, stat does not produce reliable
     results in this case.
     """
     try:
         with tempfile.TemporaryFile(dir=path):
             pass
     except OSError:
         return False
     return True
 
 
 def is_valid_repo(repo_name, base_path, expect_scm=None, explicit_scm=None, config=None):
     """
     Returns True if given path is a valid repository False otherwise.
     If expect_scm param is given also, compare if given scm is the same
     as expected from scm parameter. If explicit_scm is given don't try to
     detect the scm, just use the given one to check if repo is valid
 
     :param repo_name:
     :param base_path:
     :param expect_scm:
     :param explicit_scm:
     :param config:
 
     :return True: if given path is a valid repository
     """
     full_path = os.path.join(safe_str(base_path), safe_str(repo_name))
     log.debug('Checking if `%s` is a valid path for repository. '
               'Explicit type: %s', repo_name, explicit_scm)
 
     try:
         if explicit_scm:
             detected_scms = [get_scm_backend(explicit_scm)(
                 full_path, config=config).alias]
         else:
             detected_scms = get_scm(full_path)
 
         if expect_scm:
             return detected_scms[0] == expect_scm
         log.debug('path: %s is an vcs object:%s', full_path, detected_scms)
         return True
     except VCSError:
         log.debug('path: %s is not a valid repo !', full_path)
         return False
 
 
 def is_valid_repo_group(repo_group_name, base_path, skip_path_check=False):
     """
     Returns True if given path is a repository group, False otherwise
 
     :param repo_name:
     :param base_path:
     """
     full_path = os.path.join(safe_str(base_path), safe_str(repo_group_name))
     log.debug('Checking if `%s` is a valid path for repository group',
               repo_group_name)
 
     # check if it's not a repo
     if is_valid_repo(repo_group_name, base_path):
         log.debug('Repo called %s exist, it is not a valid repo group', repo_group_name)
         return False
 
     try:
         # we need to check bare git repos at higher level
         # since we might match branches/hooks/info/objects or possible
         # other things inside bare git repo
         maybe_repo = os.path.dirname(full_path)
         if maybe_repo == base_path:
             # skip root level repo check, we know root location CANNOT BE a repo group
             return False
 
         scm_ = get_scm(maybe_repo)
         log.debug('path: %s is a vcs object:%s, not valid repo group', full_path, scm_)
         return False
     except VCSError:
         pass
 
     # check if it's a valid path
     if skip_path_check or os.path.isdir(full_path):
         log.debug('path: %s is a valid repo group !', full_path)
         return True
 
     log.debug('path: %s is not a valid repo group !', full_path)
     return False
 
 
 def ask_ok(prompt, retries=4, complaint='[y]es or [n]o please!'):
     while True:
         ok = eval(input(prompt))
         if ok.lower() in ('y', 'ye', 'yes'):
             return True
         if ok.lower() in ('n', 'no', 'nop', 'nope'):
             return False
         retries = retries - 1
         if retries < 0:
-            raise IOError
+            raise OSError
         print(complaint)
 
 # propagated from mercurial documentation
 ui_sections = [
     'alias', 'auth',
     'decode/encode', 'defaults',
     'diff', 'email',
     'extensions', 'format',
     'merge-patterns', 'merge-tools',
     'hooks', 'http_proxy',
     'smtp', 'patch',
     'paths', 'profiling',
     'server', 'trusted',
     'ui', 'web', ]
 
 
 def config_data_from_db(clear_session=True, repo=None):
     """
     Read the configuration data from the database and return configuration
     tuples.
     """
     from rhodecode.model.settings import VcsSettingsModel
 
     config = []
 
     sa = meta.Session()
     settings_model = VcsSettingsModel(repo=repo, sa=sa)
 
     ui_settings = settings_model.get_ui_settings()
 
     ui_data = []
     for setting in ui_settings:
         if setting.active:
             ui_data.append((setting.section, setting.key, setting.value))
             config.append((
                 safe_str(setting.section), safe_str(setting.key),
                 safe_str(setting.value)))
         if setting.key == 'push_ssl':
             # force set push_ssl requirement to False, rhodecode
             # handles that
             config.append((
                 safe_str(setting.section), safe_str(setting.key), False))
     log.debug(
         'settings ui from db@repo[%s]: %s',
         repo,
         ','.join(['[{}] {}={}'.format(*s) for s in ui_data]))
     if clear_session:
         meta.Session.remove()
 
     # TODO: mikhail: probably it makes no sense to re-read hooks information.
     # It's already there and activated/deactivated
     skip_entries = []
     enabled_hook_classes = get_enabled_hook_classes(ui_settings)
     if 'pull' not in enabled_hook_classes:
         skip_entries.append(('hooks', RhodeCodeUi.HOOK_PRE_PULL))
     if 'push' not in enabled_hook_classes:
         skip_entries.append(('hooks', RhodeCodeUi.HOOK_PRE_PUSH))
         skip_entries.append(('hooks', RhodeCodeUi.HOOK_PRETX_PUSH))
         skip_entries.append(('hooks', RhodeCodeUi.HOOK_PUSH_KEY))
 
     config = [entry for entry in config if entry[:2] not in skip_entries]
 
     return config
 
 
 def make_db_config(clear_session=True, repo=None):
     """
     Create a :class:`Config` instance based on the values in the database.
     """
     config = Config()
     config_data = config_data_from_db(clear_session=clear_session, repo=repo)
     for section, option, value in config_data:
         config.set(section, option, value)
     return config
 
 
 def get_enabled_hook_classes(ui_settings):
     """
     Return the enabled hook classes.
 
     :param ui_settings: List of ui_settings as returned
         by :meth:`VcsSettingsModel.get_ui_settings`
 
     :return: a list with the enabled hook classes. The order is not guaranteed.
     :rtype: list
     """
     enabled_hooks = []
     active_hook_keys = [
         key for section, key, value, active in ui_settings
         if section == 'hooks' and active]
 
     hook_names = {
         RhodeCodeUi.HOOK_PUSH: 'push',
         RhodeCodeUi.HOOK_PULL: 'pull',
         RhodeCodeUi.HOOK_REPO_SIZE: 'repo_size'
     }
 
     for key in active_hook_keys:
         hook = hook_names.get(key)
         if hook:
             enabled_hooks.append(hook)
 
     return enabled_hooks
 
 
 def set_rhodecode_config(config):
     """
     Updates pyramid config with new settings from database
 
     :param config:
     """
     from rhodecode.model.settings import SettingsModel
     app_settings = SettingsModel().get_all_settings()
 
     for k, v in list(app_settings.items()):
         config[k] = v
 
 
 def get_rhodecode_realm():
     """
     Return the rhodecode realm from database.
     """
     from rhodecode.model.settings import SettingsModel
     realm = SettingsModel().get_setting_by_name('realm')
     return safe_str(realm.app_settings_value)
 
 
 def get_rhodecode_base_path():
     """
     Returns the base path. The base path is the filesystem path which points
     to the repository store.
     """
 
     import rhodecode
     return rhodecode.CONFIG['default_base_path']
 
 
 def map_groups(path):
     """
     Given a full path to a repository, create all nested groups that this
     repo is inside. This function creates parent-child relationships between
     groups and creates default perms for all new groups.
 
     :param paths: full path to repository
     """
     from rhodecode.model.repo_group import RepoGroupModel
     sa = meta.Session()
     groups = path.split(Repository.NAME_SEP)
     parent = None
     group = None
 
     # last element is repo in nested groups structure
     groups = groups[:-1]
     rgm = RepoGroupModel(sa)
     owner = User.get_first_super_admin()
     for lvl, group_name in enumerate(groups):
         group_name = '/'.join(groups[:lvl] + [group_name])
         group = RepoGroup.get_by_group_name(group_name)
         desc = '%s group' % group_name
 
         # skip folders that are now removed repos
         if REMOVED_REPO_PAT.match(group_name):
             break
 
         if group is None:
             log.debug('creating group level: %s group_name: %s',
                        lvl, group_name)
             group = RepoGroup(group_name, parent)
             group.group_description = desc
             group.user = owner
             sa.add(group)
             perm_obj = rgm._create_default_perms(group)
             sa.add(perm_obj)
             sa.flush()
 
         parent = group
     return group
 
 
 def repo2db_mapper(initial_repo_list, remove_obsolete=False):
     """
     maps all repos given in initial_repo_list, non existing repositories
     are created, if remove_obsolete is True it also checks for db entries
     that are not in initial_repo_list and removes them.
 
     :param initial_repo_list: list of repositories found by scanning methods
     :param remove_obsolete: check for obsolete entries in database
     """
     from rhodecode.model.repo import RepoModel
     from rhodecode.model.repo_group import RepoGroupModel
     from rhodecode.model.settings import SettingsModel
 
     sa = meta.Session()
     repo_model = RepoModel()
     user = User.get_first_super_admin()
     added = []
 
     # creation defaults
     defs = SettingsModel().get_default_repo_settings(strip_prefix=True)
     enable_statistics = defs.get('repo_enable_statistics')
     enable_locking = defs.get('repo_enable_locking')
     enable_downloads = defs.get('repo_enable_downloads')
     private = defs.get('repo_private')
 
     for name, repo in list(initial_repo_list.items()):
         group = map_groups(name)
         str_name = safe_str(name)
         db_repo = repo_model.get_by_repo_name(str_name)
         # found repo that is on filesystem not in RhodeCode database
         if not db_repo:
             log.info('repository %s not found, creating now', name)
             added.append(name)
             desc = (repo.description
                     if repo.description != 'unknown'
                     else '%s repository' % name)
 
             db_repo = repo_model._create_repo(
                 repo_name=name,
                 repo_type=repo.alias,
                 description=desc,
                 repo_group=getattr(group, 'group_id', None),
                 owner=user,
                 enable_locking=enable_locking,
                 enable_downloads=enable_downloads,
                 enable_statistics=enable_statistics,
                 private=private,
                 state=Repository.STATE_CREATED
             )
             sa.commit()
             # we added that repo just now, and make sure we updated server info
             if db_repo.repo_type == 'git':
                 git_repo = db_repo.scm_instance()
                 # update repository server-info
                 log.debug('Running update server info')
                 git_repo._update_server_info()
 
             db_repo.update_commit_cache()
 
         config = db_repo._config
         config.set('extensions', 'largefiles', '')
         repo = db_repo.scm_instance(config=config)
         repo.install_hooks()
 
     removed = []
     if remove_obsolete:
         # remove from database those repositories that are not in the filesystem
         for repo in sa.query(Repository).all():
             if repo.repo_name not in list(initial_repo_list.keys()):
                 log.debug("Removing non-existing repository found in db `%s`",
                           repo.repo_name)
                 try:
                     RepoModel(sa).delete(repo, forks='detach', fs_remove=False)
                     sa.commit()
                     removed.append(repo.repo_name)
                 except Exception:
                     # don't hold further removals on error
                     log.error(traceback.format_exc())
                     sa.rollback()
 
         def splitter(full_repo_name):
             _parts = full_repo_name.rsplit(RepoGroup.url_sep(), 1)
             gr_name = None
             if len(_parts) == 2:
                 gr_name = _parts[0]
             return gr_name
 
         initial_repo_group_list = [splitter(x) for x in
                                    list(initial_repo_list.keys()) if splitter(x)]
 
         # remove from database those repository groups that are not in the
         # filesystem due to parent child relationships we need to delete them
         # in a specific order of most nested first
         all_groups = [x.group_name for x in sa.query(RepoGroup).all()]
         def nested_sort(gr):
             return len(gr.split('/'))
         for group_name in sorted(all_groups, key=nested_sort, reverse=True):
             if group_name not in initial_repo_group_list:
                 repo_group = RepoGroup.get_by_group_name(group_name)
                 if (repo_group.children.all() or
                     not RepoGroupModel().check_exist_filesystem(
                         group_name=group_name, exc_on_failure=False)):
                     continue
 
                 log.info(
                     'Removing non-existing repository group found in db `%s`',
                     group_name)
                 try:
                     RepoGroupModel(sa).delete(group_name, fs_remove=False)
                     sa.commit()
                     removed.append(group_name)
                 except Exception:
                     # don't hold further removals on error
                     log.exception(
                         'Unable to remove repository group `%s`',
                         group_name)
                     sa.rollback()
                     raise
 
     return added, removed
 
 
 def load_rcextensions(root_path):
     import rhodecode
     from rhodecode.config import conf
 
     path = os.path.join(root_path)
     sys.path.append(path)
 
     try:
         rcextensions = __import__('rcextensions')
     except ImportError:
         if os.path.isdir(os.path.join(path, 'rcextensions')):
             log.warning('Unable to load rcextensions from %s', path)
         rcextensions = None
 
     if rcextensions:
         log.info('Loaded rcextensions from %s...', rcextensions)
         rhodecode.EXTENSIONS = rcextensions
 
         # Additional mappings that are not present in the pygments lexers
         conf.LANGUAGES_EXTENSIONS_MAP.update(
             getattr(rhodecode.EXTENSIONS, 'EXTRA_MAPPINGS', {}))
 
 
 def get_custom_lexer(extension):
     """
     returns a custom lexer if it is defined in rcextensions module, or None
     if there's no custom lexer defined
     """
     import rhodecode
     from pygments import lexers
 
     # custom override made by RhodeCode
     if extension in ['mako']:
         return lexers.get_lexer_by_name('html+mako')
 
     # check if we didn't define this extension as other lexer
     extensions = rhodecode.EXTENSIONS and getattr(rhodecode.EXTENSIONS, 'EXTRA_LEXERS', None)
     if extensions and extension in rhodecode.EXTENSIONS.EXTRA_LEXERS:
         _lexer_name = rhodecode.EXTENSIONS.EXTRA_LEXERS[extension]
         return lexers.get_lexer_by_name(_lexer_name)
 
 
 #==============================================================================
 # TEST FUNCTIONS AND CREATORS
 #==============================================================================
 def create_test_index(repo_location, config):
     """
     Makes default test index.
     """
     try:
         import rc_testdata
     except ImportError:
         raise ImportError('Failed to import rc_testdata, '
                           'please make sure this package is installed from requirements_test.txt')
     rc_testdata.extract_search_index(
         'vcs_search_index', os.path.dirname(config['search.location']))
 
 
 def create_test_directory(test_path):
     """
     Create test directory if it doesn't exist.
     """
     if not os.path.isdir(test_path):
         log.debug('Creating testdir %s', test_path)
         os.makedirs(test_path)
 
 
 def create_test_database(test_path, config):
     """
     Makes a fresh database.
     """
     from rhodecode.lib.db_manage import DbManage
     from rhodecode.lib.utils2 import get_encryption_key
 
     # PART ONE create db
     dbconf = config['sqlalchemy.db1.url']
     enc_key = get_encryption_key(config)
 
     log.debug('making test db %s', dbconf)
 
     dbmanage = DbManage(log_sql=False, dbconf=dbconf, root=config['here'],
                         tests=True, cli_args={'force_ask': True}, enc_key=enc_key)
     dbmanage.create_tables(override=True)
     dbmanage.set_db_version()
     # for tests dynamically set new root paths based on generated content
     dbmanage.create_settings(dbmanage.config_prompt(test_path))
     dbmanage.create_default_user()
     dbmanage.create_test_admin_and_users()
     dbmanage.create_permissions()
     dbmanage.populate_default_permissions()
     Session().commit()
 
 
 def create_test_repositories(test_path, config):
     """
     Creates test repositories in the temporary directory. Repositories are
     extracted from archives within the rc_testdata package.
     """
     import rc_testdata
     from rhodecode.tests import HG_REPO, GIT_REPO, SVN_REPO
 
     log.debug('making test vcs repositories')
 
     idx_path = config['search.location']
     data_path = config['cache_dir']
 
     # clean index and data
     if idx_path and os.path.exists(idx_path):
         log.debug('remove %s', idx_path)
         shutil.rmtree(idx_path)
 
     if data_path and os.path.exists(data_path):
         log.debug('remove %s', data_path)
         shutil.rmtree(data_path)
 
     rc_testdata.extract_hg_dump('vcs_test_hg', jn(test_path, HG_REPO))
     rc_testdata.extract_git_dump('vcs_test_git', jn(test_path, GIT_REPO))
 
     # Note: Subversion is in the process of being integrated with the system,
     # until we have a properly packed version of the test svn repository, this
     # tries to copy over the repo from a package "rc_testdata"
     svn_repo_path = rc_testdata.get_svn_repo_archive()
     with tarfile.open(svn_repo_path) as tar:
         tar.extractall(jn(test_path, SVN_REPO))
 
 
 def password_changed(auth_user, session):
     # Never report password change in case of default user or anonymous user.
     if auth_user.username == User.DEFAULT_USER or auth_user.user_id is None:
         return False
 
     password_hash = md5(safe_bytes(auth_user.password)) if auth_user.password else None
     rhodecode_user = session.get('rhodecode_user', {})
     session_password_hash = rhodecode_user.get('password', '')
     return password_hash != session_password_hash
 
 
 def read_opensource_licenses():
     global _license_cache
 
     if not _license_cache:
         licenses = pkg_resources.resource_string(
             'rhodecode', 'config/licenses.json')
         _license_cache = json.loads(licenses)
 
     return _license_cache
 
 
 def generate_platform_uuid():
     """
     Generates platform UUID based on it's name
     """
     import platform
 
     try:
         uuid_list = [platform.platform()]
         return sha256_safe(':'.join(uuid_list))
     except Exception as e:
         log.error('Failed to generate host uuid: %s', e)
         return 'UNDEFINED'
 
 
 def send_test_email(recipients, email_body='TEST EMAIL'):
     """
     Simple code for generating test emails.
     Usage::
 
         from rhodecode.lib import utils
         utils.send_test_email()
     """
     from rhodecode.lib.celerylib import tasks, run_task
 
     email_body = email_body_plaintext = email_body
-    subject = 'SUBJECT FROM: {}'.format(socket.gethostname())
+    subject = f'SUBJECT FROM: {socket.gethostname()}'
     tasks.send_email(recipients, subject, email_body_plaintext, email_body)

rhodecode/lib/utils2.py

@@ -1,989 +1,987 b''
-
-
 # Copyright (C) 2011-2023 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/
 
 
 """
 Some simple helper functions
 """
 
 import collections
 import datetime
 import dateutil.relativedelta
 import logging
 import re
 import sys
 import time
 import urllib.request
 import urllib.parse
 import urllib.error
 import urlobject
 import uuid
 import getpass
 import socket
 import errno
 import random
 import functools
 from contextlib import closing
 
 import pygments.lexers
 import sqlalchemy
 import sqlalchemy.event
 import sqlalchemy.engine.url
 import sqlalchemy.exc
 import sqlalchemy.sql
 import webob
 from pyramid.settings import asbool
 
 import rhodecode
 from rhodecode.translation import _, _pluralize
 from rhodecode.lib.str_utils import safe_str, safe_int, safe_bytes
 from rhodecode.lib.hash_utils import md5, md5_safe, sha1, sha1_safe
 from rhodecode.lib.type_utils import aslist, str2bool, StrictAttributeDict, AttributeDict
 
 
 def __get_lem(extra_mapping=None):
     """
     Get language extension map based on what's inside pygments lexers
     """
     d = collections.defaultdict(lambda: [])
 
     def __clean(s):
         s = s.lstrip('*')
         s = s.lstrip('.')
 
         if s.find('[') != -1:
             exts = []
             start, stop = s.find('['), s.find(']')
 
             for suffix in s[start + 1:stop]:
                 exts.append(s[:s.find('[')] + suffix)
             return [e.lower() for e in exts]
         else:
             return [s.lower()]
 
     for lx, t in sorted(pygments.lexers.LEXERS.items()):
         m = list(map(__clean, t[-2]))
         if m:
             m = functools.reduce(lambda x, y: x + y, m)
             for ext in m:
                 desc = lx.replace('Lexer', '')
                 d[ext].append(desc)
 
     data = dict(d)
 
     extra_mapping = extra_mapping or {}
     if extra_mapping:
         for k, v in list(extra_mapping.items()):
             if k not in data:
                 # register new mapping2lexer
                 data[k] = [v]
 
     return data
 
 
 def convert_line_endings(line: str, mode) -> str:
     """
     Converts a given line  "line end" accordingly to given mode
 
     Available modes are::
         0 - Unix
         1 - Mac
         2 - DOS
 
     :param line: given line to convert
     :param mode: mode to convert to
     :return: converted line according to mode
     """
     if mode == 0:
         line = line.replace('\r\n', '\n')
         line = line.replace('\r', '\n')
     elif mode == 1:
         line = line.replace('\r\n', '\r')
         line = line.replace('\n', '\r')
     elif mode == 2:
         line = re.sub('\r(?!\n)|(?<!\r)\n', '\r\n', line)
     return line
 
 
 def detect_mode(line: str, default) -> int:
     """
     Detects line break for given line, if line break couldn't be found
     given default value is returned
 
     :param line: str line
     :param default: default
     :return: value of line end on of 0 - Unix, 1 - Mac, 2 - DOS
     """
     if line.endswith('\r\n'):
         return 2
     elif line.endswith('\n'):
         return 0
     elif line.endswith('\r'):
         return 1
     else:
         return default
 
 
 def remove_suffix(s, suffix):
     if s.endswith(suffix):
         s = s[:-1 * len(suffix)]
     return s
 
 
 def remove_prefix(s, prefix):
     if s.startswith(prefix):
         s = s[len(prefix):]
     return s
 
 
 def find_calling_context(ignore_modules=None, depth=4, output_writer=None, indent=True):
     """
     Look through the calling stack and return the frame which called
     this function and is part of core module ( ie. rhodecode.* )
 
     :param ignore_modules: list of modules to ignore eg. ['rhodecode.lib']
     :param depth:
     :param output_writer:
     :param indent:
 
     usage::
 
         from rhodecode.lib.utils2 import find_calling_context
 
         calling_context = find_calling_context(ignore_modules=[
             'rhodecode.lib.caching_query',
             'rhodecode.model.settings',
         ])
 
     """
     import inspect
     if not output_writer:
         try:
             from rich import print as pprint
         except ImportError:
             pprint = print
         output_writer = pprint
 
     frame = inspect.currentframe()
     cc = []
     try:
         for i in range(depth):  # current frame + 3 callers
             frame = frame.f_back
             if not frame:
                 break
 
             info = inspect.getframeinfo(frame)
             name = frame.f_globals.get('__name__')
             if name not in ignore_modules:
                 cc.insert(0, f'CALL_CONTEXT:{i}: file {info.filename}:{info.lineno} -> {info.function}')
     finally:
         # Avoids a reference cycle
         del frame
 
     output_writer('* INFO: This code was called from: *')
     for cnt, frm_info in enumerate(cc):
         if not indent:
             cnt = 1
         output_writer(' ' * cnt + frm_info)
 
 
 def ping_connection(connection, branch):
     if branch:
         # "branch" refers to a sub-connection of a connection,
         # we don't want to bother pinging on these.
         return
 
     # turn off "close with result".  This flag is only used with
     # "connectionless" execution, otherwise will be False in any case
     save_should_close_with_result = connection.should_close_with_result
     connection.should_close_with_result = False
 
     try:
         # run a SELECT 1.   use a core select() so that
         # the SELECT of a scalar value without a table is
         # appropriately formatted for the backend
         connection.scalar(sqlalchemy.sql.select([1]))
     except sqlalchemy.exc.DBAPIError as err:
         # catch SQLAlchemy's DBAPIError, which is a wrapper
         # for the DBAPI's exception.  It includes a .connection_invalidated
         # attribute which specifies if this connection is a "disconnect"
         # condition, which is based on inspection of the original exception
         # by the dialect in use.
         if err.connection_invalidated:
             # run the same SELECT again - the connection will re-validate
             # itself and establish a new connection.  The disconnect detection
             # here also causes the whole connection pool to be invalidated
             # so that all stale connections are discarded.
             connection.scalar(sqlalchemy.sql.select([1]))
         else:
             raise
     finally:
         # restore "close with result"
         connection.should_close_with_result = save_should_close_with_result
 
 
 def engine_from_config(configuration, prefix='sqlalchemy.', **kwargs):
     """Custom engine_from_config functions."""
     log = logging.getLogger('sqlalchemy.engine')
     use_ping_connection = asbool(configuration.pop('sqlalchemy.db1.ping_connection', None))
     debug = asbool(configuration.pop('sqlalchemy.db1.debug_query', None))
 
     engine = sqlalchemy.engine_from_config(configuration, prefix, **kwargs)
 
     def color_sql(sql):
         color_seq = '\033[1;33m'  # This is yellow: code 33
         normal = '\x1b[0m'
         return ''.join([color_seq, sql, normal])
 
     if use_ping_connection:
         log.debug('Adding ping_connection on the engine config.')
         sqlalchemy.event.listen(engine, "engine_connect", ping_connection)
 
     if debug:
         # attach events only for debug configuration
         def before_cursor_execute(conn, cursor, statement,
                                   parameters, context, executemany):
             setattr(conn, 'query_start_time', time.time())
             log.info(color_sql(">>>>> STARTING QUERY >>>>>"))
             find_calling_context(ignore_modules=[
                 'rhodecode.lib.caching_query',
                 'rhodecode.model.settings',
             ], output_writer=log.info)
 
         def after_cursor_execute(conn, cursor, statement,
                                  parameters, context, executemany):
             delattr(conn, 'query_start_time')
 
         sqlalchemy.event.listen(engine, "before_cursor_execute", before_cursor_execute)
         sqlalchemy.event.listen(engine, "after_cursor_execute", after_cursor_execute)
 
     return engine
 
 
 def get_encryption_key(config) -> bytes:
     secret = config.get('rhodecode.encrypted_values.secret')
     default = config['beaker.session.secret']
     enc_key = secret or default
 
     return safe_bytes(enc_key)
 
 
 def age(prevdate, now=None, show_short_version=False, show_suffix=True, short_format=False):
     """
     Turns a datetime into an age string.
     If show_short_version is True, this generates a shorter string with
     an approximate age; ex. '1 day ago', rather than '1 day and 23 hours ago'.
 
     * IMPORTANT*
     Code of this function is written in special way so it's easier to
     backport it to javascript. If you mean to update it, please also update
     `jquery.timeago-extension.js` file
 
     :param prevdate: datetime object
     :param now: get current time, if not define we use
         `datetime.datetime.now()`
     :param show_short_version: if it should approximate the date and
         return a shorter string
     :param show_suffix:
     :param short_format: show short format, eg 2D instead of 2 days
     :rtype: unicode
     :returns: unicode words describing age
     """
 
     def _get_relative_delta(now, prevdate):
         base = dateutil.relativedelta.relativedelta(now, prevdate)
         return {
             'year': base.years,
             'month': base.months,
             'day': base.days,
             'hour': base.hours,
             'minute': base.minutes,
             'second': base.seconds,
         }
 
     def _is_leap_year(year):
         return year % 4 == 0 and (year % 100 != 0 or year % 400 == 0)
 
     def get_month(prevdate):
         return prevdate.month
 
     def get_year(prevdate):
         return prevdate.year
 
     now = now or datetime.datetime.now()
     order = ['year', 'month', 'day', 'hour', 'minute', 'second']
     deltas = {}
     future = False
 
     if prevdate > now:
         now_old = now
         now = prevdate
         prevdate = now_old
         future = True
     if future:
         prevdate = prevdate.replace(microsecond=0)
     # Get date parts deltas
     for part in order:
         rel_delta = _get_relative_delta(now, prevdate)
         deltas[part] = rel_delta[part]
 
     # Fix negative offsets (there is 1 second between 10:59:59 and 11:00:00,
     # not 1 hour, -59 minutes and -59 seconds)
     offsets = [[5, 60], [4, 60], [3, 24]]
     for element in offsets:  # seconds, minutes, hours
         num = element[0]
         length = element[1]
 
         part = order[num]
         carry_part = order[num - 1]
 
         if deltas[part] < 0:
             deltas[part] += length
             deltas[carry_part] -= 1
 
     # Same thing for days except that the increment depends on the (variable)
     # number of days in the month
     month_lengths = [31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31]
     if deltas['day'] < 0:
         if get_month(prevdate) == 2 and _is_leap_year(get_year(prevdate)):
             deltas['day'] += 29
         else:
             deltas['day'] += month_lengths[get_month(prevdate) - 1]
 
         deltas['month'] -= 1
 
     if deltas['month'] < 0:
         deltas['month'] += 12
         deltas['year'] -= 1
 
     # Format the result
     if short_format:
         fmt_funcs = {
             'year': lambda d: '%dy' % d,
             'month': lambda d: '%dm' % d,
             'day': lambda d: '%dd' % d,
             'hour': lambda d: '%dh' % d,
             'minute': lambda d: '%dmin' % d,
             'second': lambda d: '%dsec' % d,
         }
     else:
         fmt_funcs = {
             'year': lambda d: _pluralize('${num} year', '${num} years', d, mapping={'num': d}).interpolate(),
             'month': lambda d: _pluralize('${num} month', '${num} months', d, mapping={'num': d}).interpolate(),
             'day': lambda d: _pluralize('${num} day', '${num} days', d, mapping={'num': d}).interpolate(),
             'hour': lambda d: _pluralize('${num} hour', '${num} hours', d, mapping={'num': d}).interpolate(),
             'minute': lambda d: _pluralize('${num} minute', '${num} minutes', d, mapping={'num': d}).interpolate(),
             'second': lambda d: _pluralize('${num} second', '${num} seconds', d, mapping={'num': d}).interpolate(),
         }
 
     i = 0
     for part in order:
         value = deltas[part]
         if value != 0:
 
             if i < 5:
                 sub_part = order[i + 1]
                 sub_value = deltas[sub_part]
             else:
                 sub_value = 0
 
             if sub_value == 0 or show_short_version:
                 _val = fmt_funcs[part](value)
                 if future:
                     if show_suffix:
                         return _('in ${ago}', mapping={'ago': _val})
                     else:
                         return _(_val)
 
                 else:
                     if show_suffix:
                         return _('${ago} ago', mapping={'ago': _val})
                     else:
                         return _(_val)
 
             val = fmt_funcs[part](value)
             val_detail = fmt_funcs[sub_part](sub_value)
             mapping = {'val': val, 'detail': val_detail}
 
             if short_format:
                 datetime_tmpl = _('${val}, ${detail}', mapping=mapping)
                 if show_suffix:
                     datetime_tmpl = _('${val}, ${detail} ago', mapping=mapping)
                     if future:
                         datetime_tmpl = _('in ${val}, ${detail}', mapping=mapping)
             else:
                 datetime_tmpl = _('${val} and ${detail}', mapping=mapping)
                 if show_suffix:
                     datetime_tmpl = _('${val} and ${detail} ago', mapping=mapping)
                     if future:
                         datetime_tmpl = _('in ${val} and ${detail}', mapping=mapping)
 
             return datetime_tmpl
         i += 1
     return _('just now')
 
 
 def age_from_seconds(seconds):
     seconds = safe_int(seconds) or 0
     prevdate = time_to_datetime(time.time() + seconds)
     return age(prevdate, show_suffix=False, show_short_version=True)
 
 
 def cleaned_uri(uri):
     """
     Quotes '[' and ']' from uri if there is only one of them.
         according to RFC3986 we cannot use such chars in uri
     :param uri:
     :return: uri without this chars
     """
     return urllib.parse.quote(uri, safe='@$:/')
 
 
 def credentials_filter(uri):
     """
     Returns a url with removed credentials
 
     :param uri:
     """
     import urlobject
     if isinstance(uri, rhodecode.lib.encrypt.InvalidDecryptedValue):
         return 'InvalidDecryptionKey'
 
     url_obj = urlobject.URLObject(cleaned_uri(uri))
     url_obj = url_obj.without_password().without_username()
 
     return url_obj
 
 
 def get_host_info(request):
     """
     Generate host info, to obtain full url e.g https://server.com
     use this
     `{scheme}://{netloc}`
     """
     if not request:
         return {}
 
     qualified_home_url = request.route_url('home')
     parsed_url = urlobject.URLObject(qualified_home_url)
     decoded_path = safe_str(urllib.parse.unquote(parsed_url.path.rstrip('/')))
 
     return {
         'scheme': parsed_url.scheme,
         'netloc': parsed_url.netloc+decoded_path,
         'hostname': parsed_url.hostname,
     }
 
 
 def get_clone_url(request, uri_tmpl, repo_name, repo_id, repo_type, **override):
     qualified_home_url = request.route_url('home')
     parsed_url = urlobject.URLObject(qualified_home_url)
     decoded_path = safe_str(urllib.parse.unquote(parsed_url.path.rstrip('/')))
 
     args = {
         'scheme': parsed_url.scheme,
         'user': '',
         'sys_user': getpass.getuser(),
         # path if we use proxy-prefix
         'netloc': parsed_url.netloc+decoded_path,
         'hostname': parsed_url.hostname,
         'prefix': decoded_path,
         'repo': repo_name,
         'repoid': str(repo_id),
         'repo_type': repo_type
     }
     args.update(override)
     args['user'] = urllib.parse.quote(safe_str(args['user']))
 
     for k, v in list(args.items()):
         tmpl_key = '{%s}' % k
         uri_tmpl = uri_tmpl.replace(tmpl_key, v)
 
     # special case for SVN clone url
     if repo_type == 'svn':
         uri_tmpl = uri_tmpl.replace('ssh://', 'svn+ssh://')
 
     # remove leading @ sign if it's present. Case of empty user
     url_obj = urlobject.URLObject(uri_tmpl)
     url = url_obj.with_netloc(url_obj.netloc.lstrip('@'))
 
     return safe_str(url)
 
 
 def get_commit_safe(repo, commit_id=None, commit_idx=None, pre_load=None,
                     maybe_unreachable=False, reference_obj=None):
     """
     Safe version of get_commit if this commit doesn't exists for a
     repository it returns a Dummy one instead
 
     :param repo: repository instance
     :param commit_id: commit id as str
     :param commit_idx: numeric commit index
     :param pre_load: optional list of commit attributes to load
     :param maybe_unreachable: translate unreachable commits on git repos
     :param reference_obj: explicitly search via a reference obj in git. E.g "branch:123" would mean branch "123"
     """
     # TODO(skreft): remove these circular imports
     from rhodecode.lib.vcs.backends.base import BaseRepository, EmptyCommit
     from rhodecode.lib.vcs.exceptions import RepositoryError
     if not isinstance(repo, BaseRepository):
         raise Exception('You must pass an Repository '
                         'object as first argument got %s', type(repo))
 
     try:
         commit = repo.get_commit(
             commit_id=commit_id, commit_idx=commit_idx, pre_load=pre_load,
             maybe_unreachable=maybe_unreachable, reference_obj=reference_obj)
     except (RepositoryError, LookupError):
         commit = EmptyCommit()
     return commit
 
 
 def datetime_to_time(dt):
     if dt:
         return time.mktime(dt.timetuple())
 
 
 def time_to_datetime(tm):
     if tm:
         if isinstance(tm, str):
             try:
                 tm = float(tm)
             except ValueError:
                 return
         return datetime.datetime.fromtimestamp(tm)
 
 
 def time_to_utcdatetime(tm):
     if tm:
         if isinstance(tm, str):
             try:
                 tm = float(tm)
             except ValueError:
                 return
         return datetime.datetime.utcfromtimestamp(tm)
 
 
 MENTIONS_REGEX = re.compile(
     # ^@ or @ without any special chars in front
     r'(?:^@|[^a-zA-Z0-9\-\_\.]@)'
     # main body starts with letter, then can be . - _
     r'([a-zA-Z0-9]{1}[a-zA-Z0-9\-\_\.]+)',
     re.VERBOSE | re.MULTILINE)
 
 
 def extract_mentioned_users(s):
     """
     Returns unique usernames from given string s that have @mention
 
     :param s: string to get mentions
     """
     usrs = set()
     for username in MENTIONS_REGEX.findall(s):
         usrs.add(username)
 
     return sorted(list(usrs), key=lambda k: k.lower())
 
 
 def fix_PATH(os_=None):
     """
     Get current active python path, and append it to PATH variable to fix
     issues of subprocess calls and different python versions
     """
     if os_ is None:
         import os
     else:
         os = os_
 
     cur_path = os.path.split(sys.executable)[0]
     os_path = os.environ['PATH']
     if not os.environ['PATH'].startswith(cur_path):
         os.environ['PATH'] = f'{cur_path}:{os_path}'
 
 
 def obfuscate_url_pw(engine):
     _url = engine or ''
     try:
         _url = sqlalchemy.engine.url.make_url(engine)
     except Exception:
         pass
     return repr(_url)
 
 
 def get_server_url(environ):
     req = webob.Request(environ)
     return req.host_url + req.script_name
 
 
 def unique_id(hexlen=32):
     alphabet = "23456789ABCDEFGHJKLMNPQRSTUVWXYZabcdefghjklmnpqrstuvwxyz"
     return suuid(truncate_to=hexlen, alphabet=alphabet)
 
 
 def suuid(url=None, truncate_to=22, alphabet=None):
     """
     Generate and return a short URL safe UUID.
 
     If the url parameter is provided, set the namespace to the provided
     URL and generate a UUID.
 
     :param url to get the uuid for
     :truncate_to: truncate the basic 22 UUID to shorter version
 
     The IDs won't be universally unique any longer, but the probability of
     a collision will still be very low.
     """
     # Define our alphabet.
     _ALPHABET = alphabet or "23456789ABCDEFGHJKLMNPQRSTUVWXYZ"
 
     # If no URL is given, generate a random UUID.
     if url is None:
         unique_id = uuid.uuid4().int
     else:
         unique_id = uuid.uuid3(uuid.NAMESPACE_URL, url).int
 
     alphabet_length = len(_ALPHABET)
     output = []
     while unique_id > 0:
         digit = unique_id % alphabet_length
         output.append(_ALPHABET[digit])
         unique_id = int(unique_id / alphabet_length)
     return "".join(output)[:truncate_to]
 
 
 def get_current_rhodecode_user(request=None):
     """
     Gets rhodecode user from request
     """
     import pyramid.threadlocal
     pyramid_request = request or pyramid.threadlocal.get_current_request()
 
     # web case
     if pyramid_request and hasattr(pyramid_request, 'user'):
         return pyramid_request.user
 
     # api case
     if pyramid_request and hasattr(pyramid_request, 'rpc_user'):
         return pyramid_request.rpc_user
 
     return None
 
 
 def action_logger_generic(action, namespace=''):
     """
     A generic logger for actions useful to the system overview, tries to find
     an acting user for the context of the call otherwise reports unknown user
 
     :param action: logging message eg 'comment 5 deleted'
     :param type: string
 
     :param namespace: namespace of the logging message eg. 'repo.comments'
     :param type: string
 
     """
 
     logger_name = 'rhodecode.actions'
 
     if namespace:
         logger_name += '.' + namespace
 
     log = logging.getLogger(logger_name)
 
     # get a user if we can
     user = get_current_rhodecode_user()
 
     logfunc = log.info
 
     if not user:
         user = '<unknown user>'
         logfunc = log.warning
 
-    logfunc('Logging action by {}: {}'.format(user, action))
+    logfunc(f'Logging action by {user}: {action}')
 
 
 def escape_split(text, sep=',', maxsplit=-1):
     r"""
     Allows for escaping of the separator: e.g. arg='foo\, bar'
 
     It should be noted that the way bash et. al. do command line parsing, those
     single quotes are required.
     """
     escaped_sep = r'\%s' % sep
 
     if escaped_sep not in text:
         return text.split(sep, maxsplit)
 
     before, _mid, after = text.partition(escaped_sep)
     startlist = before.split(sep, maxsplit)  # a regular split is fine here
     unfinished = startlist[-1]
     startlist = startlist[:-1]
 
     # recurse because there may be more escaped separators
     endlist = escape_split(after, sep, maxsplit)
 
     # finish building the escaped value. we use endlist[0] becaue the first
     # part of the string sent in recursion is the rest of the escaped value.
     unfinished += sep + endlist[0]
 
     return startlist + [unfinished] + endlist[1:]  # put together all the parts
 
 
 class OptionalAttr(object):
     """
     Special Optional Option that defines other attribute. Example::
 
         def test(apiuser, userid=Optional(OAttr('apiuser')):
             user = Optional.extract(userid)
             # calls
 
     """
 
     def __init__(self, attr_name):
         self.attr_name = attr_name
 
     def __repr__(self):
         return '<OptionalAttr:%s>' % self.attr_name
 
     def __call__(self):
         return self
 
 
 # alias
 OAttr = OptionalAttr
 
 
 class Optional(object):
     """
     Defines an optional parameter::
 
         param = param.getval() if isinstance(param, Optional) else param
         param = param() if isinstance(param, Optional) else param
 
     is equivalent of::
 
         param = Optional.extract(param)
 
     """
 
     def __init__(self, type_):
         self.type_ = type_
 
     def __repr__(self):
         return '<Optional:%s>' % self.type_.__repr__()
 
     def __call__(self):
         return self.getval()
 
     def getval(self):
         """
         returns value from this Optional instance
         """
         if isinstance(self.type_, OAttr):
             # use params name
             return self.type_.attr_name
         return self.type_
 
     @classmethod
     def extract(cls, val):
         """
         Extracts value from Optional() instance
 
         :param val:
         :return: original value if it's not Optional instance else
             value of instance
         """
         if isinstance(val, cls):
             return val.getval()
         return val
 
 
 def glob2re(pat):
     import fnmatch
     return fnmatch.translate(pat)
 
 
 def parse_byte_string(size_str):
     match = re.match(r'(\d+)(MB|KB)', size_str, re.IGNORECASE)
     if not match:
         raise ValueError(f'Given size:{size_str} is invalid, please make sure '
                          f'to use format of <num>(MB|KB)')
 
     _parts = match.groups()
     num, type_ = _parts
     return int(num) * {'mb': 1024*1024, 'kb': 1024}[type_.lower()]
 
 
 class CachedProperty(object):
     """
     Lazy Attributes. With option to invalidate the cache by running a method
 
     >>> class Foo(object):
     ...
     ...    @CachedProperty
     ...    def heavy_func(self):
     ...        return 'super-calculation'
     ...
     ... foo = Foo()
     ... foo.heavy_func() # first computation
     ... foo.heavy_func() # fetch from cache
     ... foo._invalidate_prop_cache('heavy_func')
 
     # at this point calling foo.heavy_func() will be re-computed
     """
 
     def __init__(self, func, func_name=None):
 
         if func_name is None:
             func_name = func.__name__
         self.data = (func, func_name)
         functools.update_wrapper(self, func)
 
     def __get__(self, inst, class_):
         if inst is None:
             return self
 
         func, func_name = self.data
         value = func(inst)
         inst.__dict__[func_name] = value
         if '_invalidate_prop_cache' not in inst.__dict__:
             inst.__dict__['_invalidate_prop_cache'] = functools.partial(
                 self._invalidate_prop_cache, inst)
         return value
 
     def _invalidate_prop_cache(self, inst, name):
         inst.__dict__.pop(name, None)
 
 
 def retry(func=None, exception=Exception, n_tries=5, delay=5, backoff=1, logger=True):
     """
     Retry decorator with exponential backoff.
 
     Parameters
     ----------
     func : typing.Callable, optional
         Callable on which the decorator is applied, by default None
     exception : Exception or tuple of Exceptions, optional
         Exception(s) that invoke retry, by default Exception
     n_tries : int, optional
         Number of tries before giving up, by default 5
     delay : int, optional
         Initial delay between retries in seconds, by default 5
     backoff : int, optional
         Backoff multiplier e.g. value of 2 will double the delay, by default 1
     logger : bool, optional
         Option to log or print, by default False
 
     Returns
     -------
     typing.Callable
         Decorated callable that calls itself when exception(s) occur.
 
     Examples
     --------
     >>> import random
     >>> @retry(exception=Exception, n_tries=3)
     ... def test_random(text):
     ...    x = random.random()
     ...    if x < 0.5:
     ...        raise Exception("Fail")
     ...    else:
     ...        print("Success: ", text)
     >>> test_random("It works!")
     """
 
     if func is None:
         return functools.partial(
             retry,
             exception=exception,
             n_tries=n_tries,
             delay=delay,
             backoff=backoff,
             logger=logger,
         )
 
     @functools.wraps(func)
     def wrapper(*args, **kwargs):
         _n_tries, n_delay = n_tries, delay
         log = logging.getLogger('rhodecode.retry')
 
         while _n_tries > 1:
             try:
                 return func(*args, **kwargs)
             except exception as e:
                 e_details = repr(e)
                 msg = "Exception on calling func {func}: {e}, " \
                       "Retrying in {n_delay} seconds..."\
                     .format(func=func, e=e_details, n_delay=n_delay)
                 if logger:
                     log.warning(msg)
                 else:
                     print(msg)
                 time.sleep(n_delay)
                 _n_tries -= 1
                 n_delay *= backoff
 
         return func(*args, **kwargs)
 
     return wrapper
 
 
 def user_agent_normalizer(user_agent_raw, safe=True):
     log = logging.getLogger('rhodecode.user_agent_normalizer')
     ua = (user_agent_raw or '').strip().lower()
     ua = ua.replace('"', '')
 
     try:
         if 'mercurial/proto-1.0' in ua:
             ua = ua.replace('mercurial/proto-1.0', '')
             ua = ua.replace('(', '').replace(')', '').strip()
             ua = ua.replace('mercurial ', 'mercurial/')
         elif ua.startswith('git'):
             parts = ua.split(' ')
             if parts:
                 ua = parts[0]
             ua = re.sub(r'\.windows\.\d', '', ua).strip()
 
         return ua
     except Exception:
         log.exception('Failed to parse scm user-agent')
         if not safe:
             raise
 
     return ua
 
 
 def get_available_port(min_port=40000, max_port=55555, use_range=False):
     hostname = ''
     for _ in range(min_port, max_port):
         pick_port = 0
         if use_range:
             pick_port = random.randint(min_port, max_port)
 
         with closing(socket.socket(socket.AF_INET, socket.SOCK_STREAM)) as s:
             try:
                 s.bind((hostname, pick_port))
                 s.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)
                 return s.getsockname()[1]
             except OSError:
                 continue
             except socket.error as e:
                 if e.args[0] in [errno.EADDRINUSE, errno.ECONNREFUSED]:
                     continue
                 raise

rhodecode/lib/vcs/backends/base.py

@@ -1,1987 +1,1985 b''
-
-
 # Copyright (C) 2014-2023 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/
 
 """
 Base module for all VCS systems
 """
 import os
 import re
 import time
 import shutil
 import datetime
 import fnmatch
 import itertools
 import logging
 import dataclasses
 import warnings
 
 from zope.cachedescriptors.property import Lazy as LazyProperty
 
 
 import rhodecode
 from rhodecode.translation import lazy_ugettext
 from rhodecode.lib.utils2 import safe_str, CachedProperty
 from rhodecode.lib.vcs.utils import author_name, author_email
 from rhodecode.lib.vcs.conf import settings
 from rhodecode.lib.vcs.exceptions import (
     CommitError, EmptyRepositoryError, NodeAlreadyAddedError,
     NodeAlreadyChangedError, NodeAlreadyExistsError, NodeAlreadyRemovedError,
     NodeDoesNotExistError, NodeNotChangedError, VCSError,
     ImproperArchiveTypeError, BranchDoesNotExistError, CommitDoesNotExistError,
     RepositoryError)
 
 
 log = logging.getLogger(__name__)
 
 
 FILEMODE_DEFAULT = 0o100644
 FILEMODE_EXECUTABLE = 0o100755
 EMPTY_COMMIT_ID = '0' * 40
 
 
 @dataclasses.dataclass
 class Reference:
     type: str
     name: str
     commit_id: str
 
     def __iter__(self):
         yield self.type
         yield self.name
         yield self.commit_id
 
     @property
     def branch(self):
         if self.type == 'branch':
             return self.name
 
     @property
     def bookmark(self):
         if self.type == 'book':
             return self.name
 
     @property
     def to_str(self):
         return reference_to_unicode(self)
 
     def asdict(self):
         return dict(
             type=self.type,
             name=self.name,
             commit_id=self.commit_id
         )
 
 
 def unicode_to_reference(raw: str):
     """
     Convert a unicode (or string) to a reference object.
     If unicode evaluates to False it returns None.
     """
     if raw:
         refs = raw.split(':')
         return Reference(*refs)
     else:
         return None
 
 
 def reference_to_unicode(ref: Reference):
     """
     Convert a reference object to unicode.
     If reference is None it returns None.
     """
     if ref:
         return ':'.join(ref)
     else:
         return None
 
 
 class MergeFailureReason(object):
     """
     Enumeration with all the reasons why the server side merge could fail.
 
     DO NOT change the number of the reasons, as they may be stored in the
     database.
 
     Changing the name of a reason is acceptable and encouraged to deprecate old
     reasons.
     """
 
     # Everything went well.
     NONE = 0
 
     # An unexpected exception was raised. Check the logs for more details.
     UNKNOWN = 1
 
     # The merge was not successful, there are conflicts.
     MERGE_FAILED = 2
 
     # The merge succeeded but we could not push it to the target repository.
     PUSH_FAILED = 3
 
     # The specified target is not a head in the target repository.
     TARGET_IS_NOT_HEAD = 4
 
     # The source repository contains more branches than the target. Pushing
     # the merge will create additional branches in the target.
     HG_SOURCE_HAS_MORE_BRANCHES = 5
 
     # The target reference has multiple heads. That does not allow to correctly
     # identify the target location. This could only happen for mercurial
     # branches.
     HG_TARGET_HAS_MULTIPLE_HEADS = 6
 
     # The target repository is locked
     TARGET_IS_LOCKED = 7
 
     # Deprecated, use MISSING_TARGET_REF or MISSING_SOURCE_REF instead.
     # A involved commit could not be found.
     _DEPRECATED_MISSING_COMMIT = 8
 
     # The target repo reference is missing.
     MISSING_TARGET_REF = 9
 
     # The source repo reference is missing.
     MISSING_SOURCE_REF = 10
 
     # The merge was not successful, there are conflicts related to sub
     # repositories.
     SUBREPO_MERGE_FAILED = 11
 
 
 class UpdateFailureReason(object):
     """
     Enumeration with all the reasons why the pull request update could fail.
 
     DO NOT change the number of the reasons, as they may be stored in the
     database.
 
     Changing the name of a reason is acceptable and encouraged to deprecate old
     reasons.
     """
 
     # Everything went well.
     NONE = 0
 
     # An unexpected exception was raised. Check the logs for more details.
     UNKNOWN = 1
 
     # The pull request is up to date.
     NO_CHANGE = 2
 
     # The pull request has a reference type that is not supported for update.
     WRONG_REF_TYPE = 3
 
     # Update failed because the target reference is missing.
     MISSING_TARGET_REF = 4
 
     # Update failed because the source reference is missing.
     MISSING_SOURCE_REF = 5
 
 
 class MergeResponse(object):
 
     # uses .format(**metadata) for variables
     MERGE_STATUS_MESSAGES = {
         MergeFailureReason.NONE: lazy_ugettext(
             'This pull request can be automatically merged.'),
         MergeFailureReason.UNKNOWN: lazy_ugettext(
             'This pull request cannot be merged because of an unhandled exception. '
             '{exception}'),
         MergeFailureReason.MERGE_FAILED: lazy_ugettext(
             'This pull request cannot be merged because of merge conflicts. {unresolved_files}'),
         MergeFailureReason.PUSH_FAILED: lazy_ugettext(
             'This pull request could not be merged because push to '
             'target:`{target}@{merge_commit}` failed.'),
         MergeFailureReason.TARGET_IS_NOT_HEAD: lazy_ugettext(
             'This pull request cannot be merged because the target '
             '`{target_ref.name}` is not a head.'),
         MergeFailureReason.HG_SOURCE_HAS_MORE_BRANCHES: lazy_ugettext(
             'This pull request cannot be merged because the source contains '
             'more branches than the target.'),
         MergeFailureReason.HG_TARGET_HAS_MULTIPLE_HEADS: lazy_ugettext(
             'This pull request cannot be merged because the target `{target_ref.name}` '
             'has multiple heads: `{heads}`.'),
         MergeFailureReason.TARGET_IS_LOCKED: lazy_ugettext(
             'This pull request cannot be merged because the target repository is '
             'locked by {locked_by}.'),
 
         MergeFailureReason.MISSING_TARGET_REF: lazy_ugettext(
             'This pull request cannot be merged because the target '
             'reference `{target_ref.name}` is missing.'),
         MergeFailureReason.MISSING_SOURCE_REF: lazy_ugettext(
             'This pull request cannot be merged because the source '
             'reference `{source_ref.name}` is missing.'),
         MergeFailureReason.SUBREPO_MERGE_FAILED: lazy_ugettext(
             'This pull request cannot be merged because of conflicts related '
             'to sub repositories.'),
 
         # Deprecations
         MergeFailureReason._DEPRECATED_MISSING_COMMIT: lazy_ugettext(
             'This pull request cannot be merged because the target or the '
             'source reference is missing.'),
 
     }
 
     def __init__(self, possible, executed, merge_ref, failure_reason, metadata=None):
         self.possible = possible
         self.executed = executed
         self.merge_ref = merge_ref
         self.failure_reason = failure_reason
         self.metadata = metadata or {}
 
     def __repr__(self):
         return f'<MergeResponse:{self.label} {self.failure_reason}>'
 
     def __eq__(self, other):
         same_instance = isinstance(other, self.__class__)
         return same_instance \
                and self.possible == other.possible \
                and self.executed == other.executed \
                and self.failure_reason == other.failure_reason
 
     @property
     def label(self):
         label_dict = dict((v, k) for k, v in MergeFailureReason.__dict__.items() if
                           not k.startswith('_'))
         return label_dict.get(self.failure_reason)
 
     @property
     def merge_status_message(self):
         """
         Return a human friendly error message for the given merge status code.
         """
         msg = safe_str(self.MERGE_STATUS_MESSAGES[self.failure_reason])
 
         try:
             return msg.format(**self.metadata)
         except Exception:
             log.exception('Failed to format %s message', self)
             return msg
 
     def asdict(self):
         data = {}
         for k in ['possible', 'executed', 'merge_ref', 'failure_reason',
                   'merge_status_message']:
             data[k] = getattr(self, k)
         return data
 
 
 class TargetRefMissing(ValueError):
     pass
 
 
 class SourceRefMissing(ValueError):
     pass
 
 
 class BaseRepository(object):
     """
     Base Repository for final backends
 
     .. attribute:: DEFAULT_BRANCH_NAME
 
        name of default branch (i.e. "trunk" for svn, "master" for git etc.
 
     .. attribute:: commit_ids
 
        list of all available commit ids, in ascending order
 
     .. attribute:: path
 
        absolute path to the repository
 
     .. attribute:: bookmarks
 
        Mapping from name to :term:`Commit ID` of the bookmark. Empty in case
        there are no bookmarks or the backend implementation does not support
        bookmarks.
 
     .. attribute:: tags
 
        Mapping from name to :term:`Commit ID` of the tag.
 
     """
 
     DEFAULT_BRANCH_NAME = None
     DEFAULT_CONTACT = "Unknown"
     DEFAULT_DESCRIPTION = "unknown"
     EMPTY_COMMIT_ID = '0' * 40
     COMMIT_ID_PAT = re.compile(r'[0-9a-fA-F]{40}')
 
     path = None
 
     _is_empty = None
     _commit_ids = {}
 
     def __init__(self, repo_path, config=None, create=False, **kwargs):
         """
         Initializes repository. Raises RepositoryError if repository could
         not be find at the given ``repo_path`` or directory at ``repo_path``
         exists and ``create`` is set to True.
 
         :param repo_path: local path of the repository
         :param config: repository configuration
         :param create=False: if set to True, would try to create repository.
         :param src_url=None: if set, should be proper url from which repository
           would be cloned; requires ``create`` parameter to be set to True -
           raises RepositoryError if src_url is set and create evaluates to
           False
         """
         raise NotImplementedError
 
     def __repr__(self):
-        return '<{} at {}>'.format(self.__class__.__name__, self.path)
+        return f'<{self.__class__.__name__} at {self.path}>'
 
     def __len__(self):
         return self.count()
 
     def __eq__(self, other):
         same_instance = isinstance(other, self.__class__)
         return same_instance and other.path == self.path
 
     def __ne__(self, other):
         return not self.__eq__(other)
 
     def get_create_shadow_cache_pr_path(self, db_repo):
         path = db_repo.cached_diffs_dir
         if not os.path.exists(path):
             os.makedirs(path, 0o755)
         return path
 
     @classmethod
     def get_default_config(cls, default=None):
         config = Config()
         if default and isinstance(default, list):
             for section, key, val in default:
                 config.set(section, key, val)
         return config
 
     @LazyProperty
     def _remote(self):
         raise NotImplementedError
 
     def _heads(self, branch=None):
         return []
 
     @LazyProperty
     def EMPTY_COMMIT(self):
         return EmptyCommit(self.EMPTY_COMMIT_ID)
 
     @LazyProperty
     def alias(self):
         for k, v in settings.BACKENDS.items():
             if v.split('.')[-1] == str(self.__class__.__name__):
                 return k
 
     @LazyProperty
     def name(self):
         return safe_str(os.path.basename(self.path))
 
     @LazyProperty
     def description(self):
         raise NotImplementedError
 
     def refs(self):
         """
         returns a `dict` with branches, bookmarks, tags, and closed_branches
         for this repository
         """
         return dict(
             branches=self.branches,
             branches_closed=self.branches_closed,
             tags=self.tags,
             bookmarks=self.bookmarks
         )
 
     @LazyProperty
     def branches(self):
         """
         A `dict` which maps branch names to commit ids.
         """
         raise NotImplementedError
 
     @LazyProperty
     def branches_closed(self):
         """
         A `dict` which maps tags names to commit ids.
         """
         raise NotImplementedError
 
     @LazyProperty
     def bookmarks(self):
         """
         A `dict` which maps tags names to commit ids.
         """
         raise NotImplementedError
 
     @LazyProperty
     def tags(self):
         """
         A `dict` which maps tags names to commit ids.
         """
         raise NotImplementedError
 
     @LazyProperty
     def size(self):
         """
         Returns combined size in bytes for all repository files
         """
         tip = self.get_commit()
         return tip.size
 
     def size_at_commit(self, commit_id):
         commit = self.get_commit(commit_id)
         return commit.size
 
     def _check_for_empty(self):
         no_commits = len(self._commit_ids) == 0
         if no_commits:
             # check on remote to be sure
             return self._remote.is_empty()
         else:
             return False
 
     def is_empty(self):
         if rhodecode.is_test:
             return self._check_for_empty()
 
         if self._is_empty is None:
             # cache empty for production, but not tests
             self._is_empty = self._check_for_empty()
 
         return self._is_empty
 
     @staticmethod
     def check_url(url, config):
         """
         Function will check given url and try to verify if it's a valid
         link.
         """
         raise NotImplementedError
 
     @staticmethod
     def is_valid_repository(path):
         """
         Check if given `path` contains a valid repository of this backend
         """
         raise NotImplementedError
 
     # ==========================================================================
     # COMMITS
     # ==========================================================================
 
     @CachedProperty
     def commit_ids(self):
         raise NotImplementedError
 
     def append_commit_id(self, commit_id):
         if commit_id not in self.commit_ids:
             self._rebuild_cache(self.commit_ids + [commit_id])
 
         # clear cache
         self._invalidate_prop_cache('commit_ids')
         self._is_empty = False
 
     def get_commit(self, commit_id=None, commit_idx=None, pre_load=None,
                    translate_tag=None, maybe_unreachable=False, reference_obj=None):
         """
         Returns instance of `BaseCommit` class. If `commit_id` and `commit_idx`
         are both None, most recent commit is returned.
 
         :param pre_load: Optional. List of commit attributes to load.
 
         :raises ``EmptyRepositoryError``: if there are no commits
         """
         raise NotImplementedError
 
     def __iter__(self):
         for commit_id in self.commit_ids:
             yield self.get_commit(commit_id=commit_id)
 
     def get_commits(
             self, start_id=None, end_id=None, start_date=None, end_date=None,
             branch_name=None, show_hidden=False, pre_load=None, translate_tags=None):
         """
         Returns iterator of `BaseCommit` objects from start to end
         not inclusive. This should behave just like a list, ie. end is not
         inclusive.
 
         :param start_id: None or str, must be a valid commit id
         :param end_id: None or str, must be a valid commit id
         :param start_date:
         :param end_date:
         :param branch_name:
         :param show_hidden:
         :param pre_load:
         :param translate_tags:
         """
         raise NotImplementedError
 
     def __getitem__(self, key):
         """
         Allows index based access to the commit objects of this repository.
         """
         pre_load = ["author", "branch", "date", "message", "parents"]
         if isinstance(key, slice):
             return self._get_range(key, pre_load)
         return self.get_commit(commit_idx=key, pre_load=pre_load)
 
     def _get_range(self, slice_obj, pre_load):
         for commit_id in self.commit_ids.__getitem__(slice_obj):
             yield self.get_commit(commit_id=commit_id, pre_load=pre_load)
 
     def count(self):
         return len(self.commit_ids)
 
     def tag(self, name, user, commit_id=None, message=None, date=None, **opts):
         """
         Creates and returns a tag for the given ``commit_id``.
 
         :param name: name for new tag
         :param user: full username, i.e.: "Joe Doe <[e-mail]>"
         :param commit_id: commit id for which new tag would be created
         :param message: message of the tag's commit
         :param date: date of tag's commit
 
         :raises TagAlreadyExistError: if tag with same name already exists
         """
         raise NotImplementedError
 
     def remove_tag(self, name, user, message=None, date=None):
         """
         Removes tag with the given ``name``.
 
         :param name: name of the tag to be removed
         :param user: full username, i.e.: "Joe Doe <[e-mail]>"
         :param message: message of the tag's removal commit
         :param date: date of tag's removal commit
 
         :raises TagDoesNotExistError: if tag with given name does not exists
         """
         raise NotImplementedError
 
     def get_diff(
             self, commit1, commit2, path=None, ignore_whitespace=False,
             context=3, path1=None):
         """
         Returns (git like) *diff*, as plain text. Shows changes introduced by
         `commit2` since `commit1`.
 
         :param commit1: Entry point from which diff is shown. Can be
           ``self.EMPTY_COMMIT`` - in this case, patch showing all
           the changes since empty state of the repository until `commit2`
         :param commit2: Until which commit changes should be shown.
         :param path: Can be set to a path of a file to create a diff of that
           file. If `path1` is also set, this value is only associated to
           `commit2`.
         :param ignore_whitespace: If set to ``True``, would not show whitespace
           changes. Defaults to ``False``.
         :param context: How many lines before/after changed lines should be
           shown. Defaults to ``3``.
         :param path1: Can be set to a path to associate with `commit1`. This
           parameter works only for backends which support diff generation for
           different paths. Other backends will raise a `ValueError` if `path1`
           is set and has a different value than `path`.
         :param file_path: filter this diff by given path pattern
         """
         raise NotImplementedError
 
     def strip(self, commit_id, branch=None):
         """
         Strip given commit_id from the repository
         """
         raise NotImplementedError
 
     def get_common_ancestor(self, commit_id1, commit_id2, repo2):
         """
         Return a latest common ancestor commit if one exists for this repo
         `commit_id1` vs `commit_id2` from `repo2`.
 
         :param commit_id1: Commit it from this repository to use as a
            target for the comparison.
         :param commit_id2: Source commit id to use for comparison.
         :param repo2: Source repository to use for comparison.
         """
         raise NotImplementedError
 
     def compare(self, commit_id1, commit_id2, repo2, merge, pre_load=None):
         """
         Compare this repository's revision `commit_id1` with `commit_id2`.
 
         Returns a tuple(commits, ancestor) that would be merged from
         `commit_id2`. Doing a normal compare (``merge=False``), ``None``
         will be returned as ancestor.
 
         :param commit_id1: Commit it from this repository to use as a
            target for the comparison.
         :param commit_id2: Source commit id to use for comparison.
         :param repo2: Source repository to use for comparison.
         :param merge: If set to ``True`` will do a merge compare which also
            returns the common ancestor.
         :param pre_load: Optional. List of commit attributes to load.
         """
         raise NotImplementedError
 
     def merge(self, repo_id, workspace_id, target_ref, source_repo, source_ref,
               user_name='', user_email='', message='', dry_run=False,
               use_rebase=False, close_branch=False):
         """
         Merge the revisions specified in `source_ref` from `source_repo`
         onto the `target_ref` of this repository.
 
         `source_ref` and `target_ref` are named tupls with the following
         fields `type`, `name` and `commit_id`.
 
         Returns a MergeResponse named tuple with the following fields
         'possible', 'executed', 'source_commit', 'target_commit',
         'merge_commit'.
 
         :param repo_id: `repo_id` target repo id.
         :param workspace_id: `workspace_id` unique identifier.
         :param target_ref: `target_ref` points to the commit on top of which
             the `source_ref` should be merged.
         :param source_repo: The repository that contains the commits to be
             merged.
         :param source_ref: `source_ref` points to the topmost commit from
             the `source_repo` which should be merged.
         :param user_name: Merge commit `user_name`.
         :param user_email: Merge commit `user_email`.
         :param message: Merge commit `message`.
         :param dry_run: If `True` the merge will not take place.
         :param use_rebase: If `True` commits from the source will be rebased
             on top of the target instead of being merged.
         :param close_branch: If `True` branch will be close before merging it
         """
         if dry_run:
             message = message or settings.MERGE_DRY_RUN_MESSAGE
             user_email = user_email or settings.MERGE_DRY_RUN_EMAIL
             user_name = user_name or settings.MERGE_DRY_RUN_USER
         else:
             if not user_name:
                 raise ValueError('user_name cannot be empty')
             if not user_email:
                 raise ValueError('user_email cannot be empty')
             if not message:
                 raise ValueError('message cannot be empty')
 
         try:
             return self._merge_repo(
                 repo_id, workspace_id, target_ref, source_repo,
                 source_ref, message, user_name, user_email, dry_run=dry_run,
                 use_rebase=use_rebase, close_branch=close_branch)
         except RepositoryError as exc:
             log.exception('Unexpected failure when running merge, dry-run=%s', dry_run)
             return MergeResponse(
                 False, False, None, MergeFailureReason.UNKNOWN,
                 metadata={'exception': str(exc)})
 
     def _merge_repo(self, repo_id, workspace_id, target_ref,
                     source_repo, source_ref, merge_message,
                     merger_name, merger_email, dry_run=False,
                     use_rebase=False, close_branch=False):
         """Internal implementation of merge."""
         raise NotImplementedError
 
     def _maybe_prepare_merge_workspace(
             self, repo_id, workspace_id, target_ref, source_ref):
         """
         Create the merge workspace.
 
         :param workspace_id: `workspace_id` unique identifier.
         """
         raise NotImplementedError
 
     @classmethod
     def _get_legacy_shadow_repository_path(cls, repo_path, workspace_id):
         """
         Legacy version that was used before. We still need it for
         backward compat
         """
         return os.path.join(
             os.path.dirname(repo_path),
-            '.__shadow_{}_{}'.format(os.path.basename(repo_path), workspace_id))
+            f'.__shadow_{os.path.basename(repo_path)}_{workspace_id}')
 
     @classmethod
     def _get_shadow_repository_path(cls, repo_path, repo_id, workspace_id):
         # The name of the shadow repository must start with '.', so it is
         # skipped by 'rhodecode.lib.utils.get_filesystem_repos'.
         legacy_repository_path = cls._get_legacy_shadow_repository_path(repo_path, workspace_id)
         if os.path.exists(legacy_repository_path):
             return legacy_repository_path
         else:
             return os.path.join(
                 os.path.dirname(repo_path),
-                '.__shadow_repo_{}_{}'.format(repo_id, workspace_id))
+                f'.__shadow_repo_{repo_id}_{workspace_id}')
 
     def cleanup_merge_workspace(self, repo_id, workspace_id):
         """
         Remove merge workspace.
 
         This function MUST not fail in case there is no workspace associated to
         the given `workspace_id`.
 
         :param workspace_id: `workspace_id` unique identifier.
         """
         shadow_repository_path = self._get_shadow_repository_path(
             self.path, repo_id, workspace_id)
         shadow_repository_path_del = '{}.{}.delete'.format(
             shadow_repository_path, time.time())
 
         # move the shadow repo, so it never conflicts with the one used.
         # we use this method because shutil.rmtree had some edge case problems
         # removing symlinked repositories
         if not os.path.isdir(shadow_repository_path):
             return
 
         shutil.move(shadow_repository_path, shadow_repository_path_del)
         try:
             shutil.rmtree(shadow_repository_path_del, ignore_errors=False)
         except Exception:
             log.exception('Failed to gracefully remove shadow repo under %s',
                           shadow_repository_path_del)
             shutil.rmtree(shadow_repository_path_del, ignore_errors=True)
 
     # ========== #
     # COMMIT API #
     # ========== #
 
     @LazyProperty
     def in_memory_commit(self):
         """
         Returns :class:`InMemoryCommit` object for this repository.
         """
         raise NotImplementedError
 
     # ======================== #
     # UTILITIES FOR SUBCLASSES #
     # ======================== #
 
     def _validate_diff_commits(self, commit1, commit2):
         """
         Validates that the given commits are related to this repository.
 
         Intended as a utility for sub classes to have a consistent validation
         of input parameters in methods like :meth:`get_diff`.
         """
         self._validate_commit(commit1)
         self._validate_commit(commit2)
         if (isinstance(commit1, EmptyCommit) and
                 isinstance(commit2, EmptyCommit)):
             raise ValueError("Cannot compare two empty commits")
 
     def _validate_commit(self, commit):
         if not isinstance(commit, BaseCommit):
             raise TypeError(
                 "%s is not of type BaseCommit" % repr(commit))
         if commit.repository != self and not isinstance(commit, EmptyCommit):
             raise ValueError(
                 "Commit %s must be a valid commit from this repository %s, "
                 "related to this repository instead %s." %
                 (commit, self, commit.repository))
 
     def _validate_commit_id(self, commit_id):
         if not isinstance(commit_id, str):
             raise TypeError(f"commit_id must be a string value got {type(commit_id)} instead")
 
     def _validate_commit_idx(self, commit_idx):
         if not isinstance(commit_idx, int):
             raise TypeError(f"commit_idx must be a numeric value, got {type(commit_idx)}")
 
     def _validate_branch_name(self, branch_name):
         if branch_name and branch_name not in self.branches_all:
-            msg = ("Branch {} not found in {}".format(branch_name, self))
+            msg = (f"Branch {branch_name} not found in {self}")
             raise BranchDoesNotExistError(msg)
 
     #
     # Supporting deprecated API parts
     # TODO: johbo: consider to move this into a mixin
     #
 
     @property
     def EMPTY_CHANGESET(self):
         warnings.warn(
             "Use EMPTY_COMMIT or EMPTY_COMMIT_ID instead", DeprecationWarning)
         return self.EMPTY_COMMIT_ID
 
     @property
     def revisions(self):
         warnings.warn("Use commits attribute instead", DeprecationWarning)
         return self.commit_ids
 
     @revisions.setter
     def revisions(self, value):
         warnings.warn("Use commits attribute instead", DeprecationWarning)
         self.commit_ids = value
 
     def get_changeset(self, revision=None, pre_load=None):
         warnings.warn("Use get_commit instead", DeprecationWarning)
         commit_id = None
         commit_idx = None
         if isinstance(revision, str):
             commit_id = revision
         else:
             commit_idx = revision
         return self.get_commit(
             commit_id=commit_id, commit_idx=commit_idx, pre_load=pre_load)
 
     def get_changesets(
             self, start=None, end=None, start_date=None, end_date=None,
             branch_name=None, pre_load=None):
         warnings.warn("Use get_commits instead", DeprecationWarning)
         start_id = self._revision_to_commit(start)
         end_id = self._revision_to_commit(end)
         return self.get_commits(
             start_id=start_id, end_id=end_id, start_date=start_date,
             end_date=end_date, branch_name=branch_name, pre_load=pre_load)
 
     def _revision_to_commit(self, revision):
         """
         Translates a revision to a commit_id
 
         Helps to support the old changeset based API which allows to use
         commit ids and commit indices interchangeable.
         """
         if revision is None:
             return revision
 
         if isinstance(revision, str):
             commit_id = revision
         else:
             commit_id = self.commit_ids[revision]
         return commit_id
 
     @property
     def in_memory_changeset(self):
         warnings.warn("Use in_memory_commit instead", DeprecationWarning)
         return self.in_memory_commit
 
     def get_path_permissions(self, username):
         """
         Returns a path permission checker or None if not supported
 
         :param username: session user name
         :return: an instance of BasePathPermissionChecker or None
         """
         return None
 
     def install_hooks(self, force=False):
         return self._remote.install_hooks(force)
 
     def get_hooks_info(self):
         return self._remote.get_hooks_info()
 
     def vcsserver_invalidate_cache(self, delete=False):
         return self._remote.vcsserver_invalidate_cache(delete)
 
 
 class BaseCommit(object):
     """
     Each backend should implement it's commit representation.
 
     **Attributes**
 
         ``repository``
             repository object within which commit exists
 
         ``id``
             The commit id, may be ``raw_id`` or i.e. for mercurial's tip
             just ``tip``.
 
         ``raw_id``
             raw commit representation (i.e. full 40 length sha for git
             backend)
 
         ``short_id``
             shortened (if apply) version of ``raw_id``; it would be simple
             shortcut for ``raw_id[:12]`` for git/mercurial backends or same
             as ``raw_id`` for subversion
 
         ``idx``
             commit index
 
         ``files``
             list of ``FileNode`` (``Node`` with NodeKind.FILE) objects
 
         ``dirs``
             list of ``DirNode`` (``Node`` with NodeKind.DIR) objects
 
         ``nodes``
             combined list of ``Node`` objects
 
         ``author``
             author of the commit, as unicode
 
         ``message``
             message of the commit, as unicode
 
         ``parents``
             list of parent commits
 
     """
     repository = None
     branch = None
 
     """
     Depending on the backend this should be set to the branch name of the
     commit. Backends not supporting branches on commits should leave this
     value as ``None``.
     """
 
     _ARCHIVE_PREFIX_TEMPLATE = '{repo_name}-{short_id}'
     """
     This template is used to generate a default prefix for repository archives
     if no prefix has been specified.
     """
 
     def __repr__(self):
         return self.__str__()
 
     def __str__(self):
         return f'<{self.__class__.__name__} at {self.idx}:{self.short_id}>'
 
     def __eq__(self, other):
         same_instance = isinstance(other, self.__class__)
         return same_instance and self.raw_id == other.raw_id
 
     def __json__(self):
         parents = []
         try:
             for parent in self.parents:
                 parents.append({'raw_id': parent.raw_id})
         except NotImplementedError:
             # empty commit doesn't have parents implemented
             pass
 
         return {
             'short_id': self.short_id,
             'raw_id': self.raw_id,
             'revision': self.idx,
             'message': self.message,
             'date': self.date,
             'author': self.author,
             'parents': parents,
             'branch': self.branch
         }
 
     def __getstate__(self):
         d = self.__dict__.copy()
         d.pop('_remote', None)
         d.pop('repository', None)
         return d
 
     def get_remote(self):
         return self._remote
 
     def serialize(self):
         return self.__json__()
 
     def _get_refs(self):
         return {
             'branches': [self.branch] if self.branch else [],
             'bookmarks': getattr(self, 'bookmarks', []),
             'tags': self.tags
         }
 
     @LazyProperty
     def last(self):
         """
         ``True`` if this is last commit in repository, ``False``
         otherwise; trying to access this attribute while there is no
         commits would raise `EmptyRepositoryError`
         """
         if self.repository is None:
             raise CommitError("Cannot check if it's most recent commit")
         return self.raw_id == self.repository.commit_ids[-1]
 
     @LazyProperty
     def parents(self):
         """
         Returns list of parent commits.
         """
         raise NotImplementedError
 
     @LazyProperty
     def first_parent(self):
         """
         Returns list of parent commits.
         """
         return self.parents[0] if self.parents else EmptyCommit()
 
     @property
     def merge(self):
         """
         Returns boolean if commit is a merge.
         """
         return len(self.parents) > 1
 
     @LazyProperty
     def children(self):
         """
         Returns list of child commits.
         """
         raise NotImplementedError
 
     @LazyProperty
     def id(self):
         """
         Returns string identifying this commit.
         """
         raise NotImplementedError
 
     @LazyProperty
     def raw_id(self):
         """
         Returns raw string identifying this commit.
         """
         raise NotImplementedError
 
     @LazyProperty
     def short_id(self):
         """
         Returns shortened version of ``raw_id`` attribute, as string,
         identifying this commit, useful for presentation to users.
         """
         raise NotImplementedError
 
     @LazyProperty
     def idx(self):
         """
         Returns integer identifying this commit.
         """
         raise NotImplementedError
 
     @LazyProperty
     def committer(self):
         """
         Returns committer for this commit
         """
         raise NotImplementedError
 
     @LazyProperty
     def committer_name(self):
         """
         Returns committer name for this commit
         """
 
         return author_name(self.committer)
 
     @LazyProperty
     def committer_email(self):
         """
         Returns committer email address for this commit
         """
 
         return author_email(self.committer)
 
     @LazyProperty
     def author(self):
         """
         Returns author for this commit
         """
 
         raise NotImplementedError
 
     @LazyProperty
     def author_name(self):
         """
         Returns author name for this commit
         """
 
         return author_name(self.author)
 
     @LazyProperty
     def author_email(self):
         """
         Returns author email address for this commit
         """
 
         return author_email(self.author)
 
     def get_file_mode(self, path: bytes):
         """
         Returns stat mode of the file at `path`.
         """
         raise NotImplementedError
 
     def is_link(self, path):
         """
         Returns ``True`` if given `path` is a symlink
         """
         raise NotImplementedError
 
     def is_node_binary(self, path):
         """
         Returns ``True`` is given path is a binary file
         """
         raise NotImplementedError
 
     def node_md5_hash(self, path):
         """
         Returns md5 hash of a node data
         """
         raise NotImplementedError
 
     def get_file_content(self, path) -> bytes:
         """
         Returns content of the file at the given `path`.
         """
         raise NotImplementedError
 
     def get_file_content_streamed(self, path):
         """
         returns a streaming response from vcsserver with file content
         """
         raise NotImplementedError
 
     def get_file_size(self, path):
         """
         Returns size of the file at the given `path`.
         """
         raise NotImplementedError
 
     def get_path_commit(self, path, pre_load=None):
         """
         Returns last commit of the file at the given `path`.
 
         :param pre_load: Optional. List of commit attributes to load.
         """
         commits = self.get_path_history(path, limit=1, pre_load=pre_load)
         if not commits:
             raise RepositoryError(
                 'Failed to fetch history for path {}. '
                 'Please check if such path exists in your repository'.format(
                     path))
         return commits[0]
 
     def get_path_history(self, path, limit=None, pre_load=None):
         """
         Returns history of file as reversed list of :class:`BaseCommit`
         objects for which file at given `path` has been modified.
 
         :param limit: Optional. Allows to limit the size of the returned
            history. This is intended as a hint to the underlying backend, so
            that it can apply optimizations depending on the limit.
         :param pre_load: Optional. List of commit attributes to load.
         """
         raise NotImplementedError
 
     def get_file_annotate(self, path, pre_load=None):
         """
         Returns a generator of four element tuples with
         lineno, sha, commit lazy loader and line
 
         :param pre_load: Optional. List of commit attributes to load.
         """
         raise NotImplementedError
 
     def get_nodes(self, path, pre_load=None):
         """
         Returns combined ``DirNode`` and ``FileNode`` objects list representing
         state of commit at the given ``path``.
 
         :raises ``CommitError``: if node at the given ``path`` is not
           instance of ``DirNode``
         """
         raise NotImplementedError
 
     def get_node(self, path):
         """
         Returns ``Node`` object from the given ``path``.
 
         :raises ``NodeDoesNotExistError``: if there is no node at the given
           ``path``
         """
         raise NotImplementedError
 
     def get_largefile_node(self, path):
         """
         Returns the path to largefile from Mercurial/Git-lfs storage.
         or None if it's not a largefile node
         """
         return None
 
     def archive_repo(self, archive_name_key, kind='tgz', subrepos=None,
                      archive_dir_name=None, write_metadata=False, mtime=None,
                      archive_at_path='/', cache_config=None):
         """
         Creates an archive containing the contents of the repository.
 
         :param archive_name_key: unique key under this archive should be generated
         :param kind: one of following: ``"tbz2"``, ``"tgz"``, ``"zip"``.
         :param archive_dir_name: name of root directory in archive.
             Default is repository name and commit's short_id joined with dash:
             ``"{repo_name}-{short_id}"``.
         :param write_metadata: write a metadata file into archive.
         :param mtime: custom modification time for archive creation, defaults
             to time.time() if not given.
         :param archive_at_path: pack files at this path (default '/')
         :param cache_config: config spec to send to vcsserver to configure the backend to store files
 
         :raise VCSError: If prefix has a problem.
         """
         cache_config = cache_config or {}
         allowed_kinds = [x[0] for x in settings.ARCHIVE_SPECS]
         if kind not in allowed_kinds:
             raise ImproperArchiveTypeError(
                 'Archive kind (%s) not supported use one of %s' %
                 (kind, allowed_kinds))
 
         archive_dir_name = self._validate_archive_prefix(archive_dir_name)
         mtime = mtime is not None or time.mktime(self.date.timetuple())
         commit_id = self.raw_id
 
         return self.repository._remote.archive_repo(
             archive_name_key, kind, mtime, archive_at_path,
             archive_dir_name, commit_id, cache_config)
 
     def _validate_archive_prefix(self, archive_dir_name):
         if archive_dir_name is None:
             archive_dir_name = self._ARCHIVE_PREFIX_TEMPLATE.format(
                 repo_name=safe_str(self.repository.name),
                 short_id=self.short_id)
         elif not isinstance(archive_dir_name, str):
             raise ValueError(f"archive_dir_name is not str object but: {type(archive_dir_name)}")
         elif archive_dir_name.startswith('/'):
             raise VCSError("Prefix cannot start with leading slash")
         elif archive_dir_name.strip() == '':
             raise VCSError("Prefix cannot be empty")
         elif not archive_dir_name.isascii():
             raise VCSError("Prefix cannot contain non ascii characters")
         return archive_dir_name
 
     @LazyProperty
     def root(self):
         """
         Returns ``RootNode`` object for this commit.
         """
         return self.get_node('')
 
     def next(self, branch=None):
         """
         Returns next commit from current, if branch is gives it will return
         next commit belonging to this branch
 
         :param branch: show commits within the given named branch
         """
         indexes = range(self.idx + 1, self.repository.count())
         return self._find_next(indexes, branch)
 
     def prev(self, branch=None):
         """
         Returns previous commit from current, if branch is gives it will
         return previous commit belonging to this branch
 
         :param branch: show commit within the given named branch
         """
         indexes = range(self.idx - 1, -1, -1)
         return self._find_next(indexes, branch)
 
     def _find_next(self, indexes, branch=None):
         if branch and self.branch != branch:
             raise VCSError('Branch option used on commit not belonging '
                            'to that branch')
 
         for next_idx in indexes:
             commit = self.repository.get_commit(commit_idx=next_idx)
             if branch and branch != commit.branch:
                 continue
             return commit
         raise CommitDoesNotExistError
 
     def diff(self, ignore_whitespace=True, context=3):
         """
         Returns a `Diff` object representing the change made by this commit.
         """
         parent = self.first_parent
         diff = self.repository.get_diff(
             parent, self,
             ignore_whitespace=ignore_whitespace,
             context=context)
         return diff
 
     @LazyProperty
     def added(self):
         """
         Returns list of added ``FileNode`` objects.
         """
         raise NotImplementedError
 
     @LazyProperty
     def changed(self):
         """
         Returns list of modified ``FileNode`` objects.
         """
         raise NotImplementedError
 
     @LazyProperty
     def removed(self):
         """
         Returns list of removed ``FileNode`` objects.
         """
         raise NotImplementedError
 
     @LazyProperty
     def size(self):
         """
         Returns total number of bytes from contents of all filenodes.
         """
         return sum(node.size for node in self.get_filenodes_generator())
 
     def walk(self, topurl=''):
         """
         Similar to os.walk method. Insted of filesystem it walks through
         commit starting at given ``topurl``.  Returns generator of tuples
         (top_node, dirnodes, filenodes).
         """
         from rhodecode.lib.vcs.nodes import DirNode
 
         if isinstance(topurl, DirNode):
             top_node = topurl
         else:
             top_node = self.get_node(topurl)
 
         has_default_pre_load = False
         if isinstance(top_node, DirNode):
             # used to inject as we walk same defaults as given top_node
             default_pre_load = top_node.default_pre_load
             has_default_pre_load = True
 
         if not top_node.is_dir():
             return
         yield top_node, top_node.dirs, top_node.files
         for dir_node in top_node.dirs:
             if has_default_pre_load:
                 dir_node.default_pre_load = default_pre_load
             yield from self.walk(dir_node)
 
     def get_filenodes_generator(self):
         """
         Returns generator that yields *all* file nodes.
         """
         for topnode, dirs, files in self.walk():
             yield from files
 
     #
     # Utilities for sub classes to support consistent behavior
     #
 
     def no_node_at_path(self, path):
         return NodeDoesNotExistError(
             f"There is no file nor directory at the given path: "
             f"`{safe_str(path)}` at commit {self.short_id}")
 
     def _fix_path(self, path: str) -> str:
         """
         Paths are stored without trailing slash so we need to get rid off it if
         needed.
         """
         return safe_str(path).rstrip('/')
 
     #
     # Deprecated API based on changesets
     #
 
     @property
     def revision(self):
         warnings.warn("Use idx instead", DeprecationWarning)
         return self.idx
 
     @revision.setter
     def revision(self, value):
         warnings.warn("Use idx instead", DeprecationWarning)
         self.idx = value
 
     def get_file_changeset(self, path):
         warnings.warn("Use get_path_commit instead", DeprecationWarning)
         return self.get_path_commit(path)
 
 
 class BaseChangesetClass(type):
 
     def __instancecheck__(self, instance):
         return isinstance(instance, BaseCommit)
 
 
 class BaseChangeset(BaseCommit, metaclass=BaseChangesetClass):
 
     def __new__(cls, *args, **kwargs):
         warnings.warn(
             "Use BaseCommit instead of BaseChangeset", DeprecationWarning)
         return super().__new__(cls, *args, **kwargs)
 
 
 class BaseInMemoryCommit(object):
     """
     Represents differences between repository's state (most recent head) and
     changes made *in place*.
 
     **Attributes**
 
         ``repository``
             repository object for this in-memory-commit
 
         ``added``
             list of ``FileNode`` objects marked as *added*
 
         ``changed``
             list of ``FileNode`` objects marked as *changed*
 
         ``removed``
             list of ``FileNode`` or ``RemovedFileNode`` objects marked to be
             *removed*
 
         ``parents``
             list of :class:`BaseCommit` instances representing parents of
             in-memory commit. Should always be 2-element sequence.
 
     """
 
     def __init__(self, repository):
         self.repository = repository
         self.added = []
         self.changed = []
         self.removed = []
         self.parents = []
 
     def add(self, *filenodes):
         """
         Marks given ``FileNode`` objects as *to be committed*.
 
         :raises ``NodeAlreadyExistsError``: if node with same path exists at
            latest commit
         :raises ``NodeAlreadyAddedError``: if node with same path is already
            marked as *added*
         """
         # Check if not already marked as *added* first
         for node in filenodes:
             if node.path in (n.path for n in self.added):
                 raise NodeAlreadyAddedError(
                     "Such FileNode %s is already marked for addition"
                     % node.path)
         for node in filenodes:
             self.added.append(node)
 
     def change(self, *filenodes):
         """
         Marks given ``FileNode`` objects to be *changed* in next commit.
 
         :raises ``EmptyRepositoryError``: if there are no commits yet
         :raises ``NodeAlreadyExistsError``: if node with same path is already
            marked to be *changed*
         :raises ``NodeAlreadyRemovedError``: if node with same path is already
            marked to be *removed*
         :raises ``NodeDoesNotExistError``: if node doesn't exist in latest
            commit
         :raises ``NodeNotChangedError``: if node hasn't really be changed
         """
         for node in filenodes:
             if node.path in (n.path for n in self.removed):
                 raise NodeAlreadyRemovedError(
                     "Node at %s is already marked as removed" % node.path)
         try:
             self.repository.get_commit()
         except EmptyRepositoryError:
             raise EmptyRepositoryError(
                 "Nothing to change - try to *add* new nodes rather than "
                 "changing them")
         for node in filenodes:
             if node.path in (n.path for n in self.changed):
                 raise NodeAlreadyChangedError(
                     "Node at '%s' is already marked as changed" % node.path)
             self.changed.append(node)
 
     def remove(self, *filenodes):
         """
         Marks given ``FileNode`` (or ``RemovedFileNode``) objects to be
         *removed* in next commit.
 
         :raises ``NodeAlreadyRemovedError``: if node has been already marked to
            be *removed*
         :raises ``NodeAlreadyChangedError``: if node has been already marked to
            be *changed*
         """
         for node in filenodes:
             if node.path in (n.path for n in self.removed):
                 raise NodeAlreadyRemovedError(
                     "Node is already marked to for removal at %s" % node.path)
             if node.path in (n.path for n in self.changed):
                 raise NodeAlreadyChangedError(
                     "Node is already marked to be changed at %s" % node.path)
             # We only mark node as *removed* - real removal is done by
             # commit method
             self.removed.append(node)
 
     def reset(self):
         """
         Resets this instance to initial state (cleans ``added``, ``changed``
         and ``removed`` lists).
         """
         self.added = []
         self.changed = []
         self.removed = []
         self.parents = []
 
     def get_ipaths(self):
         """
         Returns generator of paths from nodes marked as added, changed or
         removed.
         """
         for node in itertools.chain(self.added, self.changed, self.removed):
             yield node.path
 
     def get_paths(self):
         """
         Returns list of paths from nodes marked as added, changed or removed.
         """
         return list(self.get_ipaths())
 
     def check_integrity(self, parents=None):
         """
         Checks in-memory commit's integrity. Also, sets parents if not
         already set.
 
         :raises CommitError: if any error occurs (i.e.
           ``NodeDoesNotExistError``).
         """
         if not self.parents:
             parents = parents or []
             if len(parents) == 0:
                 try:
                     parents = [self.repository.get_commit(), None]
                 except EmptyRepositoryError:
                     parents = [None, None]
             elif len(parents) == 1:
                 parents += [None]
             self.parents = parents
 
         # Local parents, only if not None
         parents = [p for p in self.parents if p]
 
         # Check nodes marked as added
         for p in parents:
             for node in self.added:
                 try:
                     p.get_node(node.path)
                 except NodeDoesNotExistError:
                     pass
                 else:
                     raise NodeAlreadyExistsError(
-                        "Node `{}` already exists at {}".format(node.path, p))
+                        f"Node `{node.path}` already exists at {p}")
 
         # Check nodes marked as changed
         missing = set(self.changed)
         not_changed = set(self.changed)
         if self.changed and not parents:
             raise NodeDoesNotExistError(str(self.changed[0].path))
         for p in parents:
             for node in self.changed:
                 try:
                     old = p.get_node(node.path)
                     missing.remove(node)
                     # if content actually changed, remove node from not_changed
                     if old.content != node.content:
                         not_changed.remove(node)
                 except NodeDoesNotExistError:
                     pass
         if self.changed and missing:
             raise NodeDoesNotExistError(
                 "Node `%s` marked as modified but missing in parents: %s"
                 % (node.path, parents))
 
         if self.changed and not_changed:
             raise NodeNotChangedError(
                 "Node `%s` wasn't actually changed (parents: %s)"
                 % (not_changed.pop().path, parents))
 
         # Check nodes marked as removed
         if self.removed and not parents:
             raise NodeDoesNotExistError(
                 "Cannot remove node at %s as there "
                 "were no parents specified" % self.removed[0].path)
         really_removed = set()
         for p in parents:
             for node in self.removed:
                 try:
                     p.get_node(node.path)
                     really_removed.add(node)
                 except CommitError:
                     pass
         not_removed = set(self.removed) - really_removed
         if not_removed:
             # TODO: johbo: This code branch does not seem to be covered
             raise NodeDoesNotExistError(
                 "Cannot remove node at %s from "
                 "following parents: %s" % (not_removed, parents))
 
     def commit(self, message, author, parents=None, branch=None, date=None, **kwargs):
         """
         Performs in-memory commit (doesn't check workdir in any way) and
         returns newly created :class:`BaseCommit`. Updates repository's
         attribute `commits`.
 
         .. note::
 
            While overriding this method each backend's should call
            ``self.check_integrity(parents)`` in the first place.
 
         :param message: message of the commit
         :param author: full username, i.e. "Joe Doe <[e-mail]>"
         :param parents: single parent or sequence of parents from which commit
            would be derived
         :param date: ``datetime.datetime`` instance. Defaults to
            ``datetime.datetime.now()``.
         :param branch: branch name, as string. If none given, default backend's
            branch would be used.
 
         :raises ``CommitError``: if any error occurs while committing
         """
         raise NotImplementedError
 
 
 class BaseInMemoryChangesetClass(type):
 
     def __instancecheck__(self, instance):
         return isinstance(instance, BaseInMemoryCommit)
 
 
 class BaseInMemoryChangeset(BaseInMemoryCommit, metaclass=BaseInMemoryChangesetClass):
 
     def __new__(cls, *args, **kwargs):
         warnings.warn(
             "Use BaseCommit instead of BaseInMemoryCommit", DeprecationWarning)
         return super().__new__(cls, *args, **kwargs)
 
 
 class EmptyCommit(BaseCommit):
     """
     An dummy empty commit. It's possible to pass hash when creating
     an EmptyCommit
     """
 
     def __init__(
             self, commit_id=EMPTY_COMMIT_ID, repo=None, alias=None, idx=-1,
             message='', author='', date=None):
         self._empty_commit_id = commit_id
         # TODO: johbo: Solve idx parameter, default value does not make
         # too much sense
         self.idx = idx
         self.message = message
         self.author = author
         self.date = date or datetime.datetime.fromtimestamp(0)
         self.repository = repo
         self.alias = alias
 
     @LazyProperty
     def raw_id(self):
         """
         Returns raw string identifying this commit, useful for web
         representation.
         """
 
         return self._empty_commit_id
 
     @LazyProperty
     def branch(self):
         if self.alias:
             from rhodecode.lib.vcs.backends import get_backend
             return get_backend(self.alias).DEFAULT_BRANCH_NAME
 
     @LazyProperty
     def short_id(self):
         return self.raw_id[:12]
 
     @LazyProperty
     def id(self):
         return self.raw_id
 
     def get_path_commit(self, path, pre_load=None):
         return self
 
     def get_file_content(self, path) -> bytes:
         return b''
 
     def get_file_content_streamed(self, path):
         yield self.get_file_content(path)
 
     def get_file_size(self, path):
         return 0
 
 
 class EmptyChangesetClass(type):
 
     def __instancecheck__(self, instance):
         return isinstance(instance, EmptyCommit)
 
 
 class EmptyChangeset(EmptyCommit, metaclass=EmptyChangesetClass):
 
     def __new__(cls, *args, **kwargs):
         warnings.warn(
             "Use EmptyCommit instead of EmptyChangeset", DeprecationWarning)
         return super(EmptyCommit, cls).__new__(cls, *args, **kwargs)
 
     def __init__(self, cs=EMPTY_COMMIT_ID, repo=None, requested_revision=None,
                  alias=None, revision=-1, message='', author='', date=None):
         if requested_revision is not None:
             warnings.warn(
                 "Parameter requested_revision not supported anymore",
                 DeprecationWarning)
         super().__init__(
             commit_id=cs, repo=repo, alias=alias, idx=revision,
             message=message, author=author, date=date)
 
         @property
         def revision(self):
             warnings.warn("Use idx instead", DeprecationWarning)
             return self.idx
 
         @revision.setter
         def revision(self, value):
             warnings.warn("Use idx instead", DeprecationWarning)
             self.idx = value
 
 
 class EmptyRepository(BaseRepository):
     def __init__(self, repo_path=None, config=None, create=False, **kwargs):
         pass
 
     def get_diff(self, *args, **kwargs):
         from rhodecode.lib.vcs.backends.git.diff import GitDiff
         return GitDiff(b'')
 
 
 class CollectionGenerator(object):
 
     def __init__(self, repo, commit_ids, collection_size=None, pre_load=None, translate_tag=None):
         self.repo = repo
         self.commit_ids = commit_ids
         self.collection_size = collection_size
         self.pre_load = pre_load
         self.translate_tag = translate_tag
 
     def __len__(self):
         if self.collection_size is not None:
             return self.collection_size
         return self.commit_ids.__len__()
 
     def __iter__(self):
         for commit_id in self.commit_ids:
             # TODO: johbo: Mercurial passes in commit indices or commit ids
             yield self._commit_factory(commit_id)
 
     def _commit_factory(self, commit_id):
         """
         Allows backends to override the way commits are generated.
         """
         return self.repo.get_commit(
             commit_id=commit_id, pre_load=self.pre_load,
             translate_tag=self.translate_tag)
 
     def __getitem__(self, key):
         """Return either a single element by index, or a sliced collection."""
 
         if isinstance(key, slice):
             commit_ids = self.commit_ids[key.start:key.stop]
 
         else:
             # single item
             commit_ids = self.commit_ids[key]
 
         return self.__class__(
             self.repo, commit_ids, pre_load=self.pre_load,
             translate_tag=self.translate_tag)
 
     def __repr__(self):
         return '<CollectionGenerator[len:%s]>' % (self.__len__())
 
 
 class Config(object):
     """
     Represents the configuration for a repository.
 
     The API is inspired by :class:`ConfigParser.ConfigParser` from the
     standard library. It implements only the needed subset.
     """
 
     def __init__(self):
         self._values = {}
 
     def copy(self):
         clone = Config()
         for section, values in self._values.items():
             clone._values[section] = values.copy()
         return clone
 
     def __repr__(self):
         return '<Config({} sections) at {}>'.format(
             len(self._values), hex(id(self)))
 
     def items(self, section):
         return self._values.get(section, {}).items()
 
     def get(self, section, option):
         return self._values.get(section, {}).get(option)
 
     def set(self, section, option, value):
         section_values = self._values.setdefault(section, {})
         section_values[option] = value
 
     def clear_section(self, section):
         self._values[section] = {}
 
     def serialize(self):
         """
         Creates a list of three tuples (section, key, value) representing
         this config object.
         """
         items = []
         for section in self._values:
             for option, value in self._values[section].items():
                 items.append(
                     (safe_str(section), safe_str(option), safe_str(value)))
         return items
 
 
 class Diff(object):
     """
     Represents a diff result from a repository backend.
 
     Subclasses have to provide a backend specific value for
     :attr:`_header_re` and :attr:`_meta_re`.
     """
     _meta_re = None
     _header_re: bytes = re.compile(br"")
 
     def __init__(self, raw_diff: bytes):
         if not isinstance(raw_diff, bytes):
             raise Exception(f'raw_diff must be bytes - got {type(raw_diff)}')
 
         self.raw = memoryview(raw_diff)
 
     def get_header_re(self):
         return self._header_re
 
     def chunks(self):
         """
         split the diff in chunks of separate --git a/file b/file chunks
         to make diffs consistent we must prepend with \n, and make sure
         we can detect last chunk as this was also has special rule
         """
 
         diff_parts = (b'\n' + bytes(self.raw)).split(b'\ndiff --git')
 
         chunks = diff_parts[1:]
         total_chunks = len(chunks)
 
         def diff_iter(_chunks):
             for cur_chunk, chunk in enumerate(_chunks, start=1):
                 yield DiffChunk(chunk, self, cur_chunk == total_chunks)
         return diff_iter(chunks)
 
 
 class DiffChunk(object):
 
     def __init__(self, chunk: bytes, diff_obj: Diff, is_last_chunk: bool):
         self.diff_obj = diff_obj
 
         # since we split by \ndiff --git that part is lost from original diff
         # we need to re-apply it at the end, EXCEPT ! if it's last chunk
         if not is_last_chunk:
             chunk += b'\n'
         header_re = self.diff_obj.get_header_re()
         match = header_re.match(chunk)
         self.header = match.groupdict()
         self.diff = chunk[match.end():]
         self.raw = chunk
 
     @property
     def header_as_str(self):
         if self.header:
             def safe_str_on_bytes(val):
                 if isinstance(val, bytes):
                     return safe_str(val)
                 return val
             return {safe_str(k): safe_str_on_bytes(v) for k, v in self.header.items()}
 
     def __repr__(self):
         return f'DiffChunk({self.header_as_str})'
 
 
 class BasePathPermissionChecker(object):
 
     @staticmethod
     def create_from_patterns(includes, excludes):
         if includes and '*' in includes and not excludes:
             return AllPathPermissionChecker()
         elif excludes and '*' in excludes:
             return NonePathPermissionChecker()
         else:
             return PatternPathPermissionChecker(includes, excludes)
 
     @property
     def has_full_access(self):
         raise NotImplementedError()
 
     def has_access(self, path):
         raise NotImplementedError()
 
 
 class AllPathPermissionChecker(BasePathPermissionChecker):
 
     @property
     def has_full_access(self):
         return True
 
     def has_access(self, path):
         return True
 
 
 class NonePathPermissionChecker(BasePathPermissionChecker):
 
     @property
     def has_full_access(self):
         return False
 
     def has_access(self, path):
         return False
 
 
 class PatternPathPermissionChecker(BasePathPermissionChecker):
 
     def __init__(self, includes, excludes):
         self.includes = includes
         self.excludes = excludes
         self.includes_re = [] if not includes else [
             re.compile(fnmatch.translate(pattern)) for pattern in includes]
         self.excludes_re = [] if not excludes else [
             re.compile(fnmatch.translate(pattern)) for pattern in excludes]
 
     @property
     def has_full_access(self):
         return '*' in self.includes and not self.excludes
 
     def has_access(self, path):
         for regex in self.excludes_re:
             if regex.match(path):
                 return False
         for regex in self.includes_re:
             if regex.match(path):
                 return True
         return False

rhodecode/lib/vcs/backends/git/repository.py

@@ -1,1053 +1,1053 b''
 # Copyright (C) 2014-2023 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/
 
 """
 GIT repository module
 """
 
 import logging
 import os
 import re
 
 from zope.cachedescriptors.property import Lazy as LazyProperty
 
 from collections import OrderedDict
 from rhodecode.lib.datelib import (
     utcdate_fromtimestamp, makedate, date_astimestamp)
 from rhodecode.lib.hash_utils import safe_str
 from rhodecode.lib.utils2 import CachedProperty
 from rhodecode.lib.vcs import connection, path as vcspath
 from rhodecode.lib.vcs.backends.base import (
     BaseRepository, CollectionGenerator, Config, MergeResponse,
     MergeFailureReason, Reference)
 from rhodecode.lib.vcs.backends.git.commit import GitCommit
 from rhodecode.lib.vcs.backends.git.diff import GitDiff
 from rhodecode.lib.vcs.backends.git.inmemory import GitInMemoryCommit
 from rhodecode.lib.vcs.exceptions import (
     CommitDoesNotExistError, EmptyRepositoryError,
     RepositoryError, TagAlreadyExistError, TagDoesNotExistError, VCSError, UnresolvedFilesInRepo)
 
 
 SHA_PATTERN = re.compile(r'^([0-9a-fA-F]{12}|[0-9a-fA-F]{40})$')
 
 log = logging.getLogger(__name__)
 
 
 class GitRepository(BaseRepository):
     """
     Git repository backend.
     """
     DEFAULT_BRANCH_NAME = os.environ.get('GIT_DEFAULT_BRANCH_NAME') or 'master'
     DEFAULT_REF = f'branch:{DEFAULT_BRANCH_NAME}'
 
     contact = BaseRepository.DEFAULT_CONTACT
 
     def __init__(self, repo_path, config=None, create=False, src_url=None,
                  do_workspace_checkout=False, with_wire=None, bare=False):
 
         self.path = safe_str(os.path.abspath(repo_path))
         self.config = config if config else self.get_default_config()
         self.with_wire = with_wire or {"cache": False}  # default should not use cache
 
         self._init_repo(create, src_url, do_workspace_checkout, bare)
 
         # caches
         self._commit_ids = {}
 
     @LazyProperty
     def _remote(self):
         repo_id = self.path
         return connection.Git(self.path, repo_id, self.config, with_wire=self.with_wire)
 
     @LazyProperty
     def bare(self):
         return self._remote.bare()
 
     @LazyProperty
     def head(self):
         return self._remote.head()
 
     @CachedProperty
     def commit_ids(self):
         """
         Returns list of commit ids, in ascending order.  Being lazy
         attribute allows external tools to inject commit ids from cache.
         """
         commit_ids = self._get_all_commit_ids()
         self._rebuild_cache(commit_ids)
         return commit_ids
 
     def _rebuild_cache(self, commit_ids):
         self._commit_ids = {commit_id: index
                                 for index, commit_id in enumerate(commit_ids)}
 
     def run_git_command(self, cmd, **opts):
         """
         Runs given ``cmd`` as git command and returns tuple
         (stdout, stderr).
 
         :param cmd: git command to be executed
         :param opts: env options to pass into Subprocess command
         """
         if not isinstance(cmd, list):
             raise ValueError(f'cmd must be a list, got {type(cmd)} instead')
 
         skip_stderr_log = opts.pop('skip_stderr_log', False)
         out, err = self._remote.run_git_command(cmd, **opts)
         if err and not skip_stderr_log:
             log.debug('Stderr output of git command "%s":\n%s', cmd, err)
         return out, err
 
     @staticmethod
     def check_url(url, config):
         """
         Function will check given url and try to verify if it's a valid
         link. Sometimes it may happened that git will issue basic
         auth request that can cause whole API to hang when used from python
         or other external calls.
 
         On failures it'll raise urllib2.HTTPError, exception is also thrown
         when the return code is non 200
         """
         # check first if it's not an url
         if os.path.isdir(url) or url.startswith('file:'):
             return True
 
         if '+' in url.split('://', 1)[0]:
             url = url.split('+', 1)[1]
 
         # Request the _remote to verify the url
         return connection.Git.check_url(url, config.serialize())
 
     @staticmethod
     def is_valid_repository(path):
         if os.path.isdir(os.path.join(path, '.git')):
             return True
         # check case of bare repository
         try:
             GitRepository(path)
             return True
         except VCSError:
             pass
         return False
 
     def _init_repo(self, create, src_url=None, do_workspace_checkout=False,
                    bare=False):
         if create and os.path.exists(self.path):
             raise RepositoryError(
                 "Cannot create repository at %s, location already exist"
                 % self.path)
 
         if bare and do_workspace_checkout:
             raise RepositoryError("Cannot update a bare repository")
         try:
 
             if src_url:
                 # check URL before any actions
                 GitRepository.check_url(src_url, self.config)
 
             if create:
                 os.makedirs(self.path, mode=0o755)
 
                 if bare:
                     self._remote.init_bare()
                 else:
                     self._remote.init()
 
                 if src_url and bare:
                     # bare repository only allows a fetch and checkout is not allowed
                     self.fetch(src_url, commit_ids=None)
                 elif src_url:
                     self.pull(src_url, commit_ids=None,
                               update_after=do_workspace_checkout)
 
             else:
                 if not self._remote.assert_correct_path():
                     raise RepositoryError(
                         'Path "%s" does not contain a Git repository' %
                         (self.path,))
 
         # TODO: johbo: check if we have to translate the OSError here
         except OSError as err:
             raise RepositoryError(err)
 
     def _get_all_commit_ids(self):
         return self._remote.get_all_commit_ids()
 
     def _get_commit_ids(self, filters=None):
         # we must check if this repo is not empty, since later command
         # fails if it is. And it's cheaper to ask than throw the subprocess
         # errors
 
         head = self._remote.head(show_exc=False)
 
         if not head:
             return []
 
         rev_filter = ['--branches', '--tags']
         extra_filter = []
 
         if filters:
             if filters.get('since'):
                 extra_filter.append('--since=%s' % (filters['since']))
             if filters.get('until'):
                 extra_filter.append('--until=%s' % (filters['until']))
             if filters.get('branch_name'):
                 rev_filter = []
                 extra_filter.append(filters['branch_name'])
         rev_filter.extend(extra_filter)
 
             # if filters.get('start') or filters.get('end'):
             #     # skip is offset, max-count is limit
             #     if filters.get('start'):
             #         extra_filter += ' --skip=%s' % filters['start']
             #     if filters.get('end'):
             #         extra_filter += ' --max-count=%s' % (filters['end'] - (filters['start'] or 0))
 
         cmd = ['rev-list', '--reverse', '--date-order'] + rev_filter
         try:
             output, __ = self.run_git_command(cmd)
         except RepositoryError:
             # Can be raised for empty repositories
             return []
         return output.splitlines()
 
     def _lookup_commit(self, commit_id_or_idx, translate_tag=True, maybe_unreachable=False, reference_obj=None):
 
         def is_null(value):
             return len(value) == commit_id_or_idx.count('0')
 
         if commit_id_or_idx in (None, '', 'tip', 'HEAD', 'head', -1):
             return self.commit_ids[-1]
 
         commit_missing_err = "Commit {} does not exist for `{}`".format(
             *map(safe_str, [commit_id_or_idx, self.name]))
 
         is_bstr = isinstance(commit_id_or_idx, str)
         is_branch = reference_obj and reference_obj.branch
 
         lookup_ok = False
         if is_bstr:
             # Need to call remote to translate id for tagging scenarios,
             # or branch that are numeric
             try:
                 remote_data = self._remote.get_object(commit_id_or_idx,
                                                       maybe_unreachable=maybe_unreachable)
                 commit_id_or_idx = remote_data["commit_id"]
                 lookup_ok = True
             except (CommitDoesNotExistError,):
                 lookup_ok = False
 
         if lookup_ok is False:
             is_numeric_idx = \
                 (is_bstr and commit_id_or_idx.isdigit() and len(commit_id_or_idx) < 12) \
                 or isinstance(commit_id_or_idx, int)
             if not is_branch and (is_numeric_idx or is_null(commit_id_or_idx)):
                 try:
                     commit_id_or_idx = self.commit_ids[int(commit_id_or_idx)]
                     lookup_ok = True
                 except Exception:
                     raise CommitDoesNotExistError(commit_missing_err)
 
         # we failed regular lookup, and by integer number lookup
         if lookup_ok is False:
             raise CommitDoesNotExistError(commit_missing_err)
 
         # Ensure we return full id
         if not SHA_PATTERN.match(str(commit_id_or_idx)):
             raise CommitDoesNotExistError(
                 "Given commit id %s not recognized" % commit_id_or_idx)
         return commit_id_or_idx
 
     def get_hook_location(self):
         """
         returns absolute path to location where hooks are stored
         """
         loc = os.path.join(self.path, 'hooks')
         if not self.bare:
             loc = os.path.join(self.path, '.git', 'hooks')
         return loc
 
     @LazyProperty
     def last_change(self):
         """
         Returns last change made on this repository as
         `datetime.datetime` object.
         """
         try:
             return self.get_commit().date
         except RepositoryError:
             tzoffset = makedate()[1]
             return utcdate_fromtimestamp(self._get_fs_mtime(), tzoffset)
 
     def _get_fs_mtime(self):
         idx_loc = '' if self.bare else '.git'
         # fallback to filesystem
         in_path = os.path.join(self.path, idx_loc, "index")
         he_path = os.path.join(self.path, idx_loc, "HEAD")
         if os.path.exists(in_path):
             return os.stat(in_path).st_mtime
         else:
             return os.stat(he_path).st_mtime
 
     @LazyProperty
     def description(self):
         description = self._remote.get_description()
         return safe_str(description or self.DEFAULT_DESCRIPTION)
 
     def _get_refs_entries(self, prefix='', reverse=False, strip_prefix=True):
         if self.is_empty():
             return OrderedDict()
 
         result = []
         for ref, sha in self._refs.items():
             if ref.startswith(prefix):
                 ref_name = ref
                 if strip_prefix:
                     ref_name = ref[len(prefix):]
                 result.append((safe_str(ref_name), sha))
 
         def get_name(entry):
             return entry[0]
 
         return OrderedDict(sorted(result, key=get_name, reverse=reverse))
 
     def _get_branches(self):
         return self._get_refs_entries(prefix='refs/heads/', strip_prefix=True)
 
     @CachedProperty
     def branches(self):
         return self._get_branches()
 
     @CachedProperty
     def branches_closed(self):
         return {}
 
     @CachedProperty
     def bookmarks(self):
         return {}
 
     @CachedProperty
     def branches_all(self):
         all_branches = {}
         all_branches.update(self.branches)
         all_branches.update(self.branches_closed)
         return all_branches
 
     @CachedProperty
     def tags(self):
         return self._get_tags()
 
     def _get_tags(self):
         return self._get_refs_entries(prefix='refs/tags/', strip_prefix=True, reverse=True)
 
     def tag(self, name, user, commit_id=None, message=None, date=None,
             **kwargs):
         # TODO: fix this method to apply annotated tags correct with message
         """
         Creates and returns a tag for the given ``commit_id``.
 
         :param name: name for new tag
         :param user: full username, i.e.: "Joe Doe <[e-mail]>"
         :param commit_id: commit id for which new tag would be created
         :param message: message of the tag's commit
         :param date: date of tag's commit
 
         :raises TagAlreadyExistError: if tag with same name already exists
         """
         if name in self.tags:
             raise TagAlreadyExistError("Tag %s already exists" % name)
         commit = self.get_commit(commit_id=commit_id)
-        message = message or "Added tag {} for commit {}".format(name, commit.raw_id)
+        message = message or f"Added tag {name} for commit {commit.raw_id}"
 
         self._remote.set_refs('refs/tags/%s' % name, commit.raw_id)
 
         self._invalidate_prop_cache('tags')
         self._invalidate_prop_cache('_refs')
 
         return commit
 
     def remove_tag(self, name, user, message=None, date=None):
         """
         Removes tag with the given ``name``.
 
         :param name: name of the tag to be removed
         :param user: full username, i.e.: "Joe Doe <[e-mail]>"
         :param message: message of the tag's removal commit
         :param date: date of tag's removal commit
 
         :raises TagDoesNotExistError: if tag with given name does not exists
         """
         if name not in self.tags:
             raise TagDoesNotExistError("Tag %s does not exist" % name)
 
         self._remote.tag_remove(name)
         self._invalidate_prop_cache('tags')
         self._invalidate_prop_cache('_refs')
 
     def _get_refs(self):
         return self._remote.get_refs()
 
     @CachedProperty
     def _refs(self):
         return self._get_refs()
 
     @property
     def _ref_tree(self):
         node = tree = {}
         for ref, sha in self._refs.items():
             path = ref.split('/')
             for bit in path[:-1]:
                 node = node.setdefault(bit, {})
             node[path[-1]] = sha
             node = tree
         return tree
 
     def get_remote_ref(self, ref_name):
         ref_key = f'refs/remotes/origin/{safe_str(ref_name)}'
         try:
             return self._refs[ref_key]
         except Exception:
             return
 
     def get_commit(self, commit_id=None, commit_idx=None, pre_load=None,
                    translate_tag=True, maybe_unreachable=False, reference_obj=None):
         """
         Returns `GitCommit` object representing commit from git repository
         at the given `commit_id` or head (most recent commit) if None given.
         """
 
         if self.is_empty():
             raise EmptyRepositoryError("There are no commits yet")
 
         if commit_id is not None:
             self._validate_commit_id(commit_id)
             try:
                 # we have cached idx, use it without contacting the remote
                 idx = self._commit_ids[commit_id]
                 return GitCommit(self, commit_id, idx, pre_load=pre_load)
             except KeyError:
                 pass
 
         elif commit_idx is not None:
             self._validate_commit_idx(commit_idx)
             try:
                 _commit_id = self.commit_ids[commit_idx]
                 if commit_idx < 0:
                     commit_idx = self.commit_ids.index(_commit_id)
                 return GitCommit(self, _commit_id, commit_idx, pre_load=pre_load)
             except IndexError:
                 commit_id = commit_idx
         else:
             commit_id = "tip"
 
         if translate_tag:
             commit_id = self._lookup_commit(
                 commit_id, maybe_unreachable=maybe_unreachable,
                 reference_obj=reference_obj)
 
         try:
             idx = self._commit_ids[commit_id]
         except KeyError:
             idx = -1
 
         return GitCommit(self, commit_id, idx, pre_load=pre_load)
 
     def get_commits(
             self, start_id=None, end_id=None, start_date=None, end_date=None,
             branch_name=None, show_hidden=False, pre_load=None, translate_tags=True):
         """
         Returns generator of `GitCommit` objects from start to end (both
         are inclusive), in ascending date order.
 
         :param start_id: None, str(commit_id)
         :param end_id: None, str(commit_id)
         :param start_date: if specified, commits with commit date less than
           ``start_date`` would be filtered out from returned set
         :param end_date: if specified, commits with commit date greater than
           ``end_date`` would be filtered out from returned set
         :param branch_name: if specified, commits not reachable from given
           branch would be filtered out from returned set
         :param show_hidden: Show hidden commits such as obsolete or hidden from
             Mercurial evolve
         :raise BranchDoesNotExistError: If given `branch_name` does not
             exist.
         :raise CommitDoesNotExistError: If commits for given `start` or
           `end` could not be found.
 
         """
         if self.is_empty():
             raise EmptyRepositoryError("There are no commits yet")
 
         self._validate_branch_name(branch_name)
 
         if start_id is not None:
             self._validate_commit_id(start_id)
         if end_id is not None:
             self._validate_commit_id(end_id)
 
         start_raw_id = self._lookup_commit(start_id)
         start_pos = self._commit_ids[start_raw_id] if start_id else None
         end_raw_id = self._lookup_commit(end_id)
         end_pos = max(0, self._commit_ids[end_raw_id]) if end_id else None
 
         if None not in [start_id, end_id] and start_pos > end_pos:
             raise RepositoryError(
                 "Start commit '%s' cannot be after end commit '%s'" %
                 (start_id, end_id))
 
         if end_pos is not None:
             end_pos += 1
 
         filter_ = []
         if branch_name:
             filter_.append({'branch_name': branch_name})
         if start_date and not end_date:
             filter_.append({'since': start_date})
         if end_date and not start_date:
             filter_.append({'until': end_date})
         if start_date and end_date:
             filter_.append({'since': start_date})
             filter_.append({'until': end_date})
 
         # if start_pos or end_pos:
         #     filter_.append({'start': start_pos})
         #     filter_.append({'end': end_pos})
 
         if filter_:
             revfilters = {
                 'branch_name': branch_name,
                 'since': start_date.strftime('%m/%d/%y %H:%M:%S') if start_date else None,
                 'until': end_date.strftime('%m/%d/%y %H:%M:%S') if end_date else None,
                 'start': start_pos,
                 'end': end_pos,
             }
             commit_ids = self._get_commit_ids(filters=revfilters)
 
         else:
             commit_ids = self.commit_ids
 
         if start_pos or end_pos:
             commit_ids = commit_ids[start_pos: end_pos]
 
         return CollectionGenerator(self, commit_ids, pre_load=pre_load,
                                    translate_tag=translate_tags)
 
     def get_diff(
             self, commit1, commit2, path='', ignore_whitespace=False,
             context=3, path1=None):
         """
         Returns (git like) *diff*, as plain text. Shows changes introduced by
         ``commit2`` since ``commit1``.
 
         :param commit1: Entry point from which diff is shown. Can be
           ``self.EMPTY_COMMIT`` - in this case, patch showing all
           the changes since empty state of the repository until ``commit2``
         :param commit2: Until which commits changes should be shown.
         :param path:
         :param ignore_whitespace: If set to ``True``, would not show whitespace
           changes. Defaults to ``False``.
         :param context: How many lines before/after changed lines should be
           shown. Defaults to ``3``.
         :param path1:
         """
         self._validate_diff_commits(commit1, commit2)
         if path1 is not None and path1 != path:
             raise ValueError("Diff of two different paths not supported.")
 
         if path:
             file_filter = path
         else:
             file_filter = None
 
         diff = self._remote.diff(
             commit1.raw_id, commit2.raw_id, file_filter=file_filter,
             opt_ignorews=ignore_whitespace,
             context=context)
 
         return GitDiff(diff)
 
     def strip(self, commit_id, branch_name):
         commit = self.get_commit(commit_id=commit_id)
         if commit.merge:
             raise Exception('Cannot reset to merge commit')
 
         # parent is going to be the new head now
         commit = commit.parents[0]
         self._remote.set_refs('refs/heads/%s' % branch_name, commit.raw_id)
 
         # clear cached properties
         self._invalidate_prop_cache('commit_ids')
         self._invalidate_prop_cache('_refs')
         self._invalidate_prop_cache('branches')
 
         return len(self.commit_ids)
 
     def get_common_ancestor(self, commit_id1, commit_id2, repo2):
         log.debug('Calculating common ancestor between %sc1:%s and %sc2:%s',
                   self, commit_id1, repo2, commit_id2)
 
         if commit_id1 == commit_id2:
             return commit_id1
 
         if self != repo2:
             commits = self._remote.get_missing_revs(
                 commit_id1, commit_id2, repo2.path)
             if commits:
                 commit = repo2.get_commit(commits[-1])
                 if commit.parents:
                     ancestor_id = commit.parents[0].raw_id
                 else:
                     ancestor_id = None
             else:
                 # no commits from other repo, ancestor_id is the commit_id2
                 ancestor_id = commit_id2
         else:
             output, __ = self.run_git_command(
                 ['merge-base', commit_id1, commit_id2])
             ancestor_id = self.COMMIT_ID_PAT.findall(output)[0]
 
         log.debug('Found common ancestor with sha: %s', ancestor_id)
 
         return ancestor_id
 
     def compare(self, commit_id1, commit_id2, repo2, merge, pre_load=None):
         repo1 = self
         ancestor_id = None
 
         if commit_id1 == commit_id2:
             commits = []
         elif repo1 != repo2:
             missing_ids = self._remote.get_missing_revs(commit_id1, commit_id2,
                                                         repo2.path)
             commits = [
                 repo2.get_commit(commit_id=commit_id, pre_load=pre_load)
                 for commit_id in reversed(missing_ids)]
         else:
             output, __ = repo1.run_git_command(
                 ['log', '--reverse', '--pretty=format: %H', '-s',
-                 '{}..{}'.format(commit_id1, commit_id2)])
+                 f'{commit_id1}..{commit_id2}'])
             commits = [
                 repo1.get_commit(commit_id=commit_id, pre_load=pre_load)
                 for commit_id in self.COMMIT_ID_PAT.findall(output)]
 
         return commits
 
     @LazyProperty
     def in_memory_commit(self):
         """
         Returns ``GitInMemoryCommit`` object for this repository.
         """
         return GitInMemoryCommit(self)
 
     def pull(self, url, commit_ids=None, update_after=False):
         """
         Pull changes from external location. Pull is different in GIT
         that fetch since it's doing a checkout
 
         :param commit_ids: Optional. Can be set to a list of commit ids
            which shall be pulled from the other repository.
         """
         refs = None
         if commit_ids is not None:
             remote_refs = self._remote.get_remote_refs(url)
             refs = [ref for ref in remote_refs if remote_refs[ref] in commit_ids]
         self._remote.pull(url, refs=refs, update_after=update_after)
         self._remote.invalidate_vcs_cache()
 
     def fetch(self, url, commit_ids=None):
         """
         Fetch all git objects from external location.
         """
         self._remote.sync_fetch(url, refs=commit_ids)
         self._remote.invalidate_vcs_cache()
 
     def push(self, url):
         refs = None
         self._remote.sync_push(url, refs=refs)
 
     def set_refs(self, ref_name, commit_id):
         self._remote.set_refs(ref_name, commit_id)
         self._invalidate_prop_cache('_refs')
 
     def remove_ref(self, ref_name):
         self._remote.remove_ref(ref_name)
         self._invalidate_prop_cache('_refs')
 
     def run_gc(self, prune=True):
         cmd = ['gc', '--aggressive']
         if prune:
             cmd += ['--prune=now']
         _stdout, stderr = self.run_git_command(cmd, fail_on_stderr=False)
         return stderr
 
     def _update_server_info(self):
         """
         runs gits update-server-info command in this repo instance
         """
         self._remote.update_server_info()
 
     def _current_branch(self):
         """
         Return the name of the current branch.
 
         It only works for non bare repositories (i.e. repositories with a
         working copy)
         """
         if self.bare:
             raise RepositoryError('Bare git repos do not have active branches')
 
         if self.is_empty():
             return None
 
         stdout, _ = self.run_git_command(['rev-parse', '--abbrev-ref', 'HEAD'])
         return stdout.strip()
 
     def _checkout(self, branch_name, create=False, force=False):
         """
         Checkout a branch in the working directory.
 
         It tries to create the branch if create is True, failing if the branch
         already exists.
 
         It only works for non bare repositories (i.e. repositories with a
         working copy)
         """
         if self.bare:
             raise RepositoryError('Cannot checkout branches in a bare git repo')
 
         cmd = ['checkout']
         if force:
             cmd.append('-f')
         if create:
             cmd.append('-b')
         cmd.append(branch_name)
         self.run_git_command(cmd, fail_on_stderr=False)
 
     def _create_branch(self, branch_name, commit_id):
         """
         creates a branch in a GIT repo
         """
         self._remote.create_branch(branch_name, commit_id)
 
     def _identify(self):
         """
         Return the current state of the working directory.
         """
         if self.bare:
             raise RepositoryError('Bare git repos do not have active branches')
 
         if self.is_empty():
             return None
 
         stdout, _ = self.run_git_command(['rev-parse', 'HEAD'])
         return stdout.strip()
 
     def _local_clone(self, clone_path, branch_name, source_branch=None):
         """
         Create a local clone of the current repo.
         """
         # N.B.(skreft): the --branch option is required as otherwise the shallow
         # clone will only fetch the active branch.
         cmd = ['clone', '--branch', branch_name,
                self.path, os.path.abspath(clone_path)]
 
         self.run_git_command(cmd, fail_on_stderr=False)
 
         # if we get the different source branch, make sure we also fetch it for
         # merge conditions
         if source_branch and source_branch != branch_name:
             # check if the ref exists.
             shadow_repo = GitRepository(os.path.abspath(clone_path))
             if shadow_repo.get_remote_ref(source_branch):
                 cmd = ['fetch', self.path, source_branch]
                 self.run_git_command(cmd, fail_on_stderr=False)
 
     def _local_fetch(self, repository_path, branch_name, use_origin=False):
         """
         Fetch a branch from a local repository.
         """
         repository_path = os.path.abspath(repository_path)
         if repository_path == self.path:
             raise ValueError('Cannot fetch from the same repository')
 
         if use_origin:
             branch_name = '+{branch}:refs/heads/{branch}'.format(
                 branch=branch_name)
 
         cmd = ['fetch', '--no-tags', '--update-head-ok',
                repository_path, branch_name]
         self.run_git_command(cmd, fail_on_stderr=False)
 
     def _local_reset(self, branch_name):
         branch_name = f'{branch_name}'
         cmd = ['reset', '--hard', branch_name, '--']
         self.run_git_command(cmd, fail_on_stderr=False)
 
     def _last_fetch_heads(self):
         """
         Return the last fetched heads that need merging.
 
         The algorithm is defined at
         https://github.com/git/git/blob/v2.1.3/git-pull.sh#L283
         """
         if not self.bare:
             fetch_heads_path = os.path.join(self.path, '.git', 'FETCH_HEAD')
         else:
             fetch_heads_path = os.path.join(self.path, 'FETCH_HEAD')
 
         heads = []
         with open(fetch_heads_path) as f:
             for line in f:
                 if '    not-for-merge   ' in line:
                     continue
                 line = re.sub('\t.*', '', line, flags=re.DOTALL)
                 heads.append(line)
 
         return heads
 
     def get_shadow_instance(self, shadow_repository_path, enable_hooks=False, cache=False):
         return GitRepository(shadow_repository_path, with_wire={"cache": cache})
 
     def _local_pull(self, repository_path, branch_name, ff_only=True):
         """
         Pull a branch from a local repository.
         """
         if self.bare:
             raise RepositoryError('Cannot pull into a bare git repository')
         # N.B.(skreft): The --ff-only option is to make sure this is a
         # fast-forward (i.e., we are only pulling new changes and there are no
         # conflicts with our current branch)
         # Additionally, that option needs to go before --no-tags, otherwise git
         # pull complains about it being an unknown flag.
         cmd = ['pull']
         if ff_only:
             cmd.append('--ff-only')
         cmd.extend(['--no-tags', repository_path, branch_name])
         self.run_git_command(cmd, fail_on_stderr=False)
 
     def _local_merge(self, merge_message, user_name, user_email, heads):
         """
         Merge the given head into the checked out branch.
 
         It will force a merge commit.
 
         Currently it raises an error if the repo is empty, as it is not possible
         to create a merge commit in an empty repo.
 
         :param merge_message: The message to use for the merge commit.
         :param heads: the heads to merge.
         """
         if self.bare:
             raise RepositoryError('Cannot merge into a bare git repository')
 
         if not heads:
             return
 
         if self.is_empty():
             # TODO(skreft): do something more robust in this case.
             raise RepositoryError('Do not know how to merge into empty repositories yet')
         unresolved = None
 
         # N.B.(skreft): the --no-ff option is used to enforce the creation of a
         # commit message. We also specify the user who is doing the merge.
         cmd = ['-c', f'user.name="{user_name}"',
                '-c', f'user.email={user_email}',
                'merge', '--no-ff', '-m', safe_str(merge_message)]
 
         merge_cmd = cmd + heads
 
         try:
             self.run_git_command(merge_cmd, fail_on_stderr=False)
         except RepositoryError:
             files = self.run_git_command(['diff', '--name-only', '--diff-filter', 'U'],
                                          fail_on_stderr=False)[0].splitlines()
             # NOTE(marcink): we add U notation for consistent with HG backend output
             unresolved = [f'U {f}' for f in files]
 
             # Cleanup any merge leftovers
             self._remote.invalidate_vcs_cache()
             self.run_git_command(['merge', '--abort'], fail_on_stderr=False)
 
             if unresolved:
                 raise UnresolvedFilesInRepo(unresolved)
             else:
                 raise
 
     def _local_push(
             self, source_branch, repository_path, target_branch,
             enable_hooks=False, rc_scm_data=None):
         """
         Push the source_branch to the given repository and target_branch.
 
         Currently it if the target_branch is not master and the target repo is
         empty, the push will work, but then GitRepository won't be able to find
         the pushed branch or the commits. As the HEAD will be corrupted (i.e.,
         pointing to master, which does not exist).
 
         It does not run the hooks in the target repo.
         """
         # TODO(skreft): deal with the case in which the target repo is empty,
         # and the target_branch is not master.
         target_repo = GitRepository(repository_path)
         if (not target_repo.bare and
                 target_repo._current_branch() == target_branch):
             # Git prevents pushing to the checked out branch, so simulate it by
             # pulling into the target repository.
             target_repo._local_pull(self.path, source_branch)
         else:
             cmd = ['push', os.path.abspath(repository_path),
-                   '{}:{}'.format(source_branch, target_branch)]
+                   f'{source_branch}:{target_branch}']
             gitenv = {}
             if rc_scm_data:
                 gitenv.update({'RC_SCM_DATA': rc_scm_data})
 
             if not enable_hooks:
                 gitenv['RC_SKIP_HOOKS'] = '1'
             self.run_git_command(cmd, fail_on_stderr=False, extra_env=gitenv)
 
     def _get_new_pr_branch(self, source_branch, target_branch):
-        prefix = 'pr_{}-{}_'.format(source_branch, target_branch)
+        prefix = f'pr_{source_branch}-{target_branch}_'
         pr_branches = []
         for branch in self.branches:
             if branch.startswith(prefix):
                 pr_branches.append(int(branch[len(prefix):]))
 
         if not pr_branches:
             branch_id = 0
         else:
             branch_id = max(pr_branches) + 1
 
         return '%s%d' % (prefix, branch_id)
 
     def _maybe_prepare_merge_workspace(
             self, repo_id, workspace_id, target_ref, source_ref):
         shadow_repository_path = self._get_shadow_repository_path(
             self.path, repo_id, workspace_id)
         if not os.path.exists(shadow_repository_path):
             self._local_clone(
                 shadow_repository_path, target_ref.name, source_ref.name)
             log.debug('Prepared %s shadow repository in %s',
                       self.alias, shadow_repository_path)
 
         return shadow_repository_path
 
     def _merge_repo(self, repo_id, workspace_id, target_ref,
                     source_repo, source_ref, merge_message,
                     merger_name, merger_email, dry_run=False,
                     use_rebase=False, close_branch=False):
 
         log.debug('Executing merge_repo with %s strategy, dry_run mode:%s',
                   'rebase' if use_rebase else 'merge', dry_run)
         if target_ref.commit_id != self.branches[target_ref.name]:
             log.warning('Target ref %s commit mismatch %s vs %s', target_ref,
                         target_ref.commit_id, self.branches[target_ref.name])
             return MergeResponse(
                 False, False, None, MergeFailureReason.TARGET_IS_NOT_HEAD,
                 metadata={'target_ref': target_ref})
 
         shadow_repository_path = self._maybe_prepare_merge_workspace(
             repo_id, workspace_id, target_ref, source_ref)
         shadow_repo = self.get_shadow_instance(shadow_repository_path)
 
         # checkout source, if it's different. Otherwise we could not
         # fetch proper commits for merge testing
         if source_ref.name != target_ref.name:
             if shadow_repo.get_remote_ref(source_ref.name):
                 shadow_repo._checkout(source_ref.name, force=True)
 
         # checkout target, and fetch changes
         shadow_repo._checkout(target_ref.name, force=True)
 
         # fetch/reset pull the target, in case it is changed
         # this handles even force changes
         shadow_repo._local_fetch(self.path, target_ref.name, use_origin=True)
         shadow_repo._local_reset(target_ref.name)
 
         # Need to reload repo to invalidate the cache, or otherwise we cannot
         # retrieve the last target commit.
         shadow_repo = self.get_shadow_instance(shadow_repository_path)
         if target_ref.commit_id != shadow_repo.branches[target_ref.name]:
             log.warning('Shadow Target ref %s commit mismatch %s vs %s',
                         target_ref, target_ref.commit_id,
                         shadow_repo.branches[target_ref.name])
             return MergeResponse(
                 False, False, None, MergeFailureReason.TARGET_IS_NOT_HEAD,
                 metadata={'target_ref': target_ref})
 
         # calculate new branch
         pr_branch = shadow_repo._get_new_pr_branch(
             source_ref.name, target_ref.name)
         log.debug('using pull-request merge branch: `%s`', pr_branch)
         # checkout to temp branch, and fetch changes
         shadow_repo._checkout(pr_branch, create=True)
         try:
             shadow_repo._local_fetch(source_repo.path, source_ref.name)
         except RepositoryError:
             log.exception('Failure when doing local fetch on '
                           'shadow repo: %s', shadow_repo)
             return MergeResponse(
                 False, False, None, MergeFailureReason.MISSING_SOURCE_REF,
                 metadata={'source_ref': source_ref})
 
         merge_ref = None
         merge_failure_reason = MergeFailureReason.NONE
         metadata = {}
         try:
             shadow_repo._local_merge(merge_message, merger_name, merger_email,
                                      [source_ref.commit_id])
             merge_possible = True
 
             # Need to invalidate the cache, or otherwise we
             # cannot retrieve the merge commit.
             shadow_repo = shadow_repo.get_shadow_instance(shadow_repository_path)
             merge_commit_id = shadow_repo.branches[pr_branch]
 
             # Set a reference pointing to the merge commit. This reference may
             # be used to easily identify the last successful merge commit in
             # the shadow repository.
             shadow_repo.set_refs('refs/heads/pr-merge', merge_commit_id)
             merge_ref = Reference('branch', 'pr-merge', merge_commit_id)
         except RepositoryError as e:
             log.exception('Failure when doing local merge on git shadow repo')
             if isinstance(e, UnresolvedFilesInRepo):
                 metadata['unresolved_files'] = '\n* conflict: ' + ('\n * conflict: '.join(e.args[0]))
 
             merge_possible = False
             merge_failure_reason = MergeFailureReason.MERGE_FAILED
 
         if merge_possible and not dry_run:
             try:
                 shadow_repo._local_push(
                     pr_branch, self.path, target_ref.name, enable_hooks=True,
                     rc_scm_data=self.config.get('rhodecode', 'RC_SCM_DATA'))
                 merge_succeeded = True
             except RepositoryError:
                 log.exception(
                     'Failure when doing local push from the shadow '
                     'repository to the target repository at %s.', self.path)
                 merge_succeeded = False
                 merge_failure_reason = MergeFailureReason.PUSH_FAILED
                 metadata['target'] = 'git shadow repo'
                 metadata['merge_commit'] = pr_branch
         else:
             merge_succeeded = False
 
         return MergeResponse(
             merge_possible, merge_succeeded, merge_ref, merge_failure_reason,
             metadata=metadata)

rhodecode/lib/vcs/backends/hg/commit.py

@@ -1,403 +1,403 b''
 # Copyright (C) 2014-2023 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/
 
 """
 HG commit module
 """
 
 import os
 
 from zope.cachedescriptors.property import Lazy as LazyProperty
 
 from rhodecode.lib.datelib import utcdate_fromtimestamp
 from rhodecode.lib.str_utils import safe_bytes, safe_str
 from rhodecode.lib.vcs import path as vcspath
 from rhodecode.lib.vcs.backends import base
 from rhodecode.lib.vcs.exceptions import CommitError
 from rhodecode.lib.vcs.nodes import (
     AddedFileNodesGenerator, ChangedFileNodesGenerator, DirNode, FileNode,
     NodeKind, RemovedFileNodesGenerator, RootNode, SubModuleNode,
     LargeFileNode)
 from rhodecode.lib.vcs.utils.paths import get_dirs_for_path
 
 
 class MercurialCommit(base.BaseCommit):
     """
     Represents state of the repository at the single commit.
     """
 
     _filter_pre_load = [
         # git specific property not supported here
         "_commit",
     ]
 
     def __init__(self, repository, raw_id, idx, pre_load=None):
         raw_id = safe_str(raw_id)
 
         self.repository = repository
         self._remote = repository._remote
 
         self.raw_id = raw_id
         self.idx = idx
 
         self._set_bulk_properties(pre_load)
 
         # caches
         self.nodes = {}
         self._stat_modes = {}  # stat info for paths
 
     def _set_bulk_properties(self, pre_load):
         if not pre_load:
             return
         pre_load = [entry for entry in pre_load
                     if entry not in self._filter_pre_load]
         if not pre_load:
             return
 
         result = self._remote.bulk_request(self.raw_id, pre_load)
 
         for attr, value in result.items():
             if attr in ["author", "branch", "message"]:
                 value = safe_str(value)
             elif attr == "affected_files":
                 value = list(map(safe_str, value))
             elif attr == "date":
                 value = utcdate_fromtimestamp(*value)
             elif attr in ["children", "parents"]:
                 value = self._make_commits(value)
             elif attr in ["phase"]:
                 value = self._get_phase_text(value)
             self.__dict__[attr] = value
 
     @LazyProperty
     def tags(self):
         tags = [name for name, commit_id in self.repository.tags.items()
                 if commit_id == self.raw_id]
         return tags
 
     @LazyProperty
     def branch(self):
         return safe_str(self._remote.ctx_branch(self.raw_id))
 
     @LazyProperty
     def bookmarks(self):
         bookmarks = [
             name for name, commit_id in self.repository.bookmarks.items()
             if commit_id == self.raw_id]
         return bookmarks
 
     @LazyProperty
     def message(self):
         return safe_str(self._remote.ctx_description(self.raw_id))
 
     @LazyProperty
     def committer(self):
         return safe_str(self.author)
 
     @LazyProperty
     def author(self):
         return safe_str(self._remote.ctx_user(self.raw_id))
 
     @LazyProperty
     def date(self):
         return utcdate_fromtimestamp(*self._remote.ctx_date(self.raw_id))
 
     @LazyProperty
     def status(self):
         """
         Returns modified, added, removed, deleted files for current commit
         """
         return self._remote.ctx_status(self.raw_id)
 
     @LazyProperty
     def _file_paths(self):
         return self._remote.ctx_list(self.raw_id)
 
     @LazyProperty
     def _dir_paths(self):
         dir_paths = ['']
         dir_paths.extend(list(set(get_dirs_for_path(*self._file_paths))))
 
         return dir_paths
 
     @LazyProperty
     def _paths(self):
         return self._dir_paths + self._file_paths
 
     @LazyProperty
     def id(self):
         if self.last:
             return 'tip'
         return self.short_id
 
     @LazyProperty
     def short_id(self):
         return self.raw_id[:12]
 
     def _make_commits(self, commit_ids, pre_load=None):
         return [self.repository.get_commit(commit_id=commit_id, pre_load=pre_load)
                 for commit_id in commit_ids]
 
     @LazyProperty
     def parents(self):
         """
         Returns list of parent commits.
         """
         parents = self._remote.ctx_parents(self.raw_id)
         return self._make_commits(parents)
 
     def _get_phase_text(self, phase_id):
         return {
                    0: 'public',
                    1: 'draft',
                    2: 'secret',
                }.get(phase_id) or ''
 
     @LazyProperty
     def phase(self):
         phase_id = self._remote.ctx_phase(self.raw_id)
         phase_text = self._get_phase_text(phase_id)
 
         return safe_str(phase_text)
 
     @LazyProperty
     def obsolete(self):
         obsolete = self._remote.ctx_obsolete(self.raw_id)
         return obsolete
 
     @LazyProperty
     def hidden(self):
         hidden = self._remote.ctx_hidden(self.raw_id)
         return hidden
 
     @LazyProperty
     def children(self):
         """
         Returns list of child commits.
         """
         children = self._remote.ctx_children(self.raw_id)
         return self._make_commits(children)
 
     def _get_kind(self, path):
         path = self._fix_path(path)
         if path in self._file_paths:
             return NodeKind.FILE
         elif path in self._dir_paths:
             return NodeKind.DIR
         else:
             raise CommitError(f"Node does not exist at the given path '{path}'")
 
     def _assert_is_path(self, path) -> str:
         path = self._fix_path(path)
         if self._get_kind(path) != NodeKind.FILE:
             raise CommitError(f"File does not exist for commit {self.raw_id} at '{path}'")
 
         return path
 
     def get_file_mode(self, path: bytes):
         """
         Returns stat mode of the file at the given ``path``.
         """
         path = self._assert_is_path(path)
 
         if path not in self._stat_modes:
             self._stat_modes[path] = self._remote.fctx_flags(self.raw_id, path)
 
         if 'x' in self._stat_modes[path]:
             return base.FILEMODE_EXECUTABLE
         return base.FILEMODE_DEFAULT
 
     def is_link(self, path):
         path = self._assert_is_path(path)
         if path not in self._stat_modes:
             self._stat_modes[path] = self._remote.fctx_flags(self.raw_id, path)
 
         return 'l' in self._stat_modes[path]
 
     def is_node_binary(self, path):
         path = self._assert_is_path(path)
         return self._remote.is_binary(self.raw_id, path)
 
     def node_md5_hash(self, path):
         path = self._assert_is_path(path)
         return self._remote.md5_hash(self.raw_id, path)
 
     def get_file_content(self, path):
         """
         Returns content of the file at given ``path``.
         """
         path = self._assert_is_path(path)
         return self._remote.fctx_node_data(self.raw_id, path)
 
     def get_file_content_streamed(self, path):
         path = self._assert_is_path(path)
         stream_method = getattr(self._remote, 'stream:fctx_node_data')
         return stream_method(self.raw_id, path)
 
     def get_file_size(self, path):
         """
         Returns size of the file at given ``path``.
         """
         path = self._assert_is_path(path)
         return self._remote.fctx_size(self.raw_id, path)
 
     def get_path_history(self, path, limit=None, pre_load=None):
         """
         Returns history of file as reversed list of `MercurialCommit` objects
         for which file at given ``path`` has been modified.
         """
         path = self._assert_is_path(path)
         hist = self._remote.node_history(self.raw_id, path, limit)
         return [
             self.repository.get_commit(commit_id=commit_id, pre_load=pre_load)
             for commit_id in hist]
 
     def get_file_annotate(self, path, pre_load=None):
         """
         Returns a generator of four element tuples with
             lineno, commit_id, commit lazy loader and line
         """
         result = self._remote.fctx_annotate(self.raw_id, path)
 
         for ln_no, commit_id, content in result:
             yield (
                 ln_no, commit_id,
                 lambda: self.repository.get_commit(commit_id=commit_id, pre_load=pre_load),
                 content)
 
     def get_nodes(self, path, pre_load=None):
         """
         Returns combined ``DirNode`` and ``FileNode`` objects list representing
         state of commit at the given ``path``. If node at the given ``path``
         is not instance of ``DirNode``, CommitError would be raised.
         """
 
         if self._get_kind(path) != NodeKind.DIR:
             raise CommitError(
-                "Directory does not exist for idx {} at '{}'".format(self.raw_id, path))
+                f"Directory does not exist for idx {self.raw_id} at '{path}'")
         path = self._fix_path(path)
 
         filenodes = [
             FileNode(safe_bytes(f), commit=self, pre_load=pre_load) for f in self._file_paths
             if os.path.dirname(f) == path]
         # TODO: johbo: Check if this can be done in a more obvious way
         dirs = path == '' and '' or [
             d for d in self._dir_paths
             if d and vcspath.dirname(d) == path]
         dirnodes = [
             DirNode(safe_bytes(d), commit=self) for d in dirs
             if os.path.dirname(d) == path]
 
         alias = self.repository.alias
         for k, vals in self._submodules.items():
             if vcspath.dirname(k) == path:
                 loc = vals[0]
                 commit = vals[1]
                 dirnodes.append(SubModuleNode(k, url=loc, commit=commit, alias=alias))
 
         nodes = dirnodes + filenodes
         for node in nodes:
             if node.path not in self.nodes:
                 self.nodes[node.path] = node
         nodes.sort()
 
         return nodes
 
     def get_node(self, path, pre_load=None):
         """
         Returns `Node` object from the given `path`. If there is no node at
         the given `path`, `NodeDoesNotExistError` would be raised.
         """
         path = self._fix_path(path)
 
         if path not in self.nodes:
             if path in self._file_paths:
                 node = FileNode(safe_bytes(path), commit=self, pre_load=pre_load)
             elif path in self._dir_paths:
                 if path == '':
                     node = RootNode(commit=self)
                 else:
                     node = DirNode(safe_bytes(path), commit=self)
             else:
                 raise self.no_node_at_path(path)
 
             # cache node
             self.nodes[path] = node
         return self.nodes[path]
 
     def get_largefile_node(self, path):
         pointer_spec = self._remote.is_large_file(self.raw_id, path)
         if pointer_spec:
             # content of that file regular FileNode is the hash of largefile
             file_id = self.get_file_content(path).strip()
 
             if self._remote.in_largefiles_store(file_id):
                 lf_path = self._remote.store_path(file_id)
                 return LargeFileNode(safe_bytes(lf_path), commit=self, org_path=path)
             elif self._remote.in_user_cache(file_id):
                 lf_path = self._remote.store_path(file_id)
                 self._remote.link(file_id, path)
                 return LargeFileNode(safe_bytes(lf_path), commit=self, org_path=path)
 
     @LazyProperty
     def _submodules(self):
         """
         Returns a dictionary with submodule information from substate file
         of hg repository.
         """
         return self._remote.ctx_substate(self.raw_id)
 
     @LazyProperty
     def affected_files(self):
         """
         Gets a fast accessible file changes for given commit
         """
         return self._remote.ctx_files(self.raw_id)
 
     @property
     def added(self):
         """
         Returns list of added ``FileNode`` objects.
         """
         return AddedFileNodesGenerator(self.added_paths, self)
 
     @LazyProperty
     def added_paths(self):
         return [n for n in self.status[1]]
 
     @property
     def changed(self):
         """
         Returns list of modified ``FileNode`` objects.
         """
         return ChangedFileNodesGenerator(self.changed_paths, self)
 
     @LazyProperty
     def changed_paths(self):
         return [n for n in self.status[0]]
 
     @property
     def removed(self):
         """
         Returns list of removed ``FileNode`` objects.
         """
         return RemovedFileNodesGenerator(self.removed_paths, self)
 
     @LazyProperty
     def removed_paths(self):
         return [n for n in self.status[2]]

rhodecode/lib/vcs/backends/hg/repository.py

@@ -1,1013 +1,1013 b''
 # Copyright (C) 2014-2023 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/
 
 """
 HG repository module
 """
 import os
 import logging
 import binascii
 import configparser
 import urllib.request
 import urllib.parse
 import urllib.error
 
 from zope.cachedescriptors.property import Lazy as LazyProperty
 
 from collections import OrderedDict
 from rhodecode.lib.datelib import (
     date_to_timestamp_plus_offset, utcdate_fromtimestamp, makedate)
 from rhodecode.lib.str_utils import safe_str
 from rhodecode.lib.utils2 import CachedProperty
 from rhodecode.lib.vcs import connection, exceptions
 from rhodecode.lib.vcs.backends.base import (
     BaseRepository, CollectionGenerator, Config, MergeResponse,
     MergeFailureReason, Reference, BasePathPermissionChecker)
 from rhodecode.lib.vcs.backends.hg.commit import MercurialCommit
 from rhodecode.lib.vcs.backends.hg.diff import MercurialDiff
 from rhodecode.lib.vcs.backends.hg.inmemory import MercurialInMemoryCommit
 from rhodecode.lib.vcs.exceptions import (
     EmptyRepositoryError, RepositoryError, TagAlreadyExistError,
     TagDoesNotExistError, CommitDoesNotExistError, SubrepoMergeError, UnresolvedFilesInRepo)
 
 hexlify = binascii.hexlify
 nullid = "\0" * 20
 
 log = logging.getLogger(__name__)
 
 
 class MercurialRepository(BaseRepository):
     """
     Mercurial repository backend
     """
     DEFAULT_BRANCH_NAME = 'default'
 
     def __init__(self, repo_path, config=None, create=False, src_url=None,
                  do_workspace_checkout=False, with_wire=None, bare=False):
         """
         Raises RepositoryError if repository could not be find at the given
         ``repo_path``.
 
         :param repo_path: local path of the repository
         :param config: config object containing the repo configuration
         :param create=False: if set to True, would try to create repository if
            it does not exist rather than raising exception
         :param src_url=None: would try to clone repository from given location
         :param do_workspace_checkout=False: sets update of working copy after
            making a clone
         :param bare: not used, compatible with other VCS
         """
 
         self.path = safe_str(os.path.abspath(repo_path))
         # mercurial since 4.4.X requires certain configuration to be present
         # because sometimes we init the repos with config we need to meet
         # special requirements
         self.config = config if config else self.get_default_config(
             default=[('extensions', 'largefiles', '1')])
         self.with_wire = with_wire or {"cache": False}  # default should not use cache
 
         self._init_repo(create, src_url, do_workspace_checkout)
 
         # caches
         self._commit_ids = {}
 
     @LazyProperty
     def _remote(self):
         repo_id = self.path
         return connection.Hg(self.path, repo_id, self.config, with_wire=self.with_wire)
 
     @CachedProperty
     def commit_ids(self):
         """
         Returns list of commit ids, in ascending order.  Being lazy
         attribute allows external tools to inject shas from cache.
         """
         commit_ids = self._get_all_commit_ids()
         self._rebuild_cache(commit_ids)
         return commit_ids
 
     def _rebuild_cache(self, commit_ids):
         self._commit_ids = {commit_id: index
                                 for index, commit_id in enumerate(commit_ids)}
 
     @CachedProperty
     def branches(self):
         return self._get_branches()
 
     @CachedProperty
     def branches_closed(self):
         return self._get_branches(active=False, closed=True)
 
     @CachedProperty
     def branches_all(self):
         all_branches = {}
         all_branches.update(self.branches)
         all_branches.update(self.branches_closed)
         return all_branches
 
     def _get_branches(self, active=True, closed=False):
         """
         Gets branches for this repository
         Returns only not closed active branches by default
 
         :param active: return also active branches
         :param closed: return also closed branches
 
         """
         if self.is_empty():
             return {}
 
         def get_name(ctx):
             return ctx[0]
 
         _branches = [(n, h,) for n, h in
                      self._remote.branches(active, closed).items()]
 
         return OrderedDict(sorted(_branches, key=get_name, reverse=False))
 
     @CachedProperty
     def tags(self):
         """
         Gets tags for this repository
         """
         return self._get_tags()
 
     def _get_tags(self):
         if self.is_empty():
             return {}
 
         def get_name(ctx):
             return ctx[0]
 
         _tags = [(n, h,) for n, h in
                  self._remote.tags().items()]
 
         return OrderedDict(sorted(_tags, key=get_name, reverse=True))
 
     def tag(self, name, user, commit_id=None, message=None, date=None, **kwargs):
         """
         Creates and returns a tag for the given ``commit_id``.
 
         :param name: name for new tag
         :param user: full username, i.e.: "Joe Doe <[e-mail]>"
         :param commit_id: commit id for which new tag would be created
         :param message: message of the tag's commit
         :param date: date of tag's commit
 
         :raises TagAlreadyExistError: if tag with same name already exists
         """
         if name in self.tags:
             raise TagAlreadyExistError("Tag %s already exists" % name)
 
         commit = self.get_commit(commit_id=commit_id)
         local = kwargs.setdefault('local', False)
 
         if message is None:
-            message = "Added tag {} for commit {}".format(name, commit.short_id)
+            message = f"Added tag {name} for commit {commit.short_id}"
 
         date, tz = date_to_timestamp_plus_offset(date)
 
         self._remote.tag(name, commit.raw_id, message, local, user, date, tz)
         self._remote.invalidate_vcs_cache()
 
         # Reinitialize tags
         self._invalidate_prop_cache('tags')
         tag_id = self.tags[name]
 
         return self.get_commit(commit_id=tag_id)
 
     def remove_tag(self, name, user, message=None, date=None):
         """
         Removes tag with the given `name`.
 
         :param name: name of the tag to be removed
         :param user: full username, i.e.: "Joe Doe <[e-mail]>"
         :param message: message of the tag's removal commit
         :param date: date of tag's removal commit
 
         :raises TagDoesNotExistError: if tag with given name does not exists
         """
         if name not in self.tags:
             raise TagDoesNotExistError("Tag %s does not exist" % name)
 
         if message is None:
             message = "Removed tag %s" % name
         local = False
 
         date, tz = date_to_timestamp_plus_offset(date)
 
         self._remote.tag(name, nullid, message, local, user, date, tz)
         self._remote.invalidate_vcs_cache()
         self._invalidate_prop_cache('tags')
 
     @LazyProperty
     def bookmarks(self):
         """
         Gets bookmarks for this repository
         """
         return self._get_bookmarks()
 
     def _get_bookmarks(self):
         if self.is_empty():
             return {}
 
         def get_name(ctx):
             return ctx[0]
 
         _bookmarks = [
             (n, h) for n, h in
             self._remote.bookmarks().items()]
 
         return OrderedDict(sorted(_bookmarks, key=get_name))
 
     def _get_all_commit_ids(self):
         return self._remote.get_all_commit_ids('visible')
 
     def get_diff(
             self, commit1, commit2, path='', ignore_whitespace=False,
             context=3, path1=None):
         """
         Returns (git like) *diff*, as plain text. Shows changes introduced by
         `commit2` since `commit1`.
 
         :param commit1: Entry point from which diff is shown. Can be
           ``self.EMPTY_COMMIT`` - in this case, patch showing all
           the changes since empty state of the repository until `commit2`
         :param commit2: Until which commit changes should be shown.
         :param ignore_whitespace: If set to ``True``, would not show whitespace
           changes. Defaults to ``False``.
         :param context: How many lines before/after changed lines should be
           shown. Defaults to ``3``.
         """
         self._validate_diff_commits(commit1, commit2)
         if path1 is not None and path1 != path:
             raise ValueError("Diff of two different paths not supported.")
 
         if path:
             file_filter = [self.path, path]
         else:
             file_filter = None
 
         diff = self._remote.diff(
             commit1.raw_id, commit2.raw_id, file_filter=file_filter,
             opt_git=True, opt_ignorews=ignore_whitespace,
             context=context)
         return MercurialDiff(diff)
 
     def strip(self, commit_id, branch=None):
         self._remote.strip(commit_id, update=False, backup="none")
 
         self._remote.invalidate_vcs_cache()
         # clear cache
         self._invalidate_prop_cache('commit_ids')
 
         return len(self.commit_ids)
 
     def verify(self):
         verify = self._remote.verify()
 
         self._remote.invalidate_vcs_cache()
         return verify
 
     def hg_update_cache(self):
         update_cache = self._remote.hg_update_cache()
 
         self._remote.invalidate_vcs_cache()
         return update_cache
 
     def hg_rebuild_fn_cache(self):
         update_cache = self._remote.hg_rebuild_fn_cache()
 
         self._remote.invalidate_vcs_cache()
         return update_cache
 
     def get_common_ancestor(self, commit_id1, commit_id2, repo2):
         log.debug('Calculating common ancestor between %sc1:%s and %sc2:%s',
                   self, commit_id1, repo2, commit_id2)
 
         if commit_id1 == commit_id2:
             return commit_id1
 
         ancestors = self._remote.revs_from_revspec(
             "ancestor(id(%s), id(%s))", commit_id1, commit_id2,
             other_path=repo2.path)
 
         ancestor_id = repo2[ancestors[0]].raw_id if ancestors else None
 
         log.debug('Found common ancestor with sha: %s', ancestor_id)
         return ancestor_id
 
     def compare(self, commit_id1, commit_id2, repo2, merge, pre_load=None):
         if commit_id1 == commit_id2:
             commits = []
         else:
             if merge:
                 indexes = self._remote.revs_from_revspec(
                     "ancestors(id(%s)) - ancestors(id(%s)) - id(%s)",
                     commit_id2, commit_id1, commit_id1, other_path=repo2.path)
             else:
                 indexes = self._remote.revs_from_revspec(
                     "id(%s)..id(%s) - id(%s)", commit_id1, commit_id2,
                     commit_id1, other_path=repo2.path)
 
             commits = [repo2.get_commit(commit_idx=idx, pre_load=pre_load)
                        for idx in indexes]
 
         return commits
 
     @staticmethod
     def check_url(url, config):
         """
         Function will check given url and try to verify if it's a valid
         link. Sometimes it may happened that mercurial will issue basic
         auth request that can cause whole API to hang when used from python
         or other external calls.
 
         On failures it'll raise urllib2.HTTPError, exception is also thrown
         when the return code is non 200
         """
         # check first if it's not an local url
         if os.path.isdir(url) or url.startswith('file:'):
             return True
 
         # Request the _remote to verify the url
         return connection.Hg.check_url(url, config.serialize())
 
     @staticmethod
     def is_valid_repository(path):
         return os.path.isdir(os.path.join(path, '.hg'))
 
     def _init_repo(self, create, src_url=None, do_workspace_checkout=False):
         """
         Function will check for mercurial repository in given path. If there
         is no repository in that path it will raise an exception unless
         `create` parameter is set to True - in that case repository would
         be created.
 
         If `src_url` is given, would try to clone repository from the
         location at given clone_point. Additionally it'll make update to
         working copy accordingly to `do_workspace_checkout` flag.
         """
         if create and os.path.exists(self.path):
             raise RepositoryError(
                 f"Cannot create repository at {self.path}, location already exist")
 
         if src_url:
             url = str(self._get_url(src_url))
             MercurialRepository.check_url(url, self.config)
 
             self._remote.clone(url, self.path, do_workspace_checkout)
 
             # Don't try to create if we've already cloned repo
             create = False
 
         if create:
             os.makedirs(self.path, mode=0o755)
 
         self._remote.localrepository(create)
 
     @LazyProperty
     def in_memory_commit(self):
         return MercurialInMemoryCommit(self)
 
     @LazyProperty
     def description(self):
         description = self._remote.get_config_value(
             'web', 'description', untrusted=True)
         return safe_str(description or self.DEFAULT_DESCRIPTION)
 
     @LazyProperty
     def contact(self):
         contact = (
             self._remote.get_config_value("web", "contact") or
             self._remote.get_config_value("ui", "username"))
         return safe_str(contact or self.DEFAULT_CONTACT)
 
     @LazyProperty
     def last_change(self):
         """
         Returns last change made on this repository as
         `datetime.datetime` object.
         """
         try:
             return self.get_commit().date
         except RepositoryError:
             tzoffset = makedate()[1]
             return utcdate_fromtimestamp(self._get_fs_mtime(), tzoffset)
 
     def _get_fs_mtime(self):
         # fallback to filesystem
         cl_path = os.path.join(self.path, '.hg', "00changelog.i")
         st_path = os.path.join(self.path, '.hg', "store")
         if os.path.exists(cl_path):
             return os.stat(cl_path).st_mtime
         else:
             return os.stat(st_path).st_mtime
 
     def _get_url(self, url):
         """
         Returns normalized url. If schema is not given, would fall
         to filesystem
         (``file:///``) schema.
         """
         if url != 'default' and '://' not in url:
             url = "file:" + urllib.request.pathname2url(url)
         return url
 
     def get_hook_location(self):
         """
         returns absolute path to location where hooks are stored
         """
         return os.path.join(self.path, '.hg', '.hgrc')
 
     def get_commit(self, commit_id=None, commit_idx=None, pre_load=None,
                    translate_tag=None, maybe_unreachable=False, reference_obj=None):
         """
         Returns ``MercurialCommit`` object representing repository's
         commit at the given `commit_id` or `commit_idx`.
         """
         if self.is_empty():
             raise EmptyRepositoryError("There are no commits yet")
 
         if commit_id is not None:
             self._validate_commit_id(commit_id)
             try:
                 # we have cached idx, use it without contacting the remote
                 idx = self._commit_ids[commit_id]
                 return MercurialCommit(self, commit_id, idx, pre_load=pre_load)
             except KeyError:
                 pass
 
         elif commit_idx is not None:
             self._validate_commit_idx(commit_idx)
             try:
                 _commit_id = self.commit_ids[commit_idx]
                 if commit_idx < 0:
                     commit_idx = self.commit_ids.index(_commit_id)
 
                 return MercurialCommit(self, _commit_id, commit_idx, pre_load=pre_load)
             except IndexError:
                 commit_id = commit_idx
         else:
             commit_id = "tip"
 
         # case here is no cached version, do an actual lookup instead
         try:
             raw_id, idx = self._remote.lookup(commit_id, both=True)
         except CommitDoesNotExistError:
             msg = "Commit {} does not exist for `{}`".format(
                 *map(safe_str, [commit_id, self.name]))
             raise CommitDoesNotExistError(msg)
 
         return MercurialCommit(self, raw_id, idx, pre_load=pre_load)
 
     def get_commits(
             self, start_id=None, end_id=None, start_date=None, end_date=None,
             branch_name=None, show_hidden=False, pre_load=None, translate_tags=None):
         """
         Returns generator of ``MercurialCommit`` objects from start to end
         (both are inclusive)
 
         :param start_id: None, str(commit_id)
         :param end_id: None, str(commit_id)
         :param start_date: if specified, commits with commit date less than
           ``start_date`` would be filtered out from returned set
         :param end_date: if specified, commits with commit date greater than
           ``end_date`` would be filtered out from returned set
         :param branch_name: if specified, commits not reachable from given
           branch would be filtered out from returned set
         :param show_hidden: Show hidden commits such as obsolete or hidden from
             Mercurial evolve
         :raise BranchDoesNotExistError: If given ``branch_name`` does not
             exist.
         :raise CommitDoesNotExistError: If commit for given ``start`` or
           ``end`` could not be found.
         """
         # actually we should check now if it's not an empty repo
         if self.is_empty():
             raise EmptyRepositoryError("There are no commits yet")
         self._validate_branch_name(branch_name)
 
         branch_ancestors = False
         if start_id is not None:
             self._validate_commit_id(start_id)
             c_start = self.get_commit(commit_id=start_id)
             start_pos = self._commit_ids[c_start.raw_id]
         else:
             start_pos = None
 
         if end_id is not None:
             self._validate_commit_id(end_id)
             c_end = self.get_commit(commit_id=end_id)
             end_pos = max(0, self._commit_ids[c_end.raw_id])
         else:
             end_pos = None
 
         if None not in [start_id, end_id] and start_pos > end_pos:
             raise RepositoryError(
                 "Start commit '%s' cannot be after end commit '%s'" %
                 (start_id, end_id))
 
         if end_pos is not None:
             end_pos += 1
 
         commit_filter = []
 
         if branch_name and not branch_ancestors:
-            commit_filter.append('branch("{}")'.format(branch_name))
+            commit_filter.append(f'branch("{branch_name}")')
         elif branch_name and branch_ancestors:
-            commit_filter.append('ancestors(branch("{}"))'.format(branch_name))
+            commit_filter.append(f'ancestors(branch("{branch_name}"))')
 
         if start_date and not end_date:
-            commit_filter.append('date(">{}")'.format(start_date))
+            commit_filter.append(f'date(">{start_date}")')
         if end_date and not start_date:
-            commit_filter.append('date("<{}")'.format(end_date))
+            commit_filter.append(f'date("<{end_date}")')
         if start_date and end_date:
             commit_filter.append(
-                'date(">{}") and date("<{}")'.format(start_date, end_date))
+                f'date(">{start_date}") and date("<{end_date}")')
 
         if not show_hidden:
             commit_filter.append('not obsolete()')
             commit_filter.append('not hidden()')
 
         # TODO: johbo: Figure out a simpler way for this solution
         collection_generator = CollectionGenerator
         if commit_filter:
             commit_filter = ' and '.join(map(safe_str, commit_filter))
             revisions = self._remote.rev_range([commit_filter])
             collection_generator = MercurialIndexBasedCollectionGenerator
         else:
             revisions = self.commit_ids
 
         if start_pos or end_pos:
             revisions = revisions[start_pos:end_pos]
 
         return collection_generator(self, revisions, pre_load=pre_load)
 
     def pull(self, url, commit_ids=None):
         """
         Pull changes from external location.
 
         :param commit_ids: Optional. Can be set to a list of commit ids
            which shall be pulled from the other repository.
         """
         url = self._get_url(url)
         self._remote.pull(url, commit_ids=commit_ids)
         self._remote.invalidate_vcs_cache()
 
     def fetch(self, url, commit_ids=None):
         """
         Backward compatibility with GIT fetch==pull
         """
         return self.pull(url, commit_ids=commit_ids)
 
     def push(self, url):
         url = self._get_url(url)
         self._remote.sync_push(url)
 
     def _local_clone(self, clone_path):
         """
         Create a local clone of the current repo.
         """
         self._remote.clone(self.path, clone_path, update_after_clone=True,
                            hooks=False)
 
     def _update(self, revision, clean=False):
         """
         Update the working copy to the specified revision.
         """
         log.debug('Doing checkout to commit: `%s` for %s', revision, self)
         self._remote.update(revision, clean=clean)
 
     def _identify(self):
         """
         Return the current state of the working directory.
         """
         return self._remote.identify().strip().rstrip('+')
 
     def _heads(self, branch=None):
         """
         Return the commit ids of the repository heads.
         """
         return self._remote.heads(branch=branch).strip().split(' ')
 
     def _ancestor(self, revision1, revision2):
         """
         Return the common ancestor of the two revisions.
         """
         return self._remote.ancestor(revision1, revision2)
 
     def _local_push(
             self, revision, repository_path, push_branches=False,
             enable_hooks=False):
         """
         Push the given revision to the specified repository.
 
         :param push_branches: allow to create branches in the target repo.
         """
         self._remote.push(
             [revision], repository_path, hooks=enable_hooks,
             push_branches=push_branches)
 
     def _local_merge(self, target_ref, merge_message, user_name, user_email,
                      source_ref, use_rebase=False, close_commit_id=None, dry_run=False):
         """
         Merge the given source_revision into the checked out revision.
 
         Returns the commit id of the merge and a boolean indicating if the
         commit needs to be pushed.
         """
         source_ref_commit_id = source_ref.commit_id
         target_ref_commit_id = target_ref.commit_id
 
         # update our workdir to target ref, for proper merge
         self._update(target_ref_commit_id, clean=True)
 
         ancestor = self._ancestor(target_ref_commit_id, source_ref_commit_id)
         is_the_same_branch = self._is_the_same_branch(target_ref, source_ref)
 
         if close_commit_id:
             # NOTE(marcink): if we get the close commit, this is our new source
             # which will include the close commit itself.
             source_ref_commit_id = close_commit_id
 
         if ancestor == source_ref_commit_id:
             # Nothing to do, the changes were already integrated
             return target_ref_commit_id, False
 
         elif ancestor == target_ref_commit_id and is_the_same_branch:
             # In this case we should force a commit message
             return source_ref_commit_id, True
 
         unresolved = None
         if use_rebase:
             try:
-                bookmark_name = 'rcbook{}{}'.format(source_ref_commit_id, target_ref_commit_id)
+                bookmark_name = f'rcbook{source_ref_commit_id}{target_ref_commit_id}'
                 self.bookmark(bookmark_name, revision=source_ref.commit_id)
                 self._remote.rebase(
                     source=source_ref_commit_id, dest=target_ref_commit_id)
                 self._remote.invalidate_vcs_cache()
                 self._update(bookmark_name, clean=True)
                 return self._identify(), True
             except RepositoryError as e:
                 # The rebase-abort may raise another exception which 'hides'
                 # the original one, therefore we log it here.
                 log.exception('Error while rebasing shadow repo during merge.')
                 if 'unresolved conflicts' in safe_str(e):
                     unresolved = self._remote.get_unresolved_files()
                     log.debug('unresolved files: %s', unresolved)
 
                 # Cleanup any rebase leftovers
                 self._remote.invalidate_vcs_cache()
                 self._remote.rebase(abort=True)
                 self._remote.invalidate_vcs_cache()
                 self._remote.update(clean=True)
                 if unresolved:
                     raise UnresolvedFilesInRepo(unresolved)
                 else:
                     raise
         else:
             try:
                 self._remote.merge(source_ref_commit_id)
                 self._remote.invalidate_vcs_cache()
                 self._remote.commit(
                     message=safe_str(merge_message),
-                    username=safe_str('{} <{}>'.format(user_name, user_email)))
+                    username=safe_str(f'{user_name} <{user_email}>'))
                 self._remote.invalidate_vcs_cache()
                 return self._identify(), True
             except RepositoryError as e:
                 # The merge-abort may raise another exception which 'hides'
                 # the original one, therefore we log it here.
                 log.exception('Error while merging shadow repo during merge.')
                 if 'unresolved merge conflicts' in safe_str(e):
                     unresolved = self._remote.get_unresolved_files()
                     log.debug('unresolved files: %s', unresolved)
 
                 # Cleanup any merge leftovers
                 self._remote.update(clean=True)
                 if unresolved:
                     raise UnresolvedFilesInRepo(unresolved)
                 else:
                     raise
 
     def _local_close(self, target_ref, user_name, user_email,
                      source_ref, close_message=''):
         """
         Close the branch of the given source_revision
 
         Returns the commit id of the close and a boolean indicating if the
         commit needs to be pushed.
         """
         self._update(source_ref.commit_id)
         message = close_message or f"Closing branch: `{source_ref.name}`"
         try:
             self._remote.commit(
                 message=safe_str(message),
-                username=safe_str('{} <{}>'.format(user_name, user_email)),
+                username=safe_str(f'{user_name} <{user_email}>'),
                 close_branch=True)
             self._remote.invalidate_vcs_cache()
             return self._identify(), True
         except RepositoryError:
             # Cleanup any commit leftovers
             self._remote.update(clean=True)
             raise
 
     def _is_the_same_branch(self, target_ref, source_ref):
         return (
             self._get_branch_name(target_ref) ==
             self._get_branch_name(source_ref))
 
     def _get_branch_name(self, ref):
         if ref.type == 'branch':
             return ref.name
         return self._remote.ctx_branch(ref.commit_id)
 
     def _maybe_prepare_merge_workspace(
             self, repo_id, workspace_id, unused_target_ref, unused_source_ref):
         shadow_repository_path = self._get_shadow_repository_path(
             self.path, repo_id, workspace_id)
         if not os.path.exists(shadow_repository_path):
             self._local_clone(shadow_repository_path)
             log.debug(
                 'Prepared shadow repository in %s', shadow_repository_path)
 
         return shadow_repository_path
 
     def _merge_repo(self, repo_id, workspace_id, target_ref,
                     source_repo, source_ref, merge_message,
                     merger_name, merger_email, dry_run=False,
                     use_rebase=False, close_branch=False):
 
         log.debug('Executing merge_repo with %s strategy, dry_run mode:%s',
                   'rebase' if use_rebase else 'merge', dry_run)
         if target_ref.commit_id not in self._heads():
             return MergeResponse(
                 False, False, None, MergeFailureReason.TARGET_IS_NOT_HEAD,
                 metadata={'target_ref': target_ref})
 
         try:
             if target_ref.type == 'branch' and len(self._heads(target_ref.name)) != 1:
                 heads_all = self._heads(target_ref.name)
                 max_heads = 10
                 if len(heads_all) > max_heads:
                     heads = '\n,'.join(
                         heads_all[:max_heads] +
-                        ['and {} more.'.format(len(heads_all)-max_heads)])
+                        [f'and {len(heads_all)-max_heads} more.'])
                 else:
                     heads = '\n,'.join(heads_all)
                 metadata = {
                     'target_ref': target_ref,
                     'source_ref': source_ref,
                     'heads': heads
                 }
                 return MergeResponse(
                     False, False, None,
                     MergeFailureReason.HG_TARGET_HAS_MULTIPLE_HEADS,
                     metadata=metadata)
         except CommitDoesNotExistError:
             log.exception('Failure when looking up branch heads on hg target')
             return MergeResponse(
                 False, False, None, MergeFailureReason.MISSING_TARGET_REF,
                 metadata={'target_ref': target_ref})
 
         shadow_repository_path = self._maybe_prepare_merge_workspace(
             repo_id, workspace_id, target_ref, source_ref)
         shadow_repo = self.get_shadow_instance(shadow_repository_path)
 
         log.debug('Pulling in target reference %s', target_ref)
         self._validate_pull_reference(target_ref)
         shadow_repo._local_pull(self.path, target_ref)
 
         try:
             log.debug('Pulling in source reference %s', source_ref)
             source_repo._validate_pull_reference(source_ref)
             shadow_repo._local_pull(source_repo.path, source_ref)
         except CommitDoesNotExistError:
             log.exception('Failure when doing local pull on hg shadow repo')
             return MergeResponse(
                 False, False, None, MergeFailureReason.MISSING_SOURCE_REF,
                 metadata={'source_ref': source_ref})
 
         merge_ref = None
         merge_commit_id = None
         close_commit_id = None
         merge_failure_reason = MergeFailureReason.NONE
         metadata = {}
 
         # enforce that close branch should be used only in case we source from
         # an actual Branch
         close_branch = close_branch and source_ref.type == 'branch'
 
         # don't allow to close branch if source and target are the same
         close_branch = close_branch and source_ref.name != target_ref.name
 
         needs_push_on_close = False
         if close_branch and not use_rebase and not dry_run:
             try:
                 close_commit_id, needs_push_on_close = shadow_repo._local_close(
                     target_ref, merger_name, merger_email, source_ref)
                 merge_possible = True
             except RepositoryError:
                 log.exception('Failure when doing close branch on '
                               'shadow repo: %s', shadow_repo)
                 merge_possible = False
                 merge_failure_reason = MergeFailureReason.MERGE_FAILED
         else:
             merge_possible = True
 
         needs_push = False
         if merge_possible:
 
             try:
                 merge_commit_id, needs_push = shadow_repo._local_merge(
                     target_ref, merge_message, merger_name, merger_email,
                     source_ref, use_rebase=use_rebase,
                     close_commit_id=close_commit_id, dry_run=dry_run)
                 merge_possible = True
 
                 # read the state of the close action, if it
                 # maybe required a push
                 needs_push = needs_push or needs_push_on_close
 
                 # Set a bookmark pointing to the merge commit. This bookmark
                 # may be used to easily identify the last successful merge
                 # commit in the shadow repository.
                 shadow_repo.bookmark('pr-merge', revision=merge_commit_id)
                 merge_ref = Reference('book', 'pr-merge', merge_commit_id)
             except SubrepoMergeError:
                 log.exception(
                     'Subrepo merge error during local merge on hg shadow repo.')
                 merge_possible = False
                 merge_failure_reason = MergeFailureReason.SUBREPO_MERGE_FAILED
                 needs_push = False
             except RepositoryError as e:
                 log.exception('Failure when doing local merge on hg shadow repo')
                 if isinstance(e, UnresolvedFilesInRepo):
                     all_conflicts = list(e.args[0])
                     max_conflicts = 20
                     if len(all_conflicts) > max_conflicts:
                         conflicts = all_conflicts[:max_conflicts] \
-                                    + ['and {} more.'.format(len(all_conflicts)-max_conflicts)]
+                                    + [f'and {len(all_conflicts)-max_conflicts} more.']
                     else:
                         conflicts = all_conflicts
                     metadata['unresolved_files'] = \
                         '\n* conflict: ' + \
                         ('\n * conflict: '.join(conflicts))
 
                 merge_possible = False
                 merge_failure_reason = MergeFailureReason.MERGE_FAILED
                 needs_push = False
 
         if merge_possible and not dry_run:
             if needs_push:
                 # In case the target is a bookmark, update it, so after pushing
                 # the bookmarks is also updated in the target.
                 if target_ref.type == 'book':
                     shadow_repo.bookmark(
                         target_ref.name, revision=merge_commit_id)
                 try:
                     shadow_repo_with_hooks = self.get_shadow_instance(
                         shadow_repository_path,
                         enable_hooks=True)
                     # This is the actual merge action, we push from shadow
                     # into origin.
                     # Note: the push_branches option will push any new branch
                     # defined in the source repository to the target. This may
                     # be dangerous as branches are permanent in Mercurial.
                     # This feature was requested in issue #441.
                     shadow_repo_with_hooks._local_push(
                         merge_commit_id, self.path, push_branches=True,
                         enable_hooks=True)
 
                     # maybe we also need to push the close_commit_id
                     if close_commit_id:
                         shadow_repo_with_hooks._local_push(
                             close_commit_id, self.path, push_branches=True,
                             enable_hooks=True)
                     merge_succeeded = True
                 except RepositoryError:
                     log.exception(
                         'Failure when doing local push from the shadow '
                         'repository to the target repository at %s.', self.path)
                     merge_succeeded = False
                     merge_failure_reason = MergeFailureReason.PUSH_FAILED
                     metadata['target'] = 'hg shadow repo'
                     metadata['merge_commit'] = merge_commit_id
             else:
                 merge_succeeded = True
         else:
             merge_succeeded = False
 
         return MergeResponse(
             merge_possible, merge_succeeded, merge_ref, merge_failure_reason,
             metadata=metadata)
 
     def get_shadow_instance(self, shadow_repository_path, enable_hooks=False, cache=False):
         config = self.config.copy()
         if not enable_hooks:
             config.clear_section('hooks')
         return MercurialRepository(shadow_repository_path, config, with_wire={"cache": cache})
 
     def _validate_pull_reference(self, reference):
         if not (reference.name in self.bookmarks or
                 reference.name in self.branches or
                 self.get_commit(reference.commit_id)):
             raise CommitDoesNotExistError(
                 'Unknown branch, bookmark or commit id')
 
     def _local_pull(self, repository_path, reference):
         """
         Fetch a branch, bookmark or commit from a local repository.
         """
         repository_path = os.path.abspath(repository_path)
         if repository_path == self.path:
             raise ValueError('Cannot pull from the same repository')
 
         reference_type_to_option_name = {
             'book': 'bookmark',
             'branch': 'branch',
         }
         option_name = reference_type_to_option_name.get(
             reference.type, 'revision')
 
         if option_name == 'revision':
             ref = reference.commit_id
         else:
             ref = reference.name
 
         options = {option_name: [ref]}
         self._remote.pull_cmd(repository_path, hooks=False, **options)
         self._remote.invalidate_vcs_cache()
 
     def bookmark(self, bookmark, revision=None):
         if isinstance(bookmark, str):
             bookmark = safe_str(bookmark)
         self._remote.bookmark(bookmark, revision=revision)
         self._remote.invalidate_vcs_cache()
 
     def get_path_permissions(self, username):
         hgacl_file = os.path.join(self.path, '.hg/hgacl')
 
         def read_patterns(suffix):
             svalue = None
             for section, option in [
                 ('narrowacl', username + suffix),
                 ('narrowacl', 'default' + suffix),
                 ('narrowhgacl', username + suffix),
                 ('narrowhgacl', 'default' + suffix)
             ]:
                 try:
                     svalue = hgacl.get(section, option)
                     break  # stop at the first value we find
                 except configparser.NoOptionError:
                     pass
             if not svalue:
                 return None
             result = ['/']
             for pattern in svalue.split():
                 result.append(pattern)
                 if '*' not in pattern and '?' not in pattern:
                     result.append(pattern + '/*')
             return result
 
         if os.path.exists(hgacl_file):
             try:
                 hgacl = configparser.RawConfigParser()
                 hgacl.read(hgacl_file)
 
                 includes = read_patterns('.includes')
                 excludes = read_patterns('.excludes')
                 return BasePathPermissionChecker.create_from_patterns(
                     includes, excludes)
             except BaseException as e:
                 msg = 'Cannot read ACL settings from {} on {}: {}'.format(
                     hgacl_file, self.name, e)
                 raise exceptions.RepositoryRequirementError(msg)
         else:
             return None
 
 
 class MercurialIndexBasedCollectionGenerator(CollectionGenerator):
 
     def _commit_factory(self, commit_id):
         if isinstance(commit_id, int):
             return self.repo.get_commit(
                 commit_idx=commit_id, pre_load=self.pre_load)
         else:
             return self.repo.get_commit(
                 commit_id=commit_id, pre_load=self.pre_load)

rhodecode/lib/vcs/backends/svn/commit.py

@@ -1,254 +1,254 b''
 # Copyright (C) 2014-2023 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/
 
 """
 SVN commit module
 """
 
 
 import dateutil.parser
 from zope.cachedescriptors.property import Lazy as LazyProperty
 
 from rhodecode.lib.str_utils import safe_bytes, safe_str
 from rhodecode.lib.vcs import nodes, path as vcspath
 from rhodecode.lib.vcs.backends import base
 from rhodecode.lib.vcs.exceptions import CommitError
 
 
 _SVN_PROP_TRUE = '*'
 
 
 class SubversionCommit(base.BaseCommit):
     """
     Subversion specific implementation of commits
 
     .. attribute:: branch
 
        The Subversion backend does not support to assign branches to
        specific commits. This attribute has always the value `None`.
 
     """
 
     def __init__(self, repository, commit_id):
         self.repository = repository
         self.idx = self.repository._get_commit_idx(commit_id)
         self._svn_rev = self.idx + 1
         self._remote = repository._remote
         # TODO: handling of raw_id should be a method on repository itself,
         # which knows how to translate commit index and commit id
         self.raw_id = commit_id
         self.short_id = commit_id
-        self.id = 'r{}'.format(commit_id)
+        self.id = f'r{commit_id}'
 
         # TODO: Implement the following placeholder attributes
         self.nodes = {}
         self.tags = []
 
     @property
     def author(self):
         return safe_str(self._properties.get('svn:author'))
 
     @property
     def date(self):
         return _date_from_svn_properties(self._properties)
 
     @property
     def message(self):
         return safe_str(self._properties.get('svn:log'))
 
     @LazyProperty
     def _properties(self):
         return self._remote.revision_properties(self._svn_rev)
 
     @LazyProperty
     def parents(self):
         parent_idx = self.idx - 1
         if parent_idx >= 0:
             parent = self.repository.get_commit(commit_idx=parent_idx)
             return [parent]
         return []
 
     @LazyProperty
     def children(self):
         child_idx = self.idx + 1
         if child_idx < len(self.repository.commit_ids):
             child = self.repository.get_commit(commit_idx=child_idx)
             return [child]
         return []
 
     def get_file_mode(self, path: bytes):
         # Note: Subversion flags files which are executable with a special
         # property `svn:executable` which is set to the value ``"*"``.
         if self._get_file_property(path, 'svn:executable') == _SVN_PROP_TRUE:
             return base.FILEMODE_EXECUTABLE
         else:
             return base.FILEMODE_DEFAULT
 
     def is_link(self, path):
         # Note: Subversion has a flag for special files, the content of the
         # file contains the type of that file.
         if self._get_file_property(path, 'svn:special') == _SVN_PROP_TRUE:
             return self.get_file_content(path).startswith(b'link')
         return False
 
     def is_node_binary(self, path):
         path = self._fix_path(path)
         return self._remote.is_binary(self._svn_rev, safe_str(path))
 
     def node_md5_hash(self, path):
         path = self._fix_path(path)
         return self._remote.md5_hash(self._svn_rev, safe_str(path))
 
     def _get_file_property(self, path, name):
         file_properties = self._remote.node_properties(
             safe_str(path), self._svn_rev)
         return file_properties.get(name)
 
     def get_file_content(self, path):
         path = self._fix_path(path)
         return self._remote.get_file_content(self._svn_rev, safe_str(path))
 
     def get_file_content_streamed(self, path):
         path = self._fix_path(path)
 
         stream_method = getattr(self._remote, 'stream:get_file_content')
         return stream_method(self._svn_rev, safe_str(path))
 
     def get_file_size(self, path):
         path = self._fix_path(path)
         return self._remote.get_file_size(self._svn_rev, safe_str(path))
 
     def get_path_history(self, path, limit=None, pre_load=None):
         path = safe_str(self._fix_path(path))
         history = self._remote.node_history(path, self._svn_rev, limit)
         return [
             self.repository.get_commit(commit_id=str(svn_rev))
             for svn_rev in history]
 
     def get_file_annotate(self, path, pre_load=None):
         result = self._remote.file_annotate(safe_str(path), self._svn_rev)
 
         for zero_based_line_no, svn_rev, content in result:
             commit_id = str(svn_rev)
             line_no = zero_based_line_no + 1
             yield (
                 line_no,
                 commit_id,
                 lambda: self.repository.get_commit(commit_id=commit_id),
                 content)
 
     def get_node(self, path, pre_load=None):
         path = self._fix_path(path)
         if path not in self.nodes:
 
             if path == '':
                 node = nodes.RootNode(commit=self)
             else:
                 node_type = self._remote.get_node_type(self._svn_rev, safe_str(path))
                 if node_type == 'dir':
                     node = nodes.DirNode(safe_bytes(path), commit=self)
                 elif node_type == 'file':
                     node = nodes.FileNode(safe_bytes(path), commit=self, pre_load=pre_load)
                 else:
                     raise self.no_node_at_path(path)
 
             self.nodes[path] = node
         return self.nodes[path]
 
     def get_nodes(self, path, pre_load=None):
         if self._get_kind(path) != nodes.NodeKind.DIR:
             raise CommitError(
                 f"Directory does not exist for commit {self.raw_id} at '{path}'")
         path = safe_str(self._fix_path(path))
 
         path_nodes = []
         for name, kind in self._remote.get_nodes(self._svn_rev, path):
             node_path = vcspath.join(path, name)
             if kind == 'dir':
                 node = nodes.DirNode(safe_bytes(node_path), commit=self)
             elif kind == 'file':
                 node = nodes.FileNode(safe_bytes(node_path), commit=self, pre_load=pre_load)
             else:
                 raise ValueError(f"Node kind {kind} not supported.")
             self.nodes[node_path] = node
             path_nodes.append(node)
 
         return path_nodes
 
     def _get_kind(self, path):
         path = self._fix_path(path)
         kind = self._remote.get_node_type(self._svn_rev, path)
         if kind == 'file':
             return nodes.NodeKind.FILE
         elif kind == 'dir':
             return nodes.NodeKind.DIR
         else:
             raise CommitError(
-                "Node does not exist at the given path '{}'".format(path))
+                f"Node does not exist at the given path '{path}'")
 
     @LazyProperty
     def _changes_cache(self):
         return self._remote.revision_changes(self._svn_rev)
 
     @LazyProperty
     def affected_files(self):
         changed_files = set()
         for files in self._changes_cache.values():
             changed_files.update(files)
         return list(changed_files)
 
     @LazyProperty
     def id(self):
         return self.raw_id
 
     @property
     def added(self):
         return nodes.AddedFileNodesGenerator(self.added_paths, self)
 
     @LazyProperty
     def added_paths(self):
         return [n for n in self._changes_cache['added']]
 
     @property
     def changed(self):
         return nodes.ChangedFileNodesGenerator(self.changed_paths, self)
 
     @LazyProperty
     def changed_paths(self):
         return [n for n in self._changes_cache['changed']]
 
     @property
     def removed(self):
         return nodes.RemovedFileNodesGenerator(self.removed_paths, self)
 
     @LazyProperty
     def removed_paths(self):
         return [n for n in self._changes_cache['removed']]
 
 
 def _date_from_svn_properties(properties):
     """
     Parses the date out of given svn properties.
 
     :return: :class:`datetime.datetime` instance. The object is naive.
     """
 
     aware_date = dateutil.parser.parse(properties.get('svn:date'))
     # final_date = aware_date.astimezone(dateutil.tz.tzlocal())
     final_date = aware_date
     return final_date.replace(tzinfo=None)

rhodecode/lib/vcs/backends/svn/repository.py

@@ -1,367 +1,367 b''
 # Copyright (C) 2014-2023 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/
 
 """
 SVN repository module
 """
 
 import logging
 import os
 import urllib.request
 import urllib.parse
 import urllib.error
 
 from zope.cachedescriptors.property import Lazy as LazyProperty
 
 from collections import OrderedDict
 from rhodecode.lib.datelib import date_astimestamp
 from rhodecode.lib.str_utils import safe_str
 from rhodecode.lib.utils2 import CachedProperty
 from rhodecode.lib.vcs import connection, path as vcspath
 from rhodecode.lib.vcs.backends import base
 from rhodecode.lib.vcs.backends.svn.commit import (
     SubversionCommit, _date_from_svn_properties)
 from rhodecode.lib.vcs.backends.svn.diff import SubversionDiff
 from rhodecode.lib.vcs.backends.svn.inmemory import SubversionInMemoryCommit
 from rhodecode.lib.vcs.conf import settings
 from rhodecode.lib.vcs.exceptions import (
     CommitDoesNotExistError, EmptyRepositoryError, RepositoryError,
     VCSError, NodeDoesNotExistError)
 
 
 log = logging.getLogger(__name__)
 
 
 class SubversionRepository(base.BaseRepository):
     """
     Subversion backend implementation
 
     .. important::
 
        It is very important to distinguish the commit index and the commit id
        which is assigned by Subversion. The first one is always handled as an
        `int` by this implementation. The commit id assigned by Subversion on
        the other side will always be a `str`.
 
        There is a specific trap since the first commit will have the index
        ``0`` but the svn id will be ``"1"``.
 
     """
 
     # Note: Subversion does not really have a default branch name.
     DEFAULT_BRANCH_NAME = None
 
     contact = base.BaseRepository.DEFAULT_CONTACT
     description = base.BaseRepository.DEFAULT_DESCRIPTION
 
     def __init__(self, repo_path, config=None, create=False, src_url=None, with_wire=None,
                  bare=False, **kwargs):
         self.path = safe_str(os.path.abspath(repo_path))
         self.config = config if config else self.get_default_config()
         self.with_wire = with_wire or {"cache": False}  # default should not use cache
 
         self._init_repo(create, src_url)
 
         # caches
         self._commit_ids = {}
 
     @LazyProperty
     def _remote(self):
         repo_id = self.path
         return connection.Svn(self.path, repo_id, self.config, with_wire=self.with_wire)
 
     def _init_repo(self, create, src_url):
         if create and os.path.exists(self.path):
             raise RepositoryError(
                 f"Cannot create repository at {self.path}, location already exist"
             )
 
         if create:
             self._remote.create_repository(settings.SVN_COMPATIBLE_VERSION)
             if src_url:
                 src_url = _sanitize_url(src_url)
                 self._remote.import_remote_repository(src_url)
         else:
             self._check_path()
 
     @CachedProperty
     def commit_ids(self):
         head = self._remote.lookup(None)
         return [str(r) for r in range(1, head + 1)]
 
     def _rebuild_cache(self, commit_ids):
         pass
 
     def run_svn_command(self, cmd, **opts):
         """
         Runs given ``cmd`` as svn command and returns tuple
         (stdout, stderr).
 
         :param cmd: full svn command to be executed
         :param opts: env options to pass into Subprocess command
         """
         if not isinstance(cmd, list):
             raise ValueError(f'cmd must be a list, got {type(cmd)} instead')
 
         skip_stderr_log = opts.pop('skip_stderr_log', False)
         out, err = self._remote.run_svn_command(cmd, **opts)
         if err and not skip_stderr_log:
             log.debug('Stderr output of svn command "%s":\n%s', cmd, err)
         return out, err
 
     @LazyProperty
     def branches(self):
         return self._tags_or_branches('vcs_svn_branch')
 
     @LazyProperty
     def branches_closed(self):
         return {}
 
     @LazyProperty
     def bookmarks(self):
         return {}
 
     @LazyProperty
     def branches_all(self):
         # TODO: johbo: Implement proper branch support
         all_branches = {}
         all_branches.update(self.branches)
         all_branches.update(self.branches_closed)
         return all_branches
 
     @LazyProperty
     def tags(self):
         return self._tags_or_branches('vcs_svn_tag')
 
     def _tags_or_branches(self, config_section):
         found_items = {}
 
         if self.is_empty():
             return {}
 
         for pattern in self._patterns_from_section(config_section):
             pattern = vcspath.sanitize(pattern)
             tip = self.get_commit()
             try:
                 if pattern.endswith('*'):
                     basedir = tip.get_node(vcspath.dirname(pattern))
                     directories = basedir.dirs
                 else:
                     directories = (tip.get_node(pattern), )
             except NodeDoesNotExistError:
                 continue
             found_items.update((safe_str(n.path), self.commit_ids[-1]) for n in directories)
 
         def get_name(item):
             return item[0]
 
         return OrderedDict(sorted(found_items.items(), key=get_name))
 
     def _patterns_from_section(self, section):
         return (pattern for key, pattern in self.config.items(section))
 
     def get_common_ancestor(self, commit_id1, commit_id2, repo2):
         if self != repo2:
             raise ValueError(
                 "Subversion does not support getting common ancestor of"
                 " different repositories.")
 
         if int(commit_id1) < int(commit_id2):
             return commit_id1
         return commit_id2
 
     def verify(self):
         verify = self._remote.verify()
 
         self._remote.invalidate_vcs_cache()
         return verify
 
     def compare(self, commit_id1, commit_id2, repo2, merge, pre_load=None):
         # TODO: johbo: Implement better comparison, this is a very naive
         # version which does not allow to compare branches, tags or folders
         # at all.
         if repo2 != self:
             raise ValueError(
                 "Subversion does not support comparison of of different "
                 "repositories.")
 
         if commit_id1 == commit_id2:
             return []
 
         commit_idx1 = self._get_commit_idx(commit_id1)
         commit_idx2 = self._get_commit_idx(commit_id2)
 
         commits = [
             self.get_commit(commit_idx=idx)
             for idx in range(commit_idx1 + 1, commit_idx2 + 1)]
 
         return commits
 
     def _get_commit_idx(self, commit_id):
         try:
             svn_rev = int(commit_id)
         except:
             # TODO: johbo: this might be only one case, HEAD, check this
             svn_rev = self._remote.lookup(commit_id)
         commit_idx = svn_rev - 1
         if commit_idx >= len(self.commit_ids):
             raise CommitDoesNotExistError(
-                "Commit at index {} does not exist.".format(commit_idx))
+                f"Commit at index {commit_idx} does not exist.")
         return commit_idx
 
     @staticmethod
     def check_url(url, config):
         """
         Check if `url` is a valid source to import a Subversion repository.
         """
         # convert to URL if it's a local directory
         if os.path.isdir(url):
             url = 'file://' + urllib.request.pathname2url(url)
         return connection.Svn.check_url(url, config.serialize())
 
     @staticmethod
     def is_valid_repository(path):
         try:
             SubversionRepository(path)
             return True
         except VCSError:
             pass
         return False
 
     def _check_path(self):
         if not os.path.exists(self.path):
-            raise VCSError('Path "{}" does not exist!'.format(self.path))
+            raise VCSError(f'Path "{self.path}" does not exist!')
         if not self._remote.is_path_valid_repository(self.path):
             raise VCSError(
                 'Path "%s" does not contain a Subversion repository' %
                 (self.path, ))
 
     @LazyProperty
     def last_change(self):
         """
         Returns last change made on this repository as
         `datetime.datetime` object.
         """
         # Subversion always has a first commit which has id "0" and contains
         # what we are looking for.
         last_id = len(self.commit_ids)
         properties = self._remote.revision_properties(last_id)
         return _date_from_svn_properties(properties)
 
     @LazyProperty
     def in_memory_commit(self):
         return SubversionInMemoryCommit(self)
 
     def get_hook_location(self):
         """
         returns absolute path to location where hooks are stored
         """
         return os.path.join(self.path, 'hooks')
 
     def get_commit(self, commit_id=None, commit_idx=None, pre_load=None,
                    translate_tag=None, maybe_unreachable=False, reference_obj=None):
         if self.is_empty():
             raise EmptyRepositoryError("There are no commits yet")
         if commit_id is not None:
             self._validate_commit_id(commit_id)
         elif commit_idx is not None:
             self._validate_commit_idx(commit_idx)
             try:
                 commit_id = self.commit_ids[commit_idx]
             except IndexError:
                 raise CommitDoesNotExistError(f'No commit with idx: {commit_idx}')
 
         commit_id = self._sanitize_commit_id(commit_id)
         commit = SubversionCommit(repository=self, commit_id=commit_id)
         return commit
 
     def get_commits(
             self, start_id=None, end_id=None, start_date=None, end_date=None,
             branch_name=None, show_hidden=False, pre_load=None, translate_tags=None):
         if self.is_empty():
             raise EmptyRepositoryError("There are no commit_ids yet")
         self._validate_branch_name(branch_name)
 
         if start_id is not None:
             self._validate_commit_id(start_id)
         if end_id is not None:
             self._validate_commit_id(end_id)
 
         start_raw_id = self._sanitize_commit_id(start_id)
         start_pos = self.commit_ids.index(start_raw_id) if start_id else None
         end_raw_id = self._sanitize_commit_id(end_id)
         end_pos = max(0, self.commit_ids.index(end_raw_id)) if end_id else None
 
         if None not in [start_id, end_id] and start_pos > end_pos:
             raise RepositoryError(
                 "Start commit '%s' cannot be after end commit '%s'" %
                 (start_id, end_id))
         if end_pos is not None:
             end_pos += 1
 
         # Date based filtering
         if start_date or end_date:
             start_raw_id, end_raw_id = self._remote.lookup_interval(
                 date_astimestamp(start_date) if start_date else None,
                 date_astimestamp(end_date) if end_date else None)
             start_pos = start_raw_id - 1
             end_pos = end_raw_id
 
         commit_ids = self.commit_ids
 
         # TODO: johbo: Reconsider impact of DEFAULT_BRANCH_NAME here
         if branch_name not in [None, self.DEFAULT_BRANCH_NAME]:
             svn_rev = int(self.commit_ids[-1])
             commit_ids = self._remote.node_history(
                 path=branch_name, revision=svn_rev, limit=None)
             commit_ids = [str(i) for i in reversed(commit_ids)]
 
         if start_pos or end_pos:
             commit_ids = commit_ids[start_pos:end_pos]
         return base.CollectionGenerator(self, commit_ids, pre_load=pre_load)
 
     def _sanitize_commit_id(self, commit_id):
         if commit_id and commit_id.isdigit():
             if int(commit_id) <= len(self.commit_ids):
                 return commit_id
             else:
                 raise CommitDoesNotExistError(
-                    "Commit {} does not exist.".format(commit_id))
+                    f"Commit {commit_id} does not exist.")
         if commit_id not in [
                 None, 'HEAD', 'tip', self.DEFAULT_BRANCH_NAME]:
             raise CommitDoesNotExistError(
-                "Commit id {} not understood.".format(commit_id))
+                f"Commit id {commit_id} not understood.")
         svn_rev = self._remote.lookup('HEAD')
         return str(svn_rev)
 
     def get_diff(
             self, commit1, commit2, path=None, ignore_whitespace=False,
             context=3, path1=None):
         self._validate_diff_commits(commit1, commit2)
         svn_rev1 = int(commit1.raw_id)
         svn_rev2 = int(commit2.raw_id)
         diff = self._remote.diff(
             svn_rev1, svn_rev2, path1=path1, path2=path,
             ignore_whitespace=ignore_whitespace, context=context)
         return SubversionDiff(diff)
 
 
 def _sanitize_url(url):
     if '://' not in url:
         url = 'file://' + urllib.request.pathname2url(url)
     return url

rhodecode/lib/vcs/utils/helpers.py

@@ -1,159 +1,159 b''
 # Copyright (C) 2014-2023 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/
 
 """
 Utilities aimed to help achieve mostly basic tasks.
 """
 
 
 
 
 import re
 import os
 import time
 import datetime
 import logging
 
 from rhodecode.lib.vcs.conf import settings
 from rhodecode.lib.vcs.exceptions import VCSError, VCSBackendNotSupportedError
 
 
 log = logging.getLogger(__name__)
 
 
 def get_scm(path):
     """
     Returns one of alias from ``ALIASES`` (in order of precedence same as
     shortcuts given in ``ALIASES``) and working dir path for the given
     argument. If no scm-specific directory is found or more than one scm is
     found at that directory, ``VCSError`` is raised.
     """
     if not os.path.isdir(path):
         raise VCSError("Given path %s is not a directory" % path)
 
     found_scms = [(scm, path) for scm in get_scms_for_path(path)]
 
     if len(found_scms) > 1:
         found = ', '.join(x[0] for x in found_scms)
         raise VCSError(
-            'More than one [{}] scm found at given path {}'.format(found, path))
+            f'More than one [{found}] scm found at given path {path}')
 
     if len(found_scms) == 0:
         raise VCSError('No scm found at given path %s' % path)
 
     return found_scms[0]
 
 
 def get_scm_backend(backend_type):
     from rhodecode.lib.vcs.backends import get_backend
     return get_backend(backend_type)
 
 
 def get_scms_for_path(path):
     """
     Returns all scm's found at the given path. If no scm is recognized
     - empty list is returned.
 
     :param path: path to directory which should be checked. May be callable.
 
     :raises VCSError: if given ``path`` is not a directory
     """
     from rhodecode.lib.vcs.backends import get_backend
     if hasattr(path, '__call__'):
         path = path()
     if not os.path.isdir(path):
         raise VCSError("Given path %r is not a directory" % path)
 
     result = []
     for key in settings.available_aliases():
         try:
             backend = get_backend(key)
         except VCSBackendNotSupportedError:
             log.warning('VCSBackendNotSupportedError: %s not supported', key)
             continue
         if backend.is_valid_repository(path):
             result.append(key)
     return result
 
 
 def parse_datetime(text):
     """
     Parses given text and returns ``datetime.datetime`` instance or raises
     ``ValueError``.
 
     :param text: string of desired date/datetime or something more verbose,
       like *yesterday*, *2weeks 3days*, etc.
     """
     if not text:
         raise ValueError('Wrong date: "%s"' % text)
 
     if isinstance(text, datetime.datetime):
         return text
 
     # we limit a format to no include microseconds e.g 2017-10-17t17:48:23.XXXX
     text = text.strip().lower()[:19]
 
     input_formats = (
         '%Y-%m-%d %H:%M:%S',
         '%Y-%m-%dt%H:%M:%S',
         '%Y-%m-%d %H:%M',
         '%Y-%m-%dt%H:%M',
         '%Y-%m-%d',
         '%m/%d/%Y %H:%M:%S',
         '%m/%d/%Yt%H:%M:%S',
         '%m/%d/%Y %H:%M',
         '%m/%d/%Yt%H:%M',
         '%m/%d/%Y',
         '%m/%d/%y %H:%M:%S',
         '%m/%d/%yt%H:%M:%S',
         '%m/%d/%y %H:%M',
         '%m/%d/%yt%H:%M',
         '%m/%d/%y',
     )
     for format_def in input_formats:
         try:
             return datetime.datetime(*time.strptime(text, format_def)[:6])
         except ValueError:
             pass
 
     # Try descriptive texts
     if text == 'tomorrow':
         future = datetime.datetime.now() + datetime.timedelta(days=1)
         args = future.timetuple()[:3] + (23, 59, 59)
         return datetime.datetime(*args)
     elif text == 'today':
         return datetime.datetime(*datetime.datetime.today().timetuple()[:3])
     elif text == 'now':
         return datetime.datetime.now()
     elif text == 'yesterday':
         past = datetime.datetime.now() - datetime.timedelta(days=1)
         return datetime.datetime(*past.timetuple()[:3])
     else:
         days = 0
         matched = re.match(
             r'^((?P<weeks>\d+) ?w(eeks?)?)? ?((?P<days>\d+) ?d(ays?)?)?$', text)
         if matched:
             groupdict = matched.groupdict()
             if groupdict['days']:
                 days += int(matched.groupdict()['days'])
             if groupdict['weeks']:
                 days += int(matched.groupdict()['weeks']) * 7
             past = datetime.datetime.now() - datetime.timedelta(days=days)
             return datetime.datetime(*past.timetuple()[:3])
 
     raise ValueError('Wrong date: "%s"' % text)

rhodecode/lib/view_utils.py

@@ -1,88 +1,87 b''
-
 # Copyright (C) 2010-2023 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/
 
 """
 Utilities to be shared by multiple controllers.
 
 Should only contain utilities to be shared in the controller layer.
 """
 
 from rhodecode.lib import helpers as h
 from rhodecode.lib.vcs.exceptions import RepositoryError
 
 
 def parse_path_ref(ref, default_path=None):
     """
     Parse out a path and reference combination and return both parts of it.
 
     This is used to allow support of path based comparisons for Subversion
     as an iterim solution in parameter handling.
     """
     if '@' in ref:
         return ref.rsplit('@', 1)
     else:
         return default_path, ref
 
 
 def get_format_ref_id(repo):
     """Returns a `repo` specific reference formatter function"""
     if h.is_svn(repo):
         return _format_ref_id_svn
     else:
         return _format_ref_id
 
 
 def _format_ref_id(name, raw_id):
     """Default formatting of a given reference `name`"""
     return name
 
 
 def _format_ref_id_svn(name, raw_id):
     """Special way of formatting a reference for Subversion including path"""
-    return '%s@%s' % (name, raw_id)
+    return '{}@{}'.format(name, raw_id)
 
 
 def get_commit_from_ref_name(repo, ref_name, ref_type=None):
     """
     Gets the commit for a `ref_name` taking into account `ref_type`.
     Needed in case a bookmark / tag share the same name.
 
     :param repo: the repo instance
     :param ref_name: the name of the ref to get
     :param ref_type: optional, used to disambiguate colliding refs
     """
     repo_scm = repo.scm_instance()
     ref_type_mapping = {
         'book': repo_scm.bookmarks,
         'bookmark': repo_scm.bookmarks,
         'tag': repo_scm.tags,
         'branch': repo_scm.branches,
     }
 
     commit_id = ref_name
     if repo_scm.alias != 'svn':  # pass svn refs straight to backend until
                                  # the branch issue with svn is fixed
         if ref_type and ref_type in ref_type_mapping:
             try:
                 commit_id = ref_type_mapping[ref_type][ref_name]
             except KeyError:
                 raise RepositoryError(
-                    '%s "%s" does not exist' % (ref_type, ref_name))
+                    '{} "{}" does not exist'.format(ref_type, ref_name))
 
     return repo_scm.get_commit(commit_id)

rhodecode/model/__init__.py

@@ -1,141 +1,140 b''
-
 # Copyright (C) 2010-2023 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 logging
 
 import rhodecode
 from rhodecode.model import meta, db
 from rhodecode.lib.utils2 import obfuscate_url_pw, get_encryption_key
 
 log = logging.getLogger(__name__)
 
 
 def init_model(engine, encryption_key: bytes = b''):
     """
     Initializes db session, bind the engine with the metadata,
     Call this before using any of the tables or classes in the model,
     preferably once in application start
 
     :param engine: engine to bind to
     :param encryption_key: key used for encryption
     """
 
     engine_str = obfuscate_url_pw(str(engine.url))
     log.info("RhodeCode %s initializing db for %s", rhodecode.__version__, engine_str)
 
     meta.bind_engine_to_session(engine)
     init_model_encryption(db, enc_key=encryption_key)
 
 
 def init_model_encryption(*db_models, enc_key: bytes = b'', config=None):
     if not enc_key:
         from pyramid.threadlocal import get_current_registry
         config = config or get_current_registry().settings
         enc_key = get_encryption_key(config)
 
     for db_model in db_models:
         log.debug('setting encryption key for model %s', db_model)
         db_model.ENCRYPTION_KEY = enc_key
 
 
 class BaseModel(object):
     """
     Base Model for all RhodeCode models, it adds sql alchemy session
     into instance of model
 
     :param sa: If passed it reuses this session instead of creating a new one
     """
 
     cls = None  # override in child class
 
     def __init__(self, sa=None):
         if sa is not None:
             self.sa = sa
         else:
             self.sa = meta.Session()
 
     def _get_instance(self, cls, instance, callback=None):
         """
         Gets instance of given cls using some simple lookup mechanism.
 
         :param cls: classes to fetch
         :param instance: int or Instance
         :param callback: callback to call if all lookups failed
         """
 
         if isinstance(instance, cls):
             return instance
         elif isinstance(instance, int):
             if isinstance(cls, tuple):
                 # if we pass multi instances we pick first to .get()
                 cls = cls[0]
             return cls.get(instance)
         else:
             if instance:
                 if callback is None:
                     raise Exception(
                         'given object must be int or Instance of %s '
                         'got %s, no callback provided' % (cls, type(instance))
                     )
                 else:
                     return callback(instance)
 
     def _get_user(self, user):
         """
         Helper method to get user by ID, or username fallback
 
         :param user: UserID, username, or User instance
         """
         return self._get_instance(
             db.User, user, callback=db.User.get_by_username)
 
     def _get_user_group(self, user_group):
         """
         Helper method to get user by ID, or username fallback
 
         :param user_group: UserGroupID, user_group_name, or UserGroup instance
         """
         return self._get_instance(
             db.UserGroup, user_group, callback=db.UserGroup.get_by_group_name)
 
     def _get_repo(self, repository):
         """
         Helper method to get repository by ID, or repository name
 
         :param repository: RepoID, repository name or Repository Instance
         """
         return self._get_instance(
             db.Repository, repository, callback=db.Repository.get_by_repo_name)
 
     def _get_perm(self, permission):
         """
         Helper method to get permission by ID, or permission name
 
         :param permission: PermissionID, permission_name or Permission instance
         """
         return self._get_instance(
             db.Permission, permission, callback=db.Permission.get_by_key)
 
     @classmethod
     def get_all(cls):
         """
         Returns all instances of what is defined in `cls` class variable
         """
         return cls.cls.getAll()

rhodecode/model/auth_token.py

@@ -1,124 +1,122 b''
-
-
 # Copyright (C) 2013-2023 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/
 
 """
 authentication tokens model for RhodeCode
 """
 
 import time
 import logging
 import traceback
 from sqlalchemy import or_
 
 from rhodecode.model import BaseModel
 from rhodecode.model.db import UserApiKeys
 from rhodecode.model.meta import Session
 
 log = logging.getLogger(__name__)
 
 
 class AuthTokenModel(BaseModel):
     cls = UserApiKeys
 
     @classmethod
     def get_lifetime_values(cls, translator):
         from rhodecode.lib import helpers as h
         _ = translator
 
         def date_after_min(mins):
             after = time.time() + (60 * mins)
             return h.format_date(h.time_to_datetime(after))
 
         return [
             (str(-1),
              _('forever')),
             (str(5),
              _('5 minutes {end_date}').format(end_date=date_after_min(5))),
             (str(60),
              _('1 hour {end_date}').format(end_date=date_after_min(60))),
             (str(60 * 24),
              _('1 day {end_date}').format(end_date=date_after_min(60 * 24))),
             (str(60 * 24 * 30),
              _('1 month {end_date}').format(end_date=date_after_min(60 * 24 * 30))),
         ]
 
     def create(self, user, description, lifetime=-1, role=UserApiKeys.ROLE_ALL):
         """
         :param user: user or user_id
         :param description: description of ApiKey
         :param lifetime: expiration time in minutes
         :param role: role for the apikey
         """
         from rhodecode.lib.auth import generate_auth_token
 
         user = self._get_user(user)
 
         new_auth_token = UserApiKeys()
         new_auth_token.api_key = generate_auth_token(user.username)
         new_auth_token.user_id = user.user_id
         new_auth_token.description = description
         new_auth_token.role = role
         new_auth_token.expires = time.time() + (lifetime * 60) \
             if lifetime != -1 else -1
         Session().add(new_auth_token)
 
         return new_auth_token
 
     def delete(self, auth_token_id, user=None):
         """
         Deletes given api_key, if user is set it also filters the object for
         deletion by given user.
         """
         auth_token = UserApiKeys.query().filter(
             UserApiKeys.user_api_key_id == auth_token_id)
 
         if user:
             user = self._get_user(user)
             auth_token = auth_token.filter(UserApiKeys.user_id == user.user_id)
             auth_token = auth_token.scalar()
 
         if auth_token:
             try:
                 Session().delete(auth_token)
             except Exception:
                 log.error(traceback.format_exc())
                 raise
 
     def get_auth_tokens(self, user, show_expired=True):
         user = self._get_user(user)
         user_auth_tokens = UserApiKeys.query()\
             .filter(UserApiKeys.user_id == user.user_id)
         if not show_expired:
             user_auth_tokens = user_auth_tokens\
                 .filter(or_(UserApiKeys.expires == -1,
                             UserApiKeys.expires >= time.time()))
         user_auth_tokens = user_auth_tokens.order_by(
             UserApiKeys.user_api_key_id)
         return user_auth_tokens
 
     def get_auth_token(self, auth_token):
         auth_token = UserApiKeys.query().filter(
             UserApiKeys.api_key == auth_token)
         auth_token = auth_token \
             .filter(or_(UserApiKeys.expires == -1,
                         UserApiKeys.expires >= time.time()))\
             .first()
 
         return auth_token

rhodecode/model/changeset_status.py

@@ -1,403 +1,402 b''
-
 # Copyright (C) 2010-2023 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 itertools
 import logging
 import collections
 
 from rhodecode.model import BaseModel
 from rhodecode.model.db import (
     ChangesetStatus, ChangesetComment, PullRequest, PullRequestReviewers, Session)
 from rhodecode.lib.exceptions import StatusChangeOnClosedPullRequestError
 from rhodecode.lib.markup_renderer import (
     DEFAULT_COMMENTS_RENDERER, RstTemplateRenderer)
 
 log = logging.getLogger(__name__)
 
 
 class ChangesetStatusModel(BaseModel):
 
     cls = ChangesetStatus
 
     def __get_changeset_status(self, changeset_status):
         return self._get_instance(ChangesetStatus, changeset_status)
 
     def __get_pull_request(self, pull_request):
         return self._get_instance(PullRequest, pull_request)
 
     def _get_status_query(self, repo, revision, pull_request,
                           with_revisions=False):
         repo = self._get_repo(repo)
 
         q = ChangesetStatus.query()\
             .filter(ChangesetStatus.repo == repo)
         if not with_revisions:
             q = q.filter(ChangesetStatus.version == 0)
 
         if revision:
             q = q.filter(ChangesetStatus.revision == revision)
         elif pull_request:
             pull_request = self.__get_pull_request(pull_request)
             # TODO: johbo: Think about the impact of this join, there must
             # be a reason why ChangesetStatus and ChanagesetComment is linked
             # to the pull request. Might be that we want to do the same for
             # the pull_request_version_id.
             q = q.join(ChangesetComment).filter(
                 ChangesetStatus.pull_request == pull_request,
                 ChangesetComment.pull_request_version_id == None)
         else:
             raise Exception('Please specify revision or pull_request')
         q = q.order_by(ChangesetStatus.version.asc())
         return q
 
     def calculate_group_vote(self, group_id, group_statuses_by_reviewers,
                              trim_votes=True):
         """
         Calculate status based on given group members, and voting rule
 
 
         group1 - 4 members, 3 required for approval
             user1 - approved
             user2 - reject
             user3 - approved
             user4 - rejected
 
         final_state: rejected, reasons not at least 3 votes
 
 
         group1 - 4 members, 2 required for approval
             user1 - approved
             user2 - reject
             user3 - approved
             user4 - rejected
 
         final_state: approved, reasons got at least 2 approvals
 
         group1 - 4 members, ALL required for approval
             user1 - approved
             user2 - reject
             user3 - approved
             user4 - rejected
 
         final_state: rejected, reasons not all approvals
 
 
         group1 - 4 members, ALL required for approval
             user1 - approved
             user2 - approved
             user3 - approved
             user4 - approved
 
         final_state: approved, reason all approvals received
 
         group1 - 4 members, 5 required for approval
             (approval should be shorted to number of actual members)
 
             user1 - approved
             user2 - approved
             user3 - approved
             user4 - approved
 
         final_state: approved, reason all approvals received
 
         """
         group_vote_data = {}
         got_rule = False
         members = collections.OrderedDict()
         for review_obj, user, reasons, mandatory, statuses \
                 in group_statuses_by_reviewers:
 
             if not got_rule:
                 group_vote_data = review_obj.rule_user_group_data()
                 got_rule = bool(group_vote_data)
 
             members[user.user_id] = statuses
 
         if not group_vote_data:
             return []
 
         required_votes = group_vote_data['vote_rule']
         if required_votes == -1:
             # -1 means all required, so we replace it with how many people
             # are in the members
             required_votes = len(members)
 
         if trim_votes and required_votes > len(members):
             # we require more votes than we have members in the group
             # in this case we trim the required votes to the number of members
             required_votes = len(members)
 
         approvals = sum([
             1 for statuses in members.values()
             if statuses and
                statuses[0][1].status == ChangesetStatus.STATUS_APPROVED])
 
         calculated_votes = []
         # we have all votes from users, now check if we have enough votes
         # to fill other
         fill_in = ChangesetStatus.STATUS_UNDER_REVIEW
         if approvals >= required_votes:
             fill_in = ChangesetStatus.STATUS_APPROVED
 
         for member, statuses in members.items():
             if statuses:
                 ver, latest = statuses[0]
                 if fill_in == ChangesetStatus.STATUS_APPROVED:
                     calculated_votes.append(fill_in)
                 else:
                     calculated_votes.append(latest.status)
             else:
                 calculated_votes.append(fill_in)
 
         return calculated_votes
 
     def calculate_status(self, statuses_by_reviewers):
         """
         Given the approval statuses from reviewers, calculates final approval
         status. There can only be 3 results, all approved, all rejected. If
         there is no consensus the PR is under review.
 
         :param statuses_by_reviewers:
         """
 
         def group_rule(element):
             _review_obj = element[0]
             rule_data = _review_obj.rule_user_group_data()
             if rule_data and rule_data['id']:
                 return rule_data['id']
             # don't return None, as we cant compare this
             return 0
 
         voting_groups = itertools.groupby(sorted(statuses_by_reviewers, key=group_rule), group_rule)
 
         voting_by_groups = [(x, list(y)) for x, y in voting_groups]
 
         reviewers_number = len(statuses_by_reviewers)
         votes = collections.defaultdict(int)
         for group, group_statuses_by_reviewers in voting_by_groups:
             if group:
                 # calculate how the "group" voted
                 for vote_status in self.calculate_group_vote(
                         group, group_statuses_by_reviewers):
                     votes[vote_status] += 1
             else:
 
                 for review_obj, user, reasons, mandatory, statuses \
                         in group_statuses_by_reviewers:
                     # individual vote
                     if statuses:
                         ver, latest = statuses[0]
                         votes[latest.status] += 1
 
         approved_votes_count = votes[ChangesetStatus.STATUS_APPROVED]
         rejected_votes_count = votes[ChangesetStatus.STATUS_REJECTED]
 
         # TODO(marcink): with group voting, how does rejected work,
         # do we ever get rejected state ?
 
         if approved_votes_count and (approved_votes_count == reviewers_number):
             return ChangesetStatus.STATUS_APPROVED
 
         if rejected_votes_count and (rejected_votes_count == reviewers_number):
             return ChangesetStatus.STATUS_REJECTED
 
         return ChangesetStatus.STATUS_UNDER_REVIEW
 
     def get_statuses(self, repo, revision=None, pull_request=None,
                      with_revisions=False):
         q = self._get_status_query(repo, revision, pull_request,
                                    with_revisions)
         return q.all()
 
     def get_status(self, repo, revision=None, pull_request=None, as_str=True):
         """
         Returns latest status of changeset for given revision or for given
         pull request. Statuses are versioned inside a table itself and
         version == 0 is always the current one
 
         :param repo:
         :param revision: 40char hash or None
         :param pull_request: pull_request reference
         :param as_str: return status as string not object
         """
         q = self._get_status_query(repo, revision, pull_request)
 
         # need to use first here since there can be multiple statuses
         # returned from pull_request
         status = q.first()
         if as_str:
             status = status.status if status else status
             st = status or ChangesetStatus.DEFAULT
             return str(st)
         return status
 
     def _render_auto_status_message(
             self, status, commit_id=None, pull_request=None):
         """
         render the message using DEFAULT_COMMENTS_RENDERER (RST renderer),
         so it's always looking the same disregarding on which default
         renderer system is using.
 
         :param status: status text to change into
         :param commit_id: the commit_id we change the status for
         :param pull_request: the pull request we change the status for
         """
 
         new_status = ChangesetStatus.get_status_lbl(status)
 
         params = {
             'new_status_label': new_status,
             'pull_request': pull_request,
             'commit_id': commit_id,
         }
         renderer = RstTemplateRenderer()
         return renderer.render('auto_status_change.mako', **params)
 
     def set_status(self, repo, status, user, comment=None, revision=None,
                    pull_request=None, dont_allow_on_closed_pull_request=False):
         """
         Creates new status for changeset or updates the old ones bumping their
         version, leaving the current status at
 
         :param repo:
         :param revision:
         :param status:
         :param user:
         :param comment:
         :param dont_allow_on_closed_pull_request: don't allow a status change
             if last status was for pull request and it's closed. We shouldn't
             mess around this manually
         """
         repo = self._get_repo(repo)
 
         q = ChangesetStatus.query()
 
         if revision:
             q = q.filter(ChangesetStatus.repo == repo)
             q = q.filter(ChangesetStatus.revision == revision)
         elif pull_request:
             pull_request = self.__get_pull_request(pull_request)
             q = q.filter(ChangesetStatus.repo == pull_request.source_repo)
             q = q.filter(ChangesetStatus.revision.in_(pull_request.revisions))
         cur_statuses = q.all()
 
         # if statuses exists and last is associated with a closed pull request
         # we need to check if we can allow this status change
         if (dont_allow_on_closed_pull_request and cur_statuses
             and getattr(cur_statuses[0].pull_request, 'status', '')
                 == PullRequest.STATUS_CLOSED):
             raise StatusChangeOnClosedPullRequestError(
                 'Changing status on closed pull request is not allowed'
             )
 
         # update all current statuses with older version
         if cur_statuses:
             for st in cur_statuses:
                 st.version += 1
                 Session().add(st)
             Session().flush()
 
         def _create_status(user, repo, status, comment, revision, pull_request):
             new_status = ChangesetStatus()
             new_status.author = self._get_user(user)
             new_status.repo = self._get_repo(repo)
             new_status.status = status
             new_status.comment = comment
             new_status.revision = revision
             new_status.pull_request = pull_request
             return new_status
 
         if not comment:
             from rhodecode.model.comment import CommentsModel
             comment = CommentsModel().create(
                 text=self._render_auto_status_message(
                     status, commit_id=revision, pull_request=pull_request),
                 repo=repo,
                 user=user,
                 pull_request=pull_request,
                 send_email=False, renderer=DEFAULT_COMMENTS_RENDERER
             )
 
         if revision:
             new_status = _create_status(
                 user=user, repo=repo, status=status, comment=comment,
                 revision=revision, pull_request=pull_request)
             Session().add(new_status)
             return new_status
         elif pull_request:
             # pull request can have more than one revision associated to it
             # we need to create new version for each one
             new_statuses = []
             repo = pull_request.source_repo
             for rev in pull_request.revisions:
                 new_status = _create_status(
                     user=user, repo=repo, status=status, comment=comment,
                     revision=rev, pull_request=pull_request)
                 new_statuses.append(new_status)
                 Session().add(new_status)
             return new_statuses
 
     def aggregate_votes_by_user(self, commit_statuses, reviewers_data, user=None):
 
         commit_statuses_map = collections.defaultdict(list)
         for st in commit_statuses:
             commit_statuses_map[st.author.username] += [st]
 
         reviewers = []
 
         def version(commit_status):
             return commit_status.version
 
         for obj in reviewers_data:
             if not obj.user:
                 continue
             if user and obj.user.username != user.username:
                 # single user filter
                 continue
 
             statuses = commit_statuses_map.get(obj.user.username, None)
             if statuses:
                 status_groups = itertools.groupby(
                     sorted(statuses, key=version), version)
                 statuses = [(x, list(y)[0]) for x, y in status_groups]
 
             reviewers.append((obj, obj.user, obj.reasons, obj.mandatory, statuses))
 
         if user:
             return reviewers[0] if reviewers else reviewers
         else:
             return reviewers
 
     def reviewers_statuses(self, pull_request, user=None):
         _commit_statuses = self.get_statuses(
             pull_request.source_repo,
             pull_request=pull_request,
             with_revisions=True)
         reviewers = pull_request.get_pull_request_reviewers(
             role=PullRequestReviewers.ROLE_REVIEWER)
         return self.aggregate_votes_by_user(_commit_statuses, reviewers, user=user)
 
     def calculated_review_status(self, pull_request):
         """
         calculate pull request status based on reviewers, it should be a list
         of two element lists.
         """
         reviewers = self.reviewers_statuses(pull_request)
         return self.calculate_status(reviewers)

rhodecode/model/comment.py

@@ -1,855 +1,852 b''
-
-
 # Copyright (C) 2011-2023 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/
 
 """
 comments model for RhodeCode
 """
 import datetime
 
 import logging
 import traceback
 import collections
 
 from pyramid.threadlocal import get_current_registry, get_current_request
 from sqlalchemy.sql.expression import null
 from sqlalchemy.sql.functions import coalesce
 
 from rhodecode.lib import helpers as h, diffs, channelstream, hooks_utils
 from rhodecode.lib import audit_logger
 from rhodecode.lib.exceptions import CommentVersionMismatch
 from rhodecode.lib.utils2 import extract_mentioned_users, safe_str, safe_int
 from rhodecode.model import BaseModel
 from rhodecode.model.db import (
     false, true,
     ChangesetComment,
     User,
     Notification,
     PullRequest,
     AttributeDict,
     ChangesetCommentHistory,
 )
 from rhodecode.model.notification import NotificationModel
 from rhodecode.model.meta import Session
 from rhodecode.model.settings import VcsSettingsModel
 from rhodecode.model.notification import EmailNotificationModel
 from rhodecode.model.validation_schema.schemas import comment_schema
 
 
 log = logging.getLogger(__name__)
 
 
 class CommentsModel(BaseModel):
 
     cls = ChangesetComment
 
     DIFF_CONTEXT_BEFORE = 3
     DIFF_CONTEXT_AFTER = 3
 
     def __get_commit_comment(self, changeset_comment):
         return self._get_instance(ChangesetComment, changeset_comment)
 
     def __get_pull_request(self, pull_request):
         return self._get_instance(PullRequest, pull_request)
 
     def _extract_mentions(self, s):
         user_objects = []
         for username in extract_mentioned_users(s):
             user_obj = User.get_by_username(username, case_insensitive=True)
             if user_obj:
                 user_objects.append(user_obj)
         return user_objects
 
     def _get_renderer(self, global_renderer='rst', request=None):
         request = request or get_current_request()
 
         try:
             global_renderer = request.call_context.visual.default_renderer
         except AttributeError:
             log.debug("Renderer not set, falling back "
                       "to default renderer '%s'", global_renderer)
         except Exception:
             log.error(traceback.format_exc())
         return global_renderer
 
     def aggregate_comments(self, comments, versions, show_version, inline=False):
         # group by versions, and count until, and display objects
 
         comment_groups = collections.defaultdict(list)
         [comment_groups[_co.pull_request_version_id].append(_co) for _co in comments]
 
         def yield_comments(pos):
-            for co in comment_groups[pos]:
-                yield co
+            yield from comment_groups[pos]
 
         comment_versions = collections.defaultdict(
             lambda: collections.defaultdict(list))
         prev_prvid = -1
         # fake last entry with None, to aggregate on "latest" version which
         # doesn't have an pull_request_version_id
         for ver in versions + [AttributeDict({'pull_request_version_id': None})]:
             prvid = ver.pull_request_version_id
             if prev_prvid == -1:
                 prev_prvid = prvid
 
             for co in yield_comments(prvid):
                 comment_versions[prvid]['at'].append(co)
 
             # save until
             current = comment_versions[prvid]['at']
             prev_until = comment_versions[prev_prvid]['until']
             cur_until = prev_until + current
             comment_versions[prvid]['until'].extend(cur_until)
 
             # save outdated
             if inline:
                 outdated = [x for x in cur_until
                             if x.outdated_at_version(show_version)]
             else:
                 outdated = [x for x in cur_until
                             if x.older_than_version(show_version)]
             display = [x for x in cur_until if x not in outdated]
 
             comment_versions[prvid]['outdated'] = outdated
             comment_versions[prvid]['display'] = display
 
             prev_prvid = prvid
 
         return comment_versions
 
     def get_repository_comments(self, repo, comment_type=None, user=None, commit_id=None):
         qry = Session().query(ChangesetComment) \
             .filter(ChangesetComment.repo == repo)
 
         if comment_type and comment_type in ChangesetComment.COMMENT_TYPES:
             qry = qry.filter(ChangesetComment.comment_type == comment_type)
 
         if user:
             user = self._get_user(user)
             if user:
                 qry = qry.filter(ChangesetComment.user_id == user.user_id)
 
         if commit_id:
             qry = qry.filter(ChangesetComment.revision == commit_id)
 
         qry = qry.order_by(ChangesetComment.created_on)
         return qry.all()
 
     def get_repository_unresolved_todos(self, repo):
         todos = Session().query(ChangesetComment) \
             .filter(ChangesetComment.repo == repo) \
             .filter(ChangesetComment.resolved_by == None) \
             .filter(ChangesetComment.comment_type
                     == ChangesetComment.COMMENT_TYPE_TODO)
         todos = todos.all()
 
         return todos
 
     def get_pull_request_unresolved_todos(self, pull_request, show_outdated=True, include_drafts=True):
 
         todos = Session().query(ChangesetComment) \
             .filter(ChangesetComment.pull_request == pull_request) \
             .filter(ChangesetComment.resolved_by == None) \
             .filter(ChangesetComment.comment_type
                     == ChangesetComment.COMMENT_TYPE_TODO)
 
         if not include_drafts:
             todos = todos.filter(ChangesetComment.draft == false())
 
         if not show_outdated:
             todos = todos.filter(
                 coalesce(ChangesetComment.display_state, '') !=
                 ChangesetComment.COMMENT_OUTDATED)
 
         todos = todos.all()
 
         return todos
 
     def get_pull_request_resolved_todos(self, pull_request, show_outdated=True, include_drafts=True):
 
         todos = Session().query(ChangesetComment) \
             .filter(ChangesetComment.pull_request == pull_request) \
             .filter(ChangesetComment.resolved_by != None) \
             .filter(ChangesetComment.comment_type
                     == ChangesetComment.COMMENT_TYPE_TODO)
 
         if not include_drafts:
             todos = todos.filter(ChangesetComment.draft == false())
 
         if not show_outdated:
             todos = todos.filter(
                 coalesce(ChangesetComment.display_state, '') !=
                 ChangesetComment.COMMENT_OUTDATED)
 
         todos = todos.all()
 
         return todos
 
     def get_pull_request_drafts(self, user_id, pull_request):
         drafts = Session().query(ChangesetComment) \
             .filter(ChangesetComment.pull_request == pull_request) \
             .filter(ChangesetComment.user_id == user_id) \
             .filter(ChangesetComment.draft == true())
         return drafts.all()
 
     def get_commit_unresolved_todos(self, commit_id, show_outdated=True, include_drafts=True):
 
         todos = Session().query(ChangesetComment) \
             .filter(ChangesetComment.revision == commit_id) \
             .filter(ChangesetComment.resolved_by == None) \
             .filter(ChangesetComment.comment_type
                     == ChangesetComment.COMMENT_TYPE_TODO)
 
         if not include_drafts:
             todos = todos.filter(ChangesetComment.draft == false())
 
         if not show_outdated:
             todos = todos.filter(
                 coalesce(ChangesetComment.display_state, '') !=
                 ChangesetComment.COMMENT_OUTDATED)
 
         todos = todos.all()
 
         return todos
 
     def get_commit_resolved_todos(self, commit_id, show_outdated=True, include_drafts=True):
 
         todos = Session().query(ChangesetComment) \
             .filter(ChangesetComment.revision == commit_id) \
             .filter(ChangesetComment.resolved_by != None) \
             .filter(ChangesetComment.comment_type
                     == ChangesetComment.COMMENT_TYPE_TODO)
 
         if not include_drafts:
             todos = todos.filter(ChangesetComment.draft == false())
 
         if not show_outdated:
             todos = todos.filter(
                 coalesce(ChangesetComment.display_state, '') !=
                 ChangesetComment.COMMENT_OUTDATED)
 
         todos = todos.all()
 
         return todos
 
     def get_commit_inline_comments(self, commit_id, include_drafts=True):
         inline_comments = Session().query(ChangesetComment) \
             .filter(ChangesetComment.line_no != None) \
             .filter(ChangesetComment.f_path != None) \
             .filter(ChangesetComment.revision == commit_id)
 
         if not include_drafts:
             inline_comments = inline_comments.filter(ChangesetComment.draft == false())
 
         inline_comments = inline_comments.all()
         return inline_comments
 
     def _log_audit_action(self, action, action_data, auth_user, comment):
         audit_logger.store(
             action=action,
             action_data=action_data,
             user=auth_user,
             repo=comment.repo)
 
     def create(self, text, repo, user, commit_id=None, pull_request=None,
                f_path=None, line_no=None, status_change=None,
                status_change_type=None, comment_type=None, is_draft=False,
                resolves_comment_id=None, closing_pr=False, send_email=True,
                renderer=None, auth_user=None, extra_recipients=None):
         """
         Creates new comment for commit or pull request.
         IF status_change is not none this comment is associated with a
         status change of commit or commit associated with pull request
 
         :param text:
         :param repo:
         :param user:
         :param commit_id:
         :param pull_request:
         :param f_path:
         :param line_no:
         :param status_change: Label for status change
         :param comment_type: Type of comment
         :param is_draft: is comment a draft only
         :param resolves_comment_id: id of comment which this one will resolve
         :param status_change_type: type of status change
         :param closing_pr:
         :param send_email:
         :param renderer: pick renderer for this comment
         :param auth_user: current authenticated user calling this method
         :param extra_recipients: list of extra users to be added to recipients
         """
 
         request = get_current_request()
         _ = request.translate
 
         if not renderer:
             renderer = self._get_renderer(request=request)
 
         repo = self._get_repo(repo)
         user = self._get_user(user)
         auth_user = auth_user or user
 
         schema = comment_schema.CommentSchema()
         validated_kwargs = schema.deserialize(dict(
             comment_body=text,
             comment_type=comment_type,
             is_draft=is_draft,
             comment_file=f_path,
             comment_line=line_no,
             renderer_type=renderer,
             status_change=status_change_type,
             resolves_comment_id=resolves_comment_id,
             repo=repo.repo_id,
             user=user.user_id,
         ))
 
         is_draft = validated_kwargs['is_draft']
 
         comment = ChangesetComment()
         comment.renderer = validated_kwargs['renderer_type']
         comment.text = validated_kwargs['comment_body']
         comment.f_path = validated_kwargs['comment_file']
         comment.line_no = validated_kwargs['comment_line']
         comment.comment_type = validated_kwargs['comment_type']
         comment.draft = is_draft
 
         comment.repo = repo
         comment.author = user
         resolved_comment = self.__get_commit_comment(
             validated_kwargs['resolves_comment_id'])
 
         # check if the comment actually belongs to this PR
         if resolved_comment and resolved_comment.pull_request and \
                 resolved_comment.pull_request != pull_request:
             log.warning('Comment tried to resolved unrelated todo comment: %s',
                         resolved_comment)
             # comment not bound to this pull request, forbid
             resolved_comment = None
 
         elif resolved_comment and resolved_comment.repo and \
                 resolved_comment.repo != repo:
             log.warning('Comment tried to resolved unrelated todo comment: %s',
                         resolved_comment)
             # comment not bound to this repo, forbid
             resolved_comment = None
 
         if resolved_comment and resolved_comment.resolved_by:
             # if this comment is already resolved, don't mark it again!
             resolved_comment = None
 
         comment.resolved_comment = resolved_comment
 
         pull_request_id = pull_request
 
         commit_obj = None
         pull_request_obj = None
 
         if commit_id:
             notification_type = EmailNotificationModel.TYPE_COMMIT_COMMENT
             # do a lookup, so we don't pass something bad here
             commit_obj = repo.scm_instance().get_commit(commit_id=commit_id)
             comment.revision = commit_obj.raw_id
 
         elif pull_request_id:
             notification_type = EmailNotificationModel.TYPE_PULL_REQUEST_COMMENT
             pull_request_obj = self.__get_pull_request(pull_request_id)
             comment.pull_request = pull_request_obj
         else:
             raise Exception('Please specify commit or pull_request_id')
 
         Session().add(comment)
         Session().flush()
         kwargs = {
             'user': user,
             'renderer_type': renderer,
             'repo_name': repo.repo_name,
             'status_change': status_change,
             'status_change_type': status_change_type,
             'comment_body': text,
             'comment_file': f_path,
             'comment_line': line_no,
             'comment_type': comment_type or 'note',
             'comment_id': comment.comment_id
         }
 
         if commit_obj:
             recipients = ChangesetComment.get_users(
                 revision=commit_obj.raw_id)
             # add commit author if it's in RhodeCode system
             cs_author = User.get_from_cs_author(commit_obj.author)
             if not cs_author:
                 # use repo owner if we cannot extract the author correctly
                 cs_author = repo.user
             recipients += [cs_author]
 
             commit_comment_url = self.get_url(comment, request=request)
             commit_comment_reply_url = self.get_url(
                 comment, request=request,
-                anchor='comment-{}/?/ReplyToComment'.format(comment.comment_id))
+                anchor=f'comment-{comment.comment_id}/?/ReplyToComment')
 
             target_repo_url = h.link_to(
                 repo.repo_name,
                 h.route_url('repo_summary', repo_name=repo.repo_name))
 
             commit_url = h.route_url('repo_commit', repo_name=repo.repo_name,
                                      commit_id=commit_id)
 
             # commit specifics
             kwargs.update({
                 'commit': commit_obj,
                 'commit_message': commit_obj.message,
                 'commit_target_repo_url': target_repo_url,
                 'commit_comment_url': commit_comment_url,
                 'commit_comment_reply_url': commit_comment_reply_url,
                 'commit_url': commit_url,
                 'thread_ids': [commit_url, commit_comment_url],
             })
 
         elif pull_request_obj:
             # get the current participants of this pull request
             recipients = ChangesetComment.get_users(
                 pull_request_id=pull_request_obj.pull_request_id)
             # add pull request author
             recipients += [pull_request_obj.author]
 
             # add the reviewers to notification
             recipients += [x.user for x in pull_request_obj.get_pull_request_reviewers()]
 
             pr_target_repo = pull_request_obj.target_repo
             pr_source_repo = pull_request_obj.source_repo
 
             pr_comment_url = self.get_url(comment, request=request)
             pr_comment_reply_url = self.get_url(
                 comment, request=request,
-                anchor='comment-{}/?/ReplyToComment'.format(comment.comment_id))
+                anchor=f'comment-{comment.comment_id}/?/ReplyToComment')
 
             pr_url = h.route_url(
                 'pullrequest_show',
                 repo_name=pr_target_repo.repo_name,
                 pull_request_id=pull_request_obj.pull_request_id, )
 
             # set some variables for email notification
             pr_target_repo_url = h.route_url(
                 'repo_summary', repo_name=pr_target_repo.repo_name)
 
             pr_source_repo_url = h.route_url(
                 'repo_summary', repo_name=pr_source_repo.repo_name)
 
             # pull request specifics
             kwargs.update({
                 'pull_request': pull_request_obj,
                 'pr_id': pull_request_obj.pull_request_id,
                 'pull_request_url': pr_url,
                 'pull_request_target_repo': pr_target_repo,
                 'pull_request_target_repo_url': pr_target_repo_url,
                 'pull_request_source_repo': pr_source_repo,
                 'pull_request_source_repo_url': pr_source_repo_url,
                 'pr_comment_url': pr_comment_url,
                 'pr_comment_reply_url': pr_comment_reply_url,
                 'pr_closing': closing_pr,
                 'thread_ids': [pr_url, pr_comment_url],
             })
 
         if send_email:
             recipients += [self._get_user(u) for u in (extra_recipients or [])]
 
             mention_recipients = set(
                 self._extract_mentions(text)).difference(recipients)
 
             # create notification objects, and emails
             NotificationModel().create(
                 created_by=user,
                 notification_subject='',  # Filled in based on the notification_type
                 notification_body='',  # Filled in based on the notification_type
                 notification_type=notification_type,
                 recipients=recipients,
                 mention_recipients=mention_recipients,
                 email_kwargs=kwargs,
             )
 
         Session().flush()
         if comment.pull_request:
             action = 'repo.pull_request.comment.create'
         else:
             action = 'repo.commit.comment.create'
 
         if not is_draft:
             comment_data = comment.get_api_data()
 
             self._log_audit_action(
                 action, {'data': comment_data}, auth_user, comment)
 
         return comment
 
     def edit(self, comment_id, text, auth_user, version):
         """
         Change existing comment for commit or pull request.
 
         :param comment_id:
         :param text:
         :param auth_user: current authenticated user calling this method
         :param version: last comment version
         """
         if not text:
             log.warning('Missing text for comment, skipping...')
             return
 
         comment = ChangesetComment.get(comment_id)
         old_comment_text = comment.text
         comment.text = text
         comment.modified_at = datetime.datetime.now()
         version = safe_int(version)
 
         # NOTE(marcink): this returns initial comment + edits, so v2 from ui
         # would return 3 here
         comment_version = ChangesetCommentHistory.get_version(comment_id)
 
         if isinstance(version, int) and (comment_version - version) != 1:
             log.warning(
                 'Version mismatch comment_version {} submitted {}, skipping'.format(
                     comment_version-1,  # -1 since note above
                     version
                 )
             )
             raise CommentVersionMismatch()
 
         comment_history = ChangesetCommentHistory()
         comment_history.comment_id = comment_id
         comment_history.version = comment_version
         comment_history.created_by_user_id = auth_user.user_id
         comment_history.text = old_comment_text
         # TODO add email notification
         Session().add(comment_history)
         Session().add(comment)
         Session().flush()
 
         if comment.pull_request:
             action = 'repo.pull_request.comment.edit'
         else:
             action = 'repo.commit.comment.edit'
 
         comment_data = comment.get_api_data()
         comment_data['old_comment_text'] = old_comment_text
         self._log_audit_action(
             action, {'data': comment_data}, auth_user, comment)
 
         return comment_history
 
     def delete(self, comment, auth_user):
         """
         Deletes given comment
         """
         comment = self.__get_commit_comment(comment)
         old_data = comment.get_api_data()
         Session().delete(comment)
 
         if comment.pull_request:
             action = 'repo.pull_request.comment.delete'
         else:
             action = 'repo.commit.comment.delete'
 
         self._log_audit_action(
             action, {'old_data': old_data}, auth_user, comment)
 
         return comment
 
     def get_all_comments(self, repo_id, revision=None, pull_request=None,
                          include_drafts=True, count_only=False):
         q = ChangesetComment.query()\
                 .filter(ChangesetComment.repo_id == repo_id)
         if revision:
             q = q.filter(ChangesetComment.revision == revision)
         elif pull_request:
             pull_request = self.__get_pull_request(pull_request)
             q = q.filter(ChangesetComment.pull_request_id == pull_request.pull_request_id)
         else:
             raise Exception('Please specify commit or pull_request')
         if not include_drafts:
             q = q.filter(ChangesetComment.draft == false())
         q = q.order_by(ChangesetComment.created_on)
         if count_only:
             return q.count()
 
         return q.all()
 
     def get_url(self, comment, request=None, permalink=False, anchor=None):
         if not request:
             request = get_current_request()
 
         comment = self.__get_commit_comment(comment)
         if anchor is None:
-            anchor = 'comment-{}'.format(comment.comment_id)
+            anchor = f'comment-{comment.comment_id}'
 
         if comment.pull_request:
             pull_request = comment.pull_request
             if permalink:
                 return request.route_url(
                     'pull_requests_global',
                     pull_request_id=pull_request.pull_request_id,
                     _anchor=anchor)
             else:
                 return request.route_url(
                     'pullrequest_show',
                     repo_name=safe_str(pull_request.target_repo.repo_name),
                     pull_request_id=pull_request.pull_request_id,
                     _anchor=anchor)
 
         else:
             repo = comment.repo
             commit_id = comment.revision
 
             if permalink:
                 return request.route_url(
                     'repo_commit', repo_name=safe_str(repo.repo_id),
                     commit_id=commit_id,
                     _anchor=anchor)
 
             else:
                 return request.route_url(
                     'repo_commit', repo_name=safe_str(repo.repo_name),
                     commit_id=commit_id,
                     _anchor=anchor)
 
     def get_comments(self, repo_id, revision=None, pull_request=None):
         """
         Gets main comments based on revision or pull_request_id
 
         :param repo_id:
         :param revision:
         :param pull_request:
         """
 
         q = ChangesetComment.query()\
                 .filter(ChangesetComment.repo_id == repo_id)\
                 .filter(ChangesetComment.line_no == None)\
                 .filter(ChangesetComment.f_path == None)
         if revision:
             q = q.filter(ChangesetComment.revision == revision)
         elif pull_request:
             pull_request = self.__get_pull_request(pull_request)
             q = q.filter(ChangesetComment.pull_request == pull_request)
         else:
             raise Exception('Please specify commit or pull_request')
         q = q.order_by(ChangesetComment.created_on)
         return q.all()
 
     def get_inline_comments(self, repo_id, revision=None, pull_request=None):
         q = self._get_inline_comments_query(repo_id, revision, pull_request)
         return self._group_comments_by_path_and_line_number(q)
 
     def get_inline_comments_as_list(self, inline_comments, skip_outdated=True,
                                     version=None):
         inline_comms = []
         for fname, per_line_comments in inline_comments.items():
             for lno, comments in per_line_comments.items():
                 for comm in comments:
                     if not comm.outdated_at_version(version) and skip_outdated:
                         inline_comms.append(comm)
 
         return inline_comms
 
     def get_outdated_comments(self, repo_id, pull_request):
         # TODO: johbo: Remove `repo_id`, it is not needed to find the comments
         # of a pull request.
         q = self._all_inline_comments_of_pull_request(pull_request)
         q = q.filter(
             ChangesetComment.display_state ==
             ChangesetComment.COMMENT_OUTDATED
         ).order_by(ChangesetComment.comment_id.asc())
 
         return self._group_comments_by_path_and_line_number(q)
 
     def _get_inline_comments_query(self, repo_id, revision, pull_request):
         # TODO: johbo: Split this into two methods: One for PR and one for
         # commit.
         if revision:
             q = Session().query(ChangesetComment).filter(
                 ChangesetComment.repo_id == repo_id,
                 ChangesetComment.line_no != null(),
                 ChangesetComment.f_path != null(),
                 ChangesetComment.revision == revision)
 
         elif pull_request:
             pull_request = self.__get_pull_request(pull_request)
             if not CommentsModel.use_outdated_comments(pull_request):
                 q = self._visible_inline_comments_of_pull_request(pull_request)
             else:
                 q = self._all_inline_comments_of_pull_request(pull_request)
 
         else:
             raise Exception('Please specify commit or pull_request_id')
         q = q.order_by(ChangesetComment.comment_id.asc())
         return q
 
     def _group_comments_by_path_and_line_number(self, q):
         comments = q.all()
         paths = collections.defaultdict(lambda: collections.defaultdict(list))
         for co in comments:
             paths[co.f_path][co.line_no].append(co)
         return paths
 
     @classmethod
     def needed_extra_diff_context(cls):
         return max(cls.DIFF_CONTEXT_BEFORE, cls.DIFF_CONTEXT_AFTER)
 
     def outdate_comments(self, pull_request, old_diff_data, new_diff_data):
         if not CommentsModel.use_outdated_comments(pull_request):
             return
 
         comments = self._visible_inline_comments_of_pull_request(pull_request)
         comments_to_outdate = comments.all()
 
         for comment in comments_to_outdate:
             self._outdate_one_comment(comment, old_diff_data, new_diff_data)
 
     def _outdate_one_comment(self, comment, old_diff_proc, new_diff_proc):
         diff_line = _parse_comment_line_number(comment.line_no)
 
         try:
             old_context = old_diff_proc.get_context_of_line(
                 path=comment.f_path, diff_line=diff_line)
             new_context = new_diff_proc.get_context_of_line(
                 path=comment.f_path, diff_line=diff_line)
         except (diffs.LineNotInDiffException,
                 diffs.FileNotInDiffException):
             if not comment.draft:
                 comment.display_state = ChangesetComment.COMMENT_OUTDATED
             return
 
         if old_context == new_context:
             return
 
         if self._should_relocate_diff_line(diff_line):
             new_diff_lines = new_diff_proc.find_context(
                 path=comment.f_path, context=old_context,
                 offset=self.DIFF_CONTEXT_BEFORE)
             if not new_diff_lines and not comment.draft:
                 comment.display_state = ChangesetComment.COMMENT_OUTDATED
             else:
                 new_diff_line = self._choose_closest_diff_line(
                     diff_line, new_diff_lines)
                 comment.line_no = _diff_to_comment_line_number(new_diff_line)
         else:
             if not comment.draft:
                 comment.display_state = ChangesetComment.COMMENT_OUTDATED
 
     def _should_relocate_diff_line(self, diff_line):
         """
         Checks if relocation shall be tried for the given `diff_line`.
 
         If a comment points into the first lines, then we can have a situation
         that after an update another line has been added on top. In this case
         we would find the context still and move the comment around. This
         would be wrong.
         """
         should_relocate = (
             (diff_line.new and diff_line.new > self.DIFF_CONTEXT_BEFORE) or
             (diff_line.old and diff_line.old > self.DIFF_CONTEXT_BEFORE))
         return should_relocate
 
     def _choose_closest_diff_line(self, diff_line, new_diff_lines):
         candidate = new_diff_lines[0]
         best_delta = _diff_line_delta(diff_line, candidate)
         for new_diff_line in new_diff_lines[1:]:
             delta = _diff_line_delta(diff_line, new_diff_line)
             if delta < best_delta:
                 candidate = new_diff_line
                 best_delta = delta
         return candidate
 
     def _visible_inline_comments_of_pull_request(self, pull_request):
         comments = self._all_inline_comments_of_pull_request(pull_request)
         comments = comments.filter(
             coalesce(ChangesetComment.display_state, '') !=
             ChangesetComment.COMMENT_OUTDATED)
         return comments
 
     def _all_inline_comments_of_pull_request(self, pull_request):
         comments = Session().query(ChangesetComment)\
             .filter(ChangesetComment.line_no != None)\
             .filter(ChangesetComment.f_path != None)\
             .filter(ChangesetComment.pull_request == pull_request)
         return comments
 
     def _all_general_comments_of_pull_request(self, pull_request):
         comments = Session().query(ChangesetComment)\
             .filter(ChangesetComment.line_no == None)\
             .filter(ChangesetComment.f_path == None)\
             .filter(ChangesetComment.pull_request == pull_request)
 
         return comments
 
     @staticmethod
     def use_outdated_comments(pull_request):
         settings_model = VcsSettingsModel(repo=pull_request.target_repo)
         settings = settings_model.get_general_settings()
         return settings.get('rhodecode_use_outdated_comments', False)
 
     def trigger_commit_comment_hook(self, repo, user, action, data=None):
         repo = self._get_repo(repo)
         target_scm = repo.scm_instance()
         if action == 'create':
             trigger_hook = hooks_utils.trigger_comment_commit_hooks
         elif action == 'edit':
             trigger_hook = hooks_utils.trigger_comment_commit_edit_hooks
         else:
             return
 
         log.debug('Handling repo %s trigger_commit_comment_hook with action %s: %s',
                   repo, action, trigger_hook)
         trigger_hook(
             username=user.username,
             repo_name=repo.repo_name,
             repo_type=target_scm.alias,
             repo=repo,
             data=data)
 
 
 def _parse_comment_line_number(line_no):
-    """
+    r"""
     Parses line numbers of the form "(o|n)\d+" and returns them in a tuple.
     """
     old_line = None
     new_line = None
     if line_no.startswith('o'):
         old_line = int(line_no[1:])
     elif line_no.startswith('n'):
         new_line = int(line_no[1:])
     else:
         raise ValueError("Comment lines have to start with either 'o' or 'n'.")
     return diffs.DiffLineNumber(old_line, new_line)
 
 
 def _diff_to_comment_line_number(diff_line):
     if diff_line.new is not None:
-        return u'n{}'.format(diff_line.new)
+        return f'n{diff_line.new}'
     elif diff_line.old is not None:
-        return u'o{}'.format(diff_line.old)
-    return u''
+        return f'o{diff_line.old}'
+    return ''
 
 
 def _diff_line_delta(a, b):
     if None not in (a.new, b.new):
         return abs(a.new - b.new)
     elif None not in (a.old, b.old):
         return abs(a.old - b.old)
     else:
         raise ValueError(
-            "Cannot compute delta between {} and {}".format(a, b))
+            f"Cannot compute delta between {a} and {b}")