##// END OF EJS Templates
vcs: Minimal change to expose the shadow repository...
vcs: Minimal change to expose the shadow repository Based on my original research, this was the "minimal" starting point. It shows that three concepts are needed for the "repo_name": * From the security standpoint we think of the shadow repository having the same ACL as the target repository of the pull request. This is because the pull request itself is considered to be a part of the target repository. Out of this thought, the variable "acl_repo_name" is used whenever we want to check permissions or when we need the database configuration of the repository. An alternative name would have been "db_repo_name", but the usage for ACL checking is the most important one. * From the web interaction perspective, we need the URL which was originally used to get to the repository. This is because based on this base URL commands can be identified. Especially for Git this is important, so that the commands are correctly recognized. Since the URL is in the focus, this is called "url_repo_name". * Finally we have to deal with the repository on the file system. This is what the VCS layer deal with normally, so this name is called "vcs_repo_name". The original repository interaction is a special case where all three names are the same. When interacting with a pull request, these three names are typically all different. This change is minimal in a sense that it just makes the interaction with a shadow repository barely work, without checking any special constraints yet. This was the starting point for further work on this topic.

File last commit:

r1:854a839a default
r887:175782be default
Show More
memory_lru_debug.py
96 lines | 3.0 KiB | text/x-python | PythonLexer
# -*- coding: utf-8 -*-
# Copyright (C) 2010-2016 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/
"""
Custom LRU memory manager for debugging purposes. It allows to track the keys
and the state of LRU dict.
inrae.cache is licensed under LRUDict is licensed under ZPL license
This software is Copyright (c) Zope Corporation (tm) and
Contributors. All rights reserved.
TODO: marcink, we might think of replacing the LRUDict with lru-dict library
written in C.
eg difference in speed:
LRUDictC Time : 0.00025 s, Memory : 110592 Kb
LRUDict Time : 0.00369 s, Memory : 147456 Kb
"""
import logging
from infrae.cache.beakerext.lru import LRUDict
from beaker.container import MemoryNamespaceManager, AbstractDictionaryNSManager
from rhodecode.lib.utils2 import safe_str
log = logging.getLogger(__name__)
class LRUDictDebug(LRUDict):
"""
Wrapper to provide some debug options
"""
def _report_keys(self):
elems_cnt = '%s/%s' % (len(self.keys()), self.size)
# trick for pformat print it more nicely
fmt = '\n'
for cnt, elem in enumerate(self.keys()):
fmt += '%s - %s\n' % (cnt+1, safe_str(elem))
log.debug('current LRU keys (%s):%s' % (elems_cnt, fmt))
def __getitem__(self, key):
self._report_keys()
return self.get(key)
class MemoryLRUNamespaceManagerBase(MemoryNamespaceManager):
default_max_items = 10000
def _get_factory(self, max_items):
def Factory():
return LRUDict(int(max_items))
return Factory
def __init__(self, namespace, **kwargs):
AbstractDictionaryNSManager.__init__(self, namespace)
if kwargs.has_key('max_items'):
max_items = kwargs['max_items']
else:
max_items = self.default_max_items
Factory = self._get_factory(max_items)
self.dictionary = MemoryNamespaceManager.namespaces.get(
self.namespace, Factory)
class MemoryLRUNamespaceManagerDebug(MemoryLRUNamespaceManagerBase):
"""
A memory namespace manager that return with LRU dicts backend,
special debug for testing
"""
default_max_items = 10000
def _get_factory(self, max_items):
def Factory():
return LRUDictDebug(int(max_items))
return Factory