cleanup: use isinstance instead of comparing types

From 2to3 idioms.

kallithea/lib/rcmail/response.py

@@ -1,454 +1,454 b''
 # The code in this module is entirely lifted from the Lamson project
 # (http://lamsonproject.org/).  Its copyright is:
 
 # Copyright (c) 2008, Zed A. Shaw
 # All rights reserved.
 
 # It is provided under this license:
 
 # Redistribution and use in source and binary forms, with or without
 # modification, are permitted provided that the following conditions are met:
 
 # * Redistributions of source code must retain the above copyright notice, this
 #   list of conditions and the following disclaimer.
 
 # * Redistributions in binary form must reproduce the above copyright notice,
 #   this list of conditions and the following disclaimer in the documentation
 #   and/or other materials provided with the distribution.
 
 # * Neither the name of the Zed A. Shaw nor the names of its contributors may
 #   be used to endorse or promote products derived from this software without
 #   specific prior written permission.
 
 # THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
 # "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
 # LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
 # FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
 # COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT,
 # INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
 # (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
 # SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
 # HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
 # STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
 # ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
 # POSSIBILITY OF SUCH DAMAGE.
 
 import mimetypes
 import os
 import string
 from email import encoders
 from email.charset import Charset
 from email.mime.base import MIMEBase
 from email.utils import parseaddr
 
 
 ADDRESS_HEADERS_WHITELIST = ['From', 'To', 'Delivered-To', 'Cc']
 DEFAULT_ENCODING = "utf-8"
 VALUE_IS_EMAIL_ADDRESS = lambda v: '@' in v
 
 
 def normalize_header(header):
     return string.capwords(header.lower(), '-')
 
 
 class EncodingError(Exception):
     """Thrown when there is an encoding error."""
     pass
 
 
 class MailBase(object):
     """MailBase is used as the basis of lamson.mail and contains the basics of
     encoding an email.  You actually can do all your email processing with this
     class, but it's more raw.
     """
     def __init__(self, items=()):
         self.headers = dict(items)
         self.parts = []
         self.body = None
         self.content_encoding = {'Content-Type': (None, {}),
                                  'Content-Disposition': (None, {}),
                                  'Content-Transfer-Encoding': (None, {})}
 
     def __getitem__(self, key):
         return self.headers.get(normalize_header(key), None)
 
     def __len__(self):
         return len(self.headers)
 
     def __iter__(self):
         return iter(self.headers)
 
     def __contains__(self, key):
         return normalize_header(key) in self.headers
 
     def __setitem__(self, key, value):
         self.headers[normalize_header(key)] = value
 
     def __delitem__(self, key):
         del self.headers[normalize_header(key)]
 
     def __nonzero__(self):
         return self.body is not None or len(self.headers) > 0 or len(self.parts) > 0
 
     def keys(self):
         """Returns the sorted keys."""
         return sorted(self.headers.keys())
 
     def attach_file(self, filename, data, ctype, disposition):
         """
         A file attachment is a raw attachment with a disposition that
         indicates the file name.
         """
         assert filename, "You can't attach a file without a filename."
         ctype = ctype.lower()
 
         part = MailBase()
         part.body = data
         part.content_encoding['Content-Type'] = (ctype, {'name': filename})
         part.content_encoding['Content-Disposition'] = (disposition,
                                                         {'filename': filename})
         self.parts.append(part)
 
     def attach_text(self, data, ctype):
         """
         This attaches a simpler text encoded part, which doesn't have a
         filename.
         """
         ctype = ctype.lower()
 
         part = MailBase()
         part.body = data
         part.content_encoding['Content-Type'] = (ctype, {})
         self.parts.append(part)
 
     def walk(self):
         for p in self.parts:
             yield p
             for x in p.walk():
                 yield x
 
 
 class MailResponse(object):
     """
     You are given MailResponse objects from the lamson.view methods, and
     whenever you want to generate an email to send to someone.  It has the
     same basic functionality as MailRequest, but it is designed to be written
     to, rather than read from (although you can do both).
 
     You can easily set a Body or Html during creation or after by passing it
     as __init__ parameters, or by setting those attributes.
 
     You can initially set the From, To, and Subject, but they are headers so
     use the dict notation to change them: msg['From'] = '[e-mail]'.
 
     The message is not fully crafted until right when you convert it with
     MailResponse.to_message.  This lets you change it and work with it, then
     send it out when it's ready.
     """
     def __init__(self, To=None, From=None, Subject=None, Body=None, Html=None,
                  separator="; "):
         self.Body = Body
         self.Html = Html
         self.base = MailBase([('To', To), ('From', From), ('Subject', Subject)])
         self.multipart = self.Body and self.Html
         self.attachments = []
         self.separator = separator
 
     def __contains__(self, key):
         return self.base.__contains__(key)
 
     def __getitem__(self, key):
         return self.base.__getitem__(key)
 
     def __setitem__(self, key, val):
         return self.base.__setitem__(key, val)
 
     def __delitem__(self, name):
         del self.base[name]
 
     def attach(self, filename=None, content_type=None, data=None,
                disposition=None):
         """
 
         Simplifies attaching files from disk or data as files.  To attach
         simple text simple give data and a content_type.  To attach a file,
         give the data/content_type/filename/disposition combination.
 
         For convenience, if you don't give data and only a filename, then it
         will read that file's contents when you call to_message() later.  If
         you give data and filename then it will assume you've filled data
         with what the file's contents are and filename is just the name to
         use.
         """
 
         assert filename or data, ("You must give a filename or some data to "
                                   "attach.")
         assert data or os.path.exists(filename), ("File doesn't exist, and no "
                                                   "data given.")
 
         self.multipart = True
 
         if filename and not content_type:
             content_type, encoding = mimetypes.guess_type(filename)
 
         assert content_type, ("No content type given, and couldn't guess "
                               "from the filename: %r" % filename)
 
         self.attachments.append({'filename': filename,
                                  'content_type': content_type,
                                  'data': data,
                                  'disposition': disposition})
 
     def attach_part(self, part):
         """
         Attaches a raw MailBase part from a MailRequest (or anywhere)
         so that you can copy it over.
         """
         self.multipart = True
 
         self.attachments.append({'filename': None,
                                  'content_type': None,
                                  'data': None,
                                  'disposition': None,
                                  'part': part,
                                  })
 
     def attach_all_parts(self, mail_request):
         """
         Used for copying the attachment parts of a mail.MailRequest
         object for mailing lists that need to maintain attachments.
         """
         for part in mail_request.all_parts():
             self.attach_part(part)
 
         self.base.content_encoding = mail_request.base.content_encoding.copy()
 
     def clear(self):
         """
         Clears out the attachments so you can redo them.  Use this to keep the
         headers for a series of different messages with different attachments.
         """
         del self.attachments[:]
         del self.base.parts[:]
         self.multipart = False
 
     def update(self, message):
         """
         Used to easily set a bunch of heading from another dict
         like object.
         """
         for k in message.keys():
             self.base[k] = message[k]
 
     def __str__(self):
         """
         Converts to a string.
         """
         return self.to_message().as_string()
 
     def _encode_attachment(self, filename=None, content_type=None, data=None,
                            disposition=None, part=None):
         """
         Used internally to take the attachments mentioned in self.attachments
         and do the actual encoding in a lazy way when you call to_message.
         """
         if part:
             self.base.parts.append(part)
         elif filename:
             if not data:
                 data = open(filename).read()
 
             self.base.attach_file(filename, data, content_type,
                                   disposition or 'attachment')
         else:
             self.base.attach_text(data, content_type)
 
         ctype = self.base.content_encoding['Content-Type'][0]
 
         if ctype and not ctype.startswith('multipart'):
             self.base.content_encoding['Content-Type'] = ('multipart/mixed', {})
 
     def to_message(self):
         """
         Figures out all the required steps to finally craft the
         message you need and return it.  The resulting message
         is also available as a self.base attribute.
 
         What is returned is a Python email API message you can
         use with those APIs.  The self.base attribute is the raw
         lamson.encoding.MailBase.
         """
         del self.base.parts[:]
 
         if self.Body and self.Html:
             self.multipart = True
             self.base.content_encoding['Content-Type'] = (
                 'multipart/alternative', {})
 
         if self.multipart:
             self.base.body = None
             if self.Body:
                 self.base.attach_text(self.Body, 'text/plain')
 
             if self.Html:
                 self.base.attach_text(self.Html, 'text/html')
 
             for args in self.attachments:
                 self._encode_attachment(**args)
 
         elif self.Body:
             self.base.body = self.Body
             self.base.content_encoding['Content-Type'] = ('text/plain', {})
 
         elif self.Html:
             self.base.body = self.Html
             self.base.content_encoding['Content-Type'] = ('text/html', {})
 
         return to_message(self.base, separator=self.separator)
 
     def all_parts(self):
         """
         Returns all the encoded parts.  Only useful for debugging
         or inspecting after calling to_message().
         """
         return self.base.parts
 
     def keys(self):
         return self.base.keys()
 
 
 def to_message(mail, separator="; "):
     """
     Given a MailBase message, this will construct a MIMEPart
     that is canonicalized for use with the Python email API.
     """
     ctype, params = mail.content_encoding['Content-Type']
 
     if not ctype:
         if mail.parts:
             ctype = 'multipart/mixed'
         else:
             ctype = 'text/plain'
     else:
         if mail.parts:
             assert ctype.startswith(("multipart", "message")), \
                    "Content type should be multipart or message, not %r" % ctype
 
     # adjust the content type according to what it should be now
     mail.content_encoding['Content-Type'] = (ctype, params)
 
     try:
         out = MIMEPart(ctype, **params)
     except TypeError as exc:  # pragma: no cover
         raise EncodingError("Content-Type malformed, not allowed: %r; "
                             "%r (Python ERROR: %s" %
                             (ctype, params, exc.message))
 
     for k in mail.keys():
         if k in ADDRESS_HEADERS_WHITELIST:
             out[k.encode('ascii')] = header_to_mime_encoding(
                                          mail[k],
                                          not_email=False,
                                          separator=separator
                                      )
         else:
             out[k.encode('ascii')] = header_to_mime_encoding(
                                          mail[k],
                                          not_email=True
                                     )
 
     out.extract_payload(mail)
 
     # go through the children
     for part in mail.parts:
         out.attach(to_message(part))
 
     return out
 
 
 class MIMEPart(MIMEBase):
     """
     A reimplementation of nearly everything in email.mime to be more useful
     for actually attaching things.  Rather than one class for every type of
     thing you'd encode, there's just this one, and it figures out how to
     encode what you ask it.
     """
     def __init__(self, type, **params):
         self.maintype, self.subtype = type.split('/')
         MIMEBase.__init__(self, self.maintype, self.subtype, **params)
 
     def add_text(self, content):
         # this is text, so encode it in canonical form
         try:
             encoded = content.encode('ascii')
             charset = 'ascii'
         except UnicodeError:
             encoded = content.encode('utf-8')
             charset = 'utf-8'
 
         self.set_payload(encoded, charset=charset)
 
     def extract_payload(self, mail):
         if mail.body is None:
             return  # only None, '' is still ok
 
         ctype, ctype_params = mail.content_encoding['Content-Type']
         cdisp, cdisp_params = mail.content_encoding['Content-Disposition']
 
         assert ctype, ("Extract payload requires that mail.content_encoding "
                        "have a valid Content-Type.")
 
         if ctype.startswith("text/"):
             self.add_text(mail.body)
         else:
             if cdisp:
                 # replicate the content-disposition settings
                 self.add_header('Content-Disposition', cdisp, **cdisp_params)
 
             self.set_payload(mail.body)
             encoders.encode_base64(self)
 
     def __repr__(self):
         return "<MIMEPart '%s/%s': %r, %r, multipart=%r>" % (
             self.subtype,
             self.maintype,
             self['Content-Type'],
             self['Content-Disposition'],
             self.is_multipart())
 
 
 def header_to_mime_encoding(value, not_email=False, separator=", "):
     if not value:
         return ""
 
     encoder = Charset(DEFAULT_ENCODING)
-    if type(value) == list:
+    if isinstance(value, list):
         return separator.join(properly_encode_header(
             v, encoder, not_email) for v in value)
     else:
         return properly_encode_header(value, encoder, not_email)
 
 
 def properly_encode_header(value, encoder, not_email):
     """
     The only thing special (weird) about this function is that it tries
     to do a fast check to see if the header value has an email address in
     it.  Since random headers could have an email address, and email addresses
     have weird special formatting rules, we have to check for it.
 
     Normally this works fine, but in Librelist, we need to "obfuscate" email
     addresses by changing the '@' to '-AT-'.  This is where
     VALUE_IS_EMAIL_ADDRESS exists.  It's a simple lambda returning True/False
     to check if a header value has an email address.  If you need to make this
     check different, then change this.
     """
     try:
         return value.encode("ascii")
     except UnicodeEncodeError:
         if not not_email and VALUE_IS_EMAIL_ADDRESS(value):
             # this could have an email address, make sure we don't screw it up
             name, address = parseaddr(value)
             return '"%s" <%s>' % (
                 encoder.header_encode(name.encode("utf-8")), address)
 
         return encoder.header_encode(value.encode("utf-8"))

kallithea/model/db.py

@@ -1,2548 +1,2548 b''
 # -*- coding: utf-8 -*-
 # This program is free software: you can redistribute it and/or modify
 # it under the terms of the GNU General Public License as published by
 # the Free Software Foundation, either version 3 of the License, or
 # (at your option) any later version.
 #
 # 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 General Public License
 # along with this program.  If not, see <http://www.gnu.org/licenses/>.
 """
 kallithea.model.db
 ~~~~~~~~~~~~~~~~~~
 
 Database Models for Kallithea
 
 This file was forked by the Kallithea project in July 2014.
 Original author and date, and relevant copyright and licensing information is below:
 :created_on: Apr 08, 2010
 :author: marcink
 :copyright: (c) 2013 RhodeCode GmbH, and others.
 :license: GPLv3, see LICENSE.md for more details.
 """
 
 import collections
 import datetime
 import functools
 import hashlib
 import logging
 import os
 import time
 import traceback
 
 import ipaddr
 import sqlalchemy
 from beaker.cache import cache_region, region_invalidate
 from sqlalchemy import *
 from sqlalchemy.ext.hybrid import hybrid_property
 from sqlalchemy.orm import class_mapper, joinedload, relationship, validates
 from tg.i18n import lazy_ugettext as _
 from webob.exc import HTTPNotFound
 
 import kallithea
 from kallithea.lib.caching_query import FromCache
 from kallithea.lib.compat import json
 from kallithea.lib.exceptions import DefaultUserException
 from kallithea.lib.utils2 import Optional, aslist, get_changeset_safe, get_clone_url, remove_prefix, safe_int, safe_str, safe_unicode, str2bool, urlreadable
 from kallithea.lib.vcs import get_backend
 from kallithea.lib.vcs.backends.base import EmptyChangeset
 from kallithea.lib.vcs.utils.helpers import get_scm
 from kallithea.lib.vcs.utils.lazy import LazyProperty
 from kallithea.model.meta import Base, Session
 
 
 URL_SEP = '/'
 log = logging.getLogger(__name__)
 
 #==============================================================================
 # BASE CLASSES
 #==============================================================================
 
 _hash_key = lambda k: hashlib.md5(safe_str(k)).hexdigest()
 
 
 class BaseDbModel(object):
     """
     Base Model for all classes
     """
 
     @classmethod
     def _get_keys(cls):
         """return column names for this model """
         return class_mapper(cls).c.keys()
 
     def get_dict(self):
         """
         return dict with keys and values corresponding
         to this model data """
 
         d = {}
         for k in self._get_keys():
             d[k] = getattr(self, k)
 
         # also use __json__() if present to get additional fields
         _json_attr = getattr(self, '__json__', None)
         if _json_attr:
             # update with attributes from __json__
             if callable(_json_attr):
                 _json_attr = _json_attr()
             for k, val in _json_attr.iteritems():
                 d[k] = val
         return d
 
     def get_appstruct(self):
         """return list with keys and values tuples corresponding
         to this model data """
 
         return [
             (k, getattr(self, k))
             for k in self._get_keys()
         ]
 
     def populate_obj(self, populate_dict):
         """populate model with data from given populate_dict"""
 
         for k in self._get_keys():
             if k in populate_dict:
                 setattr(self, k, populate_dict[k])
 
     @classmethod
     def query(cls):
         return Session().query(cls)
 
     @classmethod
     def get(cls, id_):
         if id_:
             return cls.query().get(id_)
 
     @classmethod
     def guess_instance(cls, value, callback=None):
         """Haphazardly attempt to convert `value` to a `cls` instance.
 
         If `value` is None or already a `cls` instance, return it. If `value`
         is a number (or looks like one if you squint just right), assume it's
         a database primary key and let SQLAlchemy sort things out. Otherwise,
         fall back to resolving it using `callback` (if specified); this could
         e.g. be a function that looks up instances by name (though that won't
         work if the name begins with a digit). Otherwise, raise Exception.
         """
 
         if value is None:
             return None
         if isinstance(value, cls):
             return value
         if isinstance(value, (int, long)) or safe_str(value).isdigit():
             return cls.get(value)
         if callback is not None:
             return callback(value)
 
         raise Exception(
             'given object must be int, long or Instance of %s '
             'got %s, no callback provided' % (cls, type(value))
         )
 
     @classmethod
     def get_or_404(cls, id_):
         try:
             id_ = int(id_)
         except (TypeError, ValueError):
             raise HTTPNotFound
 
         res = cls.query().get(id_)
         if res is None:
             raise HTTPNotFound
         return res
 
     @classmethod
     def delete(cls, id_):
         obj = cls.query().get(id_)
         Session().delete(obj)
 
     def __repr__(self):
         if hasattr(self, '__unicode__'):
             # python repr needs to return str
             try:
                 return safe_str(self.__unicode__())
             except UnicodeDecodeError:
                 pass
         return '<DB:%s>' % (self.__class__.__name__)
 
 
 _table_args_default_dict = {'extend_existing': True,
                             'mysql_engine': 'InnoDB',
                             'mysql_charset': 'utf8',
                             'sqlite_autoincrement': True,
                            }
 
 class Setting(Base, BaseDbModel):
     __tablename__ = 'settings'
     __table_args__ = (
         _table_args_default_dict,
     )
 
     SETTINGS_TYPES = {
         'str': safe_str,
         'int': safe_int,
         'unicode': safe_unicode,
         'bool': str2bool,
         'list': functools.partial(aslist, sep=',')
     }
     DEFAULT_UPDATE_URL = ''
 
     app_settings_id = Column(Integer(), primary_key=True)
     app_settings_name = Column(String(255), nullable=False, unique=True)
     _app_settings_value = Column("app_settings_value", Unicode(4096), nullable=False)
     _app_settings_type = Column("app_settings_type", String(255), nullable=True) # FIXME: not nullable?
 
     def __init__(self, key='', val='', type='unicode'):
         self.app_settings_name = key
         self.app_settings_value = val
         self.app_settings_type = type
 
     @validates('_app_settings_value')
     def validate_settings_value(self, key, val):
-        assert type(val) == unicode
+        assert isinstance(val, unicode)
         return val
 
     @hybrid_property
     def app_settings_value(self):
         v = self._app_settings_value
         _type = self.app_settings_type
         converter = self.SETTINGS_TYPES.get(_type) or self.SETTINGS_TYPES['unicode']
         return converter(v)
 
     @app_settings_value.setter
     def app_settings_value(self, val):
         """
         Setter that will always make sure we use unicode in app_settings_value
 
         :param val:
         """
         self._app_settings_value = safe_unicode(val)
 
     @hybrid_property
     def app_settings_type(self):
         return self._app_settings_type
 
     @app_settings_type.setter
     def app_settings_type(self, val):
         if val not in self.SETTINGS_TYPES:
             raise Exception('type must be one of %s got %s'
                             % (self.SETTINGS_TYPES.keys(), val))
         self._app_settings_type = val
 
     def __unicode__(self):
         return u"<%s('%s:%s[%s]')>" % (
             self.__class__.__name__,
             self.app_settings_name, self.app_settings_value, self.app_settings_type
         )
 
     @classmethod
     def get_by_name(cls, key):
         return cls.query() \
             .filter(cls.app_settings_name == key).scalar()
 
     @classmethod
     def get_by_name_or_create(cls, key, val='', type='unicode'):
         res = cls.get_by_name(key)
         if res is None:
             res = cls(key, val, type)
         return res
 
     @classmethod
     def create_or_update(cls, key, val=Optional(''), type=Optional('unicode')):
         """
         Creates or updates Kallithea setting. If updates are triggered, it will only
         update parameters that are explicitly set. Optional instance will be skipped.
 
         :param key:
         :param val:
         :param type:
         :return:
         """
         res = cls.get_by_name(key)
         if res is None:
             val = Optional.extract(val)
             type = Optional.extract(type)
             res = cls(key, val, type)
             Session().add(res)
         else:
             res.app_settings_name = key
             if not isinstance(val, Optional):
                 # update if set
                 res.app_settings_value = val
             if not isinstance(type, Optional):
                 # update if set
                 res.app_settings_type = type
         return res
 
     @classmethod
     def get_app_settings(cls, cache=False):
 
         ret = cls.query()
 
         if cache:
             ret = ret.options(FromCache("sql_cache_short", "get_hg_settings"))
 
         if ret is None:
             raise Exception('Could not get application settings !')
         settings = {}
         for each in ret:
             settings[each.app_settings_name] = \
                 each.app_settings_value
 
         return settings
 
     @classmethod
     def get_auth_settings(cls, cache=False):
         ret = cls.query() \
                 .filter(cls.app_settings_name.startswith('auth_')).all()
         fd = {}
         for row in ret:
             fd[row.app_settings_name] = row.app_settings_value
         return fd
 
     @classmethod
     def get_default_repo_settings(cls, cache=False, strip_prefix=False):
         ret = cls.query() \
                 .filter(cls.app_settings_name.startswith('default_')).all()
         fd = {}
         for row in ret:
             key = row.app_settings_name
             if strip_prefix:
                 key = remove_prefix(key, prefix='default_')
             fd.update({key: row.app_settings_value})
 
         return fd
 
     @classmethod
     def get_server_info(cls):
         import pkg_resources
         import platform
         from kallithea.lib.utils import check_git_version
         mods = [(p.project_name, p.version) for p in pkg_resources.working_set]
         info = {
             'modules': sorted(mods, key=lambda k: k[0].lower()),
             'py_version': platform.python_version(),
             'platform': safe_unicode(platform.platform()),
             'kallithea_version': kallithea.__version__,
             'git_version': safe_unicode(check_git_version()),
             'git_path': kallithea.CONFIG.get('git_path')
         }
         return info
 
 
 class Ui(Base, BaseDbModel):
     __tablename__ = 'ui'
     __table_args__ = (
         # FIXME: ui_key as key is wrong and should be removed when the corresponding
         # Ui.get_by_key has been replaced by the composite key
         UniqueConstraint('ui_key'),
         UniqueConstraint('ui_section', 'ui_key'),
         _table_args_default_dict,
     )
 
     HOOK_UPDATE = 'changegroup.update'
     HOOK_REPO_SIZE = 'changegroup.repo_size'
 
     ui_id = Column(Integer(), primary_key=True)
     ui_section = Column(String(255), nullable=False)
     ui_key = Column(String(255), nullable=False)
     ui_value = Column(String(255), nullable=True) # FIXME: not nullable?
     ui_active = Column(Boolean(), nullable=False, default=True)
 
     @classmethod
     def get_by_key(cls, section, key):
         """ Return specified Ui object, or None if not found. """
         return cls.query().filter_by(ui_section=section, ui_key=key).scalar()
 
     @classmethod
     def get_or_create(cls, section, key):
         """ Return specified Ui object, creating it if necessary. """
         setting = cls.get_by_key(section, key)
         if setting is None:
             setting = cls(ui_section=section, ui_key=key)
             Session().add(setting)
         return setting
 
     @classmethod
     def get_builtin_hooks(cls):
         q = cls.query()
         q = q.filter(cls.ui_key.in_([cls.HOOK_UPDATE, cls.HOOK_REPO_SIZE]))
         q = q.filter(cls.ui_section == 'hooks')
         return q.all()
 
     @classmethod
     def get_custom_hooks(cls):
         q = cls.query()
         q = q.filter(~cls.ui_key.in_([cls.HOOK_UPDATE, cls.HOOK_REPO_SIZE]))
         q = q.filter(cls.ui_section == 'hooks')
         return q.all()
 
     @classmethod
     def get_repos_location(cls):
         return cls.get_by_key('paths', '/').ui_value
 
     @classmethod
     def create_or_update_hook(cls, key, val):
         new_ui = cls.get_or_create('hooks', key)
         new_ui.ui_active = True
         new_ui.ui_value = val
 
     def __repr__(self):
         return '<%s[%s]%s=>%s]>' % (self.__class__.__name__, self.ui_section,
                                     self.ui_key, self.ui_value)
 
 
 class User(Base, BaseDbModel):
     __tablename__ = 'users'
     __table_args__ = (
         Index('u_username_idx', 'username'),
         Index('u_email_idx', 'email'),
         _table_args_default_dict,
     )
 
     DEFAULT_USER = 'default'
     DEFAULT_GRAVATAR_URL = 'https://secure.gravatar.com/avatar/{md5email}?d=identicon&s={size}'
     # The name of the default auth type in extern_type, 'internal' lives in auth_internal.py
     DEFAULT_AUTH_TYPE = 'internal'
 
     user_id = Column(Integer(), primary_key=True)
     username = Column(String(255), nullable=False, unique=True)
     password = Column(String(255), nullable=False)
     active = Column(Boolean(), nullable=False, default=True)
     admin = Column(Boolean(), nullable=False, default=False)
     name = Column("firstname", Unicode(255), nullable=False)
     lastname = Column(Unicode(255), nullable=False)
     _email = Column("email", String(255), nullable=True, unique=True) # FIXME: not nullable?
     last_login = Column(DateTime(timezone=False), nullable=True)
     extern_type = Column(String(255), nullable=True) # FIXME: not nullable?
     extern_name = Column(String(255), nullable=True) # FIXME: not nullable?
     api_key = Column(String(255), nullable=False)
     created_on = Column(DateTime(timezone=False), nullable=False, default=datetime.datetime.now)
     _user_data = Column("user_data", LargeBinary(), nullable=True)  # JSON data # FIXME: not nullable?
 
     user_log = relationship('UserLog')
     user_perms = relationship('UserToPerm', primaryjoin="User.user_id==UserToPerm.user_id", cascade='all')
 
     repositories = relationship('Repository')
     repo_groups = relationship('RepoGroup')
     user_groups = relationship('UserGroup')
     user_followers = relationship('UserFollowing', primaryjoin='UserFollowing.follows_user_id==User.user_id', cascade='all')
     followings = relationship('UserFollowing', primaryjoin='UserFollowing.user_id==User.user_id', cascade='all')
 
     repo_to_perm = relationship('UserRepoToPerm', primaryjoin='UserRepoToPerm.user_id==User.user_id', cascade='all')
     repo_group_to_perm = relationship('UserRepoGroupToPerm', primaryjoin='UserRepoGroupToPerm.user_id==User.user_id', cascade='all')
 
     group_member = relationship('UserGroupMember', cascade='all')
 
     # comments created by this user
     user_comments = relationship('ChangesetComment', cascade='all')
     # extra emails for this user
     user_emails = relationship('UserEmailMap', cascade='all')
     # extra API keys
     user_api_keys = relationship('UserApiKeys', cascade='all')
     ssh_keys = relationship('UserSshKeys', cascade='all')
 
     @hybrid_property
     def email(self):
         return self._email
 
     @email.setter
     def email(self, val):
         self._email = val.lower() if val else None
 
     @property
     def firstname(self):
         # alias for future
         return self.name
 
     @property
     def emails(self):
         other = UserEmailMap.query().filter(UserEmailMap.user == self).all()
         return [self.email] + [x.email for x in other]
 
     @property
     def api_keys(self):
         other = UserApiKeys.query().filter(UserApiKeys.user == self).all()
         return [self.api_key] + [x.api_key for x in other]
 
     @property
     def ip_addresses(self):
         ret = UserIpMap.query().filter(UserIpMap.user == self).all()
         return [x.ip_addr for x in ret]
 
     @property
     def full_name(self):
         return '%s %s' % (self.firstname, self.lastname)
 
     @property
     def full_name_or_username(self):
         """
         Show full name.
         If full name is not set, fall back to username.
         """
         return ('%s %s' % (self.firstname, self.lastname)
                 if (self.firstname and self.lastname) else self.username)
 
     @property
     def full_name_and_username(self):
         """
         Show full name and username as 'Firstname Lastname (username)'.
         If full name is not set, fall back to username.
         """
         return ('%s %s (%s)' % (self.firstname, self.lastname, self.username)
                 if (self.firstname and self.lastname) else self.username)
 
     @property
     def full_contact(self):
         return '%s %s <%s>' % (self.firstname, self.lastname, self.email)
 
     @property
     def short_contact(self):
         return '%s %s' % (self.firstname, self.lastname)
 
     @property
     def is_admin(self):
         return self.admin
 
     @hybrid_property
     def is_default_user(self):
         return self.username == User.DEFAULT_USER
 
     @hybrid_property
     def user_data(self):
         if not self._user_data:
             return {}
 
         try:
             return json.loads(self._user_data)
         except TypeError:
             return {}
 
     @user_data.setter
     def user_data(self, val):
         try:
             self._user_data = json.dumps(val)
         except Exception:
             log.error(traceback.format_exc())
 
     def __unicode__(self):
         return u"<%s('id:%s:%s')>" % (self.__class__.__name__,
                                       self.user_id, self.username)
 
     @classmethod
     def guess_instance(cls, value):
         return super(User, cls).guess_instance(value, User.get_by_username)
 
     @classmethod
     def get_or_404(cls, id_, allow_default=True):
         '''
         Overridden version of BaseDbModel.get_or_404, with an extra check on
         the default user.
         '''
         user = super(User, cls).get_or_404(id_)
         if not allow_default and user.is_default_user:
             raise DefaultUserException()
         return user
 
     @classmethod
     def get_by_username_or_email(cls, username_or_email, case_insensitive=False, cache=False):
         """
         For anything that looks like an email address, look up by the email address (matching
         case insensitively).
         For anything else, try to look up by the user name.
 
         This assumes no normal username can have '@' symbol.
         """
         if '@' in username_or_email:
             return User.get_by_email(username_or_email, cache=cache)
         else:
             return User.get_by_username(username_or_email, case_insensitive=case_insensitive, cache=cache)
 
     @classmethod
     def get_by_username(cls, username, case_insensitive=False, cache=False):
         if case_insensitive:
             q = cls.query().filter(func.lower(cls.username) == func.lower(username))
         else:
             q = cls.query().filter(cls.username == username)
 
         if cache:
             q = q.options(FromCache(
                             "sql_cache_short",
                             "get_user_%s" % _hash_key(username)
                           )
             )
         return q.scalar()
 
     @classmethod
     def get_by_api_key(cls, api_key, cache=False, fallback=True):
         if len(api_key) != 40 or not api_key.isalnum():
             return None
 
         q = cls.query().filter(cls.api_key == api_key)
 
         if cache:
             q = q.options(FromCache("sql_cache_short",
                                     "get_api_key_%s" % api_key))
         res = q.scalar()
 
         if fallback and not res:
             # fallback to additional keys
             _res = UserApiKeys.query().filter_by(api_key=api_key, is_expired=False).first()
             if _res:
                 res = _res.user
         if res is None or not res.active or res.is_default_user:
             return None
         return res
 
     @classmethod
     def get_by_email(cls, email, cache=False):
         q = cls.query().filter(func.lower(cls.email) == func.lower(email))
 
         if cache:
             q = q.options(FromCache("sql_cache_short",
                                     "get_email_key_%s" % email))
 
         ret = q.scalar()
         if ret is None:
             q = UserEmailMap.query()
             # try fetching in alternate email map
             q = q.filter(func.lower(UserEmailMap.email) == func.lower(email))
             q = q.options(joinedload(UserEmailMap.user))
             if cache:
                 q = q.options(FromCache("sql_cache_short",
                                         "get_email_map_key_%s" % email))
             ret = getattr(q.scalar(), 'user', None)
 
         return ret
 
     @classmethod
     def get_from_cs_author(cls, author):
         """
         Tries to get User objects out of commit author string
 
         :param author:
         """
         from kallithea.lib.helpers import email, author_name
         # Valid email in the attribute passed, see if they're in the system
         _email = email(author)
         if _email:
             user = cls.get_by_email(_email)
             if user is not None:
                 return user
         # Maybe we can match by username?
         _author = author_name(author)
         user = cls.get_by_username(_author, case_insensitive=True)
         if user is not None:
             return user
 
     def update_lastlogin(self):
         """Update user lastlogin"""
         self.last_login = datetime.datetime.now()
         log.debug('updated user %s lastlogin', self.username)
 
     @classmethod
     def get_first_admin(cls):
         user = User.query().filter(User.admin == True).first()
         if user is None:
             raise Exception('Missing administrative account!')
         return user
 
     @classmethod
     def get_default_user(cls, cache=False):
         user = User.get_by_username(User.DEFAULT_USER, cache=cache)
         if user is None:
             raise Exception('Missing default account!')
         return user
 
     def get_api_data(self, details=False):
         """
         Common function for generating user related data for API
         """
         user = self
         data = dict(
             user_id=user.user_id,
             username=user.username,
             firstname=user.name,
             lastname=user.lastname,
             email=user.email,
             emails=user.emails,
             active=user.active,
             admin=user.admin,
         )
         if details:
             data.update(dict(
                 extern_type=user.extern_type,
                 extern_name=user.extern_name,
                 api_key=user.api_key,
                 api_keys=user.api_keys,
                 last_login=user.last_login,
                 ip_addresses=user.ip_addresses
                 ))
         return data
 
     def __json__(self):
         data = dict(
             full_name=self.full_name,
             full_name_or_username=self.full_name_or_username,
             short_contact=self.short_contact,
             full_contact=self.full_contact
         )
         data.update(self.get_api_data())
         return data
 
 
 class UserApiKeys(Base, BaseDbModel):
     __tablename__ = 'user_api_keys'
     __table_args__ = (
         Index('uak_api_key_idx', 'api_key'),
         Index('uak_api_key_expires_idx', 'api_key', 'expires'),
         _table_args_default_dict,
     )
 
     user_api_key_id = Column(Integer(), primary_key=True)
     user_id = Column(Integer(), ForeignKey('users.user_id'), nullable=False)
     api_key = Column(String(255), nullable=False, unique=True)
     description = Column(UnicodeText(), nullable=False)
     expires = Column(Float(53), nullable=False)
     created_on = Column(DateTime(timezone=False), nullable=False, default=datetime.datetime.now)
 
     user = relationship('User')
 
     @hybrid_property
     def is_expired(self):
         return (self.expires != -1) & (time.time() > self.expires)
 
 
 class UserEmailMap(Base, BaseDbModel):
     __tablename__ = 'user_email_map'
     __table_args__ = (
         Index('uem_email_idx', 'email'),
         _table_args_default_dict,
     )
 
     email_id = Column(Integer(), primary_key=True)
     user_id = Column(Integer(), ForeignKey('users.user_id'), nullable=False)
     _email = Column("email", String(255), nullable=False, unique=True)
     user = relationship('User')
 
     @validates('_email')
     def validate_email(self, key, email):
         # check if this email is not main one
         main_email = Session().query(User).filter(User.email == email).scalar()
         if main_email is not None:
             raise AttributeError('email %s is present is user table' % email)
         return email
 
     @hybrid_property
     def email(self):
         return self._email
 
     @email.setter
     def email(self, val):
         self._email = val.lower() if val else None
 
 
 class UserIpMap(Base, BaseDbModel):
     __tablename__ = 'user_ip_map'
     __table_args__ = (
         UniqueConstraint('user_id', 'ip_addr'),
         _table_args_default_dict,
     )
 
     ip_id = Column(Integer(), primary_key=True)
     user_id = Column(Integer(), ForeignKey('users.user_id'), nullable=False)
     ip_addr = Column(String(255), nullable=False)
     active = Column(Boolean(), nullable=False, default=True)
     user = relationship('User')
 
     @classmethod
     def _get_ip_range(cls, ip_addr):
         net = ipaddr.IPNetwork(address=ip_addr)
         return [str(net.network), str(net.broadcast)]
 
     def __json__(self):
         return dict(
           ip_addr=self.ip_addr,
           ip_range=self._get_ip_range(self.ip_addr)
         )
 
     def __unicode__(self):
         return u"<%s('user_id:%s=>%s')>" % (self.__class__.__name__,
                                             self.user_id, self.ip_addr)
 
 
 class UserLog(Base, BaseDbModel):
     __tablename__ = 'user_logs'
     __table_args__ = (
         _table_args_default_dict,
     )
 
     user_log_id = Column(Integer(), primary_key=True)
     user_id = Column(Integer(), ForeignKey('users.user_id'), nullable=True)
     username = Column(String(255), nullable=False)
     repository_id = Column(Integer(), ForeignKey('repositories.repo_id'), nullable=True)
     repository_name = Column(Unicode(255), nullable=False)
     user_ip = Column(String(255), nullable=True)
     action = Column(UnicodeText(), nullable=False)
     action_date = Column(DateTime(timezone=False), nullable=False)
 
     def __unicode__(self):
         return u"<%s('id:%s:%s')>" % (self.__class__.__name__,
                                       self.repository_name,
                                       self.action)
 
     @property
     def action_as_day(self):
         return datetime.date(*self.action_date.timetuple()[:3])
 
     user = relationship('User')
     repository = relationship('Repository', cascade='')
 
 
 class UserGroup(Base, BaseDbModel):
     __tablename__ = 'users_groups'
     __table_args__ = (
         _table_args_default_dict,
     )
 
     users_group_id = Column(Integer(), primary_key=True)
     users_group_name = Column(Unicode(255), nullable=False, unique=True)
     user_group_description = Column(Unicode(10000), nullable=True) # FIXME: not nullable?
     users_group_active = Column(Boolean(), nullable=False)
     owner_id = Column('user_id', Integer(), ForeignKey('users.user_id'), nullable=False)
     created_on = Column(DateTime(timezone=False), nullable=False, default=datetime.datetime.now)
     _group_data = Column("group_data", LargeBinary(), nullable=True)  # JSON data # FIXME: not nullable?
 
     members = relationship('UserGroupMember', cascade="all, delete-orphan")
     users_group_to_perm = relationship('UserGroupToPerm', cascade='all')
     users_group_repo_to_perm = relationship('UserGroupRepoToPerm', cascade='all')
     users_group_repo_group_to_perm = relationship('UserGroupRepoGroupToPerm', cascade='all')
     user_user_group_to_perm = relationship('UserUserGroupToPerm ', cascade='all')
     user_group_user_group_to_perm = relationship('UserGroupUserGroupToPerm ', primaryjoin="UserGroupUserGroupToPerm.target_user_group_id==UserGroup.users_group_id", cascade='all')
 
     owner = relationship('User')
 
     @hybrid_property
     def group_data(self):
         if not self._group_data:
             return {}
 
         try:
             return json.loads(self._group_data)
         except TypeError:
             return {}
 
     @group_data.setter
     def group_data(self, val):
         try:
             self._group_data = json.dumps(val)
         except Exception:
             log.error(traceback.format_exc())
 
     def __unicode__(self):
         return u"<%s('id:%s:%s')>" % (self.__class__.__name__,
                                       self.users_group_id,
                                       self.users_group_name)
 
     @classmethod
     def guess_instance(cls, value):
         return super(UserGroup, cls).guess_instance(value, UserGroup.get_by_group_name)
 
     @classmethod
     def get_by_group_name(cls, group_name, cache=False,
                           case_insensitive=False):
         if case_insensitive:
             q = cls.query().filter(func.lower(cls.users_group_name) == func.lower(group_name))
         else:
             q = cls.query().filter(cls.users_group_name == group_name)
         if cache:
             q = q.options(FromCache(
                             "sql_cache_short",
                             "get_group_%s" % _hash_key(group_name)
                           )
             )
         return q.scalar()
 
     @classmethod
     def get(cls, user_group_id, cache=False):
         user_group = cls.query()
         if cache:
             user_group = user_group.options(FromCache("sql_cache_short",
                                     "get_users_group_%s" % user_group_id))
         return user_group.get(user_group_id)
 
     def get_api_data(self, with_members=True):
         user_group = self
 
         data = dict(
             users_group_id=user_group.users_group_id,
             group_name=user_group.users_group_name,
             group_description=user_group.user_group_description,
             active=user_group.users_group_active,
             owner=user_group.owner.username,
         )
         if with_members:
             data['members'] = [
                 ugm.user.get_api_data()
                 for ugm in user_group.members
             ]
 
         return data
 
 
 class UserGroupMember(Base, BaseDbModel):
     __tablename__ = 'users_groups_members'
     __table_args__ = (
         _table_args_default_dict,
     )
 
     users_group_member_id = Column(Integer(), primary_key=True)
     users_group_id = Column(Integer(), ForeignKey('users_groups.users_group_id'), nullable=False)
     user_id = Column(Integer(), ForeignKey('users.user_id'), nullable=False)
 
     user = relationship('User')
     users_group = relationship('UserGroup')
 
     def __init__(self, gr_id='', u_id=''):
         self.users_group_id = gr_id
         self.user_id = u_id
 
 
 class RepositoryField(Base, BaseDbModel):
     __tablename__ = 'repositories_fields'
     __table_args__ = (
         UniqueConstraint('repository_id', 'field_key'),  # no-multi field
         _table_args_default_dict,
     )
 
     PREFIX = 'ex_'  # prefix used in form to not conflict with already existing fields
 
     repo_field_id = Column(Integer(), primary_key=True)
     repository_id = Column(Integer(), ForeignKey('repositories.repo_id'), nullable=False)
     field_key = Column(String(250), nullable=False)
     field_label = Column(String(1024), nullable=False)
     field_value = Column(String(10000), nullable=False)
     field_desc = Column(String(1024), nullable=False)
     field_type = Column(String(255), nullable=False)
     created_on = Column(DateTime(timezone=False), nullable=False, default=datetime.datetime.now)
 
     repository = relationship('Repository')
 
     @property
     def field_key_prefixed(self):
         return 'ex_%s' % self.field_key
 
     @classmethod
     def un_prefix_key(cls, key):
         if key.startswith(cls.PREFIX):
             return key[len(cls.PREFIX):]
         return key
 
     @classmethod
     def get_by_key_name(cls, key, repo):
         row = cls.query() \
                 .filter(cls.repository == repo) \
                 .filter(cls.field_key == key).scalar()
         return row
 
 
 class Repository(Base, BaseDbModel):
     __tablename__ = 'repositories'
     __table_args__ = (
         Index('r_repo_name_idx', 'repo_name'),
         _table_args_default_dict,
     )
 
     DEFAULT_CLONE_URI = '{scheme}://{user}@{netloc}/{repo}'
     DEFAULT_CLONE_SSH = 'ssh://{system_user}@{hostname}/{repo}'
 
     STATE_CREATED = u'repo_state_created'
     STATE_PENDING = u'repo_state_pending'
     STATE_ERROR = u'repo_state_error'
 
     repo_id = Column(Integer(), primary_key=True)
     repo_name = Column(Unicode(255), nullable=False, unique=True)
     repo_state = Column(String(255), nullable=False)
 
     clone_uri = Column(String(255), nullable=True) # FIXME: not nullable?
     repo_type = Column(String(255), nullable=False) # 'hg' or 'git'
     owner_id = Column('user_id', Integer(), ForeignKey('users.user_id'), nullable=False)
     private = Column(Boolean(), nullable=False)
     enable_statistics = Column("statistics", Boolean(), nullable=False, default=True)
     enable_downloads = Column("downloads", Boolean(), nullable=False, default=True)
     description = Column(Unicode(10000), nullable=False)
     created_on = Column(DateTime(timezone=False), nullable=False, default=datetime.datetime.now)
     updated_on = Column(DateTime(timezone=False), nullable=False, default=datetime.datetime.now)
     _landing_revision = Column("landing_revision", String(255), nullable=False)
     _changeset_cache = Column("changeset_cache", LargeBinary(), nullable=True) # JSON data # FIXME: not nullable?
 
     fork_id = Column(Integer(), ForeignKey('repositories.repo_id'), nullable=True)
     group_id = Column(Integer(), ForeignKey('groups.group_id'), nullable=True)
 
     owner = relationship('User')
     fork = relationship('Repository', remote_side=repo_id)
     group = relationship('RepoGroup')
     repo_to_perm = relationship('UserRepoToPerm', cascade='all', order_by='UserRepoToPerm.repo_to_perm_id')
     users_group_to_perm = relationship('UserGroupRepoToPerm', cascade='all')
     stats = relationship('Statistics', cascade='all', uselist=False)
 
     followers = relationship('UserFollowing',
                              primaryjoin='UserFollowing.follows_repository_id==Repository.repo_id',
                              cascade='all')
     extra_fields = relationship('RepositoryField',
                                 cascade="all, delete-orphan")
 
     logs = relationship('UserLog')
     comments = relationship('ChangesetComment', cascade="all, delete-orphan")
 
     pull_requests_org = relationship('PullRequest',
                     primaryjoin='PullRequest.org_repo_id==Repository.repo_id',
                     cascade="all, delete-orphan")
 
     pull_requests_other = relationship('PullRequest',
                     primaryjoin='PullRequest.other_repo_id==Repository.repo_id',
                     cascade="all, delete-orphan")
 
     def __unicode__(self):
         return u"<%s('%s:%s')>" % (self.__class__.__name__, self.repo_id,
                                    safe_unicode(self.repo_name))
 
     @hybrid_property
     def landing_rev(self):
         # always should return [rev_type, rev]
         if self._landing_revision:
             _rev_info = self._landing_revision.split(':')
             if len(_rev_info) < 2:
                 _rev_info.insert(0, 'rev')
             return [_rev_info[0], _rev_info[1]]
         return [None, None]
 
     @landing_rev.setter
     def landing_rev(self, val):
         if ':' not in val:
             raise ValueError('value must be delimited with `:` and consist '
                              'of <rev_type>:<rev>, got %s instead' % val)
         self._landing_revision = val
 
     @hybrid_property
     def changeset_cache(self):
         try:
             cs_cache = json.loads(self._changeset_cache) # might raise on bad data
             cs_cache['raw_id'] # verify data, raise exception on error
             return cs_cache
         except (TypeError, KeyError, ValueError):
             return EmptyChangeset().__json__()
 
     @changeset_cache.setter
     def changeset_cache(self, val):
         try:
             self._changeset_cache = json.dumps(val)
         except Exception:
             log.error(traceback.format_exc())
 
     @classmethod
     def query(cls, sorted=False):
         """Add Repository-specific helpers for common query constructs.
 
         sorted: if True, apply the default ordering (name, case insensitive).
         """
         q = super(Repository, cls).query()
 
         if sorted:
             q = q.order_by(func.lower(Repository.repo_name))
 
         return q
 
     @classmethod
     def url_sep(cls):
         return URL_SEP
 
     @classmethod
     def normalize_repo_name(cls, repo_name):
         """
         Normalizes os specific repo_name to the format internally stored inside
         database using URL_SEP
 
         :param cls:
         :param repo_name:
         """
         return cls.url_sep().join(repo_name.split(os.sep))
 
     @classmethod
     def guess_instance(cls, value):
         return super(Repository, cls).guess_instance(value, Repository.get_by_repo_name)
 
     @classmethod
     def get_by_repo_name(cls, repo_name, case_insensitive=False):
         """Get the repo, defaulting to database case sensitivity.
         case_insensitive will be slower and should only be specified if necessary."""
         if case_insensitive:
             q = Session().query(cls).filter(func.lower(cls.repo_name) == func.lower(repo_name))
         else:
             q = Session().query(cls).filter(cls.repo_name == repo_name)
         q = q.options(joinedload(Repository.fork)) \
                 .options(joinedload(Repository.owner)) \
                 .options(joinedload(Repository.group))
         return q.scalar()
 
     @classmethod
     def get_by_full_path(cls, repo_full_path):
         base_full_path = os.path.realpath(cls.base_path())
         repo_full_path = os.path.realpath(repo_full_path)
         assert repo_full_path.startswith(base_full_path + os.path.sep)
         repo_name = repo_full_path[len(base_full_path) + 1:]
         repo_name = cls.normalize_repo_name(repo_name)
         return cls.get_by_repo_name(repo_name.strip(URL_SEP))
 
     @classmethod
     def get_repo_forks(cls, repo_id):
         return cls.query().filter(Repository.fork_id == repo_id)
 
     @classmethod
     def base_path(cls):
         """
         Returns base path where all repos are stored
 
         :param cls:
         """
         q = Session().query(Ui) \
             .filter(Ui.ui_key == cls.url_sep())
         q = q.options(FromCache("sql_cache_short", "repository_repo_path"))
         return q.one().ui_value
 
     @property
     def forks(self):
         """
         Return forks of this repo
         """
         return Repository.get_repo_forks(self.repo_id)
 
     @property
     def parent(self):
         """
         Returns fork parent
         """
         return self.fork
 
     @property
     def just_name(self):
         return self.repo_name.split(Repository.url_sep())[-1]
 
     @property
     def groups_with_parents(self):
         groups = []
         group = self.group
         while group is not None:
             groups.append(group)
             group = group.parent_group
             assert group not in groups, group # avoid recursion on bad db content
         groups.reverse()
         return groups
 
     @LazyProperty
     def repo_path(self):
         """
         Returns base full path for that repository means where it actually
         exists on a filesystem
         """
         q = Session().query(Ui).filter(Ui.ui_key ==
                                               Repository.url_sep())
         q = q.options(FromCache("sql_cache_short", "repository_repo_path"))
         return q.one().ui_value
 
     @property
     def repo_full_path(self):
         p = [self.repo_path]
         # we need to split the name by / since this is how we store the
         # names in the database, but that eventually needs to be converted
         # into a valid system path
         p += self.repo_name.split(Repository.url_sep())
         return os.path.join(*map(safe_unicode, p))
 
     @property
     def cache_keys(self):
         """
         Returns associated cache keys for that repo
         """
         return CacheInvalidation.query() \
             .filter(CacheInvalidation.cache_args == self.repo_name) \
             .order_by(CacheInvalidation.cache_key) \
             .all()
 
     def get_new_name(self, repo_name):
         """
         returns new full repository name based on assigned group and new new
 
         :param group_name:
         """
         path_prefix = self.group.full_path_splitted if self.group else []
         return Repository.url_sep().join(path_prefix + [repo_name])
 
     @property
     def _ui(self):
         """
         Creates an db based ui object for this repository
         """
         from kallithea.lib.utils import make_ui
         return make_ui()
 
     @classmethod
     def is_valid(cls, repo_name):
         """
         returns True if given repo name is a valid filesystem repository
 
         :param cls:
         :param repo_name:
         """
         from kallithea.lib.utils import is_valid_repo
 
         return is_valid_repo(repo_name, cls.base_path())
 
     def get_api_data(self, with_revision_names=False,
                            with_pullrequests=False):
         """
         Common function for generating repo api data.
         Optionally, also return tags, branches, bookmarks and PRs.
         """
         repo = self
         data = dict(
             repo_id=repo.repo_id,
             repo_name=repo.repo_name,
             repo_type=repo.repo_type,
             clone_uri=repo.clone_uri,
             private=repo.private,
             created_on=repo.created_on,
             description=repo.description,
             landing_rev=repo.landing_rev,
             owner=repo.owner.username,
             fork_of=repo.fork.repo_name if repo.fork else None,
             enable_statistics=repo.enable_statistics,
             enable_downloads=repo.enable_downloads,
             last_changeset=repo.changeset_cache,
         )
         if with_revision_names:
             scm_repo = repo.scm_instance_no_cache()
             data.update(dict(
                 tags=scm_repo.tags,
                 branches=scm_repo.branches,
                 bookmarks=scm_repo.bookmarks,
             ))
         if with_pullrequests:
             data['pull_requests'] = repo.pull_requests_other
         rc_config = Setting.get_app_settings()
         repository_fields = str2bool(rc_config.get('repository_fields'))
         if repository_fields:
             for f in self.extra_fields:
                 data[f.field_key_prefixed] = f.field_value
 
         return data
 
     @property
     def last_db_change(self):
         return self.updated_on
 
     @property
     def clone_uri_hidden(self):
         clone_uri = self.clone_uri
         if clone_uri:
             import urlobject
             url_obj = urlobject.URLObject(self.clone_uri)
             if url_obj.password:
                 clone_uri = url_obj.with_password('*****')
         return clone_uri
 
     def clone_url(self, clone_uri_tmpl, with_id=False, username=None):
         if '{repo}' not in clone_uri_tmpl and '_{repoid}' not in clone_uri_tmpl:
             log.error("Configured clone_uri_tmpl %r has no '{repo}' or '_{repoid}' and cannot toggle to use repo id URLs", clone_uri_tmpl)
         elif with_id:
             clone_uri_tmpl = clone_uri_tmpl.replace('{repo}', '_{repoid}')
         else:
             clone_uri_tmpl = clone_uri_tmpl.replace('_{repoid}', '{repo}')
 
         import kallithea.lib.helpers as h
         prefix_url = h.canonical_url('home')
 
         return get_clone_url(clone_uri_tmpl=clone_uri_tmpl,
                              prefix_url=prefix_url,
                              repo_name=self.repo_name,
                              repo_id=self.repo_id,
                              username=username)
 
     def set_state(self, state):
         self.repo_state = state
 
     #==========================================================================
     # SCM PROPERTIES
     #==========================================================================
 
     def get_changeset(self, rev=None):
         return get_changeset_safe(self.scm_instance, rev)
 
     def get_landing_changeset(self):
         """
         Returns landing changeset, or if that doesn't exist returns the tip
         """
         _rev_type, _rev = self.landing_rev
         cs = self.get_changeset(_rev)
         if isinstance(cs, EmptyChangeset):
             return self.get_changeset()
         return cs
 
     def update_changeset_cache(self, cs_cache=None):
         """
         Update cache of last changeset for repository, keys should be::
 
             short_id
             raw_id
             revision
             message
             date
             author
 
         :param cs_cache:
         """
         from kallithea.lib.vcs.backends.base import BaseChangeset
         if cs_cache is None:
             cs_cache = EmptyChangeset()
             # use no-cache version here
             scm_repo = self.scm_instance_no_cache()
             if scm_repo:
                 cs_cache = scm_repo.get_changeset()
 
         if isinstance(cs_cache, BaseChangeset):
             cs_cache = cs_cache.__json__()
 
         if (not self.changeset_cache or cs_cache['raw_id'] != self.changeset_cache['raw_id']):
             _default = datetime.datetime.fromtimestamp(0)
             last_change = cs_cache.get('date') or _default
             log.debug('updated repo %s with new cs cache %s',
                       self.repo_name, cs_cache)
             self.updated_on = last_change
             self.changeset_cache = cs_cache
             Session().commit()
         else:
             log.debug('changeset_cache for %s already up to date with %s',
                       self.repo_name, cs_cache['raw_id'])
 
     @property
     def tip(self):
         return self.get_changeset('tip')
 
     @property
     def author(self):
         return self.tip.author
 
     @property
     def last_change(self):
         return self.scm_instance.last_change
 
     def get_comments(self, revisions=None):
         """
         Returns comments for this repository grouped by revisions
 
         :param revisions: filter query by revisions only
         """
         cmts = ChangesetComment.query() \
             .filter(ChangesetComment.repo == self)
         if revisions is not None:
             if not revisions:
                 return {} # don't use sql 'in' on empty set
             cmts = cmts.filter(ChangesetComment.revision.in_(revisions))
         grouped = collections.defaultdict(list)
         for cmt in cmts.all():
             grouped[cmt.revision].append(cmt)
         return grouped
 
     def statuses(self, revisions):
         """
         Returns statuses for this repository.
         PRs without any votes do _not_ show up as unreviewed.
 
         :param revisions: list of revisions to get statuses for
         """
         if not revisions:
             return {}
 
         statuses = ChangesetStatus.query() \
             .filter(ChangesetStatus.repo == self) \
             .filter(ChangesetStatus.version == 0) \
             .filter(ChangesetStatus.revision.in_(revisions))
 
         grouped = {}
         for stat in statuses.all():
             pr_id = pr_nice_id = pr_repo = None
             if stat.pull_request:
                 pr_id = stat.pull_request.pull_request_id
                 pr_nice_id = PullRequest.make_nice_id(pr_id)
                 pr_repo = stat.pull_request.other_repo.repo_name
             grouped[stat.revision] = [str(stat.status), stat.status_lbl,
                                       pr_id, pr_repo, pr_nice_id,
                                       stat.author]
         return grouped
 
     def _repo_size(self):
         from kallithea.lib import helpers as h
         log.debug('calculating repository size...')
         return h.format_byte_size(self.scm_instance.size)
 
     #==========================================================================
     # SCM CACHE INSTANCE
     #==========================================================================
 
     def set_invalidate(self):
         """
         Mark caches of this repo as invalid.
         """
         CacheInvalidation.set_invalidate(self.repo_name)
 
     _scm_instance = None
 
     @property
     def scm_instance(self):
         if self._scm_instance is None:
             self._scm_instance = self.scm_instance_cached()
         return self._scm_instance
 
     def scm_instance_cached(self, valid_cache_keys=None):
         @cache_region('long_term', 'scm_instance_cached')
         def _c(repo_name): # repo_name is just for the cache key
             log.debug('Creating new %s scm_instance and populating cache', repo_name)
             return self.scm_instance_no_cache()
         rn = self.repo_name
 
         valid = CacheInvalidation.test_and_set_valid(rn, None, valid_cache_keys=valid_cache_keys)
         if not valid:
             log.debug('Cache for %s invalidated, getting new object', rn)
             region_invalidate(_c, None, 'scm_instance_cached', rn)
         else:
             log.debug('Trying to get scm_instance of %s from cache', rn)
         return _c(rn)
 
     def scm_instance_no_cache(self):
         repo_full_path = safe_str(self.repo_full_path)
         alias = get_scm(repo_full_path)[0]
         log.debug('Creating instance of %s repository from %s',
                   alias, self.repo_full_path)
         backend = get_backend(alias)
 
         if alias == 'hg':
             repo = backend(repo_full_path, create=False,
                            baseui=self._ui)
         else:
             repo = backend(repo_full_path, create=False)
 
         return repo
 
     def __json__(self):
         return dict(
             repo_id=self.repo_id,
             repo_name=self.repo_name,
             landing_rev=self.landing_rev,
         )
 
 
 class RepoGroup(Base, BaseDbModel):
     __tablename__ = 'groups'
     __table_args__ = (
         _table_args_default_dict,
     )
 
     SEP = ' &raquo; '
 
     group_id = Column(Integer(), primary_key=True)
     group_name = Column(Unicode(255), nullable=False, unique=True) # full path
     parent_group_id = Column('group_parent_id', Integer(), ForeignKey('groups.group_id'), nullable=True)
     group_description = Column(Unicode(10000), nullable=False)
     owner_id = Column('user_id', Integer(), ForeignKey('users.user_id'), nullable=False)
     created_on = Column(DateTime(timezone=False), nullable=False, default=datetime.datetime.now)
 
     repo_group_to_perm = relationship('UserRepoGroupToPerm', cascade='all', order_by='UserRepoGroupToPerm.group_to_perm_id')
     users_group_to_perm = relationship('UserGroupRepoGroupToPerm', cascade='all')
     parent_group = relationship('RepoGroup', remote_side=group_id)
     owner = relationship('User')
 
     @classmethod
     def query(cls, sorted=False):
         """Add RepoGroup-specific helpers for common query constructs.
 
         sorted: if True, apply the default ordering (name, case insensitive).
         """
         q = super(RepoGroup, cls).query()
 
         if sorted:
             q = q.order_by(func.lower(RepoGroup.group_name))
 
         return q
 
     def __init__(self, group_name='', parent_group=None):
         self.group_name = group_name
         self.parent_group = parent_group
 
     def __unicode__(self):
         return u"<%s('id:%s:%s')>" % (self.__class__.__name__, self.group_id,
                                       self.group_name)
 
     @classmethod
     def _generate_choice(cls, repo_group):
         """Return tuple with group_id and name as html literal"""
         from webhelpers2.html import literal
         if repo_group is None:
             return (-1, u'-- %s --' % _('top level'))
         return repo_group.group_id, literal(cls.SEP.join(repo_group.full_path_splitted))
 
     @classmethod
     def groups_choices(cls, groups):
         """Return tuples with group_id and name as html literal."""
         return sorted((cls._generate_choice(g) for g in groups),
                       key=lambda c: c[1].split(cls.SEP))
 
     @classmethod
     def url_sep(cls):
         return URL_SEP
 
     @classmethod
     def guess_instance(cls, value):
         return super(RepoGroup, cls).guess_instance(value, RepoGroup.get_by_group_name)
 
     @classmethod
     def get_by_group_name(cls, group_name, cache=False, case_insensitive=False):
         group_name = group_name.rstrip('/')
         if case_insensitive:
             gr = cls.query() \
                 .filter(func.lower(cls.group_name) == func.lower(group_name))
         else:
             gr = cls.query() \
                 .filter(cls.group_name == group_name)
         if cache:
             gr = gr.options(FromCache(
                             "sql_cache_short",
                             "get_group_%s" % _hash_key(group_name)
                             )
             )
         return gr.scalar()
 
     @property
     def parents(self):
         groups = []
         group = self.parent_group
         while group is not None:
             groups.append(group)
             group = group.parent_group
             assert group not in groups, group # avoid recursion on bad db content
         groups.reverse()
         return groups
 
     @property
     def children(self):
         return RepoGroup.query().filter(RepoGroup.parent_group == self)
 
     @property
     def name(self):
         return self.group_name.split(RepoGroup.url_sep())[-1]
 
     @property
     def full_path(self):
         return self.group_name
 
     @property
     def full_path_splitted(self):
         return self.group_name.split(RepoGroup.url_sep())
 
     @property
     def repositories(self):
         return Repository.query(sorted=True).filter_by(group=self)
 
     @property
     def repositories_recursive_count(self):
         cnt = self.repositories.count()
 
         def children_count(group):
             cnt = 0
             for child in group.children:
                 cnt += child.repositories.count()
                 cnt += children_count(child)
             return cnt
 
         return cnt + children_count(self)
 
     def _recursive_objects(self, include_repos=True):
         all_ = []
 
         def _get_members(root_gr):
             if include_repos:
                 for r in root_gr.repositories:
                     all_.append(r)
             childs = root_gr.children.all()
             if childs:
                 for gr in childs:
                     all_.append(gr)
                     _get_members(gr)
 
         _get_members(self)
         return [self] + all_
 
     def recursive_groups_and_repos(self):
         """
         Recursive return all groups, with repositories in those groups
         """
         return self._recursive_objects()
 
     def recursive_groups(self):
         """
         Returns all children groups for this group including children of children
         """
         return self._recursive_objects(include_repos=False)
 
     def get_new_name(self, group_name):
         """
         returns new full group name based on parent and new name
 
         :param group_name:
         """
         path_prefix = (self.parent_group.full_path_splitted if
                        self.parent_group else [])
         return RepoGroup.url_sep().join(path_prefix + [group_name])
 
     def get_api_data(self):
         """
         Common function for generating api data
 
         """
         group = self
         data = dict(
             group_id=group.group_id,
             group_name=group.group_name,
             group_description=group.group_description,
             parent_group=group.parent_group.group_name if group.parent_group else None,
             repositories=[x.repo_name for x in group.repositories],
             owner=group.owner.username
         )
         return data
 
 
 class Permission(Base, BaseDbModel):
     __tablename__ = 'permissions'
     __table_args__ = (
         Index('p_perm_name_idx', 'permission_name'),
         _table_args_default_dict,
     )
 
     PERMS = (
         ('hg.admin', _('Kallithea Administrator')),
 
         ('repository.none', _('Default user has no access to new repositories')),
         ('repository.read', _('Default user has read access to new repositories')),
         ('repository.write', _('Default user has write access to new repositories')),
         ('repository.admin', _('Default user has admin access to new repositories')),
 
         ('group.none', _('Default user has no access to new repository groups')),
         ('group.read', _('Default user has read access to new repository groups')),
         ('group.write', _('Default user has write access to new repository groups')),
         ('group.admin', _('Default user has admin access to new repository groups')),
 
         ('usergroup.none', _('Default user has no access to new user groups')),
         ('usergroup.read', _('Default user has read access to new user groups')),
         ('usergroup.write', _('Default user has write access to new user groups')),
         ('usergroup.admin', _('Default user has admin access to new user groups')),
 
         ('hg.repogroup.create.false', _('Only admins can create repository groups')),
         ('hg.repogroup.create.true', _('Non-admins can create repository groups')),
 
         ('hg.usergroup.create.false', _('Only admins can create user groups')),
         ('hg.usergroup.create.true', _('Non-admins can create user groups')),
 
         ('hg.create.none', _('Only admins can create top level repositories')),
         ('hg.create.repository', _('Non-admins can create top level repositories')),
 
         ('hg.create.write_on_repogroup.true', _('Repository creation enabled with write permission to a repository group')),
         ('hg.create.write_on_repogroup.false', _('Repository creation disabled with write permission to a repository group')),
 
         ('hg.fork.none', _('Only admins can fork repositories')),
         ('hg.fork.repository', _('Non-admins can fork repositories')),
 
         ('hg.register.none', _('Registration disabled')),
         ('hg.register.manual_activate', _('User registration with manual account activation')),
         ('hg.register.auto_activate', _('User registration with automatic account activation')),
 
         ('hg.extern_activate.manual', _('Manual activation of external account')),
         ('hg.extern_activate.auto', _('Automatic activation of external account')),
     )
 
     # definition of system default permissions for DEFAULT user
     DEFAULT_USER_PERMISSIONS = (
         'repository.read',
         'group.read',
         'usergroup.read',
         'hg.create.repository',
         'hg.create.write_on_repogroup.true',
         'hg.fork.repository',
         'hg.register.manual_activate',
         'hg.extern_activate.auto',
     )
 
     # defines which permissions are more important higher the more important
     # Weight defines which permissions are more important.
     # The higher number the more important.
     PERM_WEIGHTS = {
         'repository.none': 0,
         'repository.read': 1,
         'repository.write': 3,
         'repository.admin': 4,
 
         'group.none': 0,
         'group.read': 1,
         'group.write': 3,
         'group.admin': 4,
 
         'usergroup.none': 0,
         'usergroup.read': 1,
         'usergroup.write': 3,
         'usergroup.admin': 4,
 
         'hg.repogroup.create.false': 0,
         'hg.repogroup.create.true': 1,
 
         'hg.usergroup.create.false': 0,
         'hg.usergroup.create.true': 1,
 
         'hg.fork.none': 0,
         'hg.fork.repository': 1,
 
         'hg.create.none': 0,
         'hg.create.repository': 1,
 
         'hg.create.write_on_repogroup.false': 0,
         'hg.create.write_on_repogroup.true': 1,
 
         'hg.register.none': 0,
         'hg.register.manual_activate': 1,
         'hg.register.auto_activate': 2,
 
         'hg.extern_activate.manual': 0,
         'hg.extern_activate.auto': 1,
     }
 
     permission_id = Column(Integer(), primary_key=True)
     permission_name = Column(String(255), nullable=False)
 
     def __unicode__(self):
         return u"<%s('%s:%s')>" % (
             self.__class__.__name__, self.permission_id, self.permission_name
         )
 
     @classmethod
     def guess_instance(cls, value):
         return super(Permission, cls).guess_instance(value, Permission.get_by_key)
 
     @classmethod
     def get_by_key(cls, key):
         return cls.query().filter(cls.permission_name == key).scalar()
 
     @classmethod
     def get_default_perms(cls, default_user_id):
         q = Session().query(UserRepoToPerm, Repository, cls) \
          .join((Repository, UserRepoToPerm.repository_id == Repository.repo_id)) \
          .join((cls, UserRepoToPerm.permission_id == cls.permission_id)) \
          .filter(UserRepoToPerm.user_id == default_user_id)
 
         return q.all()
 
     @classmethod
     def get_default_group_perms(cls, default_user_id):
         q = Session().query(UserRepoGroupToPerm, RepoGroup, cls) \
          .join((RepoGroup, UserRepoGroupToPerm.group_id == RepoGroup.group_id)) \
          .join((cls, UserRepoGroupToPerm.permission_id == cls.permission_id)) \
          .filter(UserRepoGroupToPerm.user_id == default_user_id)
 
         return q.all()
 
     @classmethod
     def get_default_user_group_perms(cls, default_user_id):
         q = Session().query(UserUserGroupToPerm, UserGroup, cls) \
          .join((UserGroup, UserUserGroupToPerm.user_group_id == UserGroup.users_group_id)) \
          .join((cls, UserUserGroupToPerm.permission_id == cls.permission_id)) \
          .filter(UserUserGroupToPerm.user_id == default_user_id)
 
         return q.all()
 
 
 class UserRepoToPerm(Base, BaseDbModel):
     __tablename__ = 'repo_to_perm'
     __table_args__ = (
         UniqueConstraint('user_id', 'repository_id', 'permission_id'),
         _table_args_default_dict,
     )
 
     repo_to_perm_id = Column(Integer(), primary_key=True)
     user_id = Column(Integer(), ForeignKey('users.user_id'), nullable=False)
     permission_id = Column(Integer(), ForeignKey('permissions.permission_id'), nullable=False)
     repository_id = Column(Integer(), ForeignKey('repositories.repo_id'), nullable=False)
 
     user = relationship('User')
     repository = relationship('Repository')
     permission = relationship('Permission')
 
     @classmethod
     def create(cls, user, repository, permission):
         n = cls()
         n.user = user
         n.repository = repository
         n.permission = permission
         Session().add(n)
         return n
 
     def __unicode__(self):
         return u'<%s => %s >' % (self.user, self.repository)
 
 
 class UserUserGroupToPerm(Base, BaseDbModel):
     __tablename__ = 'user_user_group_to_perm'
     __table_args__ = (
         UniqueConstraint('user_id', 'user_group_id', 'permission_id'),
         _table_args_default_dict,
     )
 
     user_user_group_to_perm_id = Column(Integer(), primary_key=True)
     user_id = Column(Integer(), ForeignKey('users.user_id'), nullable=False)
     permission_id = Column(Integer(), ForeignKey('permissions.permission_id'), nullable=False)
     user_group_id = Column(Integer(), ForeignKey('users_groups.users_group_id'), nullable=False)
 
     user = relationship('User')
     user_group = relationship('UserGroup')
     permission = relationship('Permission')
 
     @classmethod
     def create(cls, user, user_group, permission):
         n = cls()
         n.user = user
         n.user_group = user_group
         n.permission = permission
         Session().add(n)
         return n
 
     def __unicode__(self):
         return u'<%s => %s >' % (self.user, self.user_group)
 
 
 class UserToPerm(Base, BaseDbModel):
     __tablename__ = 'user_to_perm'
     __table_args__ = (
         UniqueConstraint('user_id', 'permission_id'),
         _table_args_default_dict,
     )
 
     user_to_perm_id = Column(Integer(), primary_key=True)
     user_id = Column(Integer(), ForeignKey('users.user_id'), nullable=False)
     permission_id = Column(Integer(), ForeignKey('permissions.permission_id'), nullable=False)
 
     user = relationship('User')
     permission = relationship('Permission')
 
     def __unicode__(self):
         return u'<%s => %s >' % (self.user, self.permission)
 
 
 class UserGroupRepoToPerm(Base, BaseDbModel):
     __tablename__ = 'users_group_repo_to_perm'
     __table_args__ = (
         UniqueConstraint('repository_id', 'users_group_id', 'permission_id'),
         _table_args_default_dict,
     )
 
     users_group_to_perm_id = Column(Integer(), primary_key=True)
     users_group_id = Column(Integer(), ForeignKey('users_groups.users_group_id'), nullable=False)
     permission_id = Column(Integer(), ForeignKey('permissions.permission_id'), nullable=False)
     repository_id = Column(Integer(), ForeignKey('repositories.repo_id'), nullable=False)
 
     users_group = relationship('UserGroup')
     permission = relationship('Permission')
     repository = relationship('Repository')
 
     @classmethod
     def create(cls, users_group, repository, permission):
         n = cls()
         n.users_group = users_group
         n.repository = repository
         n.permission = permission
         Session().add(n)
         return n
 
     def __unicode__(self):
         return u'<UserGroupRepoToPerm:%s => %s >' % (self.users_group, self.repository)
 
 
 class UserGroupUserGroupToPerm(Base, BaseDbModel):
     __tablename__ = 'user_group_user_group_to_perm'
     __table_args__ = (
         UniqueConstraint('target_user_group_id', 'user_group_id', 'permission_id'),
         _table_args_default_dict,
     )
 
     user_group_user_group_to_perm_id = Column(Integer(), primary_key=True)
     target_user_group_id = Column(Integer(), ForeignKey('users_groups.users_group_id'), nullable=False)
     permission_id = Column(Integer(), ForeignKey('permissions.permission_id'), nullable=False)
     user_group_id = Column(Integer(), ForeignKey('users_groups.users_group_id'), nullable=False)
 
     target_user_group = relationship('UserGroup', primaryjoin='UserGroupUserGroupToPerm.target_user_group_id==UserGroup.users_group_id')
     user_group = relationship('UserGroup', primaryjoin='UserGroupUserGroupToPerm.user_group_id==UserGroup.users_group_id')
     permission = relationship('Permission')
 
     @classmethod
     def create(cls, target_user_group, user_group, permission):
         n = cls()
         n.target_user_group = target_user_group
         n.user_group = user_group
         n.permission = permission
         Session().add(n)
         return n
 
     def __unicode__(self):
         return u'<UserGroupUserGroup:%s => %s >' % (self.target_user_group, self.user_group)
 
 
 class UserGroupToPerm(Base, BaseDbModel):
     __tablename__ = 'users_group_to_perm'
     __table_args__ = (
         UniqueConstraint('users_group_id', 'permission_id',),
         _table_args_default_dict,
     )
 
     users_group_to_perm_id = Column(Integer(), primary_key=True)
     users_group_id = Column(Integer(), ForeignKey('users_groups.users_group_id'), nullable=False)
     permission_id = Column(Integer(), ForeignKey('permissions.permission_id'), nullable=False)
 
     users_group = relationship('UserGroup')
     permission = relationship('Permission')
 
 
 class UserRepoGroupToPerm(Base, BaseDbModel):
     __tablename__ = 'user_repo_group_to_perm'
     __table_args__ = (
         UniqueConstraint('user_id', 'group_id', 'permission_id'),
         _table_args_default_dict,
     )
 
     group_to_perm_id = Column(Integer(), primary_key=True)
     user_id = Column(Integer(), ForeignKey('users.user_id'), nullable=False)
     group_id = Column(Integer(), ForeignKey('groups.group_id'), nullable=False)
     permission_id = Column(Integer(), ForeignKey('permissions.permission_id'), nullable=False)
 
     user = relationship('User')
     group = relationship('RepoGroup')
     permission = relationship('Permission')
 
     @classmethod
     def create(cls, user, repository_group, permission):
         n = cls()
         n.user = user
         n.group = repository_group
         n.permission = permission
         Session().add(n)
         return n
 
 
 class UserGroupRepoGroupToPerm(Base, BaseDbModel):
     __tablename__ = 'users_group_repo_group_to_perm'
     __table_args__ = (
         UniqueConstraint('users_group_id', 'group_id'),
         _table_args_default_dict,
     )
 
     users_group_repo_group_to_perm_id = Column(Integer(), primary_key=True)
     users_group_id = Column(Integer(), ForeignKey('users_groups.users_group_id'), nullable=False)
     group_id = Column(Integer(), ForeignKey('groups.group_id'), nullable=False)
     permission_id = Column(Integer(), ForeignKey('permissions.permission_id'), nullable=False)
 
     users_group = relationship('UserGroup')
     permission = relationship('Permission')
     group = relationship('RepoGroup')
 
     @classmethod
     def create(cls, user_group, repository_group, permission):
         n = cls()
         n.users_group = user_group
         n.group = repository_group
         n.permission = permission
         Session().add(n)
         return n
 
 
 class Statistics(Base, BaseDbModel):
     __tablename__ = 'statistics'
     __table_args__ = (
          _table_args_default_dict,
     )
 
     stat_id = Column(Integer(), primary_key=True)
     repository_id = Column(Integer(), ForeignKey('repositories.repo_id'), nullable=False, unique=True)
     stat_on_revision = Column(Integer(), nullable=False)
     commit_activity = Column(LargeBinary(1000000), nullable=False) # JSON data
     commit_activity_combined = Column(LargeBinary(), nullable=False) # JSON data
     languages = Column(LargeBinary(1000000), nullable=False) # JSON data
 
     repository = relationship('Repository', single_parent=True)
 
 
 class UserFollowing(Base, BaseDbModel):
     __tablename__ = 'user_followings'
     __table_args__ = (
         UniqueConstraint('user_id', 'follows_repository_id', name='uq_user_followings_user_repo'),
         UniqueConstraint('user_id', 'follows_user_id', name='uq_user_followings_user_user'),
         _table_args_default_dict,
     )
 
     user_following_id = Column(Integer(), primary_key=True)
     user_id = Column(Integer(), ForeignKey('users.user_id'), nullable=False)
     follows_repository_id = Column(Integer(), ForeignKey('repositories.repo_id'), nullable=True)
     follows_user_id = Column(Integer(), ForeignKey('users.user_id'), nullable=True)
     follows_from = Column(DateTime(timezone=False), nullable=False, default=datetime.datetime.now)
 
     user = relationship('User', primaryjoin='User.user_id==UserFollowing.user_id')
 
     follows_user = relationship('User', primaryjoin='User.user_id==UserFollowing.follows_user_id')
     follows_repository = relationship('Repository', order_by=lambda: func.lower(Repository.repo_name))
 
     @classmethod
     def get_repo_followers(cls, repo_id):
         return cls.query().filter(cls.follows_repository_id == repo_id)
 
 
 class CacheInvalidation(Base, BaseDbModel):
     __tablename__ = 'cache_invalidation'
     __table_args__ = (
         Index('key_idx', 'cache_key'),
         _table_args_default_dict,
     )
 
     # cache_id, not used
     cache_id = Column(Integer(), primary_key=True)
     # cache_key as created by _get_cache_key
     cache_key = Column(Unicode(255), nullable=False, unique=True)
     # cache_args is a repo_name
     cache_args = Column(Unicode(255), nullable=False)
     # instance sets cache_active True when it is caching, other instances set
     # cache_active to False to indicate that this cache is invalid
     cache_active = Column(Boolean(), nullable=False, default=False)
 
     def __init__(self, cache_key, repo_name=''):
         self.cache_key = cache_key
         self.cache_args = repo_name
         self.cache_active = False
 
     def __unicode__(self):
         return u"<%s('%s:%s[%s]')>" % (
             self.__class__.__name__,
             self.cache_id, self.cache_key, self.cache_active)
 
     def _cache_key_partition(self):
         prefix, repo_name, suffix = self.cache_key.partition(self.cache_args)
         return prefix, repo_name, suffix
 
     def get_prefix(self):
         """
         get prefix that might have been used in _get_cache_key to
         generate self.cache_key. Only used for informational purposes
         in repo_edit.html.
         """
         # prefix, repo_name, suffix
         return self._cache_key_partition()[0]
 
     def get_suffix(self):
         """
         get suffix that might have been used in _get_cache_key to
         generate self.cache_key. Only used for informational purposes
         in repo_edit.html.
         """
         # prefix, repo_name, suffix
         return self._cache_key_partition()[2]
 
     @classmethod
     def clear_cache(cls):
         """
         Delete all cache keys from database.
         Should only be run when all instances are down and all entries thus stale.
         """
         cls.query().delete()
         Session().commit()
 
     @classmethod
     def _get_cache_key(cls, key):
         """
         Wrapper for generating a unique cache key for this instance and "key".
         key must / will start with a repo_name which will be stored in .cache_args .
         """
         prefix = kallithea.CONFIG.get('instance_id', '')
         return "%s%s" % (prefix, key)
 
     @classmethod
     def set_invalidate(cls, repo_name):
         """
         Mark all caches of a repo as invalid in the database.
         """
         inv_objs = Session().query(cls).filter(cls.cache_args == repo_name).all()
         log.debug('for repo %s got %s invalidation objects',
                   safe_str(repo_name), inv_objs)
 
         for inv_obj in inv_objs:
             log.debug('marking %s key for invalidation based on repo_name=%s',
                       inv_obj, safe_str(repo_name))
             Session().delete(inv_obj)
         Session().commit()
 
     @classmethod
     def test_and_set_valid(cls, repo_name, kind, valid_cache_keys=None):
         """
         Mark this cache key as active and currently cached.
         Return True if the existing cache registration still was valid.
         Return False to indicate that it had been invalidated and caches should be refreshed.
         """
 
         key = (repo_name + '_' + kind) if kind else repo_name
         cache_key = cls._get_cache_key(key)
 
         if valid_cache_keys and cache_key in valid_cache_keys:
             return True
 
         inv_obj = cls.query().filter(cls.cache_key == cache_key).scalar()
         if inv_obj is None:
             inv_obj = cls(cache_key, repo_name)
             Session().add(inv_obj)
         elif inv_obj.cache_active:
             return True
         inv_obj.cache_active = True
         try:
             Session().commit()
         except sqlalchemy.exc.IntegrityError:
             log.error('commit of CacheInvalidation failed - retrying')
             Session().rollback()
             inv_obj = cls.query().filter(cls.cache_key == cache_key).scalar()
             if inv_obj is None:
                 log.error('failed to create CacheInvalidation entry')
                 # TODO: fail badly?
             # else: TOCTOU - another thread added the key at the same time; no further action required
         return False
 
     @classmethod
     def get_valid_cache_keys(cls):
         """
         Return opaque object with information of which caches still are valid
         and can be used without checking for invalidation.
         """
         return set(inv_obj.cache_key for inv_obj in cls.query().filter(cls.cache_active).all())
 
 
 class ChangesetComment(Base, BaseDbModel):
     __tablename__ = 'changeset_comments'
     __table_args__ = (
         Index('cc_revision_idx', 'revision'),
         Index('cc_pull_request_id_idx', 'pull_request_id'),
         _table_args_default_dict,
     )
 
     comment_id = Column(Integer(), primary_key=True)
     repo_id = Column(Integer(), ForeignKey('repositories.repo_id'), nullable=False)
     revision = Column(String(40), nullable=True)
     pull_request_id = Column(Integer(), ForeignKey('pull_requests.pull_request_id'), nullable=True)
     line_no = Column(Unicode(10), nullable=True)
     f_path = Column(Unicode(1000), nullable=True)
     author_id = Column('user_id', Integer(), ForeignKey('users.user_id'), nullable=False)
     text = Column(UnicodeText(), nullable=False)
     created_on = Column(DateTime(timezone=False), nullable=False, default=datetime.datetime.now)
     modified_at = Column(DateTime(timezone=False), nullable=False, default=datetime.datetime.now)
 
     author = relationship('User')
     repo = relationship('Repository')
     # status_change is frequently used directly in templates - make it a lazy
     # join to avoid fetching each related ChangesetStatus on demand.
     # There will only be one ChangesetStatus referencing each comment so the join will not explode.
     status_change = relationship('ChangesetStatus',
                                  cascade="all, delete-orphan", lazy='joined')
     pull_request = relationship('PullRequest')
 
     def url(self):
         anchor = "comment-%s" % self.comment_id
         import kallithea.lib.helpers as h
         if self.revision:
             return h.url('changeset_home', repo_name=self.repo.repo_name, revision=self.revision, anchor=anchor)
         elif self.pull_request_id is not None:
             return self.pull_request.url(anchor=anchor)
 
     def __json__(self):
         return dict(
             comment_id=self.comment_id,
             username=self.author.username,
             text=self.text,
         )
 
     def deletable(self):
         return self.created_on > datetime.datetime.now() - datetime.timedelta(minutes=5)
 
 
 class ChangesetStatus(Base, BaseDbModel):
     __tablename__ = 'changeset_statuses'
     __table_args__ = (
         Index('cs_revision_idx', 'revision'),
         Index('cs_version_idx', 'version'),
         Index('cs_pull_request_id_idx', 'pull_request_id'),
         Index('cs_changeset_comment_id_idx', 'changeset_comment_id'),
         Index('cs_pull_request_id_user_id_version_idx', 'pull_request_id', 'user_id', 'version'),
         Index('cs_repo_id_pull_request_id_idx', 'repo_id', 'pull_request_id'),
         UniqueConstraint('repo_id', 'revision', 'version'),
         _table_args_default_dict,
     )
 
     STATUS_NOT_REVIEWED = DEFAULT = 'not_reviewed'
     STATUS_APPROVED = 'approved'
     STATUS_REJECTED = 'rejected' # is shown as "Not approved" - TODO: change database content / scheme
     STATUS_UNDER_REVIEW = 'under_review'
 
     STATUSES = [
         (STATUS_NOT_REVIEWED, _("Not reviewed")),  # (no icon) and default
         (STATUS_UNDER_REVIEW, _("Under review")),
         (STATUS_REJECTED, _("Not approved")),
         (STATUS_APPROVED, _("Approved")),
     ]
     STATUSES_DICT = dict(STATUSES)
 
     changeset_status_id = Column(Integer(), primary_key=True)
     repo_id = Column(Integer(), ForeignKey('repositories.repo_id'), nullable=False)
     user_id = Column(Integer(), ForeignKey('users.user_id'), nullable=False)
     revision = Column(String(40), nullable=True)
     status = Column(String(128), nullable=False, default=DEFAULT)
     comment_id = Column('changeset_comment_id', Integer(), ForeignKey('changeset_comments.comment_id'), nullable=False)
     modified_at = Column(DateTime(), nullable=False, default=datetime.datetime.now)
     version = Column(Integer(), nullable=False, default=0)
     pull_request_id = Column(Integer(), ForeignKey('pull_requests.pull_request_id'), nullable=True)
 
     author = relationship('User')
     repo = relationship('Repository')
     comment = relationship('ChangesetComment')
     pull_request = relationship('PullRequest')
 
     def __unicode__(self):
         return u"<%s('%s:%s')>" % (
             self.__class__.__name__,
             self.status, self.author
         )
 
     @classmethod
     def get_status_lbl(cls, value):
         return cls.STATUSES_DICT.get(value)
 
     @property
     def status_lbl(self):
         return ChangesetStatus.get_status_lbl(self.status)
 
     def __json__(self):
         return dict(
             status=self.status,
             modified_at=self.modified_at.replace(microsecond=0),
             reviewer=self.author.username,
             )
 
 
 class PullRequest(Base, BaseDbModel):
     __tablename__ = 'pull_requests'
     __table_args__ = (
         Index('pr_org_repo_id_idx', 'org_repo_id'),
         Index('pr_other_repo_id_idx', 'other_repo_id'),
         _table_args_default_dict,
     )
 
     # values for .status
     STATUS_NEW = u'new'
     STATUS_CLOSED = u'closed'
 
     pull_request_id = Column(Integer(), primary_key=True)
     title = Column(Unicode(255), nullable=False)
     description = Column(UnicodeText(), nullable=False)
     status = Column(Unicode(255), nullable=False, default=STATUS_NEW) # only for closedness, not approve/reject/etc
     created_on = Column(DateTime(timezone=False), nullable=False, default=datetime.datetime.now)
     updated_on = Column(DateTime(timezone=False), nullable=False, default=datetime.datetime.now)
     owner_id = Column('user_id', Integer(), ForeignKey('users.user_id'), nullable=False)
     _revisions = Column('revisions', UnicodeText(), nullable=False)
     org_repo_id = Column(Integer(), ForeignKey('repositories.repo_id'), nullable=False)
     org_ref = Column(Unicode(255), nullable=False)
     other_repo_id = Column(Integer(), ForeignKey('repositories.repo_id'), nullable=False)
     other_ref = Column(Unicode(255), nullable=False)
 
     @hybrid_property
     def revisions(self):
         return self._revisions.split(':')
 
     @revisions.setter
     def revisions(self, val):
         self._revisions = safe_unicode(':'.join(val))
 
     @property
     def org_ref_parts(self):
         return self.org_ref.split(':')
 
     @property
     def other_ref_parts(self):
         return self.other_ref.split(':')
 
     owner = relationship('User')
     reviewers = relationship('PullRequestReviewer',
                              cascade="all, delete-orphan")
     org_repo = relationship('Repository', primaryjoin='PullRequest.org_repo_id==Repository.repo_id')
     other_repo = relationship('Repository', primaryjoin='PullRequest.other_repo_id==Repository.repo_id')
     statuses = relationship('ChangesetStatus', order_by='ChangesetStatus.changeset_status_id')
     comments = relationship('ChangesetComment', order_by='ChangesetComment.comment_id',
                              cascade="all, delete-orphan")
 
     @classmethod
     def query(cls, reviewer_id=None, include_closed=True, sorted=False):
         """Add PullRequest-specific helpers for common query constructs.
 
         reviewer_id: only PRs with the specified user added as reviewer.
 
         include_closed: if False, do not include closed PRs.
 
         sorted: if True, apply the default ordering (newest first).
         """
         q = super(PullRequest, cls).query()
 
         if reviewer_id is not None:
             q = q.join(PullRequestReviewer).filter(PullRequestReviewer.user_id == reviewer_id)
 
         if not include_closed:
             q = q.filter(PullRequest.status != PullRequest.STATUS_CLOSED)
 
         if sorted:
             q = q.order_by(PullRequest.created_on.desc())
 
         return q
 
     def get_reviewer_users(self):
         """Like .reviewers, but actually returning the users"""
         return User.query() \
             .join(PullRequestReviewer) \
             .filter(PullRequestReviewer.pull_request == self) \
             .order_by(PullRequestReviewer.pull_request_reviewers_id) \
             .all()
 
     def is_closed(self):
         return self.status == self.STATUS_CLOSED
 
     def user_review_status(self, user_id):
         """Return the user's latest status votes on PR"""
         # note: no filtering on repo - that would be redundant
         status = ChangesetStatus.query() \
             .filter(ChangesetStatus.pull_request == self) \
             .filter(ChangesetStatus.user_id == user_id) \
             .order_by(ChangesetStatus.version) \
             .first()
         return str(status.status) if status else ''
 
     @classmethod
     def make_nice_id(cls, pull_request_id):
         '''Return pull request id nicely formatted for displaying'''
         return '#%s' % pull_request_id
 
     def nice_id(self):
         '''Return the id of this pull request, nicely formatted for displaying'''
         return self.make_nice_id(self.pull_request_id)
 
     def get_api_data(self):
         return self.__json__()
 
     def __json__(self):
         clone_uri_tmpl = kallithea.CONFIG.get('clone_uri_tmpl') or Repository.DEFAULT_CLONE_URI
         return dict(
             pull_request_id=self.pull_request_id,
             url=self.url(),
             reviewers=self.reviewers,
             revisions=self.revisions,
             owner=self.owner.username,
             title=self.title,
             description=self.description,
             org_repo_url=self.org_repo.clone_url(clone_uri_tmpl=clone_uri_tmpl),
             org_ref_parts=self.org_ref_parts,
             other_ref_parts=self.other_ref_parts,
             status=self.status,
             comments=self.comments,
             statuses=self.statuses,
         )
 
     def url(self, **kwargs):
         canonical = kwargs.pop('canonical', None)
         import kallithea.lib.helpers as h
         b = self.org_ref_parts[1]
         if b != self.other_ref_parts[1]:
             s = '/_/' + b
         else:
             s = '/_/' + self.title
         kwargs['extra'] = urlreadable(s)
         if canonical:
             return h.canonical_url('pullrequest_show', repo_name=self.other_repo.repo_name,
                                    pull_request_id=self.pull_request_id, **kwargs)
         return h.url('pullrequest_show', repo_name=self.other_repo.repo_name,
                      pull_request_id=self.pull_request_id, **kwargs)
 
 
 class PullRequestReviewer(Base, BaseDbModel):
     __tablename__ = 'pull_request_reviewers'
     __table_args__ = (
         Index('pull_request_reviewers_user_id_idx', 'user_id'),
         _table_args_default_dict,
     )
 
     def __init__(self, user=None, pull_request=None):
         self.user = user
         self.pull_request = pull_request
 
     pull_request_reviewers_id = Column('pull_requests_reviewers_id', Integer(), primary_key=True)
     pull_request_id = Column(Integer(), ForeignKey('pull_requests.pull_request_id'), nullable=False)
     user_id = Column(Integer(), ForeignKey('users.user_id'), nullable=False)
 
     user = relationship('User')
     pull_request = relationship('PullRequest')
 
     def __json__(self):
         return dict(
             username=self.user.username if self.user else None,
         )
 
 
 class Notification(object):
     __tablename__ = 'notifications'
 
 class UserNotification(object):
     __tablename__ = 'user_to_notification'
 
 
 class Gist(Base, BaseDbModel):
     __tablename__ = 'gists'
     __table_args__ = (
         Index('g_gist_access_id_idx', 'gist_access_id'),
         Index('g_created_on_idx', 'created_on'),
         _table_args_default_dict,
     )
 
     GIST_PUBLIC = u'public'
     GIST_PRIVATE = u'private'
     DEFAULT_FILENAME = u'gistfile1.txt'
 
     gist_id = Column(Integer(), primary_key=True)
     gist_access_id = Column(Unicode(250), nullable=False)
     gist_description = Column(UnicodeText(), nullable=False)
     owner_id = Column('user_id', Integer(), ForeignKey('users.user_id'), nullable=False)
     gist_expires = Column(Float(53), nullable=False)
     gist_type = Column(Unicode(128), nullable=False)
     created_on = Column(DateTime(timezone=False), nullable=False, default=datetime.datetime.now)
     modified_at = Column(DateTime(timezone=False), nullable=False, default=datetime.datetime.now)
 
     owner = relationship('User')
 
     @hybrid_property
     def is_expired(self):
         return (self.gist_expires != -1) & (time.time() > self.gist_expires)
 
     def __repr__(self):
         return '<Gist:[%s]%s>' % (self.gist_type, self.gist_access_id)
 
     @classmethod
     def guess_instance(cls, value):
         return super(Gist, cls).guess_instance(value, Gist.get_by_access_id)
 
     @classmethod
     def get_or_404(cls, id_):
         res = cls.query().filter(cls.gist_access_id == id_).scalar()
         if res is None:
             raise HTTPNotFound
         return res
 
     @classmethod
     def get_by_access_id(cls, gist_access_id):
         return cls.query().filter(cls.gist_access_id == gist_access_id).scalar()
 
     def gist_url(self):
         alias_url = kallithea.CONFIG.get('gist_alias_url')
         if alias_url:
             return alias_url.replace('{gistid}', self.gist_access_id)
 
         import kallithea.lib.helpers as h
         return h.canonical_url('gist', gist_id=self.gist_access_id)
 
     @classmethod
     def base_path(cls):
         """
         Returns base path where all gists are stored
 
         :param cls:
         """
         from kallithea.model.gist import GIST_STORE_LOC
         q = Session().query(Ui) \
             .filter(Ui.ui_key == URL_SEP)
         q = q.options(FromCache("sql_cache_short", "repository_repo_path"))
         return os.path.join(q.one().ui_value, GIST_STORE_LOC)
 
     def get_api_data(self):
         """
         Common function for generating gist related data for API
         """
         gist = self
         data = dict(
             gist_id=gist.gist_id,
             type=gist.gist_type,
             access_id=gist.gist_access_id,
             description=gist.gist_description,
             url=gist.gist_url(),
             expires=gist.gist_expires,
             created_on=gist.created_on,
         )
         return data
 
     def __json__(self):
         data = dict(
         )
         data.update(self.get_api_data())
         return data
     ## SCM functions
 
     @property
     def scm_instance(self):
         from kallithea.lib.vcs import get_repo
         base_path = self.base_path()
         return get_repo(os.path.join(*map(safe_str,
                                           [base_path, self.gist_access_id])))
 
 
 class UserSshKeys(Base, BaseDbModel):
     __tablename__ = 'user_ssh_keys'
     __table_args__ = (
         Index('usk_public_key_idx', 'public_key'),
         Index('usk_fingerprint_idx', 'fingerprint'),
         UniqueConstraint('fingerprint'),
         _table_args_default_dict
     )
     __mapper_args__ = {}
 
     user_ssh_key_id = Column(Integer(), primary_key=True)
     user_id = Column(Integer(), ForeignKey('users.user_id'), nullable=False)
     _public_key = Column('public_key', UnicodeText(), nullable=False)
     description = Column(UnicodeText(), nullable=False)
     fingerprint = Column(String(255), nullable=False, unique=True)
     created_on = Column(DateTime(timezone=False), nullable=False, default=datetime.datetime.now)
     last_seen = Column(DateTime(timezone=False), nullable=True)
 
     user = relationship('User')
 
     @property
     def public_key(self):
         return self._public_key
 
     @public_key.setter
     def public_key(self, full_key):
         # the full public key is too long to be suitable as database key - instead,
         # use fingerprints similar to 'ssh-keygen -E sha256 -lf ~/.ssh/id_rsa.pub'
         self._public_key = full_key
         enc_key = full_key.split(" ")[1]
         self.fingerprint = hashlib.sha256(enc_key.decode('base64')).digest().encode('base64').replace('\n', '').rstrip('=')

kallithea/tests/vcs/test_git.py

@@ -1,868 +1,868 b''
 import datetime
 import os
 import sys
 import urllib2
 
 import mock
 import pytest
 
 from kallithea.lib.vcs.backends.git import GitChangeset, GitRepository
 from kallithea.lib.vcs.exceptions import NodeDoesNotExistError, RepositoryError, VCSError
 from kallithea.lib.vcs.nodes import DirNode, FileNode, NodeKind, NodeState
 from kallithea.model.scm import ScmModel
 from kallithea.tests.vcs.base import _BackendTestMixin
 from kallithea.tests.vcs.conf import TEST_GIT_REPO, TEST_GIT_REPO_CLONE, TESTS_TMP_PATH, get_new_dir
 
 
 class TestGitRepository(object):
 
     def __check_for_existing_repo(self):
         if os.path.exists(TEST_GIT_REPO_CLONE):
             pytest.fail('Cannot test git clone repo as location %s already '
                       'exists. You should manually remove it first.'
                       % TEST_GIT_REPO_CLONE)
 
     def setup_method(self):
         self.repo = GitRepository(TEST_GIT_REPO)
 
     def test_wrong_repo_path(self):
         wrong_repo_path = os.path.join(TESTS_TMP_PATH, 'errorrepo')
         with pytest.raises(RepositoryError):
             GitRepository(wrong_repo_path)
 
     def test_git_cmd_injection(self):
         repo_inject_path = TEST_GIT_REPO + '; echo "Cake";'
         with pytest.raises(urllib2.URLError):
             # Should fail because URL will contain the parts after ; too
             GitRepository(get_new_dir('injection-repo'), src_url=repo_inject_path, update_after_clone=True, create=True)
 
         with pytest.raises(RepositoryError):
             # Should fail on direct clone call, which as of this writing does not happen outside of class
             clone_fail_repo = GitRepository(get_new_dir('injection-repo'), create=True)
             clone_fail_repo.clone(repo_inject_path, update_after_clone=True,)
 
         # Verify correct quoting of evil characters that should work on posix file systems
         if sys.platform == 'win32':
             # windows does not allow '"' in dir names
             # and some versions of the git client don't like ` and '
             tricky_path = get_new_dir("tricky-path-repo-$")
         else:
             tricky_path = get_new_dir("tricky-path-repo-$'\"`")
         successfully_cloned = GitRepository(tricky_path, src_url=TEST_GIT_REPO, update_after_clone=True, create=True)
         # Repo should have been created
         assert not successfully_cloned._repo.bare
 
         if sys.platform == 'win32':
             # windows does not allow '"' in dir names
             # and some versions of the git client don't like ` and '
             tricky_path_2 = get_new_dir("tricky-path-2-repo-$")
         else:
             tricky_path_2 = get_new_dir("tricky-path-2-repo-$'\"`")
         successfully_cloned2 = GitRepository(tricky_path_2, src_url=tricky_path, bare=True, create=True)
         # Repo should have been created and thus used correct quoting for clone
         assert successfully_cloned2._repo.bare
 
         # Should pass because URL has been properly quoted
         successfully_cloned.pull(tricky_path_2)
         successfully_cloned2.fetch(tricky_path)
 
     def test_repo_create_with_spaces_in_path(self):
         repo_path = get_new_dir("path with spaces")
         repo = GitRepository(repo_path, src_url=None, bare=True, create=True)
         # Repo should have been created
         assert repo._repo.bare
 
     def test_repo_clone(self):
         self.__check_for_existing_repo()
         repo = GitRepository(TEST_GIT_REPO)
         repo_clone = GitRepository(TEST_GIT_REPO_CLONE,
             src_url=TEST_GIT_REPO, create=True, update_after_clone=True)
         assert len(repo.revisions) == len(repo_clone.revisions)
         # Checking hashes of changesets should be enough
         for changeset in repo.get_changesets():
             raw_id = changeset.raw_id
             assert raw_id == repo_clone.get_changeset(raw_id).raw_id
 
     def test_repo_clone_with_spaces_in_path(self):
         repo_path = get_new_dir("path with spaces")
         successfully_cloned = GitRepository(repo_path, src_url=TEST_GIT_REPO, update_after_clone=True, create=True)
         # Repo should have been created
         assert not successfully_cloned._repo.bare
 
         successfully_cloned.pull(TEST_GIT_REPO)
         self.repo.fetch(repo_path)
 
     def test_repo_clone_without_create(self):
         with pytest.raises(RepositoryError):
             GitRepository(TEST_GIT_REPO_CLONE + '_wo_create', src_url=TEST_GIT_REPO)
 
     def test_repo_clone_with_update(self):
         repo = GitRepository(TEST_GIT_REPO)
         clone_path = TEST_GIT_REPO_CLONE + '_with_update'
         repo_clone = GitRepository(clone_path,
             create=True, src_url=TEST_GIT_REPO, update_after_clone=True)
         assert len(repo.revisions) == len(repo_clone.revisions)
 
         # check if current workdir was updated
         fpath = os.path.join(clone_path, 'MANIFEST.in')
         assert os.path.isfile(fpath) == True, 'Repo was cloned and updated but file %s could not be found' % fpath
 
     def test_repo_clone_without_update(self):
         repo = GitRepository(TEST_GIT_REPO)
         clone_path = TEST_GIT_REPO_CLONE + '_without_update'
         repo_clone = GitRepository(clone_path,
             create=True, src_url=TEST_GIT_REPO, update_after_clone=False)
         assert len(repo.revisions) == len(repo_clone.revisions)
         # check if current workdir was *NOT* updated
         fpath = os.path.join(clone_path, 'MANIFEST.in')
         # Make sure it's not bare repo
         assert not repo_clone._repo.bare
         assert os.path.isfile(fpath) == False, 'Repo was cloned and updated but file %s was found' % fpath
 
     def test_repo_clone_into_bare_repo(self):
         repo = GitRepository(TEST_GIT_REPO)
         clone_path = TEST_GIT_REPO_CLONE + '_bare.git'
         repo_clone = GitRepository(clone_path, create=True,
             src_url=repo.path, bare=True)
         assert repo_clone._repo.bare
 
     def test_create_repo_is_not_bare_by_default(self):
         repo = GitRepository(get_new_dir('not-bare-by-default'), create=True)
         assert not repo._repo.bare
 
     def test_create_bare_repo(self):
         repo = GitRepository(get_new_dir('bare-repo'), create=True, bare=True)
         assert repo._repo.bare
 
     def test_revisions(self):
         # there are 112 revisions (by now)
         # so we can assume they would be available from now on
         subset = set([
             'c1214f7e79e02fc37156ff215cd71275450cffc3',
             '38b5fe81f109cb111f549bfe9bb6b267e10bc557',
             'fa6600f6848800641328adbf7811fd2372c02ab2',
             '102607b09cdd60e2793929c4f90478be29f85a17',
             '49d3fd156b6f7db46313fac355dca1a0b94a0017',
             '2d1028c054665b962fa3d307adfc923ddd528038',
             'd7e0d30fbcae12c90680eb095a4f5f02505ce501',
             'ff7ca51e58c505fec0dd2491de52c622bb7a806b',
             'dd80b0f6cf5052f17cc738c2951c4f2070200d7f',
             '8430a588b43b5d6da365400117c89400326e7992',
             'd955cd312c17b02143c04fa1099a352b04368118',
             'f67b87e5c629c2ee0ba58f85197e423ff28d735b',
             'add63e382e4aabc9e1afdc4bdc24506c269b7618',
             'f298fe1189f1b69779a4423f40b48edf92a703fc',
             'bd9b619eb41994cac43d67cf4ccc8399c1125808',
             '6e125e7c890379446e98980d8ed60fba87d0f6d1',
             'd4a54db9f745dfeba6933bf5b1e79e15d0af20bd',
             '0b05e4ed56c802098dfc813cbe779b2f49e92500',
             '191caa5b2c81ed17c0794bf7bb9958f4dcb0b87e',
             '45223f8f114c64bf4d6f853e3c35a369a6305520',
             'ca1eb7957a54bce53b12d1a51b13452f95bc7c7e',
             'f5ea29fc42ef67a2a5a7aecff10e1566699acd68',
             '27d48942240f5b91dfda77accd2caac94708cc7d',
             '622f0eb0bafd619d2560c26f80f09e3b0b0d78af',
             'e686b958768ee96af8029fe19c6050b1a8dd3b2b'])
         assert subset.issubset(set(self.repo.revisions))
 
     def test_slicing(self):
         # 4 1 5 10 95
         for sfrom, sto, size in [(0, 4, 4), (1, 2, 1), (10, 15, 5),
                                  (10, 20, 10), (5, 100, 95)]:
             revs = list(self.repo[sfrom:sto])
             assert len(revs) == size
             assert revs[0] == self.repo.get_changeset(sfrom)
             assert revs[-1] == self.repo.get_changeset(sto - 1)
 
     def test_branches(self):
         # TODO: Need more tests here
         # Removed (those are 'remotes' branches for cloned repo)
         #assert 'master' in self.repo.branches
         #assert 'gittree' in self.repo.branches
         #assert 'web-branch' in self.repo.branches
         for name, id in self.repo.branches.items():
             assert isinstance(self.repo.get_changeset(id), GitChangeset)
 
     def test_tags(self):
         # TODO: Need more tests here
         assert 'v0.1.1' in self.repo.tags
         assert 'v0.1.2' in self.repo.tags
         for name, id in self.repo.tags.items():
             assert isinstance(self.repo.get_changeset(id), GitChangeset)
 
     def _test_single_changeset_cache(self, revision):
         chset = self.repo.get_changeset(revision)
         assert revision in self.repo.changesets
         assert chset is self.repo.changesets[revision]
 
     def test_initial_changeset(self):
         id = self.repo.revisions[0]
         init_chset = self.repo.get_changeset(id)
         assert init_chset.message == 'initial import\n'
         assert init_chset.author == 'Marcin Kuzminski <[e-mail]>'
         for path in ('vcs/__init__.py',
                      'vcs/backends/BaseRepository.py',
                      'vcs/backends/__init__.py'):
             assert isinstance(init_chset.get_node(path), FileNode)
         for path in ('', 'vcs', 'vcs/backends'):
             assert isinstance(init_chset.get_node(path), DirNode)
 
         with pytest.raises(NodeDoesNotExistError):
             init_chset.get_node(path='foobar')
 
         node = init_chset.get_node('vcs/')
         assert hasattr(node, 'kind')
         assert node.kind == NodeKind.DIR
 
         node = init_chset.get_node('vcs')
         assert hasattr(node, 'kind')
         assert node.kind == NodeKind.DIR
 
         node = init_chset.get_node('vcs/__init__.py')
         assert hasattr(node, 'kind')
         assert node.kind == NodeKind.FILE
 
     def test_not_existing_changeset(self):
         with pytest.raises(RepositoryError):
             self.repo.get_changeset('f' * 40)
 
     def test_changeset10(self):
 
         chset10 = self.repo.get_changeset(self.repo.revisions[9])
         readme = """===
 VCS
 ===
 
 Various Version Control System management abstraction layer for Python.
 
 Introduction
 ------------
 
 TODO: To be written...
 
 """
         node = chset10.get_node('README.rst')
         assert node.kind == NodeKind.FILE
         assert node.content == readme
 
 
 class TestGitChangeset(object):
 
     def setup_method(self):
         self.repo = GitRepository(TEST_GIT_REPO)
 
     def test_default_changeset(self):
         tip = self.repo.get_changeset()
         assert tip == self.repo.get_changeset(None)
         assert tip == self.repo.get_changeset('tip')
 
     def test_root_node(self):
         tip = self.repo.get_changeset()
         assert tip.root is tip.get_node('')
 
     def test_lazy_fetch(self):
         """
         Test if changeset's nodes expands and are cached as we walk through
         the revision. This test is somewhat hard to write as order of tests
         is a key here. Written by running command after command in a shell.
         """
         commit_id = '2a13f185e4525f9d4b59882791a2d397b90d5ddc'
         assert commit_id in self.repo.revisions
         chset = self.repo.get_changeset(commit_id)
         assert len(chset.nodes) == 0
         root = chset.root
         assert len(chset.nodes) == 1
         assert len(root.nodes) == 8
         # accessing root.nodes updates chset.nodes
         assert len(chset.nodes) == 9
 
         docs = root.get_node('docs')
         # we haven't yet accessed anything new as docs dir was already cached
         assert len(chset.nodes) == 9
         assert len(docs.nodes) == 8
         # accessing docs.nodes updates chset.nodes
         assert len(chset.nodes) == 17
 
         assert docs is chset.get_node('docs')
         assert docs is root.nodes[0]
         assert docs is root.dirs[0]
         assert docs is chset.get_node('docs')
 
     def test_nodes_with_changeset(self):
         commit_id = '2a13f185e4525f9d4b59882791a2d397b90d5ddc'
         chset = self.repo.get_changeset(commit_id)
         root = chset.root
         docs = root.get_node('docs')
         assert docs is chset.get_node('docs')
         api = docs.get_node('api')
         assert api is chset.get_node('docs/api')
         index = api.get_node('index.rst')
         assert index is chset.get_node('docs/api/index.rst')
         assert index is chset.get_node('docs') \
                              .get_node('api') \
                              .get_node('index.rst')
 
     def test_branch_and_tags(self):
         # Those tests seem to show wrong results:
         # in Git, only heads have a branch - most changesets don't
         rev0 = self.repo.revisions[0]
         chset0 = self.repo.get_changeset(rev0)
         assert chset0.branch is None # should be 'master'?
         assert chset0.branches == [] # should be 'master'?
         assert chset0.tags == []
 
         rev10 = self.repo.revisions[10]
         chset10 = self.repo.get_changeset(rev10)
         assert chset10.branch is None # should be 'master'?
         assert chset10.branches == [] # should be 'master'?
         assert chset10.tags == []
 
         rev44 = self.repo.revisions[44]
         chset44 = self.repo.get_changeset(rev44)
         assert chset44.branch is None # should be 'web-branch'?
         assert chset44.branches == [] # should be 'web-branch'?
 
         tip = self.repo.get_changeset('tip')
         assert 'tip' not in tip.tags # it should be?
         assert not tip.tags # how it is!
 
     def _test_slices(self, limit, offset):
         count = self.repo.count()
         changesets = self.repo.get_changesets(limit=limit, offset=offset)
         idx = 0
         for changeset in changesets:
             rev = offset + idx
             idx += 1
             rev_id = self.repo.revisions[rev]
             if idx > limit:
                 pytest.fail("Exceeded limit already (getting revision %s, "
                     "there are %s total revisions, offset=%s, limit=%s)"
                     % (rev_id, count, offset, limit))
             assert changeset == self.repo.get_changeset(rev_id)
         result = list(self.repo.get_changesets(limit=limit, offset=offset))
         start = offset
         end = limit and offset + limit or None
         sliced = list(self.repo[start:end])
         pytest.failUnlessEqual(result, sliced,
             msg="Comparison failed for limit=%s, offset=%s"
             "(get_changeset returned: %s and sliced: %s"
             % (limit, offset, result, sliced))
 
     def _test_file_size(self, revision, path, size):
         node = self.repo.get_changeset(revision).get_node(path)
         assert node.is_file()
         assert node.size == size
 
     def test_file_size(self):
         to_check = (
             ('c1214f7e79e02fc37156ff215cd71275450cffc3',
                 'vcs/backends/BaseRepository.py', 502),
             ('d7e0d30fbcae12c90680eb095a4f5f02505ce501',
                 'vcs/backends/hg.py', 854),
             ('6e125e7c890379446e98980d8ed60fba87d0f6d1',
                 'setup.py', 1068),
             ('d955cd312c17b02143c04fa1099a352b04368118',
                 'vcs/backends/base.py', 2921),
             ('ca1eb7957a54bce53b12d1a51b13452f95bc7c7e',
                 'vcs/backends/base.py', 3936),
             ('f50f42baeed5af6518ef4b0cb2f1423f3851a941',
                 'vcs/backends/base.py', 6189),
         )
         for revision, path, size in to_check:
             self._test_file_size(revision, path, size)
 
     def _test_dir_size(self, revision, path, size):
         node = self.repo.get_changeset(revision).get_node(path)
         assert node.size == size
 
     def test_dir_size(self):
         to_check = (
             ('5f2c6ee195929b0be80749243c18121c9864a3b3', '/', 674076),
             ('7ab37bc680b4aa72c34d07b230c866c28e9fc204', '/', 674049),
             ('6892503fb8f2a552cef5f4d4cc2cdbd13ae1cd2f', '/', 671830),
         )
         for revision, path, size in to_check:
             self._test_dir_size(revision, path, size)
 
     def test_repo_size(self):
         assert self.repo.size == 674076
 
     def test_file_history(self):
         # we can only check if those revisions are present in the history
         # as we cannot update this test every time file is changed
         files = {
             'setup.py': [
                 '54386793436c938cff89326944d4c2702340037d',
                 '51d254f0ecf5df2ce50c0b115741f4cf13985dab',
                 '998ed409c795fec2012b1c0ca054d99888b22090',
                 '5e0eb4c47f56564395f76333f319d26c79e2fb09',
                 '0115510b70c7229dbc5dc49036b32e7d91d23acd',
                 '7cb3fd1b6d8c20ba89e2264f1c8baebc8a52d36e',
                 '2a13f185e4525f9d4b59882791a2d397b90d5ddc',
                 '191caa5b2c81ed17c0794bf7bb9958f4dcb0b87e',
                 'ff7ca51e58c505fec0dd2491de52c622bb7a806b',
             ],
             'vcs/nodes.py': [
                 '33fa3223355104431402a888fa77a4e9956feb3e',
                 'fa014c12c26d10ba682fadb78f2a11c24c8118e1',
                 'e686b958768ee96af8029fe19c6050b1a8dd3b2b',
                 'ab5721ca0a081f26bf43d9051e615af2cc99952f',
                 'c877b68d18e792a66b7f4c529ea02c8f80801542',
                 '4313566d2e417cb382948f8d9d7c765330356054',
                 '6c2303a793671e807d1cfc70134c9ca0767d98c2',
                 '54386793436c938cff89326944d4c2702340037d',
                 '54000345d2e78b03a99d561399e8e548de3f3203',
                 '1c6b3677b37ea064cb4b51714d8f7498f93f4b2b',
                 '2d03ca750a44440fb5ea8b751176d1f36f8e8f46',
                 '2a08b128c206db48c2f0b8f70df060e6db0ae4f8',
                 '30c26513ff1eb8e5ce0e1c6b477ee5dc50e2f34b',
                 'ac71e9503c2ca95542839af0ce7b64011b72ea7c',
                 '12669288fd13adba2a9b7dd5b870cc23ffab92d2',
                 '5a0c84f3e6fe3473e4c8427199d5a6fc71a9b382',
                 '12f2f5e2b38e6ff3fbdb5d722efed9aa72ecb0d5',
                 '5eab1222a7cd4bfcbabc218ca6d04276d4e27378',
                 'f50f42baeed5af6518ef4b0cb2f1423f3851a941',
                 'd7e390a45f6aa96f04f5e7f583ad4f867431aa25',
                 'f15c21f97864b4f071cddfbf2750ec2e23859414',
                 'e906ef056cf539a4e4e5fc8003eaf7cf14dd8ade',
                 'ea2b108b48aa8f8c9c4a941f66c1a03315ca1c3b',
                 '84dec09632a4458f79f50ddbbd155506c460b4f9',
                 '0115510b70c7229dbc5dc49036b32e7d91d23acd',
                 '2a13f185e4525f9d4b59882791a2d397b90d5ddc',
                 '3bf1c5868e570e39569d094f922d33ced2fa3b2b',
                 'b8d04012574729d2c29886e53b1a43ef16dd00a1',
                 '6970b057cffe4aab0a792aa634c89f4bebf01441',
                 'dd80b0f6cf5052f17cc738c2951c4f2070200d7f',
                 'ff7ca51e58c505fec0dd2491de52c622bb7a806b',
             ],
             'vcs/backends/git.py': [
                 '4cf116ad5a457530381135e2f4c453e68a1b0105',
                 '9a751d84d8e9408e736329767387f41b36935153',
                 'cb681fb539c3faaedbcdf5ca71ca413425c18f01',
                 '428f81bb652bcba8d631bce926e8834ff49bdcc6',
                 '180ab15aebf26f98f714d8c68715e0f05fa6e1c7',
                 '2b8e07312a2e89e92b90426ab97f349f4bce2a3a',
                 '50e08c506174d8645a4bb517dd122ac946a0f3bf',
                 '54000345d2e78b03a99d561399e8e548de3f3203',
             ],
         }
         for path, revs in files.items():
             node = self.repo.get_changeset(revs[0]).get_node(path)
             node_revs = [chset.raw_id for chset in node.history]
             assert set(revs).issubset(set(node_revs)), "We assumed that %s is subset of revisions for which file %s " \
                 "has been changed, and history of that node returned: %s" \
                 % (revs, path, node_revs)
 
     def test_file_annotate(self):
         files = {
             'vcs/backends/__init__.py': {
                 'c1214f7e79e02fc37156ff215cd71275450cffc3': {
                     'lines_no': 1,
                     'changesets': [
                         'c1214f7e79e02fc37156ff215cd71275450cffc3',
                     ],
                 },
                 '16fba1ae9334d79b66d7afed2c2dfbfa2ae53647': {
                     'lines_no': 21,
                     'changesets': [
                         '49d3fd156b6f7db46313fac355dca1a0b94a0017',
                         '16fba1ae9334d79b66d7afed2c2dfbfa2ae53647',
                         '16fba1ae9334d79b66d7afed2c2dfbfa2ae53647',
                         '16fba1ae9334d79b66d7afed2c2dfbfa2ae53647',
                         '16fba1ae9334d79b66d7afed2c2dfbfa2ae53647',
                         '16fba1ae9334d79b66d7afed2c2dfbfa2ae53647',
                         '16fba1ae9334d79b66d7afed2c2dfbfa2ae53647',
                         '16fba1ae9334d79b66d7afed2c2dfbfa2ae53647',
                         '16fba1ae9334d79b66d7afed2c2dfbfa2ae53647',
                         '16fba1ae9334d79b66d7afed2c2dfbfa2ae53647',
                         '16fba1ae9334d79b66d7afed2c2dfbfa2ae53647',
                         '16fba1ae9334d79b66d7afed2c2dfbfa2ae53647',
                         '16fba1ae9334d79b66d7afed2c2dfbfa2ae53647',
                         '16fba1ae9334d79b66d7afed2c2dfbfa2ae53647',
                         '16fba1ae9334d79b66d7afed2c2dfbfa2ae53647',
                         '16fba1ae9334d79b66d7afed2c2dfbfa2ae53647',
                         '16fba1ae9334d79b66d7afed2c2dfbfa2ae53647',
                         '16fba1ae9334d79b66d7afed2c2dfbfa2ae53647',
                         '16fba1ae9334d79b66d7afed2c2dfbfa2ae53647',
                         '16fba1ae9334d79b66d7afed2c2dfbfa2ae53647',
                         '16fba1ae9334d79b66d7afed2c2dfbfa2ae53647',
                     ],
                 },
                 'e29b67bd158580fc90fc5e9111240b90e6e86064': {
                     'lines_no': 32,
                     'changesets': [
                         '16fba1ae9334d79b66d7afed2c2dfbfa2ae53647',
                         '16fba1ae9334d79b66d7afed2c2dfbfa2ae53647',
                         '5eab1222a7cd4bfcbabc218ca6d04276d4e27378',
                         '16fba1ae9334d79b66d7afed2c2dfbfa2ae53647',
                         '16fba1ae9334d79b66d7afed2c2dfbfa2ae53647',
                         '992f38217b979d0b0987d0bae3cc26dac85d9b19',
                         '16fba1ae9334d79b66d7afed2c2dfbfa2ae53647',
                         '54000345d2e78b03a99d561399e8e548de3f3203',
                         '16fba1ae9334d79b66d7afed2c2dfbfa2ae53647',
                         '16fba1ae9334d79b66d7afed2c2dfbfa2ae53647',
                         '78c3f0c23b7ee935ec276acb8b8212444c33c396',
                         '992f38217b979d0b0987d0bae3cc26dac85d9b19',
                         '992f38217b979d0b0987d0bae3cc26dac85d9b19',
                         '992f38217b979d0b0987d0bae3cc26dac85d9b19',
                         '992f38217b979d0b0987d0bae3cc26dac85d9b19',
                         '2a13f185e4525f9d4b59882791a2d397b90d5ddc',
                         '992f38217b979d0b0987d0bae3cc26dac85d9b19',
                         '78c3f0c23b7ee935ec276acb8b8212444c33c396',
                         '992f38217b979d0b0987d0bae3cc26dac85d9b19',
                         '992f38217b979d0b0987d0bae3cc26dac85d9b19',
                         '992f38217b979d0b0987d0bae3cc26dac85d9b19',
                         '16fba1ae9334d79b66d7afed2c2dfbfa2ae53647',
                         '16fba1ae9334d79b66d7afed2c2dfbfa2ae53647',
                         '16fba1ae9334d79b66d7afed2c2dfbfa2ae53647',
                         '16fba1ae9334d79b66d7afed2c2dfbfa2ae53647',
                         '992f38217b979d0b0987d0bae3cc26dac85d9b19',
                         '16fba1ae9334d79b66d7afed2c2dfbfa2ae53647',
                         '992f38217b979d0b0987d0bae3cc26dac85d9b19',
                         '992f38217b979d0b0987d0bae3cc26dac85d9b19',
                         '16fba1ae9334d79b66d7afed2c2dfbfa2ae53647',
                         '16fba1ae9334d79b66d7afed2c2dfbfa2ae53647',
                         '16fba1ae9334d79b66d7afed2c2dfbfa2ae53647',
                     ],
                 },
             },
         }
 
         for fname, revision_dict in files.items():
             for rev, data in revision_dict.items():
                 cs = self.repo.get_changeset(rev)
 
                 l1_1 = [x[1] for x in cs.get_file_annotate(fname)]
                 l1_2 = [x[2]().raw_id for x in cs.get_file_annotate(fname)]
                 assert l1_1 == l1_2
                 l1 = l1_1
                 l2 = files[fname][rev]['changesets']
                 assert l1 == l2, "The lists of revision for %s@rev %s" \
                     "from annotation list should match each other, " \
                     "got \n%s \nvs \n%s " % (fname, rev, l1, l2)
 
     def test_files_state(self):
         """
         Tests state of FileNodes.
         """
         node = self.repo \
             .get_changeset('e6ea6d16e2f26250124a1f4b4fe37a912f9d86a0') \
             .get_node('vcs/utils/diffs.py')
         assert node.state, NodeState.ADDED
         assert node.added
         assert not node.changed
         assert not node.not_changed
         assert not node.removed
 
         node = self.repo \
             .get_changeset('33fa3223355104431402a888fa77a4e9956feb3e') \
             .get_node('.hgignore')
         assert node.state, NodeState.CHANGED
         assert not node.added
         assert node.changed
         assert not node.not_changed
         assert not node.removed
 
         node = self.repo \
             .get_changeset('e29b67bd158580fc90fc5e9111240b90e6e86064') \
             .get_node('setup.py')
         assert node.state, NodeState.NOT_CHANGED
         assert not node.added
         assert not node.changed
         assert node.not_changed
         assert not node.removed
 
         # If node has REMOVED state then trying to fetch it would raise
         # ChangesetError exception
         chset = self.repo.get_changeset(
             'fa6600f6848800641328adbf7811fd2372c02ab2')
         path = 'vcs/backends/BaseRepository.py'
         with pytest.raises(NodeDoesNotExistError):
             chset.get_node(path)
         # but it would be one of ``removed`` (changeset's attribute)
         assert path in [rf.path for rf in chset.removed]
 
         chset = self.repo.get_changeset(
             '54386793436c938cff89326944d4c2702340037d')
         changed = ['setup.py', 'tests/test_nodes.py', 'vcs/backends/hg.py',
             'vcs/nodes.py']
         assert set(changed) == set([f.path for f in chset.changed])
 
     def test_commit_message_is_unicode(self):
         for cs in self.repo:
-            assert type(cs.message) == unicode
+            assert isinstance(cs.message, unicode)
 
     def test_changeset_author_is_unicode(self):
         for cs in self.repo:
-            assert type(cs.author) == unicode
+            assert isinstance(cs.author, unicode)
 
     def test_repo_files_content_is_unicode(self):
         changeset = self.repo.get_changeset()
         for node in changeset.get_node('/'):
             if node.is_file():
-                assert type(node.content) == unicode
+                assert isinstance(node.content, unicode)
 
     def test_wrong_path(self):
         # There is 'setup.py' in the root dir but not there:
         path = 'foo/bar/setup.py'
         tip = self.repo.get_changeset()
         with pytest.raises(VCSError):
             tip.get_node(path)
 
     def test_author_email(self):
         assert '[e-mail]' == self.repo.get_changeset('c1214f7e79e02fc37156ff215cd71275450cffc3').author_email
         assert '[e-mail]' == self.repo.get_changeset('ff7ca51e58c505fec0dd2491de52c622bb7a806b').author_email
         assert '' == self.repo.get_changeset('8430a588b43b5d6da365400117c89400326e7992').author_email
 
     def test_author_username(self):
         assert 'Marcin Kuzminski' == self.repo.get_changeset('c1214f7e79e02fc37156ff215cd71275450cffc3').author_name
         assert 'Lukasz Balcerzak' == self.repo.get_changeset('ff7ca51e58c505fec0dd2491de52c622bb7a806b').author_name
         assert 'marcink none@none' == self.repo.get_changeset('8430a588b43b5d6da365400117c89400326e7992').author_name
 
 
 class TestGitSpecific():
 
     def test_error_is_raised_for_added_if_diff_name_status_is_wrong(self):
         repo = mock.MagicMock()
         changeset = GitChangeset(repo, 'foobar')
         changeset._diff_name_status = 'foobar'
         with pytest.raises(VCSError):
             changeset.added
 
     def test_error_is_raised_for_changed_if_diff_name_status_is_wrong(self):
         repo = mock.MagicMock()
         changeset = GitChangeset(repo, 'foobar')
         changeset._diff_name_status = 'foobar'
         with pytest.raises(VCSError):
             changeset.added
 
     def test_error_is_raised_for_removed_if_diff_name_status_is_wrong(self):
         repo = mock.MagicMock()
         changeset = GitChangeset(repo, 'foobar')
         changeset._diff_name_status = 'foobar'
         with pytest.raises(VCSError):
             changeset.added
 
 
 class TestGitSpecificWithRepo(_BackendTestMixin):
     backend_alias = 'git'
 
     @classmethod
     def _get_commits(cls):
         return [
             {
                 'message': 'Initial',
                 'author': 'Joe Doe <[e-mail]>',
                 'date': datetime.datetime(2010, 1, 1, 20),
                 'added': [
                     FileNode('foobar/static/js/admin/base.js', content='base'),
                     FileNode('foobar/static/admin', content='admin',
                         mode=0120000), # this is a link
                     FileNode('foo', content='foo'),
                 ],
             },
             {
                 'message': 'Second',
                 'author': 'Joe Doe <[e-mail]>',
                 'date': datetime.datetime(2010, 1, 1, 22),
                 'added': [
                     FileNode('foo2', content='foo2'),
                 ],
             },
         ]
 
     def test_paths_slow_traversing(self):
         cs = self.repo.get_changeset()
         assert cs.get_node('foobar').get_node('static').get_node('js').get_node('admin').get_node('base.js').content == 'base'
 
     def test_paths_fast_traversing(self):
         cs = self.repo.get_changeset()
         assert cs.get_node('foobar/static/js/admin/base.js').content == 'base'
 
     def test_workdir_get_branch(self):
         self.repo.run_git_command(['checkout', '-b', 'production'])
         # Regression test: one of following would fail if we don't check
         # .git/HEAD file
         self.repo.run_git_command(['checkout', 'production'])
         assert self.repo.workdir.get_branch() == 'production'
         self.repo.run_git_command(['checkout', 'master'])
         assert self.repo.workdir.get_branch() == 'master'
 
     def test_get_diff_runs_git_command_with_hashes(self):
         self.repo.run_git_command = mock.Mock(return_value=['', ''])
         self.repo.get_diff(0, 1)
         self.repo.run_git_command.assert_called_once_with(
             ['diff', '-U3', '--full-index', '--binary', '-p', '-M', '--abbrev=40',
              self.repo._get_revision(0), self.repo._get_revision(1)])
 
     def test_get_diff_runs_git_command_with_str_hashes(self):
         self.repo.run_git_command = mock.Mock(return_value=['', ''])
         self.repo.get_diff(self.repo.EMPTY_CHANGESET, 1)
         self.repo.run_git_command.assert_called_once_with(
             ['show', '-U3', '--full-index', '--binary', '-p', '-M', '--abbrev=40',
              self.repo._get_revision(1)])
 
     def test_get_diff_runs_git_command_with_path_if_its_given(self):
         self.repo.run_git_command = mock.Mock(return_value=['', ''])
         self.repo.get_diff(0, 1, 'foo')
         self.repo.run_git_command.assert_called_once_with(
             ['diff', '-U3', '--full-index', '--binary', '-p', '-M', '--abbrev=40',
              self.repo._get_revision(0), self.repo._get_revision(1), '--', 'foo'])
 
     def test_get_diff_does_not_sanitize_valid_context(self):
         almost_overflowed_long_int = 2**31-1
 
         self.repo.run_git_command = mock.Mock(return_value=['', ''])
         self.repo.get_diff(0, 1, 'foo', context=almost_overflowed_long_int)
         self.repo.run_git_command.assert_called_once_with(
             ['diff', '-U' + str(almost_overflowed_long_int), '--full-index', '--binary', '-p', '-M', '--abbrev=40',
              self.repo._get_revision(0), self.repo._get_revision(1), '--', 'foo'])
 
     def test_get_diff_sanitizes_overflowing_context(self):
         overflowed_long_int = 2**31
         sanitized_overflowed_long_int = overflowed_long_int-1
 
         self.repo.run_git_command = mock.Mock(return_value=['', ''])
         self.repo.get_diff(0, 1, 'foo', context=overflowed_long_int)
 
         self.repo.run_git_command.assert_called_once_with(
             ['diff', '-U' + str(sanitized_overflowed_long_int), '--full-index', '--binary', '-p', '-M', '--abbrev=40',
              self.repo._get_revision(0), self.repo._get_revision(1), '--', 'foo'])
 
     def test_get_diff_does_not_sanitize_zero_context(self):
         zero_context = 0
 
         self.repo.run_git_command = mock.Mock(return_value=['', ''])
         self.repo.get_diff(0, 1, 'foo', context=zero_context)
 
         self.repo.run_git_command.assert_called_once_with(
             ['diff', '-U' + str(zero_context), '--full-index', '--binary', '-p', '-M', '--abbrev=40',
              self.repo._get_revision(0), self.repo._get_revision(1), '--', 'foo'])
 
     def test_get_diff_sanitizes_negative_context(self):
         negative_context = -10
 
         self.repo.run_git_command = mock.Mock(return_value=['', ''])
         self.repo.get_diff(0, 1, 'foo', context=negative_context)
 
         self.repo.run_git_command.assert_called_once_with(
             ['diff', '-U0', '--full-index', '--binary', '-p', '-M', '--abbrev=40',
              self.repo._get_revision(0), self.repo._get_revision(1), '--', 'foo'])
 
 
 class TestGitRegression(_BackendTestMixin):
     backend_alias = 'git'
 
     @classmethod
     def _get_commits(cls):
         return [
             {
                 'message': 'Initial',
                 'author': 'Joe Doe <[e-mail]>',
                 'date': datetime.datetime(2010, 1, 1, 20),
                 'added': [
                     FileNode('bot/__init__.py', content='base'),
                     FileNode('bot/templates/404.html', content='base'),
                     FileNode('bot/templates/500.html', content='base'),
                 ],
             },
             {
                 'message': 'Second',
                 'author': 'Joe Doe <[e-mail]>',
                 'date': datetime.datetime(2010, 1, 1, 22),
                 'added': [
                     FileNode('bot/build/migrations/1.py', content='foo2'),
                     FileNode('bot/build/migrations/2.py', content='foo2'),
                     FileNode('bot/build/static/templates/f.html', content='foo2'),
                     FileNode('bot/build/static/templates/f1.html', content='foo2'),
                     FileNode('bot/build/templates/err.html', content='foo2'),
                     FileNode('bot/build/templates/err2.html', content='foo2'),
                 ],
             },
         ]
 
     def test_similar_paths(self):
         cs = self.repo.get_changeset()
         paths = lambda *n: [x.path for x in n]
         assert paths(*cs.get_nodes('bot')) == ['bot/build', 'bot/templates', 'bot/__init__.py']
         assert paths(*cs.get_nodes('bot/build')) == ['bot/build/migrations', 'bot/build/static', 'bot/build/templates']
         assert paths(*cs.get_nodes('bot/build/static')) == ['bot/build/static/templates']
         # this get_nodes below causes troubles !
         assert paths(*cs.get_nodes('bot/build/static/templates')) == ['bot/build/static/templates/f.html', 'bot/build/static/templates/f1.html']
         assert paths(*cs.get_nodes('bot/build/templates')) == ['bot/build/templates/err.html', 'bot/build/templates/err2.html']
         assert paths(*cs.get_nodes('bot/templates/')) == ['bot/templates/404.html', 'bot/templates/500.html']
 
 
 class TestGitHooks(object):
     """
     Tests related to hook functionality of Git repositories.
     """
 
     def setup_method(self):
         # For each run we want a fresh repo.
         self.repo_directory = get_new_dir("githookrepo")
         self.repo = GitRepository(self.repo_directory, create=True)
 
         # Create a dictionary where keys are hook names, and values are paths to
         # them. Deduplicates code in tests a bit.
         self.hook_directory = self.repo.get_hook_location()
         self.kallithea_hooks = dict((h, os.path.join(self.hook_directory, h)) for h in ("pre-receive", "post-receive"))
 
     def test_hooks_created_if_missing(self):
         """
         Tests if hooks are installed in repository if they are missing.
         """
 
         for hook, hook_path in self.kallithea_hooks.iteritems():
             if os.path.exists(hook_path):
                 os.remove(hook_path)
 
         ScmModel().install_git_hooks(repo=self.repo)
 
         for hook, hook_path in self.kallithea_hooks.iteritems():
             assert os.path.exists(hook_path)
 
     def test_kallithea_hooks_updated(self):
         """
         Tests if hooks are updated if they are Kallithea hooks already.
         """
 
         for hook, hook_path in self.kallithea_hooks.iteritems():
             with open(hook_path, "w") as f:
                 f.write("KALLITHEA_HOOK_VER=0.0.0\nJUST_BOGUS")
 
         ScmModel().install_git_hooks(repo=self.repo)
 
         for hook, hook_path in self.kallithea_hooks.iteritems():
             with open(hook_path) as f:
                 assert "JUST_BOGUS" not in f.read()
 
     def test_custom_hooks_untouched(self):
         """
         Tests if hooks are left untouched if they are not Kallithea hooks.
         """
 
         for hook, hook_path in self.kallithea_hooks.iteritems():
             with open(hook_path, "w") as f:
                 f.write("#!/bin/bash\n#CUSTOM_HOOK")
 
         ScmModel().install_git_hooks(repo=self.repo)
 
         for hook, hook_path in self.kallithea_hooks.iteritems():
             with open(hook_path) as f:
                 assert "CUSTOM_HOOK" in f.read()
 
     def test_custom_hooks_forced_update(self):
         """
         Tests if hooks are forcefully updated even though they are custom hooks.
         """
 
         for hook, hook_path in self.kallithea_hooks.iteritems():
             with open(hook_path, "w") as f:
                 f.write("#!/bin/bash\n#CUSTOM_HOOK")
 
         ScmModel().install_git_hooks(repo=self.repo, force_create=True)
 
         for hook, hook_path in self.kallithea_hooks.iteritems():
             with open(hook_path) as f:
                 assert "KALLITHEA_HOOK_VER" in f.read()

kallithea/tests/vcs/test_hg.py

@@ -1,592 +1,592 b''
 import os
 
 import mock
 import pytest
 
 from kallithea.lib.utils2 import safe_str
 from kallithea.lib.vcs.backends.hg import MercurialChangeset, MercurialRepository
 from kallithea.lib.vcs.exceptions import NodeDoesNotExistError, RepositoryError, VCSError
 from kallithea.lib.vcs.nodes import NodeKind, NodeState
 from kallithea.tests.vcs.conf import TEST_HG_REPO, TEST_HG_REPO_CLONE, TEST_HG_REPO_PULL, TESTS_TMP_PATH
 
 
 class TestMercurialRepository(object):
 
     def __check_for_existing_repo(self):
         if os.path.exists(TEST_HG_REPO_CLONE):
             pytest.fail('Cannot test mercurial clone repo as location %s already '
                       'exists. You should manually remove it first.'
                       % TEST_HG_REPO_CLONE)
 
     def setup_method(self):
         self.repo = MercurialRepository(safe_str(TEST_HG_REPO))
 
     def test_wrong_repo_path(self):
         wrong_repo_path = os.path.join(TESTS_TMP_PATH, 'errorrepo')
         with pytest.raises(RepositoryError):
             MercurialRepository(wrong_repo_path)
 
     def test_unicode_path_repo(self):
         with pytest.raises(VCSError):
             MercurialRepository(u'iShouldFail')
 
     def test_repo_clone(self):
         self.__check_for_existing_repo()
         repo = MercurialRepository(safe_str(TEST_HG_REPO))
         repo_clone = MercurialRepository(TEST_HG_REPO_CLONE,
             src_url=TEST_HG_REPO, update_after_clone=True)
         assert len(repo.revisions) == len(repo_clone.revisions)
         # Checking hashes of changesets should be enough
         for changeset in repo.get_changesets():
             raw_id = changeset.raw_id
             assert raw_id == repo_clone.get_changeset(raw_id).raw_id
 
     def test_repo_clone_with_update(self):
         repo = MercurialRepository(safe_str(TEST_HG_REPO))
         repo_clone = MercurialRepository(TEST_HG_REPO_CLONE + '_w_update',
             src_url=TEST_HG_REPO, update_after_clone=True)
         assert len(repo.revisions) == len(repo_clone.revisions)
 
         # check if current workdir was updated
         assert os.path.isfile(
             os.path.join(
                 TEST_HG_REPO_CLONE + '_w_update', 'MANIFEST.in'
             )
         )
 
     def test_repo_clone_without_update(self):
         repo = MercurialRepository(safe_str(TEST_HG_REPO))
         repo_clone = MercurialRepository(TEST_HG_REPO_CLONE + '_wo_update',
             src_url=TEST_HG_REPO, update_after_clone=False)
         assert len(repo.revisions) == len(repo_clone.revisions)
         assert not os.path.isfile(
             os.path.join(
                 TEST_HG_REPO_CLONE + '_wo_update', 'MANIFEST.in'
             )
         )
 
     def test_pull(self):
         if os.path.exists(TEST_HG_REPO_PULL):
             pytest.fail('Cannot test mercurial pull command as location %s '
                       'already exists. You should manually remove it first'
                       % TEST_HG_REPO_PULL)
         repo_new = MercurialRepository(TEST_HG_REPO_PULL, create=True)
         assert len(self.repo.revisions) > len(repo_new.revisions)
 
         repo_new.pull(self.repo.path)
         repo_new = MercurialRepository(TEST_HG_REPO_PULL)
         assert len(self.repo.revisions) == len(repo_new.revisions)
 
     def test_revisions(self):
         # there are 21 revisions at bitbucket now
         # so we can assume they would be available from now on
         subset = set(['b986218ba1c9b0d6a259fac9b050b1724ed8e545',
                  '3d8f361e72ab303da48d799ff1ac40d5ac37c67e',
                  '6cba7170863a2411822803fa77a0a264f1310b35',
                  '56349e29c2af3ac913b28bde9a2c6154436e615b',
                  '2dda4e345facb0ccff1a191052dd1606dba6781d',
                  '6fff84722075f1607a30f436523403845f84cd9e',
                  '7d4bc8ec6be56c0f10425afb40b6fc315a4c25e7',
                  '3803844fdbd3b711175fc3da9bdacfcd6d29a6fb',
                  'dc5d2c0661b61928834a785d3e64a3f80d3aad9c',
                  'be90031137367893f1c406e0a8683010fd115b79',
                  'db8e58be770518cbb2b1cdfa69146e47cd481481',
                  '84478366594b424af694a6c784cb991a16b87c21',
                  '17f8e105dddb9f339600389c6dc7175d395a535c',
                  '20a662e756499bde3095ffc9bc0643d1def2d0eb',
                  '2e319b85e70a707bba0beff866d9f9de032aa4f9',
                  '786facd2c61deb9cf91e9534735124fb8fc11842',
                  '94593d2128d38210a2fcd1aabff6dda0d6d9edf8',
                  'aa6a0de05b7612707db567078e130a6cd114a9a7',
                  'eada5a770da98ab0dd7325e29d00e0714f228d09'
                 ])
         assert subset.issubset(set(self.repo.revisions))
 
         # check if we have the proper order of revisions
         org = ['b986218ba1c9b0d6a259fac9b050b1724ed8e545',
                 '3d8f361e72ab303da48d799ff1ac40d5ac37c67e',
                 '6cba7170863a2411822803fa77a0a264f1310b35',
                 '56349e29c2af3ac913b28bde9a2c6154436e615b',
                 '2dda4e345facb0ccff1a191052dd1606dba6781d',
                 '6fff84722075f1607a30f436523403845f84cd9e',
                 '7d4bc8ec6be56c0f10425afb40b6fc315a4c25e7',
                 '3803844fdbd3b711175fc3da9bdacfcd6d29a6fb',
                 'dc5d2c0661b61928834a785d3e64a3f80d3aad9c',
                 'be90031137367893f1c406e0a8683010fd115b79',
                 'db8e58be770518cbb2b1cdfa69146e47cd481481',
                 '84478366594b424af694a6c784cb991a16b87c21',
                 '17f8e105dddb9f339600389c6dc7175d395a535c',
                 '20a662e756499bde3095ffc9bc0643d1def2d0eb',
                 '2e319b85e70a707bba0beff866d9f9de032aa4f9',
                 '786facd2c61deb9cf91e9534735124fb8fc11842',
                 '94593d2128d38210a2fcd1aabff6dda0d6d9edf8',
                 'aa6a0de05b7612707db567078e130a6cd114a9a7',
                 'eada5a770da98ab0dd7325e29d00e0714f228d09',
                 '2c1885c735575ca478bf9e17b0029dca68824458',
                 'd9bcd465040bf869799b09ad732c04e0eea99fe9',
                 '469e9c847fe1f6f7a697b8b25b4bc5b48780c1a7',
                 '4fb8326d78e5120da2c7468dcf7098997be385da',
                 '62b4a097164940bd66030c4db51687f3ec035eed',
                 '536c1a19428381cfea92ac44985304f6a8049569',
                 '965e8ab3c44b070cdaa5bf727ddef0ada980ecc4',
                 '9bb326a04ae5d98d437dece54be04f830cf1edd9',
                 'f8940bcb890a98c4702319fbe36db75ea309b475',
                 'ff5ab059786ebc7411e559a2cc309dfae3625a3b',
                 '6b6ad5f82ad5bb6190037671bd254bd4e1f4bf08',
                 'ee87846a61c12153b51543bf860e1026c6d3dcba', ]
         assert org == self.repo.revisions[:31]
 
     def test_iter_slice(self):
         sliced = list(self.repo[:10])
         itered = list(self.repo)[:10]
         assert sliced == itered
 
     def test_slicing(self):
         # 4 1 5 10 95
         for sfrom, sto, size in [(0, 4, 4), (1, 2, 1), (10, 15, 5),
                                  (10, 20, 10), (5, 100, 95)]:
             revs = list(self.repo[sfrom:sto])
             assert len(revs) == size
             assert revs[0] == self.repo.get_changeset(sfrom)
             assert revs[-1] == self.repo.get_changeset(sto - 1)
 
     def test_branches(self):
         # TODO: Need more tests here
 
         # active branches
         assert 'default' in self.repo.branches
         assert 'stable' in self.repo.branches
 
         # closed
         assert 'git' in self.repo._get_branches(closed=True)
         assert 'web' in self.repo._get_branches(closed=True)
 
         for name, id in self.repo.branches.items():
             assert isinstance(self.repo.get_changeset(id), MercurialChangeset)
 
     def test_tip_in_tags(self):
         # tip is always a tag
         assert 'tip' in self.repo.tags
 
     def test_tip_changeset_in_tags(self):
         tip = self.repo.get_changeset()
         assert self.repo.tags['tip'] == tip.raw_id
 
     def test_initial_changeset(self):
 
         init_chset = self.repo.get_changeset(0)
         assert init_chset.message == 'initial import'
         assert init_chset.author == 'Marcin Kuzminski <[e-mail]>'
         assert sorted(init_chset._file_paths) == sorted([
             'vcs/__init__.py',
             'vcs/backends/BaseRepository.py',
             'vcs/backends/__init__.py',
         ])
 
         assert sorted(init_chset._dir_paths) == sorted(['', 'vcs', 'vcs/backends'])
 
         with pytest.raises(NodeDoesNotExistError):
             init_chset.get_node(path='foobar')
 
         node = init_chset.get_node('vcs/')
         assert hasattr(node, 'kind')
         assert node.kind == NodeKind.DIR
 
         node = init_chset.get_node('vcs')
         assert hasattr(node, 'kind')
         assert node.kind == NodeKind.DIR
 
         node = init_chset.get_node('vcs/__init__.py')
         assert hasattr(node, 'kind')
         assert node.kind == NodeKind.FILE
 
     def test_not_existing_changeset(self):
         # rawid
         with pytest.raises(RepositoryError):
             self.repo.get_changeset('abcd' * 10)
         # shortid
         with pytest.raises(RepositoryError):
             self.repo.get_changeset('erro' * 4)
         # numeric
         with pytest.raises(RepositoryError):
             self.repo.get_changeset(self.repo.count() + 1)
 
         # Small chance we ever get to this one
         revision = pow(2, 30)
         with pytest.raises(RepositoryError):
             self.repo.get_changeset(revision)
 
     def test_changeset10(self):
 
         chset10 = self.repo.get_changeset(10)
         readme = """===
 VCS
 ===
 
 Various Version Control System management abstraction layer for Python.
 
 Introduction
 ------------
 
 TODO: To be written...
 
 """
         node = chset10.get_node('README.rst')
         assert node.kind == NodeKind.FILE
         assert node.content == readme
 
     @mock.patch('kallithea.lib.vcs.backends.hg.repository.diffopts')
     def test_get_diff_does_not_sanitize_zero_context(self, mock_diffopts):
         zero_context = 0
 
         self.repo.get_diff(0, 1, 'foo', context=zero_context)
 
         mock_diffopts.assert_called_once_with(git=True, showfunc=True, ignorews=False, context=zero_context)
 
     @mock.patch('kallithea.lib.vcs.backends.hg.repository.diffopts')
     def test_get_diff_sanitizes_negative_context(self, mock_diffopts):
         negative_context = -10
         zero_context = 0
 
         self.repo.get_diff(0, 1, 'foo', context=negative_context)
 
         mock_diffopts.assert_called_once_with(git=True, showfunc=True, ignorews=False, context=zero_context)
 
 
 class TestMercurialChangeset(object):
 
     def setup_method(self):
         self.repo = MercurialRepository(safe_str(TEST_HG_REPO))
 
     def _test_equality(self, changeset):
         revision = changeset.revision
         assert changeset == self.repo.get_changeset(revision)
 
     def test_equality(self):
         revs = [0, 10, 20]
         changesets = [self.repo.get_changeset(rev) for rev in revs]
         for changeset in changesets:
             self._test_equality(changeset)
 
     def test_default_changeset(self):
         tip = self.repo.get_changeset('tip')
         assert tip == self.repo.get_changeset()
         assert tip == self.repo.get_changeset(revision=None)
         assert tip == list(self.repo[-1:])[0]
 
     def test_root_node(self):
         tip = self.repo.get_changeset('tip')
         assert tip.root is tip.get_node('')
 
     def test_lazy_fetch(self):
         """
         Test if changeset's nodes expands and are cached as we walk through
         the revision. This test is somewhat hard to write as order of tests
         is a key here. Written by running command after command in a shell.
         """
         chset = self.repo.get_changeset(45)
         assert len(chset.nodes) == 0
         root = chset.root
         assert len(chset.nodes) == 1
         assert len(root.nodes) == 8
         # accessing root.nodes updates chset.nodes
         assert len(chset.nodes) == 9
 
         docs = root.get_node('docs')
         # we haven't yet accessed anything new as docs dir was already cached
         assert len(chset.nodes) == 9
         assert len(docs.nodes) == 8
         # accessing docs.nodes updates chset.nodes
         assert len(chset.nodes) == 17
 
         assert docs is chset.get_node('docs')
         assert docs is root.nodes[0]
         assert docs is root.dirs[0]
         assert docs is chset.get_node('docs')
 
     def test_nodes_with_changeset(self):
         chset = self.repo.get_changeset(45)
         root = chset.root
         docs = root.get_node('docs')
         assert docs is chset.get_node('docs')
         api = docs.get_node('api')
         assert api is chset.get_node('docs/api')
         index = api.get_node('index.rst')
         assert index is chset.get_node('docs/api/index.rst')
         assert index is chset.get_node('docs').get_node('api').get_node('index.rst')
 
     def test_branch_and_tags(self):
         chset0 = self.repo.get_changeset(0)
         assert chset0.branch == 'default'
         assert chset0.branches == ['default']
         assert chset0.tags == []
 
         chset10 = self.repo.get_changeset(10)
         assert chset10.branch == 'default'
         assert chset10.branches == ['default']
         assert chset10.tags == []
 
         chset44 = self.repo.get_changeset(44)
         assert chset44.branch == 'web'
         assert chset44.branches == ['web']
 
         tip = self.repo.get_changeset('tip')
         assert 'tip' in tip.tags
 
     def _test_file_size(self, revision, path, size):
         node = self.repo.get_changeset(revision).get_node(path)
         assert node.is_file()
         assert node.size == size
 
     def test_file_size(self):
         to_check = (
             (10, 'setup.py', 1068),
             (20, 'setup.py', 1106),
             (60, 'setup.py', 1074),
 
             (10, 'vcs/backends/base.py', 2921),
             (20, 'vcs/backends/base.py', 3936),
             (60, 'vcs/backends/base.py', 6189),
         )
         for revision, path, size in to_check:
             self._test_file_size(revision, path, size)
 
     def _test_dir_size(self, revision, path, size):
         node = self.repo.get_changeset(revision).get_node(path)
         assert not node.is_file()
         assert node.size == size
 
     def test_dir_size(self):
         to_check = (
             ('96507bd11ecc', '/', 682421),
             ('a53d9201d4bc', '/', 682410),
             ('90243de06161', '/', 682006),
         )
         for revision, path, size in to_check:
             self._test_dir_size(revision, path, size)
 
     def test_repo_size(self):
         assert self.repo.size == 682421
 
     def test_file_history(self):
         # we can only check if those revisions are present in the history
         # as we cannot update this test every time file is changed
         files = {
             'setup.py': [7, 18, 45, 46, 47, 69, 77],
             'vcs/nodes.py': [7, 8, 24, 26, 30, 45, 47, 49, 56, 57, 58, 59, 60,
                 61, 73, 76],
             'vcs/backends/hg.py': [4, 5, 6, 11, 12, 13, 14, 15, 16, 21, 22, 23,
                 26, 27, 28, 30, 31, 33, 35, 36, 37, 38, 39, 40, 41, 44, 45, 47,
                 48, 49, 53, 54, 55, 58, 60, 61, 67, 68, 69, 70, 73, 77, 78, 79,
                 82],
         }
         for path, revs in files.items():
             tip = self.repo.get_changeset(revs[-1])
             node = tip.get_node(path)
             node_revs = [chset.revision for chset in node.history]
             assert set(revs).issubset(set(node_revs)), \
                 "We assumed that %s is subset of revisions for which file %s " \
                 "has been changed, and history of that node returned: %s" \
                 % (revs, path, node_revs)
 
     def test_file_annotate(self):
         files = {
                  'vcs/backends/__init__.py':
                   {89: {'lines_no': 31,
                         'changesets': [32, 32, 61, 32, 32, 37, 32, 32, 32, 44,
                                        37, 37, 37, 37, 45, 37, 44, 37, 37, 37,
                                        32, 32, 32, 32, 37, 32, 37, 37, 32,
                                        32, 32]},
                    20: {'lines_no': 1,
                         'changesets': [4]},
                    55: {'lines_no': 31,
                         'changesets': [32, 32, 45, 32, 32, 37, 32, 32, 32, 44,
                                        37, 37, 37, 37, 45, 37, 44, 37, 37, 37,
                                        32, 32, 32, 32, 37, 32, 37, 37, 32,
                                        32, 32]}},
                  'vcs/exceptions.py':
                  {89: {'lines_no': 18,
                        'changesets': [16, 16, 16, 16, 16, 16, 16, 16, 16, 16,
                                       16, 16, 17, 16, 16, 18, 18, 18]},
                   20: {'lines_no': 18,
                        'changesets': [16, 16, 16, 16, 16, 16, 16, 16, 16, 16,
                                       16, 16, 17, 16, 16, 18, 18, 18]},
                   55: {'lines_no': 18, 'changesets': [16, 16, 16, 16, 16, 16,
                                                       16, 16, 16, 16, 16, 16,
                                                       17, 16, 16, 18, 18, 18]}},
                  'MANIFEST.in': {89: {'lines_no': 5,
                                       'changesets': [7, 7, 7, 71, 71]},
                                  20: {'lines_no': 3,
                                       'changesets': [7, 7, 7]},
                                  55: {'lines_no': 3,
                                      'changesets': [7, 7, 7]}}}
 
         for fname, revision_dict in files.items():
             for rev, data in revision_dict.items():
                 cs = self.repo.get_changeset(rev)
                 l1_1 = [x[1] for x in cs.get_file_annotate(fname)]
                 l1_2 = [x[2]().raw_id for x in cs.get_file_annotate(fname)]
                 assert l1_1 == l1_2
                 l1 = l1_2 = [x[2]().revision for x in cs.get_file_annotate(fname)]
                 l2 = files[fname][rev]['changesets']
                 assert l1 == l2, "The lists of revision for %s@rev%s" \
                     "from annotation list should match each other," \
                     "got \n%s \nvs \n%s " % (fname, rev, l1, l2)
 
     def test_changeset_state(self):
         """
         Tests which files have been added/changed/removed at particular revision
         """
 
         # rev 46ad32a4f974:
         # hg st --rev 46ad32a4f974
         #    changed: 13
         #    added:   20
         #    removed: 1
         changed = set(['.hgignore'
             , 'README.rst', 'docs/conf.py', 'docs/index.rst', 'setup.py'
             , 'tests/test_hg.py', 'tests/test_nodes.py', 'vcs/__init__.py'
             , 'vcs/backends/__init__.py', 'vcs/backends/base.py'
             , 'vcs/backends/hg.py', 'vcs/nodes.py', 'vcs/utils/__init__.py'])
 
         added = set(['docs/api/backends/hg.rst'
             , 'docs/api/backends/index.rst', 'docs/api/index.rst'
             , 'docs/api/nodes.rst', 'docs/api/web/index.rst'
             , 'docs/api/web/simplevcs.rst', 'docs/installation.rst'
             , 'docs/quickstart.rst', 'setup.cfg', 'vcs/utils/baseui_config.py'
             , 'vcs/utils/web.py', 'vcs/web/__init__.py', 'vcs/web/exceptions.py'
             , 'vcs/web/simplevcs/__init__.py', 'vcs/web/simplevcs/exceptions.py'
             , 'vcs/web/simplevcs/middleware.py', 'vcs/web/simplevcs/models.py'
             , 'vcs/web/simplevcs/settings.py', 'vcs/web/simplevcs/utils.py'
             , 'vcs/web/simplevcs/views.py'])
 
         removed = set(['docs/api.rst'])
 
         chset64 = self.repo.get_changeset('46ad32a4f974')
         assert set((node.path for node in chset64.added)) == added
         assert set((node.path for node in chset64.changed)) == changed
         assert set((node.path for node in chset64.removed)) == removed
 
         # rev b090f22d27d6:
         # hg st --rev b090f22d27d6
         #    changed: 13
         #    added:   20
         #    removed: 1
         chset88 = self.repo.get_changeset('b090f22d27d6')
         assert set((node.path for node in chset88.added)) == set()
         assert set((node.path for node in chset88.changed)) == set(['.hgignore'])
         assert set((node.path for node in chset88.removed)) == set()
 
         # 85:
         #    added:   2 ['vcs/utils/diffs.py', 'vcs/web/simplevcs/views/diffs.py']
         #    changed: 4 ['vcs/web/simplevcs/models.py', ...]
         #    removed: 1 ['vcs/utils/web.py']
         chset85 = self.repo.get_changeset(85)
         assert set((node.path for node in chset85.added)) == set([
             'vcs/utils/diffs.py',
             'vcs/web/simplevcs/views/diffs.py'
         ])
 
         assert set((node.path for node in chset85.changed)) == set([
             'vcs/web/simplevcs/models.py',
             'vcs/web/simplevcs/utils.py',
             'vcs/web/simplevcs/views/__init__.py',
             'vcs/web/simplevcs/views/repository.py',
         ])
 
         assert set((node.path for node in chset85.removed)) == set([
             'vcs/utils/web.py'
         ])
 
 
     def test_files_state(self):
         """
         Tests state of FileNodes.
         """
         chset = self.repo.get_changeset(85)
         node = chset.get_node('vcs/utils/diffs.py')
         assert node.state, NodeState.ADDED
         assert node.added
         assert not node.changed
         assert not node.not_changed
         assert not node.removed
 
         chset = self.repo.get_changeset(88)
         node = chset.get_node('.hgignore')
         assert node.state, NodeState.CHANGED
         assert not node.added
         assert node.changed
         assert not node.not_changed
         assert not node.removed
 
         chset = self.repo.get_changeset(85)
         node = chset.get_node('setup.py')
         assert node.state, NodeState.NOT_CHANGED
         assert not node.added
         assert not node.changed
         assert node.not_changed
         assert not node.removed
 
         # If node has REMOVED state then trying to fetch it would raise
         # ChangesetError exception
         chset = self.repo.get_changeset(2)
         path = 'vcs/backends/BaseRepository.py'
         with pytest.raises(NodeDoesNotExistError):
             chset.get_node(path)
         # but it would be one of ``removed`` (changeset's attribute)
         assert path in [rf.path for rf in chset.removed]
 
     def test_commit_message_is_unicode(self):
         for cm in self.repo:
-            assert type(cm.message) == unicode
+            assert isinstance(cm.message, unicode)
 
     def test_changeset_author_is_unicode(self):
         for cm in self.repo:
-            assert type(cm.author) == unicode
+            assert isinstance(cm.author, unicode)
 
     def test_repo_files_content_is_unicode(self):
         test_changeset = self.repo.get_changeset(100)
         for node in test_changeset.get_node('/'):
             if node.is_file():
-                assert type(node.content) == unicode
+                assert isinstance(node.content, unicode)
 
     def test_wrong_path(self):
         # There is 'setup.py' in the root dir but not there:
         path = 'foo/bar/setup.py'
         with pytest.raises(VCSError):
             self.repo.get_changeset().get_node(path)
 
     def test_archival_file(self):
         # TODO:
         pass
 
     def test_archival_as_generator(self):
         # TODO:
         pass
 
     def test_archival_wrong_kind(self):
         tip = self.repo.get_changeset()
         with pytest.raises(VCSError):
             tip.fill_archive(kind='error')
 
     def test_archival_empty_prefix(self):
         # TODO:
         pass
 
     def test_author_email(self):
         assert '[e-mail]' == self.repo.get_changeset('b986218ba1c9').author_email
         assert '[e-mail]' == self.repo.get_changeset('3803844fdbd3').author_email
         assert '' == self.repo.get_changeset('84478366594b').author_email
 
     def test_author_username(self):
         assert 'Marcin Kuzminski' == self.repo.get_changeset('b986218ba1c9').author_name
         assert 'Lukasz Balcerzak' == self.repo.get_changeset('3803844fdbd3').author_name
         assert 'marcink' == self.repo.get_changeset('84478366594b').author_name
 
     def test_successors(self):
         init_chset = self.repo.get_changeset(0)
         assert init_chset.successors == []
 
     def test_predecessors(self):
         init_chset = self.repo.get_changeset(0)
         assert len(init_chset.predecessors) == 0