[devtools] change the way we start/stop postgresql
Instead of having the test db handler start a cluster on demand, use the
test module's setUp/tearDown callbacks to do it. This allows to have
one data directory (and thus cluster) per test module, allowing
different test modules to run in parallel, each using its own database
cluster whose path is based on the test module.
# copyright 2003-2012 LOGILAB S.A. (Paris, FRANCE), all rights reserved.
# contact http://www.logilab.fr/ -- mailto:contact@logilab.fr
#
# This file is part of CubicWeb.
#
# CubicWeb is free software: you can redistribute it and/or modify it under the
# terms of the GNU Lesser General Public License as published by the Free
# Software Foundation, either version 2.1 of the License, or (at your option)
# any later version.
#
# CubicWeb 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 Lesser General Public License for more
# details.
#
# You should have received a copy of the GNU Lesser General Public License along
# with CubicWeb. If not, see <http://www.gnu.org/licenses/>.
"""Base class for request/session"""
__docformat__ = "restructuredtext en"
from warnings import warn
from urlparse import urlsplit, urlunsplit
from urllib import quote as urlquote, unquote as urlunquote
from datetime import time, datetime, timedelta
from cgi import parse_qs, parse_qsl
from logilab.common.decorators import cached
from logilab.common.deprecation import deprecated
from logilab.common.date import ustrftime, strptime, todate, todatetime
from rql.utils import rqlvar_maker
from cubicweb import (Unauthorized, NoSelectableObject, NoResultError,
MultipleResultsError, uilib)
from cubicweb.rset import ResultSet
ONESECOND = timedelta(0, 1, 0)
CACHE_REGISTRY = {}
class FindEntityError(Exception):
"""raised when find_one_entity() can not return one and only one entity"""
class Cache(dict):
def __init__(self):
super(Cache, self).__init__()
_now = datetime.now()
self.cache_creation_date = _now
self.latest_cache_lookup = _now
class RequestSessionBase(object):
"""base class containing stuff shared by server session and web request
request/session is the main resources accessor, mainly through it's vreg
attribute:
:attribute vreg: the instance's registry
:attribute vreg.schema: the instance's schema
:attribute vreg.config: the instance's configuration
"""
is_request = True # False for repository session
def __init__(self, vreg):
self.vreg = vreg
try:
encoding = vreg.property_value('ui.encoding')
except Exception: # no vreg or property not registered
encoding = 'utf-8'
self.encoding = encoding
# cache result of execution for (rql expr / eids),
# should be emptied on commit/rollback of the server session / web
# connection
self.user = None
self.local_perm_cache = {}
self._ = unicode
def _set_user(self, orig_user):
"""set the user for this req_session_base
A special method is needed to ensure the linked user is linked to the
connection too.
"""
rset = self.eid_rset(orig_user.eid, 'CWUser')
user_cls = self.vreg['etypes'].etype_class('CWUser')
user = user_cls(self, rset, row=0, groups=orig_user.groups,
properties=orig_user.properties)
user.cw_attr_cache['login'] = orig_user.login # cache login
self.user = user
self.set_entity_cache(user)
self.set_language(user.prefered_language())
def set_language(self, lang):
"""install i18n configuration for `lang` translation.
Raises :exc:`KeyError` if translation doesn't exist.
"""
self.lang = lang
gettext, pgettext = self.vreg.config.translations[lang]
# use _cw.__ to translate a message without registering it to the catalog
self._ = self.__ = gettext
self.pgettext = pgettext
def get_option_value(self, option):
raise NotImplementedError
def property_value(self, key):
"""return value of the property with the given key, giving priority to
user specific value if any, else using site value
"""
if self.user:
val = self.user.property_value(key)
if val is not None:
return val
return self.vreg.property_value(key)
def etype_rset(self, etype, size=1):
"""return a fake result set for a particular entity type"""
rset = ResultSet([('A',)]*size, '%s X' % etype,
description=[(etype,)]*size)
def get_entity(row, col=0, etype=etype, req=self, rset=rset):
return req.vreg['etypes'].etype_class(etype)(req, rset, row, col)
rset.get_entity = get_entity
rset.req = self
return rset
def eid_rset(self, eid, etype=None):
"""return a result set for the given eid without doing actual query
(we have the eid, we can suppose it exists and user has access to the
entity)
"""
eid = int(eid)
if etype is None:
etype = self.entity_metas(eid)['type']
rset = ResultSet([(eid,)], 'Any X WHERE X eid %(x)s', {'x': eid},
[(etype,)])
rset.req = self
return rset
def empty_rset(self):
""" return a guaranteed empty result """
rset = ResultSet([], 'Any X WHERE X eid -1')
rset.req = self
return rset
def entity_from_eid(self, eid, etype=None):
"""return an entity instance for the given eid. No query is done"""
try:
return self.entity_cache(eid)
except KeyError:
rset = self.eid_rset(eid, etype)
entity = rset.get_entity(0, 0)
self.set_entity_cache(entity)
return entity
def entity_cache(self, eid):
raise KeyError
def set_entity_cache(self, entity):
pass
def create_entity(self, etype, **kwargs):
"""add a new entity of the given type
Example (in a shell session):
>>> c = create_entity('Company', name=u'Logilab')
>>> create_entity('Person', firstname=u'John', surname=u'Doe',
... works_for=c)
"""
cls = self.vreg['etypes'].etype_class(etype)
return cls.cw_instantiate(self.execute, **kwargs)
@deprecated('[3.18] use find(etype, **kwargs).entities()')
def find_entities(self, etype, **kwargs):
"""find entities of the given type and attribute values.
>>> users = find_entities('CWGroup', name=u'users')
>>> groups = find_entities('CWGroup')
"""
return self.find(etype, **kwargs).entities()
@deprecated('[3.18] use find(etype, **kwargs).one()')
def find_one_entity(self, etype, **kwargs):
"""find one entity of the given type and attribute values.
raise :exc:`FindEntityError` if can not return one and only one entity.
>>> users = find_one_entity('CWGroup', name=u'users')
>>> groups = find_one_entity('CWGroup')
Exception()
"""
try:
return self.find(etype, **kwargs).one()
except (NoResultError, MultipleResultsError) as e:
raise FindEntityError("%s: (%s, %s)" % (str(e), etype, kwargs))
def find(self, etype, **kwargs):
"""find entities of the given type and attribute values.
:returns: A :class:`ResultSet`
>>> users = find('CWGroup', name=u"users").one()
>>> groups = find('CWGroup').entities()
"""
parts = ['Any X WHERE X is %s' % etype]
varmaker = rqlvar_maker(defined='X')
eschema = self.vreg.schema.eschema(etype)
for attr, value in kwargs.items():
if isinstance(value, list) or isinstance(value, tuple):
raise NotImplementedError("List of values are not supported")
if hasattr(value, 'eid'):
kwargs[attr] = value.eid
if attr.startswith('reverse_'):
attr = attr[8:]
assert attr in eschema.objrels, \
'%s not in %s object relations' % (attr, eschema)
parts.append(
'%(varname)s %(attr)s X, '
'%(varname)s eid %%(reverse_%(attr)s)s'
% {'attr': attr, 'varname': varmaker.next()})
else:
assert attr in eschema.subjrels, \
'%s not in %s subject relations' % (attr, eschema)
parts.append('X %(attr)s %%(%(attr)s)s' % {'attr': attr})
rql = ', '.join(parts)
return self.execute(rql, kwargs)
def ensure_ro_rql(self, rql):
"""raise an exception if the given rql is not a select query"""
first = rql.split(None, 1)[0].lower()
if first in ('insert', 'set', 'delete'):
raise Unauthorized(self._('only select queries are authorized'))
def get_cache(self, cachename):
"""cachename should be dotted names as in :
- cubicweb.mycache
- cubes.blog.mycache
- etc.
"""
warn.warning('[3.19] .get_cache will disappear soon. '
'Distributed caching mechanisms are being introduced instead.'
'Other caching mechanism can be used more reliably '
'to the same effect.',
DeprecationWarning)
if cachename in CACHE_REGISTRY:
cache = CACHE_REGISTRY[cachename]
else:
cache = CACHE_REGISTRY[cachename] = Cache()
_now = datetime.now()
if _now > cache.latest_cache_lookup + ONESECOND:
ecache = self.execute(
'Any C,T WHERE C is CWCache, C name %(name)s, C timestamp T',
{'name':cachename}).get_entity(0,0)
cache.latest_cache_lookup = _now
if not ecache.valid(cache.cache_creation_date):
cache.clear()
cache.cache_creation_date = _now
return cache
# url generation methods ##################################################
def build_url(self, *args, **kwargs):
"""return an absolute URL using params dictionary key/values as URL
parameters. Values are automatically URL quoted, and the
publishing method to use may be specified or will be guessed.
if ``__secure__`` argument is True, the request will try to build a
https url.
raises :exc:`ValueError` if None is found in arguments
"""
# use *args since we don't want first argument to be "anonymous" to
# avoid potential clash with kwargs
method = None
if args:
assert len(args) == 1, 'only 0 or 1 non-named-argument expected'
method = args[0]
if method is None:
method = 'view'
# XXX I (adim) think that if method is passed explicitly, we should
# not try to process it and directly call req.build_url()
base_url = kwargs.pop('base_url', None)
if base_url is None:
secure = kwargs.pop('__secure__', None)
base_url = self.base_url(secure=secure)
if '_restpath' in kwargs:
assert method == 'view', repr(method)
path = kwargs.pop('_restpath')
else:
path = method
if not kwargs:
return u'%s%s' % (base_url, path)
return u'%s%s?%s' % (base_url, path, self.build_url_params(**kwargs))
def build_url_params(self, **kwargs):
"""return encoded params to incorporate them in a URL"""
args = []
for param, values in kwargs.iteritems():
if not isinstance(values, (list, tuple)):
values = (values,)
for value in values:
assert value is not None
args.append(u'%s=%s' % (param, self.url_quote(value)))
return '&'.join(args)
def url_quote(self, value, safe=''):
"""urllib.quote is not unicode safe, use this method to do the
necessary encoding / decoding. Also it's designed to quote each
part of a url path and so the '/' character will be encoded as well.
"""
if isinstance(value, unicode):
quoted = urlquote(value.encode(self.encoding), safe=safe)
return unicode(quoted, self.encoding)
return urlquote(str(value), safe=safe)
def url_unquote(self, quoted):
"""returns a unicode unquoted string
decoding is based on `self.encoding` which is the encoding
used in `url_quote`
"""
if isinstance(quoted, unicode):
quoted = quoted.encode(self.encoding)
try:
return unicode(urlunquote(quoted), self.encoding)
except UnicodeDecodeError: # might occurs on manually typed URLs
return unicode(urlunquote(quoted), 'iso-8859-1')
def url_parse_qsl(self, querystring):
"""return a list of (key, val) found in the url quoted query string"""
if isinstance(querystring, unicode):
querystring = querystring.encode(self.encoding)
for key, val in parse_qsl(querystring):
try:
yield unicode(key, self.encoding), unicode(val, self.encoding)
except UnicodeDecodeError: # might occurs on manually typed URLs
yield unicode(key, 'iso-8859-1'), unicode(val, 'iso-8859-1')
def rebuild_url(self, url, **newparams):
"""return the given url with newparams inserted. If any new params
is already specified in the url, it's overriden by the new value
newparams may only be mono-valued.
"""
if isinstance(url, unicode):
url = url.encode(self.encoding)
schema, netloc, path, query, fragment = urlsplit(url)
query = parse_qs(query)
# sort for testing predictability
for key, val in sorted(newparams.iteritems()):
query[key] = (self.url_quote(val),)
query = '&'.join(u'%s=%s' % (param, value)
for param, values in sorted(query.items())
for value in values)
return urlunsplit((schema, netloc, path, query, fragment))
# bound user related methods ###############################################
@cached
def user_data(self):
"""returns a dictionary with this user's information.
The keys are :
login
The user login
name
The user name, returned by user.name()
email
The user principal email
"""
userinfo = {}
user = self.user
userinfo['login'] = user.login
userinfo['name'] = user.name()
userinfo['email'] = user.cw_adapt_to('IEmailable').get_email()
return userinfo
# formating methods #######################################################
def view(self, __vid, rset=None, __fallback_oid=None, __registry='views',
initargs=None, w=None, **kwargs):
"""Select object with the given id (`__oid`) then render it. If the
object isn't selectable, try to select fallback object if
`__fallback_oid` is specified.
If specified `initargs` is expected to be a dictionary containing
arguments that should be given to selection (hence to object's __init__
as well), but not to render(). Other arbitrary keyword arguments will be
given to selection *and* to render(), and so should be handled by
object's call or cell_call method..
"""
if initargs is None:
initargs = kwargs
else:
initargs.update(kwargs)
try:
view = self.vreg[__registry].select(__vid, self, rset=rset, **initargs)
except NoSelectableObject:
if __fallback_oid is None:
raise
view = self.vreg[__registry].select(__fallback_oid, self,
rset=rset, **initargs)
return view.render(w=w, **kwargs)
def printable_value(self, attrtype, value, props=None, displaytime=True,
formatters=uilib.PRINTERS):
"""return a displayablye value (i.e. unicode string)"""
if value is None:
return u''
try:
as_string = formatters[attrtype]
except KeyError:
self.error('given bad attrtype %s', attrtype)
return unicode(value)
return as_string(value, self, props, displaytime)
def format_date(self, date, date_format=None, time=False):
"""return a string for a date time according to instance's
configuration
"""
if date is not None:
if date_format is None:
if time:
date_format = self.property_value('ui.datetime-format')
else:
date_format = self.property_value('ui.date-format')
return ustrftime(date, date_format)
return u''
def format_time(self, time):
"""return a string for a time according to instance's
configuration
"""
if time is not None:
return ustrftime(time, self.property_value('ui.time-format'))
return u''
def format_float(self, num):
"""return a string for floating point number according to instance's
configuration
"""
if num is not None:
return self.property_value('ui.float-format') % num
return u''
def parse_datetime(self, value, etype='Datetime'):
"""get a datetime or time from a string (according to etype)
Datetime formatted as Date are accepted
"""
assert etype in ('Datetime', 'Date', 'Time'), etype
# XXX raise proper validation error
if etype == 'Datetime':
format = self.property_value('ui.datetime-format')
try:
return todatetime(strptime(value, format))
except ValueError:
pass
elif etype == 'Time':
format = self.property_value('ui.time-format')
try:
# (adim) I can't find a way to parse a Time with a custom format
date = strptime(value, format) # this returns a DateTime
return time(date.hour, date.minute, date.second)
except ValueError:
raise ValueError(self._('can\'t parse %(value)r (expected %(format)s)')
% {'value': value, 'format': format})
try:
format = self.property_value('ui.date-format')
dt = strptime(value, format)
if etype == 'Datetime':
return todatetime(dt)
return todate(dt)
except ValueError:
raise ValueError(self._('can\'t parse %(value)r (expected %(format)s)')
% {'value': value, 'format': format})
def _base_url(self, secure=None):
if secure:
return self.vreg.config.get('https-url') or self.vreg.config['base-url']
return self.vreg.config['base-url']
def base_url(self, secure=None):
"""return the root url of the instance
"""
url = self._base_url(secure=secure)
return url if url is None else url.rstrip('/') + '/'
# abstract methods to override according to the web front-end #############
def describe(self, eid, asdict=False):
"""return a tuple (type, sourceuri, extid) for the entity with id <eid>"""
raise NotImplementedError