fix optimisation with super session that may lead to integrity loss
at some point I've decided to stop ensuring ?1 cardinality was respected
when adding a new relation using a super session, to avoid the cost of
the delete query. That was yet discussable because it introduced unexpected
difference between execute and unsafe_execute, which is imo not worth it.
Also, now that rql() in migration script default to unsafe_execute, we
definitly don't want that implicit behaviour change (which already cause
bug when for instance adding another default workflow for an entity type:
without that fix we end up with *two* default workflows while the schema
tells we can have only one.
IMO we should go to the direction that super session skip all security
check, but nothing else, unless explicitly asked.
"""
:organization: Logilab
:copyright: 2001-2010 LOGILAB S.A. (Paris, FRANCE), license is LGPL v2.
:contact: http://www.logilab.fr/ -- mailto:contact@logilab.fr
:license: GNU Lesser General Public License, v2.1 - http://www.gnu.org/licenses
"""
from cubicweb.devtools.testlib import CubicWebTC
from cubicweb.web.facet import insert_attr_select_relation, prepare_facets_rqlst
class InsertAttrRelationTC(CubicWebTC):
def parse(self, query):
rqlst = self.vreg.parse(self.session, query)
select = rqlst.children[0]
# XXX done in real life?
select.remove_groups()
return rqlst
def _generate(self, rqlst, rel, role, attr):
mainvar = prepare_facets_rqlst(rqlst)[0]
insert_attr_select_relation(rqlst.children[0], mainvar, rel, role, attr)
return rqlst.as_string()
@property
def select(self):
return self.parse('Any B,(NOW - CD),S,V,U,GROUP_CONCAT(TN),VN,P,CD,BMD '
'GROUPBY B,CD,S,V,U,VN,P,BMD '
'WHERE B in_state S, B creation_date CD, '
'B modification_date BMD, T? tags B, T name TN, '
'V? bookmarked_by B, V title VN, B created_by U?, '
'B in_group P, P name "managers"')
def test_1(self):
self.assertEquals(self._generate(self.select, 'in_state', 'subject', 'name'),
"DISTINCT Any A,C ORDERBY C WHERE B in_group P, P name 'managers', "
"B in_state A, A name C, B is CWUser")
def test_2(self):
self.assertEquals(self._generate(self.select, 'tags', 'object', 'name'),
"DISTINCT Any A,C ORDERBY C WHERE B in_group P, P name 'managers', "
"A tags B, A name C, B is CWUser")
def test_3(self):
self.assertEquals(self._generate(self.select, 'created_by', 'subject', 'login'),
"DISTINCT Any A,C ORDERBY C WHERE B in_group P, P name 'managers', "
"B created_by A, A login C, B is CWUser")
def test_4(self):
self.assertEquals(self._generate(self.parse('Any X WHERE X is CWUser'), 'created_by', 'subject', 'login'),
"DISTINCT Any A,B ORDERBY B WHERE X is CWUser, X created_by A, A login B")
def test_5(self):
self.assertEquals(self._generate(self.parse('Any X,L WHERE X is CWUser, X login L'), 'created_by', 'subject', 'login'),
"DISTINCT Any A,B ORDERBY B WHERE X is CWUser, X created_by A, A login B")
def test_nonregr1(self):
select = self.parse('Any T,V WHERE T bookmarked_by V?, '
'V in_state VS, VS name "published", T created_by U')
self.assertEquals(self._generate(select, 'created_by', 'subject', 'login'),
"DISTINCT Any A,B ORDERBY B WHERE T created_by U, "
"T created_by A, A login B, T is Bookmark")
def test_nonregr2(self):
#'DISTINCT Any X,TMP,N WHERE P name TMP, X version_of P, P is Project, X is Version, not X in_state S,S name "published", X num N ORDERBY TMP,N'
select = self.parse('DISTINCT Any V,TN,L ORDERBY TN,L WHERE T nom TN, V connait T, T is Personne, V is CWUser,'
'NOT V in_state VS, VS name "published", V login L')
rschema = self.schema['connait']
for rdefs in rschema.rdefs.values():
rdefs.cardinality = '++'
try:
self.assertEquals(self._generate(select, 'in_state', 'subject', 'name'),
"DISTINCT Any A,B ORDERBY B WHERE V is CWUser, "
"NOT V in_state VS, VS name 'published', "
"V in_state A, A name B")
finally:
for rdefs in rschema.rdefs.values():
rdefs.cardinality = '**'
def test_nonregr3(self):
#'DISTINCT Any X,TMP,N WHERE P name TMP, X version_of P, P is Project, X is Version, not X in_state S,S name "published", X num N ORDERBY TMP,N'
select = self.parse('DISTINCT Any X, MAX(Y) GROUPBY X WHERE X is CWUser, Y is Bookmark, X in_group A')
self.assertEquals(self._generate(select, 'in_group', 'subject', 'name'),
"DISTINCT Any B,C ORDERBY C WHERE X is CWUser, X in_group B, B name C")
if __name__ == '__main__':
from logilab.common.testlib import unittest_main
unittest_main()