|
1 # copyright 2003-2010 LOGILAB S.A. (Paris, FRANCE), all rights reserved. |
|
2 # contact http://www.logilab.fr/ -- mailto:contact@logilab.fr |
|
3 # |
|
4 # This file is part of CubicWeb. |
|
5 # |
|
6 # CubicWeb is free software: you can redistribute it and/or modify it under the |
|
7 # terms of the GNU Lesser General Public License as published by the Free |
|
8 # Software Foundation, either version 2.1 of the License, or (at your option) |
|
9 # any later version. |
|
10 # |
|
11 # CubicWeb is distributed in the hope that it will be useful, but WITHOUT |
|
12 # ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS |
|
13 # FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more |
|
14 # details. |
|
15 # |
|
16 # You should have received a copy of the GNU Lesser General Public License along |
|
17 # with CubicWeb. If not, see <http://www.gnu.org/licenses/>. |
|
18 """SPARQL -> RQL translator""" |
|
19 |
|
20 from logilab.common import make_domains |
|
21 from rql import TypeResolverException |
|
22 from fyzz.yappsparser import parse |
|
23 from fyzz import ast |
|
24 |
|
25 from cubicweb.xy import xy |
|
26 |
|
27 |
|
28 class UnsupportedQuery(Exception): pass |
|
29 |
|
30 def order_limit_offset(sparqlst): |
|
31 addons = '' |
|
32 if sparqlst.orderby: |
|
33 sortterms = ', '.join('%s %s' % (var.name.upper(), ascdesc.upper()) |
|
34 for var, ascdesc in sparqlst.orderby) |
|
35 addons += ' ORDERBY %s' % sortterms |
|
36 if sparqlst.limit: |
|
37 addons += ' LIMIT %s' % sparqlst.limit |
|
38 if sparqlst.offset: |
|
39 addons += ' OFFSET %s' % sparqlst.offset |
|
40 return addons |
|
41 |
|
42 |
|
43 class QueryInfo(object): |
|
44 """wrapper class containing necessary information to generate a RQL query |
|
45 from a sparql syntax tree |
|
46 """ |
|
47 def __init__(self, sparqlst): |
|
48 self.sparqlst = sparqlst |
|
49 if sparqlst.selected == ['*']: |
|
50 self.selection = [var.upper() for var in sparqlst.variables] |
|
51 else: |
|
52 self.selection = [var.name.upper() for var in sparqlst.selected] |
|
53 self.possible_types = {} |
|
54 self.infer_types_info = [] |
|
55 self.union_params = [] |
|
56 self.restrictions = [] |
|
57 self.literals = {} |
|
58 self._litcount = 0 |
|
59 |
|
60 def add_literal(self, value): |
|
61 key = chr(ord('a') + self._litcount) |
|
62 self._litcount += 1 |
|
63 self.literals[key] = value |
|
64 return key |
|
65 |
|
66 def set_possible_types(self, var, varpossibletypes): |
|
67 """set/restrict possible types for the given variable. |
|
68 |
|
69 :return: True if something changed, else false. |
|
70 :raise: TypeResolverException if no more type allowed |
|
71 """ |
|
72 varpossibletypes = set(varpossibletypes) |
|
73 try: |
|
74 ctypes = self.possible_types[var] |
|
75 nbctypes = len(ctypes) |
|
76 ctypes &= varpossibletypes |
|
77 if not ctypes: |
|
78 raise TypeResolverException('No possible type') |
|
79 return len(ctypes) != nbctypes |
|
80 except KeyError: |
|
81 self.possible_types[var] = varpossibletypes |
|
82 return True |
|
83 |
|
84 def infer_types(self): |
|
85 # XXX should use something similar to rql.analyze for proper type inference |
|
86 modified = True |
|
87 # loop to infer types until nothing changed |
|
88 while modified: |
|
89 modified = False |
|
90 for yams_predicates, subjvar, obj in self.infer_types_info: |
|
91 nbchoices = len(yams_predicates) |
|
92 # get possible types for the subject variable, according to the |
|
93 # current predicate |
|
94 svptypes = set(s for s, r, o in yams_predicates) |
|
95 if not '*' in svptypes: |
|
96 if self.set_possible_types(subjvar, svptypes): |
|
97 modified = True |
|
98 # restrict predicates according to allowed subject var types |
|
99 if subjvar in self.possible_types: |
|
100 yams_predicates[:] = [(s, r, o) for s, r, o in yams_predicates |
|
101 if s == '*' or s in self.possible_types[subjvar]] |
|
102 if isinstance(obj, ast.SparqlVar): |
|
103 # make a valid rql var name |
|
104 objvar = obj.name.upper() |
|
105 # get possible types for the object variable, according to |
|
106 # the current predicate |
|
107 ovptypes = set(o for s, r, o in yams_predicates) |
|
108 if not '*' in ovptypes: |
|
109 if self.set_possible_types(objvar, ovptypes): |
|
110 modified = True |
|
111 # restrict predicates according to allowed object var types |
|
112 if objvar in self.possible_types: |
|
113 yams_predicates[:] = [(s, r, o) for s, r, o in yams_predicates |
|
114 if o == '*' or o in self.possible_types[objvar]] |
|
115 # ensure this still make sense |
|
116 if not yams_predicates: |
|
117 raise TypeResolverException('No yams predicate') |
|
118 if len(yams_predicates) != nbchoices: |
|
119 modified = True |
|
120 |
|
121 def build_restrictions(self): |
|
122 # now, for each predicate |
|
123 for yams_predicates, subjvar, obj in self.infer_types_info: |
|
124 rel = yams_predicates[0] |
|
125 # if there are several yams relation type equivalences, we will have |
|
126 # to generate several unioned rql queries |
|
127 for s, r, o in yams_predicates[1:]: |
|
128 if r != rel[1]: |
|
129 self.union_params.append((yams_predicates, subjvar, obj)) |
|
130 break |
|
131 # else we can simply add it to base rql restrictions |
|
132 else: |
|
133 restr = self.build_restriction(subjvar, rel[1], obj) |
|
134 self.restrictions.append(restr) |
|
135 |
|
136 def build_restriction(self, subjvar, rtype, obj): |
|
137 if isinstance(obj, ast.SparqlLiteral): |
|
138 key = self.add_literal(obj.value) |
|
139 objvar = '%%(%s)s' % key |
|
140 else: |
|
141 assert isinstance(obj, ast.SparqlVar) |
|
142 # make a valid rql var name |
|
143 objvar = obj.name.upper() |
|
144 # else we can simply add it to base rql restrictions |
|
145 return '%s %s %s' % (subjvar, rtype, objvar) |
|
146 |
|
147 def finalize(self): |
|
148 """return corresponding rql query (string) / args (dict)""" |
|
149 for varname, ptypes in self.possible_types.items(): |
|
150 if len(ptypes) == 1: |
|
151 self.restrictions.append('%s is %s' % (varname, next(iter(ptypes)))) |
|
152 unions = [] |
|
153 for releq, subjvar, obj in self.union_params: |
|
154 thisunions = [] |
|
155 for st, rt, ot in releq: |
|
156 thisunions.append([self.build_restriction(subjvar, rt, obj)]) |
|
157 if st != '*': |
|
158 thisunions[-1].append('%s is %s' % (subjvar, st)) |
|
159 if isinstance(obj, ast.SparqlVar) and ot != '*': |
|
160 objvar = obj.name.upper() |
|
161 thisunions[-1].append('%s is %s' % (objvar, objvar)) |
|
162 if not unions: |
|
163 unions = thisunions |
|
164 else: |
|
165 unions = zip(*make_domains([unions, thisunions])) |
|
166 selection = 'Any ' + ', '.join(self.selection) |
|
167 sparqlst = self.sparqlst |
|
168 if sparqlst.distinct: |
|
169 selection = 'DISTINCT ' + selection |
|
170 if unions: |
|
171 baserql = '%s WHERE %s' % (selection, ', '.join(self.restrictions)) |
|
172 rqls = ['(%s, %s)' % (baserql, ', '.join(unionrestrs)) |
|
173 for unionrestrs in unions] |
|
174 rql = ' UNION '.join(rqls) |
|
175 if sparqlst.orderby or sparqlst.limit or sparqlst.offset: |
|
176 rql = '%s%s WITH %s BEING (%s)' % ( |
|
177 selection, order_limit_offset(sparqlst), |
|
178 ', '.join(self.selection), rql) |
|
179 else: |
|
180 rql = '%s%s WHERE %s' % (selection, order_limit_offset(sparqlst), |
|
181 ', '.join(self.restrictions)) |
|
182 return rql, self.literals |
|
183 |
|
184 |
|
185 class Sparql2rqlTranslator(object): |
|
186 def __init__(self, yschema): |
|
187 self.yschema = yschema |
|
188 |
|
189 def translate(self, sparql): |
|
190 sparqlst = parse(sparql) |
|
191 if sparqlst.type != 'select': |
|
192 raise UnsupportedQuery() |
|
193 qi = QueryInfo(sparqlst) |
|
194 for subj, predicate, obj in sparqlst.where: |
|
195 if not isinstance(subj, ast.SparqlVar): |
|
196 raise UnsupportedQuery() |
|
197 # make a valid rql var name |
|
198 subjvar = subj.name.upper() |
|
199 if predicate in [('', 'a'), |
|
200 ('http://www.w3.org/1999/02/22-rdf-syntax-ns#', 'type')]: |
|
201 # special 'is' relation |
|
202 if not isinstance(obj, tuple): |
|
203 raise UnsupportedQuery() |
|
204 # restrict possible types for the subject variable |
|
205 qi.set_possible_types( |
|
206 subjvar, xy.yeq(':'.join(obj), isentity=True)) |
|
207 else: |
|
208 # 'regular' relation (eg not 'is') |
|
209 if not isinstance(predicate, tuple): |
|
210 raise UnsupportedQuery() |
|
211 # list of 3-uple |
|
212 # (yams etype (subject), yams rtype, yams etype (object)) |
|
213 # where subject / object entity type may '*' if not specified |
|
214 yams_predicates = xy.yeq(':'.join(predicate)) |
|
215 qi.infer_types_info.append((yams_predicates, subjvar, obj)) |
|
216 if not isinstance(obj, (ast.SparqlLiteral, ast.SparqlVar)): |
|
217 raise UnsupportedQuery() |
|
218 qi.infer_types() |
|
219 qi.build_restrictions() |
|
220 return qi |