1 |
|
/****************************************************************************** |
2 |
|
* Top contributors (to current version): |
3 |
|
* Andrew Reynolds, Andres Noetzli, Mudathir Mohamed |
4 |
|
* |
5 |
|
* This file is part of the cvc5 project. |
6 |
|
* |
7 |
|
* Copyright (c) 2009-2021 by the authors listed in the file AUTHORS |
8 |
|
* in the top-level source directory and their institutional affiliations. |
9 |
|
* All rights reserved. See the file COPYING in the top-level source |
10 |
|
* directory for licensing information. |
11 |
|
* **************************************************************************** |
12 |
|
* |
13 |
|
* Base solver for term indexing and constant inference for the |
14 |
|
* theory of strings. |
15 |
|
*/ |
16 |
|
|
17 |
|
#include "cvc5_private.h" |
18 |
|
|
19 |
|
#ifndef CVC5__THEORY__STRINGS__BASE_SOLVER_H |
20 |
|
#define CVC5__THEORY__STRINGS__BASE_SOLVER_H |
21 |
|
|
22 |
|
#include "context/cdhashset.h" |
23 |
|
#include "context/cdlist.h" |
24 |
|
#include "smt/env_obj.h" |
25 |
|
#include "theory/strings/infer_info.h" |
26 |
|
#include "theory/strings/inference_manager.h" |
27 |
|
#include "theory/strings/normal_form.h" |
28 |
|
#include "theory/strings/skolem_cache.h" |
29 |
|
#include "theory/strings/solver_state.h" |
30 |
|
|
31 |
|
namespace cvc5 { |
32 |
|
namespace theory { |
33 |
|
namespace strings { |
34 |
|
|
35 |
|
/** The base solver for the theory of strings |
36 |
|
* |
37 |
|
* This implements techniques for inferring when terms are congruent in the |
38 |
|
* current context, and techniques for inferring when equivalence classes |
39 |
|
* are equivalent to constants. |
40 |
|
*/ |
41 |
|
class BaseSolver : protected EnvObj |
42 |
|
{ |
43 |
|
using NodeSet = context::CDHashSet<Node>; |
44 |
|
|
45 |
|
public: |
46 |
|
BaseSolver(Env& env, SolverState& s, InferenceManager& im); |
47 |
|
~BaseSolver(); |
48 |
|
|
49 |
|
//-----------------------inference steps |
50 |
|
/** check initial |
51 |
|
* |
52 |
|
* This function initializes term indices for each strings function symbol. |
53 |
|
* One key aspect of this construction is that concat terms are indexed by |
54 |
|
* their list of non-empty components. For example, if x = "" is an equality |
55 |
|
* asserted in this SAT context, then y ++ x ++ z may be indexed by (y,z). |
56 |
|
* This method may infer various facts while building these term indices, for |
57 |
|
* instance, based on congruence. An example would be inferring: |
58 |
|
* y ++ x ++ z = y ++ z |
59 |
|
* if both terms are registered in this SAT context. |
60 |
|
* |
61 |
|
* This function should be called as a first step of any strategy. |
62 |
|
*/ |
63 |
|
void checkInit(); |
64 |
|
/** check constant equivalence classes |
65 |
|
* |
66 |
|
* This function infers whether CONCAT terms can be simplified to constants. |
67 |
|
* For example, if x = "a" and y = "b" are equalities in the current SAT |
68 |
|
* context, then we may infer x ++ "c" ++ y is equivalent to "acb". In this |
69 |
|
* case, we infer the fact x ++ "c" ++ y = "acb". |
70 |
|
*/ |
71 |
|
void checkConstantEquivalenceClasses(); |
72 |
|
/** check cardinality |
73 |
|
* |
74 |
|
* This function checks whether a cardinality inference needs to be applied |
75 |
|
* to a set of equivalence classes. For details, see Step 5 of the proof |
76 |
|
* procedure from Liang et al, CAV 2014. |
77 |
|
*/ |
78 |
|
void checkCardinality(); |
79 |
|
//-----------------------end inference steps |
80 |
|
|
81 |
|
//-----------------------query functions |
82 |
|
/** |
83 |
|
* Is n congruent to another term in the current context that has not been |
84 |
|
* marked congruent? If so, we can ignore n. |
85 |
|
* |
86 |
|
* Note this and the functions in this block below are valid during a full |
87 |
|
* effort check after a call to checkInit. |
88 |
|
*/ |
89 |
|
bool isCongruent(Node n); |
90 |
|
/** |
91 |
|
* Get the constant that the equivalence class eqc is entailed to be equal |
92 |
|
* to, or null if none exist. |
93 |
|
*/ |
94 |
|
Node getConstantEqc(Node eqc); |
95 |
|
/** |
96 |
|
* Same as above, where the explanation for n = c is added to exp if c is |
97 |
|
* the (non-null) return value of this function, where n is a term in the |
98 |
|
* equivalence class of eqc. |
99 |
|
*/ |
100 |
|
Node explainConstantEqc(Node n, Node eqc, std::vector<Node>& exp); |
101 |
|
/** |
102 |
|
* Same as above, for "best content" terms. |
103 |
|
*/ |
104 |
|
Node explainBestContentEqc(Node n, Node eqc, std::vector<Node>& exp); |
105 |
|
/** |
106 |
|
* Get the set of equivalence classes of type string. |
107 |
|
*/ |
108 |
|
const std::vector<Node>& getStringEqc() const; |
109 |
|
//-----------------------end query functions |
110 |
|
|
111 |
|
private: |
112 |
|
/** |
113 |
|
* The information that we associated with each equivalence class. |
114 |
|
* |
115 |
|
* Example 1. Consider the equivalence class { r, x++"a"++y, x++z }, and |
116 |
|
* assume x = "" and y = "bb" in the current context. We have that |
117 |
|
* d_bestContent = "abb", |
118 |
|
* d_base = x++"a"++y |
119 |
|
* d_exp = ( x = "" AND y = "bb" ) |
120 |
|
* |
121 |
|
* Example 2. Consider the equivalence class { r, x++"a"++w++y, x++z }, and |
122 |
|
* assume x = "" and y = "bb" in the current context. We have that |
123 |
|
* d_bestContent = "a" ++ w ++ "bb", |
124 |
|
* d_bestScore = 3 |
125 |
|
* d_base = x++"a"++w++y |
126 |
|
* d_exp = ( x = "" AND y = "bb" ) |
127 |
|
* |
128 |
|
* This information is computed during checkInit and is used during various |
129 |
|
* inference schemas for deriving inferences. |
130 |
|
*/ |
131 |
1192666 |
struct BaseEqcInfo |
132 |
|
{ |
133 |
|
/** |
134 |
|
* Either a constant or a concatentation of constants and variables that |
135 |
|
* this equivalence class is entailed to be equal to. If it is a |
136 |
|
* concatenation, this is the concatenation that is currently known to have |
137 |
|
* the highest score (see `d_bestScore`). |
138 |
|
*/ |
139 |
|
Node d_bestContent; |
140 |
|
/** |
141 |
|
* The sum of the number of characters in the string literals of |
142 |
|
* `d_bestContent`. |
143 |
|
*/ |
144 |
|
size_t d_bestScore; |
145 |
|
/** |
146 |
|
* The term in the equivalence class that is entailed to be equal to |
147 |
|
* `d_bestContent`. |
148 |
|
*/ |
149 |
|
Node d_base; |
150 |
|
/** This term explains why `d_bestContent` is equal to `d_base`. */ |
151 |
|
Node d_exp; |
152 |
|
}; |
153 |
|
|
154 |
|
/** |
155 |
|
* A term index that considers terms modulo flattening and constant merging |
156 |
|
* for concatenation terms. |
157 |
|
*/ |
158 |
4278542 |
class TermIndex |
159 |
|
{ |
160 |
|
public: |
161 |
|
/** Add n to this trie |
162 |
|
* |
163 |
|
* A term is indexed by flattening arguments of concatenation terms, |
164 |
|
* and replacing them by (non-empty) constants when possible, for example |
165 |
|
* if n is (str.++ x y z) and x = "abc" and y = "" are asserted, then n is |
166 |
|
* indexed by ("abc", z). |
167 |
|
* |
168 |
|
* index: the child of n we are currently processing, |
169 |
|
* s : reference to solver state, |
170 |
|
* er : the representative of the empty equivalence class. |
171 |
|
* |
172 |
|
* We store the vector of terms that n was indexed by in the vector c. |
173 |
|
*/ |
174 |
|
Node add(TNode n, |
175 |
|
unsigned index, |
176 |
|
const SolverState& s, |
177 |
|
Node er, |
178 |
|
std::vector<Node>& c); |
179 |
|
/** Clear this trie */ |
180 |
|
void clear() { d_children.clear(); } |
181 |
|
/** The data at this node of the trie */ |
182 |
|
Node d_data; |
183 |
|
/** The children of this node of the trie */ |
184 |
|
std::map<TNode, TermIndex> d_children; |
185 |
|
}; |
186 |
|
/** |
187 |
|
* This method is called as we are traversing the term index ti, where vecc |
188 |
|
* accumulates the list of constants in the path to ti. If ti has a non-null |
189 |
|
* data n, then we have inferred that d_data is equivalent to the |
190 |
|
* constant specified by vecc. |
191 |
|
* |
192 |
|
* @param ti The term index for string concatenations |
193 |
|
* @param vecc The list of constants in the path to ti |
194 |
|
* @param ensureConst If true, require that each element in the path is |
195 |
|
* constant |
196 |
|
* @param isConst If true, the path so far only includes constants |
197 |
|
*/ |
198 |
|
void checkConstantEquivalenceClasses(TermIndex* ti, |
199 |
|
std::vector<Node>& vecc, |
200 |
|
bool ensureConst = true, |
201 |
|
bool isConst = true); |
202 |
|
/** |
203 |
|
* Check cardinality for type tn. This adds a lemma corresponding to |
204 |
|
* cardinality for terms of type tn, if applicable. |
205 |
|
* |
206 |
|
* @param tn The string-like type of terms we are considering, |
207 |
|
* @param cols The list of collections of equivalence classes. This is a |
208 |
|
* partition of all string equivalence classes, grouped by those with equal |
209 |
|
* lengths. |
210 |
|
* @param lts The length of each of the collections in cols. |
211 |
|
*/ |
212 |
|
void checkCardinalityType(TypeNode tn, |
213 |
|
std::vector<std::vector<Node> >& cols, |
214 |
|
std::vector<Node>& lts); |
215 |
|
/** The solver state object */ |
216 |
|
SolverState& d_state; |
217 |
|
/** The (custom) output channel of the theory of strings */ |
218 |
|
InferenceManager& d_im; |
219 |
|
/** Commonly used constants */ |
220 |
|
Node d_emptyString; |
221 |
|
Node d_false; |
222 |
|
/** |
223 |
|
* A congruence class is a set of terms f( t1 ), ..., f( tn ) where |
224 |
|
* t1 = ... = tn. Congruence classes are important since all but |
225 |
|
* one of the above terms (the representative of the congruence class) |
226 |
|
* can be ignored by the solver. |
227 |
|
* |
228 |
|
* This set contains a set of nodes that are not representatives of their |
229 |
|
* congruence class. This set is used to skip reasoning about terms in |
230 |
|
* various inference schemas implemented by this class. |
231 |
|
*/ |
232 |
|
NodeSet d_congruent; |
233 |
|
/** |
234 |
|
* Maps equivalence classes to their info, see description of `BaseEqcInfo` |
235 |
|
* for more information. |
236 |
|
*/ |
237 |
|
std::map<Node, BaseEqcInfo> d_eqcInfo; |
238 |
|
/** The list of equivalence classes of type string */ |
239 |
|
std::vector<Node> d_stringsEqc; |
240 |
|
/** A term index for each type, function kind pair */ |
241 |
|
std::map<TypeNode, std::map<Kind, TermIndex> > d_termIndex; |
242 |
|
/** the cardinality of the alphabet */ |
243 |
|
uint32_t d_cardSize; |
244 |
|
}; /* class BaseSolver */ |
245 |
|
|
246 |
|
} // namespace strings |
247 |
|
} // namespace theory |
248 |
|
} // namespace cvc5 |
249 |
|
|
250 |
|
#endif /* CVC5__THEORY__STRINGS__BASE_SOLVER_H */ |