1 | // Copyright 2009 The RE2 Authors. All Rights Reserved. |
2 | // Use of this source code is governed by a BSD-style |
3 | // license that can be found in the LICENSE file. |
4 | |
5 | #ifndef RE2_PREFILTER_H_ |
6 | #define RE2_PREFILTER_H_ |
7 | |
8 | // Prefilter is the class used to extract string guards from regexps. |
9 | // Rather than using Prefilter class directly, use FilteredRE2. |
10 | // See filtered_re2.h |
11 | |
12 | #include <set> |
13 | #include <string> |
14 | #include <vector> |
15 | |
16 | #include "util/logging.h" |
17 | |
18 | namespace re2 { |
19 | |
20 | class RE2; |
21 | |
22 | class Regexp; |
23 | |
24 | class Prefilter { |
25 | // Instead of using Prefilter directly, use FilteredRE2; see filtered_re2.h |
26 | public: |
27 | enum Op { |
28 | ALL = 0, // Everything matches |
29 | NONE, // Nothing matches |
30 | ATOM, // The string atom() must match |
31 | AND, // All in subs() must match |
32 | OR, // One of subs() must match |
33 | }; |
34 | |
35 | explicit Prefilter(Op op); |
36 | ~Prefilter(); |
37 | |
38 | Op op() { return op_; } |
39 | const std::string& atom() const { return atom_; } |
40 | void set_unique_id(int id) { unique_id_ = id; } |
41 | int unique_id() const { return unique_id_; } |
42 | |
43 | // The children of the Prefilter node. |
44 | std::vector<Prefilter*>* subs() { |
45 | DCHECK(op_ == AND || op_ == OR); |
46 | return subs_; |
47 | } |
48 | |
49 | // Set the children vector. Prefilter takes ownership of subs and |
50 | // subs_ will be deleted when Prefilter is deleted. |
51 | void set_subs(std::vector<Prefilter*>* subs) { subs_ = subs; } |
52 | |
53 | // Given a RE2, return a Prefilter. The caller takes ownership of |
54 | // the Prefilter and should deallocate it. Returns NULL if Prefilter |
55 | // cannot be formed. |
56 | static Prefilter* FromRE2(const RE2* re2); |
57 | |
58 | // Returns a readable debug string of the prefilter. |
59 | std::string DebugString() const; |
60 | |
61 | private: |
62 | class Info; |
63 | |
64 | // Combines two prefilters together to create an AND. The passed |
65 | // Prefilters will be part of the returned Prefilter or deleted. |
66 | static Prefilter* And(Prefilter* a, Prefilter* b); |
67 | |
68 | // Combines two prefilters together to create an OR. The passed |
69 | // Prefilters will be part of the returned Prefilter or deleted. |
70 | static Prefilter* Or(Prefilter* a, Prefilter* b); |
71 | |
72 | // Generalized And/Or |
73 | static Prefilter* AndOr(Op op, Prefilter* a, Prefilter* b); |
74 | |
75 | static Prefilter* FromRegexp(Regexp* a); |
76 | |
77 | static Prefilter* FromString(const std::string& str); |
78 | |
79 | static Prefilter* OrStrings(std::set<std::string>* ss); |
80 | |
81 | static Info* BuildInfo(Regexp* re); |
82 | |
83 | Prefilter* Simplify(); |
84 | |
85 | // Kind of Prefilter. |
86 | Op op_; |
87 | |
88 | // Sub-matches for AND or OR Prefilter. |
89 | std::vector<Prefilter*>* subs_; |
90 | |
91 | // Actual string to match in leaf node. |
92 | std::string atom_; |
93 | |
94 | // If different prefilters have the same string atom, or if they are |
95 | // structurally the same (e.g., OR of same atom strings) they are |
96 | // considered the same unique nodes. This is the id for each unique |
97 | // node. This field is populated with a unique id for every node, |
98 | // and -1 for duplicate nodes. |
99 | int unique_id_; |
100 | |
101 | Prefilter(const Prefilter&) = delete; |
102 | Prefilter& operator=(const Prefilter&) = delete; |
103 | }; |
104 | |
105 | } // namespace re2 |
106 | |
107 | #endif // RE2_PREFILTER_H_ |
108 | |