1 #ifndef MSP_STRINGS_REGEX_H_
2 #define MSP_STRINGS_REGEX_H_
10 class bad_regex: public std::logic_error
13 bad_regex(const std::string &, const std::string &, const std::string::const_iterator &);
16 std::string make_where(const std::string &, const std::string::const_iterator &);
21 This class provides regular expression matching. It supports a subset of
22 POSIX.2 extended regex syntax. Character classes, equivalence classes and
23 collating elements are not supported. Refer to the regex(7) manpage for more
26 A description of the internal workings of the class follows. You may skip this
27 if you are not a developer or otherwise interested.
29 The core of this class is a virtual machine running a non-deterministic finite
30 automaton (NFA). The state of the automaton is represented by an iterator into
31 the code. Transitions between states are represented by instructions.
32 Instructions may take multiple bytes, so a valid code iterator may not be a
35 The virtual machine may have any number of execution contexts at any one time.
36 On every cycle, each context is advanced until an input-consuming instruction
37 is encountered, keeping the input positions in sync with each other. Execution
38 continues as long as there's at least one context remaining.
40 The GROUP_BEGIN and GROUP_END instructions record the begin and end offset of a
41 match group, respectively. The GROUP_END instruction also marks the group as
42 successfully matched. If the target group was already matched, these
43 instructions do nothing.
45 The JUMP instruction causes the execution iterator to be modified by an offset.
47 The ND_JUMP instruction causes the execution context to be split in two. One
48 continues directly after the instruction and the other continues at an offset.
50 The NEGATE instruction causes the result of the next match instruction to be
53 Match instructions compare the input against a condition. If the match
54 succeeds, execution continues at the next instruction. If the match fails,
55 execution of that context is terminated.
57 The MATCH_BEGIN and MATCH_END instructions match the beginning and end of
58 the input string, respectively. They do not consume input.
60 The MATCH_CHAR instruction consumes the input character and matches it against
61 a single character. Since regexes often match sequences of printable character,
62 a match for a non-opcode character may be encoded as the character itself.
64 The MATCH_RANGE instruction consumes the input character and matches it against
65 an inclusive character range.
67 The MATCH_MASK instruction consumes the input character and matches it against
70 The MATCH_ANY instruction consumes the input character and always succeeds.
75 typedef std::basic_string<unsigned char> Code;
76 typedef unsigned short Count;
78 typedef unsigned short Index;
82 FIRST_INSTRUCTION_ = 0,
99 LAST_INSTRUCTION_ = 31
104 Code::const_iterator citer;
105 std::vector<RegMatch::Group> groups;
109 unsigned n_groups = 0;
112 /** Constructs a new Regex object from a string representation. */
113 Regex(const std::string &expr);
116 /** Compiles a regular expression into NFA bytecode. , 2011The iterator will be
117 advanced to the first unused character in the string. */
118 Code compile(const std::string &expr, std::string::const_iterator &iter, unsigned &group, bool branch);
120 Code parse_atom(const std::string &, std::string::const_iterator &i, unsigned &);
121 Code parse_brackets(const std::string &, std::string::const_iterator &);
122 bool parse_repeat(const std::string &, std::string::const_iterator &, Count &, Count &);
125 /** Matches the regex against a string. Refer to RegMatch documentation for
126 more information on the resulting object. */
127 RegMatch match(const std::string &str) const;
130 bool run(const std::string &, const std::string::const_iterator &, std::vector<RegMatch::Group> &) const;
131 bool group_compare(const RegMatch::Group &, const RegMatch::Group &) const;
134 /** Returns a disassembled representation of the NFA bytecode. For debugging
136 std::string disassemble() const;
138 std::string disassemble_instruction(Code::const_iterator &) const;