Namespace mata::parser

namespace parser

Parser from .mata format to automata (currently Nfa and Afa are supported).

This includes parsing either from files or from other streams (strings, etc.).

Typedefs

using KeyListStore = std::map<std::string, std::vector<std::string>>
using BodyLine = std::vector<std::string>
using Parsed = std::vector<ParsedSection>

Parsed data.

Functions

Parsed parse_mf(const std::string &input, bool keepQuotes = false)

Parses a string into an intermediary structure.

Parsed parse_mf(std::istream &input, bool keepQuotes = false)

Parses a stream into an intermediary structure.

ParsedSection parse_mf_section(std::istream &input, bool keepQuotes = false)

Parses one section from a stream into an intermediary structure.

ParsedSection parse_mf_section(const std::string &input, bool keepQuotes = false)

Parses one section from a string into an intermediary structure.

void init()

registers dispatcher

nfa::Nfa create_nfa(const std::string &pattern, bool use_epsilon = false, mata::Symbol epsilon_value = 306, bool use_reduce = true, const Encoding encoding = Encoding::Latin1)

Creates NFA from regular expression using RE2 parser.

At https://github.com/google/re2/wiki/Syntax, you can find the syntax of regular expressions with following futher limitations: 1) If you use UTF8 encoding, the created NFA will have the values of bytes instead of full symbols. For example, the character Ā whose Unicode code point is U+0100 and is represented in UTF8 as two bytes c4 80 will have two transitions, one with c4 followed with by 80, to encode it. 2) The created automaton represents the language of the regex and is not expected to be used in regex matching. Therefore, stuff like ^, $, , etc. are ignored in the regex.

See also

mata::nfa::builder::create_from_regex()

Parameters:

  • pattern – regex as a string

  • use_epsilon – whether to keep epsilon transitions in created NFA

  • epsilon_value – symbol representing epsilon

  • use_reduce – if set to true the result is trimmed and reduced using simulation reduction

  • encoding – encoding of the regex, default is Latin1

Returns:

Nfa corresponding to pattern

void create_nfa(nfa::Nfa *nfa, const std::string &pattern, bool use_epsilon = false, mata::Symbol epsilon_value = 306, bool use_reduce = true, const Encoding encoding = Encoding::Latin1)
struct ParsedSection
#include <parser.hh>

Parsed data (single section)

Public Functions

inline bool empty() const

Is the section empty?

bool operator==(const ParsedSection &rhs) const

Equality operator.

const std::vector<std::string> &operator[](const std::string &key) const

subscript operator for the key-value store

inline bool haskey(const std::string &key) const

check whether the key-value store contains a key