| 1 | // Copyright (c) 2001-2010 Hartmut Kaiser |
| 2 | // |
| 3 | // Distributed under the Boost Software License, Version 1.0. (See accompanying |
| 4 | // file LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) |
| 5 | |
| 6 | // This example is the equivalent to the following lex program: |
| 7 | /* |
| 8 | //[wcp_flex_version |
| 9 | %{ |
| 10 | int c = 0, w = 0, l = 0; |
| 11 | %} |
| 12 | word [^ \t\n]+ |
| 13 | eol \n |
| 14 | %% |
| 15 | {word} { ++w; c += yyleng; } |
| 16 | {eol} { ++c; ++l; } |
| 17 | . { ++c; } |
| 18 | %% |
| 19 | main() |
| 20 | { |
| 21 | yylex(); |
| 22 | printf("%d %d %d\n", l, w, c); |
| 23 | } |
| 24 | //] |
| 25 | */ |
| 26 | // Its purpose is to do the word count function of the wc command in UNIX. It |
| 27 | // prints the number of lines, words and characters in a file. |
| 28 | // |
| 29 | // The example additionally demonstrates how to use the add_pattern(...)(...) |
| 30 | // syntax to define lexer patterns. These patterns are essentially parameter- |
| 31 | // less 'macros' for regular expressions, allowing to simplify their |
| 32 | // definition. |
| 33 | |
| 34 | // #define BOOST_SPIRIT_LEXERTL_DEBUG |
| 35 | #define BOOST_VARIANT_MINIMIZE_SIZE |
| 36 | |
| 37 | //[wcp_includes |
| 38 | #include <boost/spirit/include/qi.hpp> |
| 39 | #include <boost/spirit/include/lex_lexertl.hpp> |
| 40 | #include <boost/phoenix/operator.hpp> |
| 41 | #include <boost/phoenix/statement.hpp> |
| 42 | #include <boost/phoenix/stl/container.hpp> |
| 43 | //] |
| 44 | |
| 45 | #include <iostream> |
| 46 | #include <string> |
| 47 | |
| 48 | #include "example.hpp" |
| 49 | |
| 50 | //[wcp_namespaces |
| 51 | using namespace boost::spirit; |
| 52 | using namespace boost::spirit::ascii; |
| 53 | //] |
| 54 | |
| 55 | /////////////////////////////////////////////////////////////////////////////// |
| 56 | // Token definition: We use the lexertl based lexer engine as the underlying |
| 57 | // lexer type. |
| 58 | /////////////////////////////////////////////////////////////////////////////// |
| 59 | //[wcp_token_ids |
| 60 | enum tokenids |
| 61 | { |
| 62 | IDANY = lex::min_token_id + 10 |
| 63 | }; |
| 64 | //] |
| 65 | |
| 66 | //[wcp_token_definition |
| 67 | template <typename Lexer> |
| 68 | struct word_count_tokens : lex::lexer<Lexer> |
| 69 | { |
| 70 | word_count_tokens() |
| 71 | { |
| 72 | // define patterns (lexer macros) to be used during token definition |
| 73 | // below |
| 74 | this->self.add_pattern |
| 75 | ("WORD" , "[^ \t\n]+" ) |
| 76 | ; |
| 77 | |
| 78 | // define tokens and associate them with the lexer |
| 79 | word = "{WORD}" ; // reference the pattern 'WORD' as defined above |
| 80 | |
| 81 | // this lexer will recognize 3 token types: words, newlines, and |
| 82 | // everything else |
| 83 | this->self.add |
| 84 | (word) // no token id is needed here |
| 85 | ('\n') // characters are usable as tokens as well |
| 86 | ("." , IDANY) // string literals will not be escaped by the library |
| 87 | ; |
| 88 | } |
| 89 | |
| 90 | // the token 'word' exposes the matched string as its parser attribute |
| 91 | lex::token_def<std::string> word; |
| 92 | }; |
| 93 | //] |
| 94 | |
| 95 | /////////////////////////////////////////////////////////////////////////////// |
| 96 | // Grammar definition |
| 97 | /////////////////////////////////////////////////////////////////////////////// |
| 98 | //[wcp_grammar_definition |
| 99 | template <typename Iterator> |
| 100 | struct word_count_grammar : qi::grammar<Iterator> |
| 101 | { |
| 102 | template <typename TokenDef> |
| 103 | word_count_grammar(TokenDef const& tok) |
| 104 | : word_count_grammar::base_type(start) |
| 105 | , c(0), w(0), l(0) |
| 106 | { |
| 107 | using boost::phoenix::ref; |
| 108 | using boost::phoenix::size; |
| 109 | |
| 110 | start = *( tok.word [++ref(w), ref(c) += size(_1)] |
| 111 | | lit('\n') [++ref(c), ++ref(l)] |
| 112 | | qi::token(IDANY) [++ref(c)] |
| 113 | ) |
| 114 | ; |
| 115 | } |
| 116 | |
| 117 | std::size_t c, w, l; |
| 118 | qi::rule<Iterator> start; |
| 119 | }; |
| 120 | //] |
| 121 | |
| 122 | /////////////////////////////////////////////////////////////////////////////// |
| 123 | //[wcp_main |
| 124 | int main(int argc, char* argv[]) |
| 125 | { |
| 126 | /*< Define the token type to be used: `std::string` is available as the |
| 127 | type of the token attribute |
| 128 | >*/ typedef lex::lexertl::token< |
| 129 | char const*, boost::mpl::vector<std::string> |
| 130 | > token_type; |
| 131 | |
| 132 | /*< Define the lexer type to use implementing the state machine |
| 133 | >*/ typedef lex::lexertl::lexer<token_type> lexer_type; |
| 134 | |
| 135 | /*< Define the iterator type exposed by the lexer type |
| 136 | >*/ typedef word_count_tokens<lexer_type>::iterator_type iterator_type; |
| 137 | |
| 138 | // now we use the types defined above to create the lexer and grammar |
| 139 | // object instances needed to invoke the parsing process |
| 140 | word_count_tokens<lexer_type> word_count; // Our lexer |
| 141 | word_count_grammar<iterator_type> g (word_count); // Our parser |
| 142 | |
| 143 | // read in the file int memory |
| 144 | std::string str (read_from_file(infile: 1 == argc ? "word_count.input" : argv[1])); |
| 145 | char const* first = str.c_str(); |
| 146 | char const* last = &first[str.size()]; |
| 147 | |
| 148 | /*< Parsing is done based on the token stream, not the character |
| 149 | stream read from the input. The function `tokenize_and_parse()` wraps |
| 150 | the passed iterator range `[first, last)` by the lexical analyzer and |
| 151 | uses its exposed iterators to parse the token stream. |
| 152 | >*/ bool r = lex::tokenize_and_parse(first, last, lex: word_count, xpr: g); |
| 153 | |
| 154 | if (r) { |
| 155 | std::cout << "lines: " << g.l << ", words: " << g.w |
| 156 | << ", characters: " << g.c << "\n" ; |
| 157 | } |
| 158 | else { |
| 159 | std::string rest(first, last); |
| 160 | std::cerr << "Parsing failed\n" << "stopped at: \"" |
| 161 | << rest << "\"\n" ; |
| 162 | } |
| 163 | return 0; |
| 164 | } |
| 165 | //] |
| 166 | |