Commit d5e439c5 authored by Dominik Charousset's avatar Dominik Charousset

Add an FSM-based one-pass string parser

parent 350b1941
...@@ -70,6 +70,11 @@ ...@@ -70,6 +70,11 @@
goto s_##target; \ goto s_##target; \
} }
#define default_action(target, statement) \
statement; \
ch = ps.next(); \
goto s_##target;
#define checked_action(predicate, target, statement, error_code) \ #define checked_action(predicate, target, statement, error_code) \
if (predicate(ch)) { \ if (predicate(ch)) { \
if (statement) { \ if (statement) { \
...@@ -81,5 +86,15 @@ ...@@ -81,5 +86,15 @@
} \ } \
} }
#define invalid_input(predicate, error_code) \
if (predicate(ch)) { \
ps.code = error_code; \
return; \
}
#define default_failure(error_code) \
ps.code = error_code; \
return;
#define epsilon(target) goto e_##target; #define epsilon(target) goto e_##target;
/******************************************************************************
* ____ _ _____ *
* / ___| / \ | ___| C++ *
* | | / _ \ | |_ Actor *
* | |___ / ___ \| _| Framework *
* \____/_/ \_|_| *
* *
* Copyright 2011-2018 Dominik Charousset *
* *
* Distributed under the terms and conditions of the BSD 3-Clause License or *
* (at your option) under the terms and conditions of the Boost Software *
* License 1.0. See accompanying files LICENSE and LICENSE_ALTERNATIVE. *
* *
* If you did not receive a copy of the license files, see *
* http://opensource.org/licenses/BSD-3-Clause and *
* http://www.boost.org/LICENSE_1_0.txt. *
******************************************************************************/
#pragma once
#include <cstdint>
#include <string>
#include "caf/detail/scope_guard.hpp"
#include "caf/detail/parser/ec.hpp"
#include "caf/detail/parser/fsm.hpp"
#include "caf/detail/parser/is_char.hpp"
#include "caf/detail/parser/state.hpp"
namespace caf {
namespace detail {
namespace parser {
/// Reads a number, i.e., on success produces either an `int64_t` or a
/// `double`.
template <class Iterator, class Sentinel, class Consumer>
void read_string(state<Iterator, Sentinel>& ps, Consumer& consumer) {
std::string res;
auto g = caf::detail::make_scope_guard([&] {
if (ps.code <= ec::trailing_character)
consumer.value(std::move(res));
});
start();
state(init) {
input(is_char<' '>, init)
input(is_char<'\t'>, init)
input(is_char<'"'>, read_chars)
}
state(read_chars) {
input(is_char<'\\'>, escape)
input(is_char<'"'>, done)
invalid_input(is_char<'\n'>, ec::unexpected_newline)
default_action(read_chars, res += ch)
}
state(escape) {
action(is_char<'n'>, read_chars, res += '\n')
action(is_char<'r'>, read_chars, res += '\r')
action(is_char<'t'>, read_chars, res += '\t')
action(is_char<'\\'>, read_chars, res += '\\')
action(is_char<'"'>, read_chars, res += '"')
default_failure(ec::illegal_escape_sequence)
}
term_state(done) {
input(is_char<' '>, done)
input(is_char<'\t'>, done)
}
fin();
}
} // namespace parser
} // namespace detail
} // namespace caf
/******************************************************************************
* ____ _ _____ *
* / ___| / \ | ___| C++ *
* | | / _ \ | |_ Actor *
* | |___ / ___ \| _| Framework *
* \____/_/ \_|_| *
* *
* Copyright 2011-2018 Dominik Charousset *
* *
* Distributed under the terms and conditions of the BSD 3-Clause License or *
* (at your option) under the terms and conditions of the Boost Software *
* License 1.0. See accompanying files LICENSE and LICENSE_ALTERNATIVE. *
* *
* If you did not receive a copy of the license files, see *
* http://opensource.org/licenses/BSD-3-Clause and *
* http://www.boost.org/LICENSE_1_0.txt. *
******************************************************************************/
#define CAF_SUITE read_string
#include <string>
#include "caf/test/unit_test.hpp"
#include "caf/variant.hpp"
#include "caf/detail/parser/read_string.hpp"
using namespace caf;
using detail::parser::ec;
namespace {
struct string_parser_consumer {
std::string x;
inline void value(std::string y) {
x = std::move(y);
}
};
using res_t = variant<ec, std::string>;
struct string_parser {
res_t operator()(std::string str) {
detail::parser::state<std::string::iterator> res;
string_parser_consumer f;
res.i = str.begin();
res.e = str.end();
detail::parser::read_string(res, f);
if (res.code == ec::success)
return f.x;
return res.code;
}
};
struct fixture {
string_parser p;
};
// TODO: remove and use "..."s from the STL when switching to C++14
std::string operator "" _s(const char* str, size_t str_len) {
std::string result;
result.assign(str, str_len);
return result;
}
} // namespace <anonymous>
CAF_TEST_FIXTURE_SCOPE(read_string_tests, fixture)
CAF_TEST(empty string) {
CAF_CHECK_EQUAL(p(R"("")"), ""_s);
CAF_CHECK_EQUAL(p(R"( "")"), ""_s);
CAF_CHECK_EQUAL(p(R"( "")"), ""_s);
CAF_CHECK_EQUAL(p(R"("" )"), ""_s);
CAF_CHECK_EQUAL(p(R"("" )"), ""_s);
CAF_CHECK_EQUAL(p(R"( "" )"), ""_s);
CAF_CHECK_EQUAL(p("\t \"\" \t\t\t "), ""_s);
}
CAF_TEST(non-empty string) {
CAF_CHECK_EQUAL(p(R"("abc")"), "abc"_s);
CAF_CHECK_EQUAL(p(R"("a b c")"), "a b c"_s);
CAF_CHECK_EQUAL(p(R"( "abcdefABCDEF" )"), "abcdefABCDEF"_s);
}
CAF_TEST(string with escaped characters) {
CAF_CHECK_EQUAL(p(R"("a\tb\tc")"), "a\tb\tc"_s);
CAF_CHECK_EQUAL(p(R"("a\nb\r\nc")"), "a\nb\r\nc"_s);
CAF_CHECK_EQUAL(p(R"("a\\b")"), "a\\b"_s);
CAF_CHECK_EQUAL(p(R"("foo = \"bar\"")"), "foo = \"bar\""_s);
}
CAF_TEST(invalid strings) {
CAF_CHECK_EQUAL(p(R"("abc)"), ec::unexpected_eof);
CAF_CHECK_EQUAL(p("\"ab\nc\""), ec::unexpected_newline);
CAF_CHECK_EQUAL(p(R"("foo \i bar")"), ec::illegal_escape_sequence);
CAF_CHECK_EQUAL(p(R"(foo)"), ec::unexpected_character);
CAF_CHECK_EQUAL(p(R"("abc" def)"), ec::trailing_character);
}
CAF_TEST_FIXTURE_SCOPE_END()
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment