I am writing a parser for a type of input file. The input file looks something like:
[CalculationBlock]
CalculationTitle="Test Parser Input System" , MatchingRadius=25.0, StepSize=0.01,ProblemType=RelSchroedingerEqn
MaxPartialWaveJ=800, SMatConv=10E-8
PartialWaveConv= 10E-8, SmallValueLimit = 10E-8
PotentialRadType=HeavyIon
[end]
Essentially it is divided into blocks that start with [BlockName] and then have a set of named parameters within. The named parameters can be separated by ',' or '\n' characters.
Using the incomplete input file I gave above, I wanted to write a parser for it that would serve as a jumping off point for a more complete input file. I did so but the parser has a weakness that I am not sure how to address. It is not parameter order independent. For example, if a user were to put the parameter PartialWaveConv= 10E-8 before SMatConv=10E-8 it would fail.
I briefly contemplated enumerating each possible order of parameters in a block but I discarded it since there are n! permutations of n parameter value pairs. So my question is: Is there any way to make the parser independent of parameter ordering?
The toy parser I wrote is below, I apologize if it is amateurish, this is my first foray into boost, let alone boost.spirit.
#include<string>
#include<iostream>
#include<cstdlib>
#include<fstream>
#include<boost/config/warning_disable.hpp>
#include<boost/spirit/include/qi.hpp>
#include<boost/spirit/include/phoenix_core.hpp>
#include<boost/spirit/include/phoenix_operator.hpp>
#include<boost/spirit/include/phoenix_object.hpp>
#include<boost/fusion/include/adapt_struct.hpp>
#include<boost/fusion/include/io.hpp>
#include<boost/spirit/include/support_istream_iterator.hpp>
namespace blocks
{
namespace qi = boost::spirit::qi;
namespace ascii = boost::spirit::ascii;
namespace phoenix = boost::phoenix;
struct CalcBlock
{
std::string calculationTitle;
float matchingRad;
float stepSize;
std::string problemType;
int maxPartialWaveJ;
float sMatrixConvergenceValue;
float partialWaveConvergenceValue;
float smallValueLimit;
std::string potentialRadType;
};
}
//tell fusion about the block structure
BOOST_FUSION_ADAPT_STRUCT(blocks::CalcBlock,
(std::string, calculationTitle)
(float, matchingRad)
(float, stepSize)
(std::string, problemType)
(int, maxPartialWaveJ)
(float, sMatrixConvergenceValue)
(float, partialWaveConvergenceValue)
(float, smallValueLimit)
(std::string, potentialRadType)
)
namespace blocks
{
template <typename Iterator>
struct CalcBlockParser : qi::grammar<Iterator, CalcBlock(), boost::spirit::ascii::blank_type>
{
CalcBlockParser() : CalcBlockParser::base_type(start)
{
using qi::int_;
using qi::lit;
using qi::float_;
using qi::lexeme;
using ascii::char_;
quotedString %= lexeme['"' >> +(char_ - '"' - '\n') >> '"'];
plainString %= lexeme[ +(char_ - ' ' - ',' - '\n') ];
start %=
lit("[CalculationBlock]") >> '\n'
>> lit("CalculationTitle") >> '=' >> quotedString >> (lit(',') | lit('\n'))
>> lit("MatchingRadius") >> '=' >> float_ >> (lit(',') | lit('\n'))
>> lit("StepSize") >> '=' >> float_ >> (lit(',') | lit('\n'))
>> lit("ProblemType") >> '=' >> plainString >> (lit(',') | lit('\n'))
>> lit("MaxPartialWaveJ") >> '=' >> int_ >> (lit(',') | lit('\n'))
>> lit("SMatConv") >> '=' >> float_ >> (lit(',') | lit('\n'))
>> lit("PartialWaveConv") >> '=' >> float_ >> (lit(',') | lit('\n'))
>> lit("SmallValueLimit") >> '=' >> float_ >> (lit(',') | lit('\n'))
>> lit("PotentialRadType") >> '=' >> plainString
>> lit("\n[end]\n");
}
qi::rule<Iterator, std::string(), boost::spirit::ascii::blank_type> quotedString;
qi::rule<Iterator, std::string(), boost::spirit::ascii::blank_type> plainString;
qi::rule<Iterator, CalcBlock(), boost::spirit::ascii::blank_type> start;
};
}
using std::cout;
using std::endl;
namespace spirit = boost::spirit;
int main(int argc, char *argv[])
{
if (argc != 2)
{
cout << "\nUsage:\n\t./echos InputFileName\n" << endl;
return EXIT_FAILURE;
}
std::string inputFileName(argv[1]);
cout << "Reading input from the file: " << inputFileName << endl;
std::ifstream input(inputFileName);
input.unsetf(std::ios::skipws);
spirit::istream_iterator start(input);
spirit::istream_iterator stop;
typedef blocks::CalcBlockParser<spirit::istream_iterator> CalcBlockParser;
CalcBlockParser cbParser;
blocks::CalcBlock cb;
bool success = phrase_parse(start, stop, cbParser, boost::spirit::ascii::blank, cb);
if (success && start == stop)
{
std::cout << boost::fusion::tuple_open('[');
std::cout << boost::fusion::tuple_close(']');
std::cout << boost::fusion::tuple_delimiter(", ");
std::cout << "-------------------------\n";
std::cout << "Parsing succeeded\n";
std::cout << "got: " << boost::fusion::as_vector(cb) << std::endl;
std::cout << "\n-------------------------\n";
}
else
{
std::cout << boost::fusion::tuple_open('[');
std::cout << boost::fusion::tuple_close(']');
std::cout << boost::fusion::tuple_delimiter(", ");
std::cout << "-------------------------\n";
std::cout << "Parsing failed\n";
std::cout << "got: " << boost::fusion::as_vector(cb) << std::endl;
std::cout << "\n-------------------------\n";
}
return EXIT_SUCCESS;
}
Just for fun/completeness I reviewed the grammar and came up with the following test.
I have made a few improvement suggestions left and right (as the OP witnessed on the live stream), and the resulting code, test and output are here:
#include <boost/fusion/include/adapt_struct.hpp>
#include <boost/spirit/include/qi.hpp>
#include <fstream>
#include <iostream>
namespace blocks {
struct CalcBlock {
std::string calculationTitle;
float matchingRad;
float stepSize;
std::string problemType;
int maxPartialWaveJ;
float sMatrixConvergenceValue;
float partialWaveConvergenceValue;
float smallValueLimit;
std::string potentialRadType;
};
}
BOOST_FUSION_ADAPT_STRUCT(blocks::CalcBlock, // Boost 1.58+ style adapt-struct
calculationTitle, matchingRad, stepSize, problemType, maxPartialWaveJ,
sMatrixConvergenceValue, partialWaveConvergenceValue, smallValueLimit,
potentialRadType)
namespace blocks {
namespace qi = boost::spirit::qi;
template <typename Iterator>
struct CalcBlockParser : qi::grammar<Iterator, CalcBlock()> {
CalcBlockParser() : CalcBlockParser::base_type(start) {
using namespace qi;
auto eol_ = copy((',' >> *eol) | +eol); // http://stackoverflow.com/a/26411266/85371 (!)
quotedString = '"' >> +~char_("\"\n") >> '"';
plainString = +~char_(" ,\n");
start = skip(blank) [cbRule];
cbRule = lexeme["[CalculationBlock]"] >> eol
>> (
(lexeme["CalculationTitle"] >> '=' >> quotedString >> eol_)
^ (lexeme["MatchingRadius"] >> '=' >> float_ >> eol_)
^ (lexeme["StepSize"] >> '=' >> float_ >> eol_)
^ (lexeme["ProblemType"] >> '=' >> plainString >> eol_)
^ (lexeme["MaxPartialWaveJ"] >> '=' >> int_ >> eol_)
^ (lexeme["SMatConv"] >> '=' >> float_ >> eol_)
^ (lexeme["PartialWaveConv"] >> '=' >> float_ >> eol_)
^ (lexeme["SmallValueLimit"] >> '=' >> float_ >> eol_)
^ (lexeme["PotentialRadType"] >> '=' >> plainString >> eol_)
)
>> lexeme["[end]"]
>> *eol
>> eoi;
}
private:
qi::rule<Iterator, CalcBlock()> start;
qi::rule<Iterator, CalcBlock(), qi::blank_type> cbRule;
// lexemes:
qi::rule<Iterator, std::string()> quotedString, plainString;
};
}
using boost::fusion::as_vector;
typedef boost::spirit::istream_iterator It;
int main(int argc, char **argv) {
if (argc != 2) {
std::cout << "Usage:\n\t" << argv[0] << " InputFileName" << std::endl;
return 1;
}
std::string inputFileName(argv[1]);
std::cout << "Reading input from the file: " << inputFileName << std::endl;
std::ifstream input(inputFileName);
input.unsetf(std::ios::skipws);
It start(input), stop;
blocks::CalcBlock cb;
blocks::CalcBlockParser<It> cbParser;
bool success = parse(start, stop, cbParser, cb);
{
using namespace boost::fusion;
std::cout << tuple_open('[') << tuple_close(']') << tuple_delimiter(", ");
}
std::cout << "-------------------------\n";
std::cout << "Parsing " << (success?"succeeded":"failed") << "\n";
std::cout << "got: " << as_vector(cb) << "\n";
std::cout << "-------------------------\n";
}
Input:
[CalculationBlock]
CalculationTitle="Test Parser Input System"
SMatConv=10E-8,
PartialWaveConv= 10E-8, MaxPartialWaveJ=800, SmallValueLimit = 10E-8
PotentialRadType=HeavyIon , MatchingRadius=25.0, StepSize=0.01,ProblemType=RelSchroedingerEqn
[end]
Output:
Reading input from the file: input.txt
-------------------------
Parsing succeeded
got: [Test Parser Input System, 25, 0.01, RelSchroedingerEqn, 800, 1e-07, 1e-07, 1e-07, HeavyIon]
-------------------------