grammar.g

header {
	//don't try to add all the headers inside our namespace
	ANTLR_END_NAMESPACE

	#include <libwccl/parser/ParserException.h>

	#include <cstdio>
	#include <antlr/Token.hpp>
	#include <boost/lexical_cast.hpp>

	// values/variables
	#include <libwccl/variables.h>
	#include <libwccl/values/bool.h>
	#include <libwccl/values/tset.h>
	#include <libwccl/values/strset.h>
	#include <libwccl/values/position.h>
	
	// sentence context
	#include <libwccl/sentencecontext.h>

	// operators
	#include <libwccl/ops/operator.h>

	#include <libwccl/ops/functions/constant.h>
	#include <libwccl/ops/functions/vargetter.h>
	#include <libwccl/ops/functions/conditional.h>

	#include <libwccl/ops/functions/bool/varsetter.h>
	#include <libwccl/ops/functions/bool/predicates/or.h>
	#include <libwccl/ops/functions/bool/predicates/nor.h>
	#include <libwccl/ops/functions/bool/predicates/and.h>
	#include <libwccl/ops/functions/bool/predicates/regex.h>
	#include <libwccl/ops/functions/bool/predicates/intersects.h>
	#include <libwccl/ops/functions/bool/predicates/issubsetof.h>
	#include <libwccl/ops/functions/bool/predicates/isinside.h>
	#include <libwccl/ops/functions/bool/predicates/isoutside.h>
	#include <libwccl/ops/functions/bool/predicates/equals.h>
	#include <libwccl/ops/functions/bool/predicates/weakagreement.h>
	#include <libwccl/ops/functions/bool/predicates/pointagreement.h>
	#include <libwccl/ops/functions/bool/predicates/strongagreement.h>

	#include <libwccl/ops/functions/strset/affix.h>
	#include <libwccl/ops/functions/strset/getorth.h>
	#include <libwccl/ops/functions/strset/toupper.h>
	#include <libwccl/ops/functions/strset/tolower.h>
	#include <libwccl/ops/functions/strset/getlemmas.h>
	
	#include <libwccl/ops/functions/tset/agrfilter.h>
	#include <libwccl/ops/functions/tset/catfilter.h>
	#include <libwccl/ops/functions/tset/getsymbols.h>
	#include <libwccl/ops/functions/tset/getwordclass.h>
	#include <libwccl/ops/functions/tset/getsymbolsinrange.h>

	#include <libwccl/ops/functions/position/relativeposition.h>

	#include <libwccl/ops/functions/bool/iterations/only.h>
	#include <libwccl/ops/functions/bool/iterations/atleast.h>
	#include <libwccl/ops/functions/bool/iterations/leftlook.h>
	#include <libwccl/ops/functions/bool/iterations/rightlook.h>
	
	// Rules, actions
	#include <libwccl/ops/tagrule.h>
	#include <libwccl/ops/rulesequence.h>
	//
	#include <libwccl/ops/tagactions/unify.h>
	#include <libwccl/ops/tagactions/delete.h>
	#include <libwccl/ops/tagactions/select.h>
	#include <libwccl/ops/tagactions/relabel.h>


	// Unicode String
	#include <unicode/uniset.h>
	#include <unicode/unistr.h>

	// start our namespace again
	ANTLR_BEGIN_NAMESPACE(Wccl)
}

options {
	language = "Cpp";
	genHashLines = false;
	namespace = "Wccl";
//	genHashLines = true;
}

// ----------------------------------------------------------------------------
// ANTLR PARSER
// ----------------------------------------------------------------------------
class ANTLRParser extends Parser;
options {
	k = 1;
	buildAST = false;
	exportVocab = ANTLRExpr;
	defaultErrorHandler = false;
}
{
private:
	// 
	const UnicodeString token_ref_to_ustring(antlr::RefToken& rstr) const { 
		return UnicodeString::fromUTF8(((antlr::Token*)rstr)->getText()).unescape();
	}
	/*
	const UnicodeString str_token_ref_to_ustring(antlr::RefToken& rstr) const { 
		UnicodeString ret_ustr, ustr = token_ref_to_ustring(rstr);

		if (ustr.length() < 3) {
			return "";
		}

		ustr.extract(1, ustr.length() - 2, ret_ustr);

		return ret_ustr;
	}
	*/
	//
	const std::string str_token_rem_grav(antlr::RefToken& rstr) const {
		size_t len = 0;
		std::string ret = token_ref_to_std_string(rstr);

		if ((len = ret.length()) < 2) {
			return ret;
		}

		if (ret[0] == '`' && ret[len - 1] == '`') {
			return ret.substr(1, len - 2);
		}

		return ret;
	}
	//
	const std::string token_ref_to_std_string(antlr::RefToken& rstr) const { 
		return (((antlr::Token*)rstr)->getText());
	}
	//
	int token_ref_to_int(antlr::RefToken& rstr) { 
		return atoi(((antlr::Token*)rstr)->getText().c_str());
	}
}

///////////////////////////////////////////////////////////////////////////////
///////////////////////////////////////////////////////////////////////////////
// "GLOBAL" RULES
///////////////////////////////////////////////////////////////////////////////
///////////////////////////////////////////////////////////////////////////////

// ----------------------------------------------------------------------------
// Rule for parsing string set operator with scope. 
// Returns boost::shared_ptr<Operator<StrSet> >
parse_strset_operator
	[const Corpus2::Tagset &tagset]
	returns [boost::shared_ptr<Operator<StrSet> > res]
{
	Variables vars;
	boost::shared_ptr<Function<StrSet> > body;
}
	: body = strset_operator [tagset, vars] {
			res.reset(new Operator<StrSet>(body, vars));
	}
	EOF
;

// ----------------------------------------------------------------------------
// Rule for parsing bool operator with scope. 
// Returns boost::shared_ptr<Operator<Bool> > 
parse_bool_operator 
	[const Corpus2::Tagset &tagset]
	returns [boost::shared_ptr<Operator<Bool> > res]
{
	Variables vars;
	boost::shared_ptr<Function<Bool> > body;
}
	: body = bool_operator [tagset, vars] {
			res.reset(new Operator<Bool>(body, vars));
	}
	EOF
;

// ----------------------------------------------------------------------------
// Rule for parsing symbol set operator with scope.
// Returns boost::shared_ptr<Operator<TSet> >
parse_symset_operator
	[const Corpus2::Tagset &tagset]
	returns [boost::shared_ptr<Operator<TSet> > res]
{
	Variables vars;
	boost::shared_ptr<Function<TSet> > body;
}
	: body = symset_operator [tagset, vars] {
			res.reset(new Operator<TSet>(body, vars));
	}
	EOF
;

// ----------------------------------------------------------------------------
// Rule for parsing position operator with scope.
// Returns boost::shared_ptr<Operator<Position> >
parse_position_operator
	[const Corpus2::Tagset &tagset]
	returns [boost::shared_ptr<Operator<Position> > res]
{