diff --git a/libwccl/parser/grammar.g b/libwccl/parser/grammar.g
index 0fb8f619a7b5439728c0c44b263618313b5270e4..e0af47afc57497ed8812b0f811080208bd7b211d 100644
--- a/libwccl/parser/grammar.g
+++ b/libwccl/parser/grammar.g
@@ -57,7 +57,7 @@ options {
// ----------------------------------------------------------------------------
class ANTLRParser extends Parser;
options {
- k = 4;
+ k = 1;
buildAST = false;
exportVocab = ANTLRExpr;
defaultErrorHandler = false;
@@ -115,9 +115,6 @@ parse_string_operator
: op = string_operators [tagset, *res->variables.get()] {
res->op = op;
}
- | op = condit_str [tagset, *res->variables.get()] {
- res->op = op;
- }
;
// ----------------------------------------------------------------------------
@@ -133,9 +130,6 @@ parse_predicates
: op = logical_predicates [tagset, *res->variables.get()] {
res->op = op;
}
- | op = condit_bool [tagset, *res->variables.get()] {
- res->op = op;
- }
;
// ----------------------------------------------------------------------------
@@ -151,9 +145,6 @@ parse_sym_set_operator
: op = sym_set_operators [tagset, *res->variables.get()] {
res->op = op;
}
- | op = condit_sym [tagset, *res->variables.get()] {
- res->op = op;
- }
;
///////////////////////////////////////////////////////////////////////////////
@@ -170,15 +161,16 @@ str_set_literal
{
s_set.reset(new Wccl::StrSet());
}
- : LBRACKET RBRACKET
- | LBRACKET s1: STRING {
- s_set->insert(str_token_ref_to_ustring(s1));
- }
- (
- COMMA s2: STRING {
+ : LBRACKET
+ ( s1: STRING {
+ s_set->insert(str_token_ref_to_ustring(s1));
+ }
+ ( COMMA s2: STRING {
s_set->insert(str_token_ref_to_ustring(s2));
}
- )* RBRACKET
+ )*
+ )?
+ RBRACKET
;
// Constrant string set
// Returns boost::shared_ptr<Wccl::Constant<Wccl::StrSet> >
@@ -216,10 +208,9 @@ sym_set_literal
{
t_set.reset(new Wccl::TSet());
}
- : LCURLY RCURLY
- | LCURLY
- sym_set_elem[tagset, t_set] (COMMA sym_set_elem[tagset, t_set]) *
- RCURLY
+ : LCURLY
+ ( sym_set_elem[tagset, t_set] (COMMA sym_set_elem[tagset, t_set])* )?
+ RCURLY
;
// Constant symbol set
// Returns boost::shared_ptr<Wccl::Constant<Wccl::TSet> >
@@ -419,64 +410,71 @@ boolean_variable
setvar_op
[const Corpus2::Tagset& tagset, Wccl::Variables& vars]
returns [boost::shared_ptr<Wccl::Function<Wccl::Bool> > ret]
- : ret = setvar_pos [tagset, vars]
- | ret = setvar_bool [tagset, vars]
- | ret = setvar_sset [tagset, vars]
- | ret = setvar_tset [tagset, vars]
+ : "setvar" LPAREN
+ (
+ ret = setvar_body_pos [tagset, vars]
+ | ret = setvar_body_bool [tagset, vars]
+ | ret = setvar_body_sset [tagset, vars]
+ | ret = setvar_body_tset [tagset, vars]
+ )
+ RPAREN
;
// Implementations of setvar:
// ----------------------------------------------------------------------------
-setvar_pos
+setvar_body_pos
[const Corpus2::Tagset& tagset, Wccl::Variables& vars]
returns [boost::shared_ptr<Wccl::Function<Wccl::Bool> > op]
{
boost::shared_ptr<Wccl::Function<Wccl::Position> > ret_op;
boost::shared_ptr<Wccl::VariableAccessor<Wccl::Position> > ret_acc;
}
- : "setvar" LPAREN ret_acc = position_variable_acc [vars] COMMA
- ret_op = op_position [vars] RPAREN {
- op.reset(new Wccl::VarSetter<Wccl::Position>(*ret_acc.get(), ret_op));
- }
+ : ret_acc = position_variable_acc [vars]
+ COMMA
+ ret_op = op_position [vars] {
+ op.reset(new Wccl::VarSetter<Wccl::Position>(*ret_acc.get(), ret_op));
+ }
;
// ----------------------------------------------------------------------------
-setvar_bool
+setvar_body_bool
[const Corpus2::Tagset& tagset, Wccl::Variables& vars]
returns [boost::shared_ptr<Wccl::Function<Wccl::Bool> > op]
{
boost::shared_ptr<Wccl::Function<Wccl::Bool> > ret_op;
boost::shared_ptr<Wccl::VariableAccessor<Wccl::Bool> > ret_acc;
}
- : "setvar" LPAREN ret_acc = boolean_variable_acc [vars] COMMA
- ret_op = logical_predicates [tagset, vars] RPAREN {
- op.reset(new Wccl::VarSetter<Wccl::Bool>(*ret_acc.get(), ret_op));
- }
+ : ret_acc = boolean_variable_acc [vars]
+ COMMA
+ ret_op = logical_predicates [tagset, vars] {
+ op.reset(new Wccl::VarSetter<Wccl::Bool>(*ret_acc.get(), ret_op));
+ }
;
// ----------------------------------------------------------------------------
-setvar_sset
+setvar_body_sset
[const Corpus2::Tagset& tagset, Wccl::Variables& vars]
returns [boost::shared_ptr<Wccl::Function<Wccl::Bool> > op]
{
boost::shared_ptr<Wccl::Function<Wccl::StrSet> > ret_op;
boost::shared_ptr<Wccl::VariableAccessor<Wccl::StrSet> > ret_acc;
}
- : "setvar" LPAREN ret_acc = str_set_variable_acc [vars] COMMA
- ret_op = string_operators [tagset, vars] RPAREN {
+ : ret_acc = str_set_variable_acc [vars]
+ COMMA ret_op = string_operators [tagset, vars] {
op.reset(new Wccl::VarSetter<Wccl::StrSet>(*ret_acc.get(), ret_op));
}
;
// ----------------------------------------------------------------------------
-setvar_tset
+setvar_body_tset
[const Corpus2::Tagset& tagset, Wccl::Variables& vars]
returns [boost::shared_ptr<Wccl::Function<Wccl::Bool> > op]
{
boost::shared_ptr<Wccl::Function<Wccl::TSet> > ret_op;
boost::shared_ptr<Wccl::VariableAccessor<Wccl::TSet> > ret_acc;
}
- : "setvar" LPAREN ret_acc = sym_set_variable_acc [vars] COMMA
- ret_op = sym_set_operators [tagset, vars] RPAREN {
+ : ret_acc = sym_set_variable_acc [vars]
+ COMMA
+ ret_op = sym_set_operators [tagset, vars] {
op.reset(new Wccl::VarSetter<Wccl::TSet>(*ret_acc.get(), ret_op));
}
;
@@ -490,7 +488,7 @@ sym_set_operators
[const Corpus2::Tagset& tagset, Wccl::Variables& vars]
returns [boost::shared_ptr<Wccl::Function<Wccl::TSet> > ret]
: ret = op_sym_set [tagset, vars]
-// | ret = condit_sym [vars]
+ | ret = condit_sym [tagset, vars]
;
// Implementations of symbol set operators:
// ----------------------------------------------------------------------------
@@ -523,12 +521,9 @@ condit_sym
}
}
| Q_MARK
- (p_true = sym_set_operators [tagset, vars] |
- p_true = condit_sym [tagset, vars])
+ (p_true = sym_set_operators [tagset, vars])
Q_MARK
- (test = logical_predicates [tagset, vars] |
- test = condit_bool [tagset, vars])
- {
+ (test = logical_predicates [tagset, vars]) {
op.reset(new Wccl::Conditional<Wccl::TSet>(test, p_true));
}
;
@@ -565,7 +560,7 @@ string_operators
| ret = op_upper [tagset, vars]
| ret = op_affix [tagset, vars]
| ret = op_str_set [tagset, vars]
-// | ret = condit_str [vars]
+ | ret = condit_str [tagset, vars]
;
// Implementations of string operators:
// ----------------------------------------------------------------------------
@@ -654,12 +649,9 @@ condit_str
}
}
| Q_MARK
- (p_true = string_operators [tagset, vars] |
- p_true = condit_str [tagset, vars])
+ p_true = string_operators [tagset, vars]
Q_MARK
- (test = logical_predicates [tagset, vars] |
- test = condit_bool [tagset, vars])
- {
+ test = logical_predicates [tagset, vars] {
op.reset(new Wccl::Conditional<Wccl::StrSet>(test, p_true));
}
;
@@ -677,12 +669,12 @@ logical_predicates
| ret = lpred_nor [tagset, vars]
| ret = lpred_bool [tagset, vars]
| ret = lpred_in [tagset, vars]
- | ret = lpred_inter [tagset, vars]
+ | ret = lpred_inter [tagset, vars]
| ret = lpred_eq [tagset, vars]
| ret = lpred_regex [tagset, vars]
| ret = setvar_op [tagset, vars]
| ret = lpred_inout [tagset, vars]
-// | ret = condit_bool [vars]
+ | ret = condit_bool [tagset, vars]
;
// ----------------------------------------------------------------------------
// comma-separated predicates
@@ -758,14 +750,26 @@ lpred_in
boost::shared_ptr<Wccl::Function<Wccl::TSet> > t1, t2;
boost::shared_ptr<Wccl::Function<Wccl::StrSet> > s1, s2;
}
- : "in" LPAREN t1 = sym_set_operators [tagset, vars] COMMA
- t2 = sym_set_operators [tagset, vars] RPAREN {
- op.reset(new Wccl::IsSubsetOf<Wccl::TSet>(t1, t2));
- }
- | "in" LPAREN s1 = string_operators [tagset, vars] COMMA
- s2 = string_operators [tagset, vars] RPAREN {
- op.reset(new Wccl::IsSubsetOf<Wccl::StrSet>(s1, s2));
- }
+:
+ "in" LPAREN
+ (
+ (sym_set_operators [tagset, vars]) =>
+ (
+ t1 = sym_set_operators [tagset, vars] COMMA
+ t2 = sym_set_operators [tagset, vars] {
+ op.reset(new Wccl::IsSubsetOf<Wccl::TSet>(t1, t2));
+ }
+ )
+ |
+ (
+ s1 = string_operators [tagset, vars] COMMA
+ s2 = string_operators [tagset, vars] {
+ op.reset(new Wccl::IsSubsetOf<Wccl::StrSet>(s1, s2));
+ }
+ )
+ )
+ RPAREN
+
;
// ----------------------------------------------------------------------------
@@ -776,14 +780,25 @@ lpred_inter
boost::shared_ptr<Wccl::Function<Wccl::TSet> > t1, t2;
boost::shared_ptr<Wccl::Function<Wccl::StrSet> > s1, s2;
}
- : "inter" LPAREN s1 = string_operators [tagset, vars] COMMA
- s2 = string_operators [tagset, vars] RPAREN {
- op.reset(new Wccl::Intersects<Wccl::StrSet>(s1, s2));
- }
- | "inter" LPAREN t1 = sym_set_operators [tagset, vars] COMMA
- t2 = sym_set_operators [tagset, vars] RPAREN {
- op.reset(new Wccl::Intersects<Wccl::TSet>(t1, t2));
- }
+ :
+ "inter" LPAREN
+ (
+ (sym_set_operators [tagset, vars]) =>
+ (
+ t1 = sym_set_operators [tagset, vars] COMMA
+ t2 = sym_set_operators [tagset, vars] {
+ op.reset(new Wccl::Intersects<Wccl::TSet>(t1, t2));
+ }
+ )
+ |
+ (
+ s1 = string_operators [tagset, vars] COMMA
+ s2 = string_operators [tagset, vars] {
+ op.reset(new Wccl::Intersects<Wccl::StrSet>(s1, s2));
+ }
+ )
+ )
+ RPAREN
;
// ----------------------------------------------------------------------------
@@ -796,22 +811,39 @@ lpred_eq
boost::shared_ptr<Wccl::Function<Wccl::StrSet> > s1, s2;
boost::shared_ptr<Wccl::Function<Wccl::Position> > p1, p2;
}
- : "equal" LPAREN s1 = string_operators [tagset, vars] COMMA
- s2 = string_operators [tagset, vars] RPAREN {
- op.reset(new Wccl::Equals<Wccl::StrSet>(s1, s2));
- }
- | "equal" LPAREN t1 = sym_set_operators [tagset, vars] COMMA
- t2 = sym_set_operators [tagset, vars] RPAREN {
- op.reset(new Wccl::Equals<Wccl::TSet>(t1, t2));
- }
- | "equal" LPAREN p1 = position_operators [tagset, vars] COMMA
- p2 = position_operators [tagset, vars] RPAREN {
- op.reset(new Wccl::Equals<Wccl::Position>(p1, p2));
- }
- | "equal" LPAREN b1 = logical_predicates [tagset, vars] COMMA
- b2 = logical_predicates [tagset, vars] RPAREN {
- op.reset(new Wccl::Equals<Wccl::Bool>(b1, b2));
- }
+ : "equal" LPAREN
+ (
+ (
+ p1 = position_operators [tagset, vars] COMMA
+ p2 = position_operators [tagset, vars] {
+ op.reset(new Wccl::Equals<Wccl::Position>(p1, p2));
+ }
+ )
+ |
+ (sym_set_operators [tagset, vars]) =>
+ (
+ t1 = sym_set_operators [tagset, vars] COMMA
+ t2 = sym_set_operators [tagset, vars] {
+ op.reset(new Wccl::Equals<Wccl::TSet>(t1, t2));
+ }
+ )
+ |
+ (string_operators [tagset, vars]) =>
+ (
+ s1 = string_operators [tagset, vars] COMMA
+ s2 = string_operators [tagset, vars] {
+ op.reset(new Wccl::Equals<Wccl::StrSet>(s1, s2));
+ }
+ )
+ |
+ (
+ b1 = logical_predicates [tagset, vars] COMMA
+ b2 = logical_predicates [tagset, vars] {
+ op.reset(new Wccl::Equals<Wccl::Bool>(b1, b2));
+ }
+ )
+ )
+ RPAREN
;
// ----------------------------------------------------------------------------
@@ -862,12 +894,9 @@ condit_bool
}
}
| Q_MARK
- (p_true = logical_predicates [tagset, vars] |
- p_true = condit_bool [tagset, vars])
+ p_true = logical_predicates [tagset, vars]
Q_MARK
- (test = logical_predicates [tagset, vars] |
- test = condit_bool [tagset, vars])
- {
+ test = logical_predicates [tagset, vars] {
op.reset(new Wccl::Conditional<Wccl::Bool>(test, p_true));
}
;
@@ -881,13 +910,13 @@ class ANTLRLexer extends Lexer;
options {
exportVocab = ANTLRExpr;
charVocabulary = '\3'..'\377';
- k = 3;
+ testLiterals = false;
+ k = 2;
}
STRING
options {
paraphrase = "a string";
- testLiterals = true;
}
: '"' (~'"')* '"'
| '\'' (~'\'')* '\''
@@ -896,7 +925,6 @@ options {
INT
options {
paraphrase = "Integer";
- testLiterals = true;
}
: ('-'|'+')? ('0'..'9')+
;
@@ -904,7 +932,6 @@ options {
QUOT_MARK
options {
paraphrase = "Quota mark";
- testLiterals = true;
}
: '\''
;
@@ -912,7 +939,6 @@ options {
APOS_MARK
options {
paraphrase = "Aposptrophe mark";
- testLiterals = true;
}
: '"'
;
@@ -920,7 +946,6 @@ options {
Q_MARK
options {
paraphrase = "Query mark";
- testLiterals = true;
}
: '?'
;
@@ -928,7 +953,6 @@ options {
E_MARK
options {
paraphrase = "Exclamanation mark";
- testLiterals = true;
}
: '!'
;
@@ -936,7 +960,6 @@ options {
STR_PREFIX
options {
paraphrase = "String prefix";
- testLiterals = true;
}
: "$s:"
;
@@ -944,7 +967,6 @@ options {
TST_PREFIX
options {
paraphrase = "Tag set (symbol) prefix";
- testLiterals = true;
}
: "$t:"
;
@@ -952,7 +974,6 @@ options {
BOOL_PREFIX
options {
paraphrase = "Bool prefix";
- testLiterals = true;
}
: "$b:"
;
@@ -967,7 +988,6 @@ options {
LBRACKET
options {
paraphrase = "'['";
- testLiterals = true;
}
: '['
;
@@ -975,7 +995,6 @@ options {
RBRACKET
options {
paraphrase = "']'";
- testLiterals = true;
}
: ']'
;
@@ -983,7 +1002,6 @@ options {
LPAREN
options {
paraphrase = "'('";
- testLiterals = true;
}
: '('
;
@@ -991,7 +1009,6 @@ options {
RPAREN
options {
paraphrase = "')'";
- testLiterals = true;
}
: ')'
;
@@ -999,7 +1016,6 @@ options {
LCURLY
options {
paraphrase = "'{'";
- testLiterals = true;
}
: '{'
;
@@ -1007,7 +1023,6 @@ options {
RCURLY
options {
paraphrase = "'}'";
- testLiterals = true;
}
: '}'
;
@@ -1015,15 +1030,13 @@ options {
AT_MARK
options {
paraphrase = "'@'";
- testLiterals = true;
}
: '@'
;
COMMA
options {
- paraphrase = "','";
- testLiterals = true;
+ paraphrase = "','";
}
: ','
;
@@ -1039,25 +1052,50 @@ options {
WS
: ( ' '
- | '\t'
- | '\r' '\n' { newline(); }
- | '\n' { newline(); }
- ) { $setType(antlr::Token::SKIP); }
+ | '\t'
+ | '\f'
+ | ( "\r\n"
+ | '\r'
+ | '\n'
+ )
+ { newline(); }
+ )
+ { $setType(antlr::Token::SKIP); }
;
COMMENT
options {
- paraphrase = "Comment";
- testLiterals = true;
+ paraphrase = "Single line comment";
+}
+ : "//" (~('\n'|'\r'))* { $setType(antlr::Token::SKIP); }
+;
+
+ML_COMMENT
+options {
+ paraphrase = "Multi line comment";
}
- : "//" (~'\n')* '\n'{ $setType(antlr::Token::SKIP); newline(); }
- | "/*" (.)* "*/" { $setType(antlr::Token::SKIP); }
+ : "/*"
+ ( /* This actually works OK despite the ambiguity that
+ '\r' '\n' can be matched in one alternative or by matching
+ '\r' in one iteration and '\n' in another.. But
+ this is really matched just by one rule per (...)*
+ loop iteration, so it's OK.
+ This is exactly how they do it all over the web - just
+ turn off the warning for this particular token.*/
+ options { generateAmbigWarnings=false; }
+ : { LA(2)!='/' }? '*'
+ | '\r' '\n' { newline(); }
+ | '\r' { newline(); }
+ | '\n' { newline(); }
+ | ~('*'|'\n'|'\r')
+ )*
+ "*/"
+ {$setType(antlr::Token::SKIP);}
;
HASH
options {
paraphrase = "'#'";
- testLiterals = true;
}
: '#'
;
@@ -1065,7 +1103,6 @@ options {
DSEPARATOR
options {
paraphrase = "':-'";
- testLiterals = true;
}
: ":-"
;
diff --git a/wcclparser/main.cpp b/wcclparser/main.cpp
index 86e850b554c590845e54c561568a75316322ed85..c048ec83d51f0d002a710b34f804af51966c185d 100644
--- a/wcclparser/main.cpp
+++ b/wcclparser/main.cpp
@@ -26,6 +26,7 @@ namespace {
void std_read_loop(boost::function<bool (const std::string&)>& line_cb)
{
while (std::cin.good()) {
+ std::cout << _prompt << std::flush;
std::string s;
getline(std::cin, s);
if (line_cb(s)) {