/* [The "BSD licence"] Copyright (c) 2005-2006 Terence Parr All rights reserved. Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met: 1. Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer. 2. Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution. 3. The name of the author may not be used to endorse or promote products derived from this software without specific prior written permission. THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. */ interface ANTLRCore; /** The overall file structure of a recognizer; stores methods for rules * and cyclic DFAs plus support code. */ outputFile(LEXER,PARSER,TREE_PARSER, actionScope, actions, docComment, recognizer, name, tokens, tokenNames, rules, cyclicDFAs, bitsets, buildTemplate, buildAST, rewriteMode, profile, backtracking, synpreds, memoize, numRules, fileName, ANTLRVersion, generatedTimestamp, trace, scopes, superClass, literals); /** The header file; make sure to define headerFileExtension() below */ optional headerFile(LEXER,PARSER,TREE_PARSER, actionScope, actions, docComment, recognizer, name, tokens, tokenNames, rules, cyclicDFAs, bitsets, buildTemplate, buildAST, rewriteMode, profile, backtracking, synpreds, memoize, numRules, fileName, ANTLRVersion, generatedTimestamp, trace, scopes, superClass, literals); lexer(grammar, name, tokens, scopes, rules, numRules, labelType, filterMode, superClass); parser(grammar, name, scopes, tokens, tokenNames, rules, numRules, bitsets, ASTLabelType, superClass, labelType, members); /** How to generate a tree parser; same as parser except the input * stream is a different type. */ treeParser(grammar, name, scopes, tokens, tokenNames, globalAction, rules, numRules, bitsets, labelType, ASTLabelType, superClass, members); /** A simpler version of a rule template that is specific to the imaginary * rules created for syntactic predicates. As they never have return values * nor parameters etc..., just give simplest possible method. Don't do * any of the normal memoization stuff in here either; it's a waste. * As predicates cannot be inlined into the invoking rule, they need to * be in a rule by themselves. */ synpredRule(ruleName, ruleDescriptor, block, description, nakedBlock); /** How to generate code for a rule. This includes any return type * data aggregates required for multiple return values. */ rule(ruleName,ruleDescriptor,block,emptyRule,description,exceptions,finally,memoize); /** How to generate a rule in the lexer; naked blocks are used for * fragment rules. */ lexerRule(ruleName,nakedBlock,ruleDescriptor,block,memoize); /** How to generate code for the implicitly-defined lexer grammar rule * that chooses between lexer rules. */ tokensRule(ruleName,nakedBlock,args,block,ruleDescriptor); filteringNextToken(); filteringActionGate(); // S U B R U L E S /** A (...) subrule with multiple alternatives */ block(alts,decls,decision,enclosingBlockLevel,blockLevel,decisionNumber,maxK,maxAlt,description); /** A rule block with multiple alternatives */ ruleBlock(alts,decls,decision,enclosingBlockLevel,blockLevel,decisionNumber,maxK,maxAlt,description); ruleBlockSingleAlt(alts,decls,decision,enclosingBlockLevel,blockLevel,decisionNumber,description); /** A special case of a (...) subrule with a single alternative */ blockSingleAlt(alts,decls,decision,enclosingBlockLevel,blockLevel,decisionNumber,description); /** A (..)+ block with 0 or more alternatives */ positiveClosureBlock(alts,decls,decision,enclosingBlockLevel,blockLevel,decisionNumber,maxK,maxAlt,description); positiveClosureBlockSingleAlt(alts,decls,decision,enclosingBlockLevel,blockLevel,decisionNumber,maxK,maxAlt,description); /** A (..)* block with 0 or more alternatives */ closureBlock(alts,decls,decision,enclosingBlockLevel,blockLevel,decisionNumber,maxK,maxAlt,description); closureBlockSingleAlt(alts,decls,decision,enclosingBlockLevel,blockLevel,decisionNumber,maxK,maxAlt,description); /** Optional blocks (x)? are translated to (x|) by before code generation * so we can just use the normal block template */ optionalBlock(alts,decls,decision,enclosingBlockLevel,blockLevel,decisionNumber,maxK,maxAlt,description); optionalBlockSingleAlt(alts,decls,decision,enclosingBlockLevel,blockLevel,decisionNumber,maxK,maxAlt,description); /** An alternative is just a list of elements; at outermost level */ alt(elements,altNum,description,autoAST,outerAlt,treeLevel,rew); // E L E M E N T S /** match a token optionally with a label in front */ tokenRef(token,label,elementIndex,hetero); /** ids+=ID */ tokenRefAndListLabel(token,label,elementIndex,hetero); listLabel(label,elem); /** match a character */ charRef(char,label); /** match a character range */ charRangeRef(a,b,label); /** For now, sets are interval tests and must be tested inline */ matchSet(s,label,elementIndex,postmatchCode); matchSetAndListLabel(s,label,elementIndex,postmatchCode); /** Match a string literal */ lexerStringRef(string,label); wildcard(label,elementIndex); wildcardAndListLabel(label,elementIndex); /** Match . wildcard in lexer */ wildcardChar(label, elementIndex); wildcardCharListLabel(label, elementIndex); /** Match a rule reference by invoking it possibly with arguments * and a return value or values. */ ruleRef(rule,label,elementIndex,args,scope); /** ids+=ID */ ruleRefAndListLabel(rule,label,elementIndex,args,scope); /** A lexer rule reference */ lexerRuleRef(rule,label,args,elementIndex,scope); /** i+=INT in lexer */ lexerRuleRefAndListLabel(rule,label,args,elementIndex,scope); /** EOF in the lexer */ lexerMatchEOF(label,elementIndex); /** match ^(root children) in tree parser */ tree(root, actionsAfterRoot, children, nullableChildList, enclosingTreeLevel, treeLevel); /** Every predicate is used as a validating predicate (even when it is * also hoisted into a prediction expression). */ validateSemanticPredicate(pred,description); // F i x e d D F A (if-then-else) dfaState(k,edges,eotPredictsAlt,description,stateNumber,semPredState); /** Same as a normal DFA state except that we don't examine lookahead * for the bypass alternative. It delays error detection but this * is faster, smaller, and more what people expect. For (X)? people * expect "if ( LA(1)==X ) match(X);" and that's it. * * If a semPredState, don't force lookahead lookup; preds might not * need. */ dfaOptionalBlockState(k,edges,eotPredictsAlt,description,stateNumber,semPredState); /** A DFA state that is actually the loopback decision of a closure * loop. If end-of-token (EOT) predicts any of the targets then it * should act like a default clause (i.e., no error can be generated). * This is used only in the lexer so that for ('a')* on the end of a * rule anything other than 'a' predicts exiting. * * If a semPredState, don't force lookahead lookup; preds might not * need. */ dfaLoopbackState(k,edges,eotPredictsAlt,description,stateNumber,semPredState); /** An accept state indicates a unique alternative has been predicted */ dfaAcceptState(alt); /** A simple edge with an expression. If the expression is satisfied, * enter to the target state. To handle gated productions, we may * have to evaluate some predicates for this edge. */ dfaEdge(labelExpr, targetState, predicates); // F i x e d D F A (switch case) /** A DFA state where a SWITCH may be generated. The code generator * decides if this is possible: CodeGenerator.canGenerateSwitch(). */ dfaStateSwitch(k,edges,eotPredictsAlt,description,stateNumber,semPredState); dfaOptionalBlockStateSwitch(k,edges,eotPredictsAlt,description,stateNumber,semPredState); dfaLoopbackStateSwitch(k, edges,eotPredictsAlt,description,stateNumber,semPredState); dfaEdgeSwitch(labels, targetState); // C y c l i c D F A /** The code to initiate execution of a cyclic DFA; this is used * in the rule to predict an alt just like the fixed DFA case. * The attribute is inherited via the parser, lexer, ... */ dfaDecision(decisionNumber,description); /** Generate the tables and support code needed for the DFAState object * argument. Unless there is a semantic predicate (or syn pred, which * become sem preds), all states should be encoded in the state tables. * Consequently, cyclicDFAState/cyclicDFAEdge,eotDFAEdge templates are * not used except for special DFA states that cannot be encoded as * a transition table. */ cyclicDFA(dfa); /** A special state in a cyclic DFA; special means has a semantic predicate * or it's a huge set of symbols to check. */ cyclicDFAState(decisionNumber,stateNumber,edges,needErrorClause,semPredState); /** Just like a fixed DFA edge, test the lookahead and indicate what * state to jump to next if successful. Again, this is for special * states. */ cyclicDFAEdge(labelExpr, targetStateNumber, edgeNumber, predicates); /** An edge pointing at end-of-token; essentially matches any char; * always jump to the target. */ eotDFAEdge(targetStateNumber,edgeNumber, predicates); // D F A E X P R E S S I O N S andPredicates(left,right); orPredicates(operands); notPredicate(pred); evalPredicate(pred,description); evalSynPredicate(pred,description); lookaheadTest(atom,k,atomAsInt); /** Sometimes a lookahead test cannot assume that LA(k) is in a temp variable * somewhere. Must ask for the lookahead directly. */ isolatedLookaheadTest(atom,k,atomAsInt); lookaheadRangeTest(lower,upper,k,rangeNumber,lowerAsInt,upperAsInt); isolatedLookaheadRangeTest(lower,upper,k,rangeNumber,lowerAsInt,upperAsInt); setTest(ranges); // A T T R I B U T E S parameterAttributeRef(attr); parameterSetAttributeRef(attr,expr); scopeAttributeRef(scope,attr,index,negIndex); scopeSetAttributeRef(scope,attr,expr,index,negIndex); /** $x is either global scope or x is rule with dynamic scope; refers * to stack itself not top of stack. This is useful for predicates * like {$function.size()>0 && $function::name.equals("foo")}? */ isolatedDynamicScopeRef(scope); /** reference an attribute of rule; might only have single return value */ ruleLabelRef(referencedRule,scope,attr); returnAttributeRef(ruleDescriptor,attr); returnSetAttributeRef(ruleDescriptor,attr,expr); /** How to translate $tokenLabel */ tokenLabelRef(label); /** ids+=ID {$ids} or e+=expr {$e} */ listLabelRef(label); // not sure the next are the right approach; and they are evaluated early; // they cannot see TREE_PARSER or PARSER attributes for example. :( tokenLabelPropertyRef_text(scope,attr); tokenLabelPropertyRef_type(scope,attr); tokenLabelPropertyRef_line(scope,attr); tokenLabelPropertyRef_pos(scope,attr); tokenLabelPropertyRef_channel(scope,attr); tokenLabelPropertyRef_index(scope,attr); tokenLabelPropertyRef_tree(scope,attr); ruleLabelPropertyRef_start(scope,attr); ruleLabelPropertyRef_stop(scope,attr); ruleLabelPropertyRef_tree(scope,attr); ruleLabelPropertyRef_text(scope,attr); ruleLabelPropertyRef_st(scope,attr); /** Isolated $RULE ref ok in lexer as it's a Token */ lexerRuleLabel(label); lexerRuleLabelPropertyRef_type(scope,attr); lexerRuleLabelPropertyRef_line(scope,attr); lexerRuleLabelPropertyRef_pos(scope,attr); lexerRuleLabelPropertyRef_channel(scope,attr); lexerRuleLabelPropertyRef_index(scope,attr); lexerRuleLabelPropertyRef_text(scope,attr); // Somebody may ref $template or $tree or $stop within a rule: rulePropertyRef_start(scope,attr); rulePropertyRef_stop(scope,attr); rulePropertyRef_tree(scope,attr); rulePropertyRef_text(scope,attr); rulePropertyRef_st(scope,attr); lexerRulePropertyRef_text(scope,attr); lexerRulePropertyRef_type(scope,attr); lexerRulePropertyRef_line(scope,attr); lexerRulePropertyRef_pos(scope,attr); /** Undefined, but present for consistency with Token attributes; set to -1 */ lexerRulePropertyRef_index(scope,attr); lexerRulePropertyRef_channel(scope,attr); lexerRulePropertyRef_start(scope,attr); lexerRulePropertyRef_stop(scope,attr); ruleSetPropertyRef_tree(scope,attr,expr); ruleSetPropertyRef_st(scope,attr,expr); /** How to execute an action */ execAction(action); // M I S C (properties, etc...) codeFileExtension(); /** Your language needs a header file; e.g., ".h" */ optional headerFileExtension(); true(); false();