/*
 * [The "BSD license"]
 *  Copyright (c) 2010 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.
 */
package org.antlr.tool;

import org.antlr.analysis.DFA;
import org.antlr.analysis.NFAState;
import org.antlr.grammar.v3.ANTLRParser;
import org.antlr.misc.IntSet;
import org.antlr.runtime.CommonToken;
import org.antlr.runtime.Token;
import org.antlr.runtime.tree.CommonTree;
import org.antlr.runtime.tree.Tree;
import org.stringtemplate.v4.ST;

import java.util.*;

Grammars are first converted to ASTs using this class and then are converted to NFAs via a tree walker. The reader may notice that I have made a very non-OO decision in this class to track variables for many different kinds of nodes. It wastes space for nodes that don't need the values and OO principles cry out for a new class type for each kind of node in my tree. I am doing this on purpose for a variety of reasons. I don't like using the type system for different node types; it yields too many damn class files which I hate. Perhaps if I put them all in one file. Most importantly though I hate all the type casting that would have to go on. I would have all sorts of extra work to do. Ick. Anyway, I'm doing all this on purpose, not out of ignorance. ;)
/** Grammars are first converted to ASTs using this class and then are * converted to NFAs via a tree walker. * * The reader may notice that I have made a very non-OO decision in this * class to track variables for many different kinds of nodes. It wastes * space for nodes that don't need the values and OO principles cry out * for a new class type for each kind of node in my tree. I am doing this * on purpose for a variety of reasons. I don't like using the type * system for different node types; it yields too many damn class files * which I hate. Perhaps if I put them all in one file. Most importantly * though I hate all the type casting that would have to go on. I would * have all sorts of extra work to do. Ick. Anyway, I'm doing all this * on purpose, not out of ignorance. ;) */
public class GrammarAST extends CommonTree { static int count = 0; public int ID = ++count; private String textOverride; public String enclosingRuleName;
If this is a decision node, what is the lookahead DFA?
/** If this is a decision node, what is the lookahead DFA? */
public DFA lookaheadDFA = null;
What NFA start state was built from this node?
/** What NFA start state was built from this node? */
public NFAState NFAStartState = null;
This is used for TREE_BEGIN nodes to point into the NFA. TREE_BEGINs point at left edge of DOWN for LOOK computation purposes (Nullable tree child list needs special code gen when matching).
/** This is used for TREE_BEGIN nodes to point into * the NFA. TREE_BEGINs point at left edge of DOWN for LOOK computation * purposes (Nullable tree child list needs special code gen when matching). */
public NFAState NFATreeDownState = null;
Rule ref nodes, token refs, set, and NOT set refs need to track their location in the generated NFA so that local FOLLOW sets can be computed during code gen for automatic error recovery.
/** Rule ref nodes, token refs, set, and NOT set refs need to track their * location in the generated NFA so that local FOLLOW sets can be * computed during code gen for automatic error recovery. */
public NFAState followingNFAState = null;
If this is a SET node, what are the elements?
/** If this is a SET node, what are the elements? */
protected IntSet setValue = null;
If this is a BLOCK node, track options here
/** If this is a BLOCK node, track options here */
protected Map<String,Object> blockOptions;
If this is a BLOCK node for a rewrite rule, track referenced elements here. Don't track elements in nested subrules.
/** If this is a BLOCK node for a rewrite rule, track referenced * elements here. Don't track elements in nested subrules. */
public Set<GrammarAST> rewriteRefsShallow; /* If REWRITE node, track EVERY element and label ref to right of -> * for this rewrite rule. There could be multiple of these per * rule: * * a : ( ... -> ... | ... -> ... ) -> ... ; * * We may need a list of all refs to do definitions for whole rewrite * later. * * If BLOCK then tracks every element at that level and below. */ public Set<GrammarAST> rewriteRefsDeep; public Map<String,Object> terminalOptions;
if this is an ACTION node, this is the outermost enclosing alt num in rule. For actions, define.g sets these (used to be codegen.g). We need these set so we can examine actions early, before code gen, for refs to rule predefined properties and rule labels. For most part define.g sets outerAltNum, but codegen.g does the ones for %foo(a={$ID.text}) type refs as the {$ID...} is not seen as an action until code gen pulls apart.
/** if this is an ACTION node, this is the outermost enclosing * alt num in rule. For actions, define.g sets these (used to * be codegen.g). We need these set so we can examine actions * early, before code gen, for refs to rule predefined properties * and rule labels. For most part define.g sets outerAltNum, but * codegen.g does the ones for %foo(a={$ID.text}) type refs as * the {$ID...} is not seen as an action until code gen pulls apart. */
public int outerAltNum;
if this is a TOKEN_REF or RULE_REF node, this is the code ST generated for this node. We need to update it later to add a label if someone does $tokenref or $ruleref in an action.
/** if this is a TOKEN_REF or RULE_REF node, this is the code ST * generated for this node. We need to update it later to add * a label if someone does $tokenref or $ruleref in an action. */
public ST code; /** * */ public Map<String, Object> getBlockOptions() { return blockOptions; }
Params:
  • blockOptions –
/** * * @param blockOptions */
public void setBlockOptions(Map<String, Object> blockOptions) { this.blockOptions = blockOptions; } public GrammarAST() {} @SuppressWarnings("OverridableMethodCallInConstructor") public GrammarAST(int t, String txt) { initialize(t,txt); } @SuppressWarnings("OverridableMethodCallInConstructor") public GrammarAST(Token token) { initialize(token); } public void initialize(int i, String s) { token = new CommonToken(i,s); token.setTokenIndex(-1); } public void initialize(Tree ast) { GrammarAST t = ((GrammarAST)ast); this.startIndex = t.startIndex; this.stopIndex = t.stopIndex; this.token = t.token; this.enclosingRuleName = t.enclosingRuleName; this.setValue = t.setValue; this.blockOptions = t.blockOptions; this.outerAltNum = t.outerAltNum; } public void initialize(Token token) { this.token = token; if ( token!=null ) { startIndex = token.getTokenIndex(); stopIndex = startIndex; } } public DFA getLookaheadDFA() { return lookaheadDFA; } public void setLookaheadDFA(DFA lookaheadDFA) { this.lookaheadDFA = lookaheadDFA; } public NFAState getNFAStartState() { return NFAStartState; } public void setNFAStartState(NFAState nfaStartState) { this.NFAStartState = nfaStartState; }
Save the option key/value pair and process it; return the key or null if invalid option.
/** Save the option key/value pair and process it; return the key * or null if invalid option. */
public String setBlockOption(Grammar grammar, String key, Object value) { if ( blockOptions == null ) { blockOptions = new HashMap<String, Object>(); } return setOption(blockOptions, Grammar.legalBlockOptions, grammar, key, value); } public String setTerminalOption(Grammar grammar, String key, Object value) { if ( terminalOptions == null ) { terminalOptions = new HashMap<String,Object>(); } return setOption(terminalOptions, Grammar.legalTokenOptions, grammar, key, value); } public String setOption(Map<String, Object> options, Set<String> legalOptions, Grammar grammar, String key, Object value) { if ( !legalOptions.contains(key) ) { ErrorManager.grammarError(ErrorManager.MSG_ILLEGAL_OPTION, grammar, token, key); return null; } if ( value instanceof String ) { String vs = (String)value; if ( vs.charAt(0)=='"' ) { value = vs.substring(1,vs.length()-1); // strip quotes } } if ( key.equals("k") ) { grammar.numberOfManualLookaheadOptions++; } if ( key.equals("backtrack") && value.toString().equals("true") ) { grammar.composite.getRootGrammar().atLeastOneBacktrackOption = true; } options.put(key, value); return key; } public Object getBlockOption(String key) { Object value = null; if ( blockOptions != null ) { value = blockOptions.get(key); } return value; } public void setOptions(Grammar grammar, Map<String, Object> options) { if ( options==null ) { this.blockOptions = null; return; } String[] keys = options.keySet().toArray(new String[options.size()]); for (String optionName : keys) { String stored= setBlockOption(grammar, optionName, options.get(optionName)); if ( stored==null ) { options.remove(optionName); } } } @Override public String getText() { if ( textOverride!=null ) return textOverride; if ( token!=null ) { return token.getText(); } return ""; } public void setType(int type) { token.setType(type); } public void setText(String text) { textOverride = text; // don't alt tokens as others might see } @Override public int getType() { if ( token!=null ) { return token.getType(); } return -1; } @Override public int getLine() { int line=0; if ( token!=null ) { line = token.getLine(); } if ( line==0 ) { Tree child = getChild(0); if ( child!=null ) { line = child.getLine(); } } return line; } @Override public int getCharPositionInLine(){ int col=0; if ( token!=null ) { col = token.getCharPositionInLine(); } if ( col==0 ) { Tree child = getChild(0); if ( child!=null ) { col = child.getCharPositionInLine(); } } return col; } public void setLine(int line) { token.setLine(line); } public void setCharPositionInLine(int value){ token.setCharPositionInLine(value); } public IntSet getSetValue() { return setValue; } public void setSetValue(IntSet setValue) { this.setValue = setValue; } public GrammarAST getLastChild() { if (getChildCount() == 0) return null; return (GrammarAST)getChild(getChildCount() - 1); } public GrammarAST getNextSibling() { return (GrammarAST)getParent().getChild(getChildIndex() + 1); } public GrammarAST getLastSibling() { Tree parent = getParent(); if ( parent==null ) { return null; } return (GrammarAST)parent.getChild(parent.getChildCount() - 1); } public GrammarAST[] getChildrenAsArray() { List<? extends Object> children = getChildren(); if (children == null) { return new GrammarAST[0]; } return children.toArray(new GrammarAST[children.size()]); } private static final GrammarAST DescendantDownNode = new GrammarAST(Token.DOWN, "DOWN"); private static final GrammarAST DescendantUpNode = new GrammarAST(Token.UP, "UP"); public static List<Tree> descendants(Tree root){ return descendants(root, false); } public static List<Tree> descendants(Tree root, boolean insertDownUpNodes){ List<Tree> result = new ArrayList<Tree>(); int count = root.getChildCount(); if (insertDownUpNodes){ result.add(root); result.add(DescendantDownNode); for (int i = 0 ; i < count ; i++){ Tree child = root.getChild(i); for (Tree subchild : descendants(child, true)) result.add(subchild); } result.add(DescendantUpNode); }else{ result.add(root); for (int i = 0 ; i < count ; i++){ Tree child = root.getChild(i); for (Tree subchild : descendants(child, false)) result.add(subchild); } } return result; } public GrammarAST findFirstType(int ttype) { // check this node (the root) first if ( this.getType()==ttype ) { return this; } // else check children List<Tree> descendants = descendants(this); for (Tree child : descendants) { if ( child.getType()==ttype ) { return (GrammarAST)child; } } return null; } public List<GrammarAST> findAllType(int ttype) { List<GrammarAST> nodes = new ArrayList<GrammarAST>(); _findAllType(ttype, nodes); return nodes; } public void _findAllType(int ttype, List<GrammarAST> nodes) { // check this node (the root) first if ( this.getType()==ttype ) nodes.add(this); // check children for (int i = 0; i < getChildCount(); i++){ GrammarAST child = (GrammarAST)getChild(i); child._findAllType(ttype, nodes); } }
Make nodes unique based upon Token so we can add them to a Set; if not a GrammarAST, check type.
/** Make nodes unique based upon Token so we can add them to a Set; if * not a GrammarAST, check type. */
@Override public boolean equals(Object ast) { if ( this == ast ) { return true; } if ( !(ast instanceof GrammarAST) ) { return this.getType() == ((Tree)ast).getType(); } GrammarAST t = (GrammarAST)ast; return token.getLine() == t.getLine() && token.getCharPositionInLine() == t.getCharPositionInLine(); }
Make nodes unique based upon Token so we can add them to a Set; if not a GrammarAST, check type.
/** Make nodes unique based upon Token so we can add them to a Set; if * not a GrammarAST, check type. */
@Override public int hashCode(){ if (token == null) return 0; return token.hashCode(); }
See if tree has exact token types and structure; no text
/** See if tree has exact token types and structure; no text */
public boolean hasSameTreeStructure(Tree other) { // check roots first. if (this.getType() != other.getType()) return false; // if roots match, do full list match test on children. Iterator<Tree> thisDescendants = descendants(this, true).iterator(); Iterator<Tree> otherDescendants = descendants(other, true).iterator(); while (thisDescendants.hasNext()) { if (!otherDescendants.hasNext()) return false; if (thisDescendants.next().getType() != otherDescendants.next().getType()) return false; } return !otherDescendants.hasNext(); } public static GrammarAST dup(Tree t) { if ( t==null ) { return null; } GrammarAST dup_t = new GrammarAST(); dup_t.initialize(t); return dup_t; } @Override public Tree dupNode(){ return dup(this); }
Duplicate a tree, assuming this is a root node of a tree-- duplicate that node and what's below; ignore siblings of root node.
/**Duplicate a tree, assuming this is a root node of a tree-- * duplicate that node and what's below; ignore siblings of root node. */
public static GrammarAST dupTreeNoActions(GrammarAST t, GrammarAST parent) { if ( t==null ) { return null; } GrammarAST result = (GrammarAST)t.dupNode(); for (GrammarAST subchild : getChildrenForDupTree(t)) { result.addChild(dupTreeNoActions(subchild, result)); } return result; } private static List<GrammarAST> getChildrenForDupTree(GrammarAST t) { List<GrammarAST> result = new ArrayList<GrammarAST>(); for (int i = 0; i < t.getChildCount(); i++){ GrammarAST child = (GrammarAST)t.getChild(i); int ttype = child.getType(); if (ttype == ANTLRParser.REWRITES || ttype == ANTLRParser.REWRITE || ttype==ANTLRParser.ACTION) { continue; } if (ttype == ANTLRParser.BANG || ttype == ANTLRParser.ROOT) { for (GrammarAST subchild : getChildrenForDupTree(child)) result.add(subchild); } else { result.add(child); } } if ( result.size()==1 && result.get(0).getType()==ANTLRParser.EOA && t.getType()==ANTLRParser.ALT ) { // can't have an empty alt, insert epsilon result.add(0, new GrammarAST(ANTLRParser.EPSILON, "epsilon")); } return result; } public static GrammarAST dupTree(GrammarAST t) { if ( t==null ) { return null; } GrammarAST root = dup(t); // make copy of root // copy all children of root. for (int i= 0; i < t.getChildCount(); i++) { GrammarAST child = (GrammarAST)t.getChild(i); root.addChild(dupTree(child)); } return root; } public void setTreeEnclosingRuleNameDeeply(String rname) { enclosingRuleName = rname; if (getChildCount() == 0) return; for (Object child : getChildren()) { if (!(child instanceof GrammarAST)) { continue; } GrammarAST grammarAST = (GrammarAST)child; grammarAST.setTreeEnclosingRuleNameDeeply(rname); } } public String toStringList() { String result = toStringTree(); if (this.getNextSibling() != null) { result += ' ' + getNextSibling().toStringList(); } return result; }
Track start/stop token for subtree root created for a rule. Only works with Tree nodes. For rules that match nothing, seems like this will yield start=i and stop=i-1 in a nil node. Might be useful info so I'll not force to be i..i.
/** Track start/stop token for subtree root created for a rule. * Only works with Tree nodes. For rules that match nothing, * seems like this will yield start=i and stop=i-1 in a nil node. * Might be useful info so I'll not force to be i..i. */
public void setTokenBoundaries(Token startToken, Token stopToken) { if ( startToken!=null ) startIndex = startToken.getTokenIndex(); if ( stopToken!=null ) stopIndex = stopToken.getTokenIndex(); } public GrammarAST getBlockALT(int i) { if ( this.getType()!=ANTLRParser.BLOCK ) return null; int alts = 0; for (int j =0 ; j < getChildCount(); j++) { if (getChild(j).getType() == ANTLRParser.ALT) { alts++; } if (alts == i) { return (GrammarAST)getChild(j); } } return null; } }