1 /* 2 [The "BSD license"] 3 Copyright (c) 2005-2009 Terence Parr 4 All rights reserved. 5 6 Redistribution and use in source and binary forms, with or without 7 modification, are permitted provided that the following conditions 8 are met: 9 1. Redistributions of source code must retain the above copyright 10 notice, this list of conditions and the following disclaimer. 11 2. Redistributions in binary form must reproduce the above copyright 12 notice, this list of conditions and the following disclaimer in the 13 documentation and/or other materials provided with the distribution. 14 3. The name of the author may not be used to endorse or promote products 15 derived from this software without specific prior written permission. 16 17 THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR 18 IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES 19 OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. 20 IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, 21 INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT 22 NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, 23 DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY 24 THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT 25 (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF 26 THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. 27 */ 28 package org.antlr.runtime.tree; 29 30 import org.antlr.runtime.RecognizerSharedState; 31 import org.antlr.runtime.RecognitionException; 32 import org.antlr.runtime.TokenStream; 33 34 /** 35 Cut-n-paste from material I'm not using in the book anymore (edit later 36 to make sense): 37 38 Now, how are we going to test these tree patterns against every 39 subtree in our original tree? In what order should we visit nodes? 40 For this application, it turns out we need a simple ``apply once'' 41 rule application strategy and a ``down then up'' tree traversal 42 strategy. Let's look at rule application first. 43 44 As we visit each node, we need to see if any of our patterns match. If 45 a pattern matches, we execute the associated tree rewrite and move on 46 to the next node. In other words, we only look for a single rule 47 application opportunity (we'll see below that we sometimes need to 48 repeatedly apply rules). The following method applies a rule in a @cl 49 TreeParser (derived from a tree grammar) to a tree: 50 51 here is where weReferenced code/walking/patterns/TreePatternMatcher.java 52 53 It uses reflection to lookup the appropriate rule within the generated 54 tree parser class (@cl Simplify in this case). Most of the time, the 55 rule will not match the tree. To avoid issuing syntax errors and 56 attempting error recovery, it bumps up the backtracking level. Upon 57 failure, the invoked rule immediately returns. If you don't plan on 58 using this technique in your own ANTLR-based application, don't sweat 59 the details. This method boils down to ``call a rule to match a tree, 60 executing any embedded actions and rewrite rules.'' 61 62 At this point, we know how to define tree grammar rules and how to 63 apply them to a particular subtree. The final piece of the tree 64 pattern matcher is the actual tree traversal. We have to get the 65 correct node visitation order. In particular, we need to perform the 66 scalar-vector multiply transformation on the way down (preorder) and 67 we need to reduce multiply-by-zero subtrees on the way up (postorder). 68 69 To implement a top-down visitor, we do a depth first walk of the tree, 70 executing an action in the preorder position. To get a bottom-up 71 visitor, we execute an action in the postorder position. ANTLR 72 provides a standard @cl TreeVisitor class with a depth first search @v 73 visit method. That method executes either a @m pre or @m post method 74 or both. In our case, we need to call @m applyOnce in both. On the way 75 down, we'll look for @r vmult patterns. On the way up, 76 we'll look for @r mult0 patterns. 77 */ 78 public class TreeFilter extends TreeParser { 79 public interface fptr { 80 public void rule() throws RecognitionException; 81 } 82 83 protected TokenStream originalTokenStream; 84 protected TreeAdaptor originalAdaptor; 85 86 public TreeFilter(TreeNodeStream input) { 87 this(input, new RecognizerSharedState()); 88 } 89 public TreeFilter(TreeNodeStream input, RecognizerSharedState state) { 90 super(input, state); 91 originalAdaptor = input.getTreeAdaptor(); 92 originalTokenStream = input.getTokenStream(); 93 } 94 95 public void applyOnce(Object t, fptr whichRule) { 96 if ( t==null ) return; 97 try { 98 // share TreeParser object but not parsing-related state 99 state = new RecognizerSharedState(); 100 input = new CommonTreeNodeStream(originalAdaptor, t); 101 ((CommonTreeNodeStream)input).setTokenStream(originalTokenStream); 102 setBacktrackingLevel(1); 103 whichRule.rule(); 104 setBacktrackingLevel(0); 105 } 106 catch (RecognitionException e) { ; } 107 } 108 109 public void downup(Object t) { 110 TreeVisitor v = new TreeVisitor(new CommonTreeAdaptor()); 111 TreeVisitorAction actions = new TreeVisitorAction() { 112 public Object pre(Object t) { applyOnce(t, topdown_fptr); return t; } 113 public Object post(Object t) { applyOnce(t, bottomup_fptr); return t; } 114 }; 115 v.visit(t, actions); 116 } 117 118 fptr topdown_fptr = new fptr() { 119 public void rule() throws RecognitionException { 120 topdown(); 121 } 122 }; 123 124 fptr bottomup_fptr = new fptr() { 125 public void rule() throws RecognitionException { 126 bottomup(); 127 } 128 }; 129 130 // methods the downup strategy uses to do the up and down rules. 131 // to override, just define tree grammar rule topdown and turn on 132 // filter=true. 133 public void topdown() throws RecognitionException {;} 134 public void bottomup() throws RecognitionException {;} 135 } 136
