InstructionFinder.java

  1. /*
  2.  * Licensed to the Apache Software Foundation (ASF) under one or more
  3.  * contributor license agreements.  See the NOTICE file distributed with
  4.  * this work for additional information regarding copyright ownership.
  5.  * The ASF licenses this file to You under the Apache License, Version 2.0
  6.  * (the "License"); you may not use this file except in compliance with
  7.  * the License.  You may obtain a copy of the License at
  8.  *
  9.  *      http://www.apache.org/licenses/LICENSE-2.0
  10.  *
  11.  *  Unless required by applicable law or agreed to in writing, software
  12.  *  distributed under the License is distributed on an "AS IS" BASIS,
  13.  *  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
  14.  *  See the License for the specific language governing permissions and
  15.  *  limitations under the License.
  16.  */
  17. package org.apache.bcel.util;

  19. import java.util.ArrayList;
  20. import java.util.Arrays;
  21. import java.util.HashMap;
  22. import java.util.Iterator;
  23. import java.util.List;
  24. import java.util.Locale;
  25. import java.util.Map;
  26. import java.util.regex.Matcher;
  27. import java.util.regex.Pattern;

  29. import org.apache.bcel.Const;
  30. import org.apache.bcel.generic.ClassGenException;
  31. import org.apache.bcel.generic.InstructionHandle;
  32. import org.apache.bcel.generic.InstructionList;

  34. /**
  35.  * InstructionFinder is a tool to search for given instructions patterns, i.e., match sequences of instructions in an
  36.  * instruction list via regular expressions. This can be used, e.g., in order to implement a peep hole optimizer that
  37.  * looks for code patterns and replaces them with faster equivalents.
  38.  *
  39.  * <p>
  40.  * This class internally uses the java.util.regex package to search for regular expressions.
  41.  *
  42.  * A typical application would look like this:
  43.  *
  44.  * <pre>
  45.  *
  46.  *
  47.  *   InstructionFinder f   = new InstructionFinder(il);
  48.  *   String            pat = &quot;IfInstruction ICONST_0 GOTO ICONST_1 NOP (IFEQ|IFNE)&quot;;
  49.  *
  50.  *   for (Iterator i = f.search(pat, constraint); i.hasNext(); ) {
  51.  *   InstructionHandle[] match = (InstructionHandle[])i.next();
  52.  *   ...
  53.  *   il.delete(match[1], match[5]);
  54.  *   ...
  55.  *   }
  56.  *
  57.  *
  58.  * </pre>
  59.  *
  60.  * @see org.apache.bcel.generic.Instruction
  61.  * @see InstructionList
  62.  */
  63. public class InstructionFinder {

  65.     /**
  66.      * Code patterns found may be checked using an additional user-defined constraint object whether they really match the
  67.      * needed criterion. I.e., check constraints that can not expressed with regular expressions.
  68.      */
  69.     public interface CodeConstraint {

  71.         /**
  72.          * @param match array of instructions matching the requested pattern
  73.          * @return true if the matched area is really useful
  74.          */
  75.         boolean checkCode(InstructionHandle[] match);
  76.     }

  78.     private static final int OFFSET = 32767; // char + OFFSET is outside of LATIN-1
  79.     private static final int NO_OPCODES = 256; // Potential number, some are not used
  80.     private static final Map<String, String> map = new HashMap<>();

  82.     // Initialize pattern map
  83.     static {
  84.         map.put("arithmeticinstruction",
  85.             "(irem|lrem|iand|ior|ineg|isub|lneg|fneg|fmul|ldiv|fadd|lxor|frem|idiv|land|ixor|ishr|fsub|lshl|fdiv|iadd|lor|dmul|lsub|ishl|imul|lmul|lushr|dneg|iushr|lshr|ddiv|drem|dadd|ladd|dsub)");
  86.         map.put("invokeinstruction", "(invokevirtual|invokeinterface|invokestatic|invokespecial|invokedynamic)");
  87.         map.put("arrayinstruction",
  88.             "(baload|aastore|saload|caload|fastore|lastore|iaload|castore|iastore|aaload|bastore|sastore|faload|laload|daload|dastore)");
  89.         map.put("gotoinstruction", "(goto|goto_w)");
  90.         map.put("conversioninstruction", "(d2l|l2d|i2s|d2i|l2i|i2b|l2f|d2f|f2i|i2d|i2l|f2d|i2c|f2l|i2f)");
  91.         map.put("localvariableinstruction", "(fstore|iinc|lload|dstore|dload|iload|aload|astore|istore|fload|lstore)");
  92.         map.put("loadinstruction", "(fload|dload|lload|iload|aload)");
  93.         map.put("fieldinstruction", "(getfield|putstatic|getstatic|putfield)");
  94.         map.put("cpinstruction",
  95.             "(ldc2_w|invokeinterface|invokedynamic|multianewarray|putstatic|instanceof|getstatic|checkcast|getfield|invokespecial|ldc_w|invokestatic|invokevirtual|putfield|ldc|new|anewarray)");
  96.         map.put("stackinstruction", "(dup2|swap|dup2_x2|pop|pop2|dup|dup2_x1|dup_x2|dup_x1)");
  97.         map.put("branchinstruction",
  98.             "(ifle|if_acmpne|if_icmpeq|if_acmpeq|ifnonnull|goto_w|iflt|ifnull|if_icmpne|tableswitch|if_icmple|ifeq|if_icmplt|jsr_w|if_icmpgt|ifgt|jsr|goto|ifne|ifge|lookupswitch|if_icmpge)");
  99.         map.put("returninstruction", "(lreturn|ireturn|freturn|dreturn|areturn|return)");
  100.         map.put("storeinstruction", "(istore|fstore|dstore|astore|lstore)");
  101.         map.put("select", "(tableswitch|lookupswitch)");
  102.         map.put("ifinstruction",
  103.             "(ifeq|ifgt|if_icmpne|if_icmpeq|ifge|ifnull|ifne|if_icmple|if_icmpge|if_acmpeq|if_icmplt|if_acmpne|ifnonnull|iflt|if_icmpgt|ifle)");
  104.         map.put("jsrinstruction", "(jsr|jsr_w)");
  105.         map.put("variablelengthinstruction", "(tableswitch|jsr|goto|lookupswitch)");
  106.         map.put("unconditionalbranch", "(goto|jsr|jsr_w|athrow|goto_w)");
  107.         map.put("constantpushinstruction", "(dconst|bipush|sipush|fconst|iconst|lconst)");
  108.         map.put("typedinstruction",
  109.             "(imul|lsub|aload|fload|lor|new|aaload|fcmpg|iand|iaload|lrem|idiv|d2l|isub|dcmpg|dastore|ret|f2d|f2i|drem|iinc|i2c|checkcast|frem|lreturn|astore|lushr|daload|dneg|fastore|istore|lshl|ldiv|lstore|areturn|ishr|ldc_w|invokeinterface|invokedynamic|aastore|lxor|ishl|l2d|i2f|return|faload|sipush|iushr|caload|instanceof|invokespecial|putfield|fmul|ireturn|laload|d2f|lneg|ixor|i2l|fdiv|lastore|multianewarray|i2b|getstatic|i2d|putstatic|fcmpl|saload|ladd|irem|dload|jsr_w|dconst|dcmpl|fsub|freturn|ldc|aconst_null|castore|lmul|ldc2_w|dadd|iconst|f2l|ddiv|dstore|land|jsr|anewarray|dmul|bipush|dsub|sastore|d2i|i2s|lshr|iadd|l2i|lload|bastore|fstore|fneg|iload|fadd|baload|fconst|ior|ineg|dreturn|l2f|lconst|getfield|invokevirtual|invokestatic|iastore)");
  110.         map.put("popinstruction", "(fstore|dstore|pop|pop2|astore|putstatic|istore|lstore)");
  111.         map.put("allocationinstruction", "(multianewarray|new|anewarray|newarray)");
  112.         map.put("indexedinstruction",
  113.             "(lload|lstore|fload|ldc2_w|invokeinterface|invokedynamic|multianewarray|astore|dload|putstatic|instanceof|getstatic|checkcast|getfield|invokespecial|dstore|istore|iinc|ldc_w|ret|fstore|invokestatic|iload|putfield|invokevirtual|ldc|new|aload|anewarray)");
  114.         map.put("pushinstruction", "(dup|lload|dup2|bipush|fload|ldc2_w|sipush|lconst|fconst|dload|getstatic|ldc_w|aconst_null|dconst|iload|ldc|iconst|aload)");
  115.         map.put("stackproducer",
  116.             "(imul|lsub|aload|fload|lor|new|aaload|fcmpg|iand|iaload|lrem|idiv|d2l|isub|dcmpg|dup|f2d|f2i|drem|i2c|checkcast|frem|lushr|daload|dneg|lshl|ldiv|ishr|ldc_w|invokeinterface|invokedynamic|lxor|ishl|l2d|i2f|faload|sipush|iushr|caload|instanceof|invokespecial|fmul|laload|d2f|lneg|ixor|i2l|fdiv|getstatic|i2b|swap|i2d|dup2|fcmpl|saload|ladd|irem|dload|jsr_w|dconst|dcmpl|fsub|ldc|arraylength|aconst_null|tableswitch|lmul|ldc2_w|iconst|dadd|f2l|ddiv|land|jsr|anewarray|dmul|bipush|dsub|d2i|newarray|i2s|lshr|iadd|lload|l2i|fneg|iload|fadd|baload|fconst|lookupswitch|ior|ineg|lconst|l2f|getfield|invokevirtual|invokestatic)");
  117.         map.put("stackconsumer",
  118.             "(imul|lsub|lor|iflt|fcmpg|if_icmpgt|iand|ifeq|if_icmplt|lrem|ifnonnull|idiv|d2l|isub|dcmpg|dastore|if_icmpeq|f2d|f2i|drem|i2c|checkcast|frem|lreturn|astore|lushr|pop2|monitorexit|dneg|fastore|istore|lshl|ldiv|lstore|areturn|if_icmpge|ishr|monitorenter|invokeinterface|invokedynamic|aastore|lxor|ishl|l2d|i2f|return|iushr|instanceof|invokespecial|fmul|ireturn|d2f|lneg|ixor|pop|i2l|ifnull|fdiv|lastore|i2b|if_acmpeq|ifge|swap|i2d|putstatic|fcmpl|ladd|irem|dcmpl|fsub|freturn|ifgt|castore|lmul|dadd|f2l|ddiv|dstore|land|if_icmpne|if_acmpne|dmul|dsub|sastore|ifle|d2i|i2s|lshr|iadd|l2i|bastore|fstore|fneg|fadd|ior|ineg|ifne|dreturn|l2f|if_icmple|getfield|invokevirtual|invokestatic|iastore)");
  119.         map.put("exceptionthrower",
  120.             "(irem|lrem|laload|putstatic|baload|dastore|areturn|getstatic|ldiv|anewarray|iastore|castore|idiv|saload|lastore|fastore|putfield|lreturn|caload|getfield|return|aastore|freturn|newarray|instanceof|multianewarray|athrow|faload|iaload|aaload|dreturn|monitorenter|checkcast|bastore|arraylength|new|invokevirtual|sastore|ldc_w|ireturn|invokespecial|monitorexit|invokeinterface|invokedynamic|ldc|invokestatic|daload)");
  121.         map.put("loadclass",
  122.             "(multianewarray|invokeinterface|invokedynamic|instanceof|invokespecial|putfield|checkcast|putstatic|invokevirtual|new|getstatic|invokestatic|getfield|anewarray)");
  123.         map.put("instructiontargeter",
  124.             "(ifle|if_acmpne|if_icmpeq|if_acmpeq|ifnonnull|goto_w|iflt|ifnull|if_icmpne|tableswitch|if_icmple|ifeq|if_icmplt|jsr_w|if_icmpgt|ifgt|jsr|goto|ifne|ifge|lookupswitch|if_icmpge)");
  125.         // Some aliases
  126.         map.put("if_icmp", "(if_icmpne|if_icmpeq|if_icmple|if_icmpge|if_icmplt|if_icmpgt)");
  127.         map.put("if_acmp", "(if_acmpeq|if_acmpne)");
  128.         map.put("if", "(ifeq|ifne|iflt|ifge|ifgt|ifle)");
  129.         // Precompile some aliases first
  130.         map.put("iconst", precompile(Const.ICONST_0, Const.ICONST_5, Const.ICONST_M1));
  131.         map.put("lconst", new String(new char[] {'(', makeChar(Const.LCONST_0), '|', makeChar(Const.LCONST_1), ')'}));
  132.         map.put("dconst", new String(new char[] {'(', makeChar(Const.DCONST_0), '|', makeChar(Const.DCONST_1), ')'}));
  133.         map.put("fconst", new String(new char[] {'(', makeChar(Const.FCONST_0), '|', makeChar(Const.FCONST_1), '|', makeChar(Const.FCONST_2), ')'}));
  134.         map.put("lload", precompile(Const.LLOAD_0, Const.LLOAD_3, Const.LLOAD));
  135.         map.put("iload", precompile(Const.ILOAD_0, Const.ILOAD_3, Const.ILOAD));
  136.         map.put("dload", precompile(Const.DLOAD_0, Const.DLOAD_3, Const.DLOAD));
  137.         map.put("fload", precompile(Const.FLOAD_0, Const.FLOAD_3, Const.FLOAD));
  138.         map.put("aload", precompile(Const.ALOAD_0, Const.ALOAD_3, Const.ALOAD));
  139.         map.put("lstore", precompile(Const.LSTORE_0, Const.LSTORE_3, Const.LSTORE));
  140.         map.put("istore", precompile(Const.ISTORE_0, Const.ISTORE_3, Const.ISTORE));
  141.         map.put("dstore", precompile(Const.DSTORE_0, Const.DSTORE_3, Const.DSTORE));
  142.         map.put("fstore", precompile(Const.FSTORE_0, Const.FSTORE_3, Const.FSTORE));
  143.         map.put("astore", precompile(Const.ASTORE_0, Const.ASTORE_3, Const.ASTORE));
  144.         // Compile strings
  145.         map.forEach((key, value) -> {
  146.             final char ch = value.charAt(1); // Omit already precompiled patterns
  147.             if (ch < OFFSET) {
  148.                 map.put(key, compilePattern(value)); // precompile all patterns
  149.             }
  150.         });
  151.         // Add instruction alias to match anything
  152.         final StringBuilder buf = new StringBuilder("(");
  153.         for (short i = 0; i < NO_OPCODES; i++) {
  154.             if (Const.getNoOfOperands(i) != Const.UNDEFINED) { // Not an invalid opcode
  155.                 buf.append(makeChar(i));
  156.                 if (i < NO_OPCODES - 1) {
  157.                     buf.append('|');
  158.                 }
  159.             }
  160.         }
  161.         buf.append(')');
  162.         map.put("instruction", buf.toString());
  163.     }

  165.     /**
  166.      * Replace symbolic names of instructions with the appropriate character and remove all white space from string. Meta
  167.      * characters such as +, * are ignored.
  168.      *
  169.      * @param pattern The pattern to compile
  170.      * @return translated regular expression string
  171.      */
  172.     private static String compilePattern(final String pattern) {
  173.         // Bug: BCEL-77 - Instructions are assumed to be english, to avoid odd Locale issues
  174.         final String lower = pattern.toLowerCase(Locale.ENGLISH);
  175.         final StringBuilder buf = new StringBuilder();
  176.         final int size = pattern.length();
  177.         for (int i = 0; i < size; i++) {
  178.             char ch = lower.charAt(i);
  179.             if (Character.isLetterOrDigit(ch)) {
  180.                 final StringBuilder name = new StringBuilder();
  181.                 while ((Character.isLetterOrDigit(ch) || ch == '_') && i < size) {
  182.                     name.append(ch);
  183.                     if (++i >= size) {
  184.                         break;
  185.                     }
  186.                     ch = lower.charAt(i);
  187.                 }
  188.                 i--;
  189.                 buf.append(mapName(name.toString()));
  190.             } else if (!Character.isWhitespace(ch)) {
  191.                 buf.append(ch);
  192.             }
  193.         }
  194.         return buf.toString();
  195.     }

  197.     /**
  198.      * Convert opcode number to char.
  199.      */
  200.     private static char makeChar(final short opcode) {
  201.         return (char) (opcode + OFFSET);
  202.     }

  204.     /**
  205.      * Map symbolic instruction names like "getfield" to a single character.
  206.      *
  207.      * @param pattern instruction pattern in lower case
  208.      * @return encoded string for a pattern such as "BranchInstruction".
  209.      */
  210.     private static String mapName(final String pattern) {
  211.         final String result = map.get(pattern);
  212.         if (result != null) {
  213.             return result;
  214.         }
  215.         for (short i = 0; i < NO_OPCODES; i++) {
  216.             if (pattern.equals(Const.getOpcodeName(i))) {
  217.                 return String.valueOf(makeChar(i));
  218.             }
  219.         }
  220.         throw new IllegalArgumentException("Instruction unknown: " + pattern);
  221.     }

  223.     private static String precompile(final short from, final short to, final short extra) {
  224.         final StringBuilder buf = new StringBuilder("(");
  225.         for (short i = from; i <= to; i++) {
  226.             buf.append(makeChar(i));
  227.             buf.append('|');
  228.         }
  229.         buf.append(makeChar(extra));
  230.         buf.append(")");
  231.         return buf.toString();
  232.     }

  234.     private final InstructionList il;

  236.     private String ilString; // instruction list as string

  238.     private InstructionHandle[] handles; // map instruction

  240.     // list to array
  241.     /**
  242.      * @param il instruction list to search for given patterns
  243.      */
  244.     public InstructionFinder(final InstructionList il) {
  245.         this.il = il;
  246.         reread();
  247.     }

  249.     /**
  250.      * @return the inquired instruction list
  251.      */
  252.     public final InstructionList getInstructionList() {
  253.         return il;
  254.     }

  256.     /**
  257.      * @return the matched piece of code as an array of instruction (handles)
  258.      */
  259.     private InstructionHandle[] getMatch(final int matchedFrom, final int matchLength) {
  260.         return Arrays.copyOfRange(handles, matchedFrom, matchedFrom + matchLength);
  261.     }

  263.     /**
  264.      * Reread the instruction list, e.g., after you've altered the list upon a match.
  265.      */
  266.     public final void reread() {
  267.         final int size = il.getLength();
  268.         final char[] buf = new char[size]; // Create a string with length equal to il length
  269.         handles = il.getInstructionHandles();
  270.         // Map opcodes to characters
  271.         for (int i = 0; i < size; i++) {
  272.             buf[i] = makeChar(handles[i].getInstruction().getOpcode());
  273.         }
  274.         ilString = new String(buf);
  275.     }

  277.     /**
  278.      * Start search beginning from the start of the given instruction list.
  279.      *
  280.      * @param pattern the instruction pattern to search for, where case is ignored
  281.      * @return iterator of matches where e.nextElement() returns an array of instruction handles describing the matched area
  282.      */
  283.     public final Iterator<InstructionHandle[]> search(final String pattern) {
  284.         return search(pattern, il.getStart(), null);
  285.     }

  287.     /**
  288.      * Start search beginning from the start of the given instruction list. Check found matches with the constraint object.
  289.      *
  290.      * @param pattern the instruction pattern to search for, case is ignored
  291.      * @param constraint constraints to be checked on matching code
  292.      * @return instruction handle or 'null' if the match failed
  293.      */
  294.     public final Iterator<InstructionHandle[]> search(final String pattern, final CodeConstraint constraint) {
  295.         return search(pattern, il.getStart(), constraint);
  296.     }

  298.     /**
  299.      * Start search beginning from 'from'.
  300.      *
  301.      * @param pattern the instruction pattern to search for, where case is ignored
  302.      * @param from where to start the search in the instruction list
  303.      * @return iterator of matches where e.nextElement() returns an array of instruction handles describing the matched area
  304.      */
  305.     public final Iterator<InstructionHandle[]> search(final String pattern, final InstructionHandle from) {
  306.         return search(pattern, from, null);
  307.     }

  309.     /**
  310.      * Search for the given pattern in the instruction list. You can search for any valid opcode via its symbolic name, e.g.
  311.      * "istore". You can also use a super class or an interface name to match a whole set of instructions, e.g.
  312.      * "BranchInstruction" or "LoadInstruction". "istore" is also an alias for all "istore_x" instructions. Additional
  313.      * aliases are "if" for "ifxx", "if_icmp" for "if_icmpxx", "if_acmp" for "if_acmpxx".
  314.      *
  315.      * Consecutive instruction names must be separated by white space which will be removed during the compilation of the
  316.      * pattern.
  317.      *
  318.      * For the rest the usual pattern matching rules for regular expressions apply.
  319.      * <P>
  320.      * Example pattern:
  321.      *
  322.      * <pre>
  323.      * search(&quot;BranchInstruction NOP ((IfInstruction|GOTO)+ ISTORE Instruction)*&quot;);
  324.      * </pre>
  325.      *
  326.      * <p>
  327.      * If you alter the instruction list upon a match such that other matching areas are affected, you should call reread()
  328.      * to update the finder and call search() again, because the matches are cached.
  329.      *
  330.      * @param pattern the instruction pattern to search for, where case is ignored
  331.      * @param from where to start the search in the instruction list
  332.      * @param constraint optional CodeConstraint to check the found code pattern for user-defined constraints
  333.      * @return iterator of matches where e.nextElement() returns an array of instruction handles describing the matched area
  334.      */
  335.     public final Iterator<InstructionHandle[]> search(final String pattern, final InstructionHandle from, final CodeConstraint constraint) {
  336.         final String search = compilePattern(pattern);
  337.         int start = -1;
  338.         for (int i = 0; i < handles.length; i++) {
  339.             if (handles[i] == from) {
  340.                 start = i; // Where to start search from (index)
  341.                 break;
  342.             }
  343.         }
  344.         if (start == -1) {
  345.             throw new ClassGenException("Instruction handle " + from + " not found in instruction list.");
  346.         }
  347.         final Pattern regex = Pattern.compile(search);
  348.         final List<InstructionHandle[]> matches = new ArrayList<>();
  349.         final Matcher matcher = regex.matcher(ilString);
  350.         while (start < ilString.length() && matcher.find(start)) {
  351.             final int startExpr = matcher.start();
  352.             final int endExpr = matcher.end();
  353.             final int lenExpr = endExpr - startExpr;
  354.             final InstructionHandle[] match = getMatch(startExpr, lenExpr);
  355.             if (constraint == null || constraint.checkCode(match)) {
  356.                 matches.add(match);
  357.             }
  358.             start = endExpr;
  359.         }
  360.         return matches.iterator();
  361.     }

  363.     /*
  364.      * Internal debugging routines.
  365.      */
  366. //    private static final String pattern2string( String pattern ) {
  367. //        return pattern2string(pattern, true);
  368. //    }

  370. //    private static final String pattern2string( String pattern, boolean make_string ) {
  371. //        StringBuilder buf = new StringBuilder();
  372. //        for (int i = 0; i < pattern.length(); i++) {
  373. //            char ch = pattern.charAt(i);
  374. //            if (ch >= OFFSET) {
  375. //                if (make_string) {
  376. //                    buf.append(Constants.getOpcodeName(ch - OFFSET));
  377. //                } else {
  378. //                    buf.append((ch - OFFSET));
  379. //                }
  380. //            } else {
  381. //                buf.append(ch);
  382. //            }
  383. //        }
  384. //        return buf.toString();
  385. //    }
  386. }
Created with JaCoCo 0.8.12.202403310830