re.java

linux下建立JAVA虚拟机的源码KAFFE

      if (m.end[0] == 0) {   // handle pathological case of zero-length match
	index++;
	input.move(1);
      } else {
	input.move(m.end[0]);
      }
      if (!input.isValid()) break;
    }
    REMatch[] mset = new REMatch[all.size()];
    all.copyInto(mset);
    return mset;
  }
  
    /* Implements abstract method REToken.match() */
    boolean match(CharIndexed input, REMatch mymatch) {
	if (firstToken == null) {
	    return next(input, mymatch);
	}

	// Note the start of this subexpression
	mymatch.start[subIndex] = mymatch.index;

	return firstToken.match(input, mymatch);
    }

    REMatch findMatch(CharIndexed input, REMatch mymatch) {
        if (mymatch.backtrackStack == null)
	  mymatch.backtrackStack = new BacktrackStack();
	boolean b = match(input, mymatch);
	if (b) {
	    // mymatch.backtrackStack.push(new REMatch.Backtrack(
	    //     this, input, mymatch, null));
	    return mymatch;
	}
	return null;
    }

  /**
   * Returns the first match found in the input.  If no match is found,
   * null is returned.
   *
   * @param input The input text.
   * @return An REMatch instance referencing the match, or null if none.
   */
  public REMatch getMatch(Object input) {
    return getMatch(input,0,0);
  }
  
  /**
   * Returns the first match found in the input, beginning
   * the search at the specified index.  If no match is found,
   * returns null.
   *
   * @param input The input text.
   * @param index The offset within the text to begin looking for a match.
   * @return An REMatch instance referencing the match, or null if none.
   */
  public REMatch getMatch(Object input, int index) {
    return getMatch(input,index,0);
  }
  
  /**
   * Returns the first match found in the input, beginning
   * the search at the specified index, and using the specified
   * execution flags.  If no match is found, returns null.
   *
   * @param input The input text.
   * @param index The offset index at which the search should be begin.
   * @param eflags The logical OR of any execution flags above.
   * @return An REMatch instance referencing the match, or null if none.
   */
  public REMatch getMatch(Object input, int index, int eflags) {
    return getMatch(input,index,eflags,null);
  }

  /**
   * Returns the first match found in the input, beginning the search
   * at the specified index, and using the specified execution flags.
   * If no match is found, returns null.  If a StringBuffer is
   * provided and is non-null, the contents of the input text from the
   * index to the beginning of the match (or to the end of the input,
   * if there is no match) are appended to the StringBuffer.
   *
   * @param input The input text.
   * @param index The offset index at which the search should be begin.
   * @param eflags The logical OR of any execution flags above.
   * @param buffer The StringBuffer to save pre-match text in.
   * @return An REMatch instance referencing the match, or null if none.  */
  public REMatch getMatch(Object input, int index, int eflags, StringBuffer buffer) {
    return getMatchImpl(makeCharIndexed(input,index),index,eflags,buffer);
  }

  REMatch getMatchImpl(CharIndexed input, int anchor, int eflags, StringBuffer buffer) {
      boolean tryEntireMatch = ((eflags & REG_TRY_ENTIRE_MATCH) != 0);
      RE re = (tryEntireMatch ? (RE) this.clone() : this);
      if (tryEntireMatch) {
	  re.chain(new RETokenEnd(0, null));
      }
      // Create a new REMatch to hold results
      REMatch mymatch = new REMatch(numSubs, anchor, eflags);
      do {
	  // Optimization: check if anchor + minimumLength > length
	  if (minimumLength == 0 || input.charAt(minimumLength-1) != CharIndexed.OUT_OF_BOUNDS) {
	      if (re.match(input, mymatch)) {
		  REMatch best = mymatch;
		  // We assume that the match that coms first is the best.
		  // And the following "The longer, the better" rule has
		  // been commented out. The longest is not neccesarily
		  // the best. For example, "a" out of "aaa" is the best
		  // match for /a+?/.
		  /*
		  // Find best match of them all to observe leftmost longest
		  while ((mymatch = mymatch.next) != null) {
		      if (mymatch.index > best.index) {
		   	best = mymatch;
		      }
		  }
		  */
		  best.end[0] = best.index;
		  best.finish(input);
		  return best;
	      }
	  }
	  mymatch.clear(++anchor);
	  // Append character to buffer if needed
	  if (buffer != null && input.charAt(0) != CharIndexed.OUT_OF_BOUNDS) {
	      buffer.append(input.charAt(0));
	  }
      } while (input.move(1));
      
      // Special handling at end of input for e.g. "$"
      if (minimumLength == 0) {
	  if (match(input, mymatch)) {
	      mymatch.finish(input);
	      return mymatch;
	  }
      }

      return null;
  }

  /**
   * Returns an REMatchEnumeration that can be used to iterate over the
   * matches found in the input text.
   *
   * @param input The input text.
   * @return A non-null REMatchEnumeration instance.
   */
  public REMatchEnumeration getMatchEnumeration(Object input) {
    return getMatchEnumeration(input,0,0);
  }


  /**
   * Returns an REMatchEnumeration that can be used to iterate over the
   * matches found in the input text.
   *
   * @param input The input text.
   * @param index The offset index at which the search should be begin.
   * @return A non-null REMatchEnumeration instance, with its input cursor
   *  set to the index position specified.
   */
  public REMatchEnumeration getMatchEnumeration(Object input, int index) {
    return getMatchEnumeration(input,index,0);
  }

  /**
   * Returns an REMatchEnumeration that can be used to iterate over the
   * matches found in the input text.
   *
   * @param input The input text.
   * @param index The offset index at which the search should be begin.
   * @param eflags The logical OR of any execution flags above.
   * @return A non-null REMatchEnumeration instance, with its input cursor
   *  set to the index position specified.
   */
  public REMatchEnumeration getMatchEnumeration(Object input, int index, int eflags) {
    return new REMatchEnumeration(this,makeCharIndexed(input,index),index,eflags);
  }


  /**
   * Substitutes the replacement text for the first match found in the input.
   *
   * @param input The input text.
   * @param replace The replacement text, which may contain $x metacharacters (see REMatch.substituteInto).
   * @return A String interpolating the substituted text.
   * @see REMatch#substituteInto
   */
  public String substitute(Object input,String replace) {
    return substitute(input,replace,0,0);
  }

  /**
   * Substitutes the replacement text for the first match found in the input
   * beginning at the specified index position.  Specifying an index
   * effectively causes the regular expression engine to throw away the
   * specified number of characters. 
   *
   * @param input The input text.
   * @param replace The replacement text, which may contain $x metacharacters (see REMatch.substituteInto).
   * @param index The offset index at which the search should be begin.
   * @return A String containing the substring of the input, starting
   *   at the index position, and interpolating the substituted text.
   * @see REMatch#substituteInto
   */
  public String substitute(Object input,String replace,int index) {
    return substitute(input,replace,index,0);
  }

  /**
   * Substitutes the replacement text for the first match found in the input
   * string, beginning at the specified index position and using the
   * specified execution flags.
   *
   * @param input The input text.
   * @param replace The replacement text, which may contain $x metacharacters (see REMatch.substituteInto).
   * @param index The offset index at which the search should be begin.
   * @param eflags The logical OR of any execution flags above.
   * @return A String containing the substring of the input, starting
   *   at the index position, and interpolating the substituted text.
   * @see REMatch#substituteInto
   */
  public String substitute(Object input,String replace,int index,int eflags) {
    return substituteImpl(makeCharIndexed(input,index),replace,index,eflags);
  }

  private String substituteImpl(CharIndexed input,String replace,int index,int eflags) {
    StringBuffer buffer = new StringBuffer();
    REMatch m = getMatchImpl(input,index,eflags,buffer);
    if (m==null) return buffer.toString();
    buffer.append(getReplacement(replace, m, eflags));
    if (input.move(m.end[0])) {
      do {
	buffer.append(input.charAt(0));
      } while (input.move(1));
    }
    return buffer.toString();
  }
  
  /**
   * Substitutes the replacement text for each non-overlapping match found 
   * in the input text.
   *
   * @param input The input text.
   * @param replace The replacement text, which may contain $x metacharacters (see REMatch.substituteInto).
   * @return A String interpolating the substituted text.
   * @see REMatch#substituteInto
   */
  public String substituteAll(Object input,String replace) {
    return substituteAll(input,replace,0,0);
  }

  /**
   * Substitutes the replacement text for each non-overlapping match found 
   * in the input text, starting at the specified index.
   *
   * If the regular expression allows the empty string to match, it will
   * substitute matches at all positions except the end of the input.
   *
   * @param input The input text.
   * @param replace The replacement text, which may contain $x metacharacters (see REMatch.substituteInto).
   * @param index The offset index at which the search should be begin.
   * @return A String containing the substring of the input, starting
   *   at the index position, and interpolating the substituted text.
   * @see REMatch#substituteInto
   */
  public String substituteAll(Object input,String replace,int index) {
    return substituteAll(input,replace,index,0);
  }
 
  /**
   * Substitutes the replacement text for each non-overlapping match found 
   * in the input text, starting at the specified index and using the
   * specified execution flags.
   *
   * @param input The input text.
   * @param replace The replacement text, which may contain $x metacharacters (see REMatch.substituteInto).
   * @param index The offset index at which the search should be begin.
   * @param eflags The logical OR of any execution flags above.
   * @return A String containing the substring of the input, starting
   *   at the index position, and interpolating the substituted text.
   * @see REMatch#substituteInto
   */
  public String substituteAll(Object input,String replace,int index,int eflags) {
    return substituteAllImpl(makeCharIndexed(input,index),replace,index,eflags);
  }

  private String substituteAllImpl(CharIndexed input,String replace,int index,int eflags) {
    StringBuffer buffer = new StringBuffer();
    REMatch m;
    while ((m = getMatchImpl(input,index,eflags,buffer)) != null) {
      buffer.append(getReplacement(replace, m, eflags));
      index = m.getEndIndex();
      if (m.end[0] == 0) {
	char ch = input.charAt(0);
	if (ch != CharIndexed.OUT_OF_BOUNDS) 
	    buffer.append(ch);
	input.move(1);
      } else {
	  input.move(m.end[0]);
      }

      if (!input.isValid()) break;
    }
    return buffer.toString();
  }

  public static String getReplacement(String replace, REMatch m, int eflags) {
    if ((eflags & REG_NO_INTERPOLATE) > 0)
      return replace;
    else {
      if ((eflags & REG_REPLACE_USE_BACKSLASHESCAPE) > 0) {
        StringBuffer sb = new StringBuffer();
        int l = replace.length();
        for (int i = 0; i < l; i++) {
	    char c = replace.charAt(i);
            switch(c) {
            case '\\':
              i++;
              // Let StringIndexOutOfBoundsException be thrown.
              sb.append(replace.charAt(i));
              break;
            case '$':
	      int i1 = i + 1;
	      while (i1 < replace.length() &&
		Character.isDigit(replace.charAt(i1))) i1++;
              sb.append(m.substituteInto(replace.substring(i, i1)));
              i = i1 - 1;
              break;
            default:
              sb.append(c);
            }
        }
        return sb.toString();
      }
      else
        return m.substituteInto(replace);
    }
  }	
  
  /* Helper function for constructor */
  private void addToken(REToken next) {
    if (next == null) return;
    minimumLength += next.getMinimumLength();
    int nmax = next.getMaximumLength();
    if (nmax < Integer.MAX_VALUE && maximumLength < Integer.MAX_VALUE)
	maximumLength += nmax;
    else 
	maximumLength = Integer.MAX_VALUE;

    if (firstToken == null) {
	lastToken = firstToken = next;
    } else {
      // if chain returns false, it "rejected" the token due to
      // an optimization, and next was combined with lastToken
      if (lastToken.chain(next)) {
	  lastToken = next;
      }
    }
  }

  private static REToken setRepeated(REToken current, int min, int max, int index) throws REException {
    if (current == null) throw new REException(getLocalizedMessage("repeat.no.token"),REException.REG_BADRPT,index);
    return new RETokenRepeated(current.subIndex,current,min,max);
  }

  private static int getPosixSet(char[] pattern,int index,StringBuffer buf) {
    // Precondition: pattern[index-1] == ':'
    // we will return pos