perl5matcher.java

jakarta-oro-2.0.8 正则表达式 引擎 源代码

      input = _toLower(input);

    __initInterpreterGlobals(expression, input, 0, input.length, offset);

    __lastSuccess = __tryExpression(offset);
    __lastMatchResult = null;

    return __lastSuccess;
  }


  /**
   * Determines if a prefix of a string (represented as a char[])
   * matches a given pattern.
   * If a prefix of the string matches the pattern, a MatchResult instance
   * representing the match is made accesible via
   * {@link #getMatch()}.
   * <p>
   * This method is useful for certain common token identification tasks
   * that are made more difficult without this functionality.
   * <p>
   * @param input  The char[] to test for a prefix match.
   * @param pattern  The Pattern to be matched.
   * @return True if input matches pattern, false otherwise.
   */
  public boolean matchesPrefix(char[] input, Pattern pattern) {
    return matchesPrefix(input, pattern, 0);
  }


  /**
   * Determines if a prefix of a string matches a given pattern.
   * If a prefix of the string matches the pattern, a MatchResult instance
   * representing the match is made accesible via
   * {@link #getMatch()}.
   * <p>
   * This method is useful for certain common token identification tasks
   * that are made more difficult without this functionality.
   * <p>
   * @param input  The String to test for a prefix match.
   * @param pattern  The Pattern to be matched.
   * @return True if input matches pattern, false otherwise.
   */
  public boolean matchesPrefix(String input, Pattern pattern) {
    return matchesPrefix(input.toCharArray(), pattern, 0);
  }

  /**
   * Determines if a prefix of a PatternMatcherInput instance
   * matches a given pattern.  If there is a match, a MatchResult instance
   * representing the match is made accesible via
   * {@link #getMatch()}.  Unlike the
   * {@link #contains(PatternMatcherInput, Pattern)}
   * method, the current offset of the PatternMatcherInput argument
   * is not updated.  However, unlike the
   * {@link #matches matches(PatternMatcherInput, Pattern)} method,
   * matchesPrefix() will start its search from the current offset
   * rather than the begin offset of the PatternMatcherInput.
   * <p>
   * This method is useful for certain common token identification tasks
   * that are made more difficult without this functionality.
   * <p>
   * @param input  The PatternMatcherInput to test for a prefix match.
   * @param pattern  The Pattern to be matched.
   * @return True if input matches pattern, false otherwise.
   */
  public boolean matchesPrefix(PatternMatcherInput input, Pattern pattern) {
    char[] inp;
    Perl5Pattern expression;

    expression = (Perl5Pattern)pattern;

    __originalInput = input._originalBuffer;
    if(expression._isCaseInsensitive) {
      if(input._toLowerBuffer == null)
	input._toLowerBuffer = _toLower(__originalInput);
      inp = input._toLowerBuffer;
    } else
      inp = __originalInput;

    __initInterpreterGlobals(expression, inp, input._beginOffset,
			     input._endOffset, input._currentOffset);
    __lastSuccess = __tryExpression(input._currentOffset);
    __lastMatchResult = null;

    return __lastSuccess;
  }



  /**
   * Determines if a string (represented as a char[]) exactly 
   * matches a given pattern.  If
   * there is an exact match, a MatchResult instance
   * representing the match is made accesible via
   * {@link #getMatch()}.  The pattern must be
   * a Perl5Pattern instance, otherwise a ClassCastException will
   * be thrown.  You are not required to, and indeed should NOT try to
   * (for performance reasons), catch a ClassCastException because it
   * will never be thrown as long as you use a Perl5Pattern as the pattern
   * parameter.
   * <p>
   * <b>Note:</b> matches() is not the same as sticking a ^ in front of
   * your expression and a $ at the end of your expression in Perl5
   * and using the =~ operator, even though in many cases it will be
   * equivalent.  matches() literally looks for an exact match according
   * to the rules of Perl5 expression matching.  Therefore, if you have
   * a pattern <em>foo|foot</em> and are matching the input <em>foot</em>
   * it will not produce an exact match.  But <em>foot|foo</em> will
   * produce an exact match for either <em>foot</em> or <em>foo</em>.
   * Remember, Perl5 regular expressions do not match the longest
   * possible match.  From the perlre manpage:
   * <blockquote>
   *   Alternatives are tried from left to right, so the first
   *   alternative found for which the entire expression matches,
   *   is the one that is chosen. This means that alternatives
   *   are not necessarily greedy. For example: when matching
   *   foo|foot against "barefoot", only the "foo" part will
   *   match, as that is the first alternative tried, and it
   *   successfully matches the target string.
   * </blockquote>
   * <p>
   * @param input  The char[] to test for an exact match.
   * @param pattern  The Perl5Pattern to be matched.
   * @return True if input matches pattern, false otherwise.
   * @exception ClassCastException If a Pattern instance other than a
   *         Perl5Pattern is passed as the pattern parameter.
   */
  public boolean matches(char[] input, Pattern pattern) {
    Perl5Pattern expression;

    expression = (Perl5Pattern)pattern;
    __originalInput = input;
    if(expression._isCaseInsensitive)
      input = _toLower(input);

    __initInterpreterGlobals(expression, input, 0, input.length, 0);
    __lastSuccess = (__tryExpression(0) &&
		     __endMatchOffsets[0] == input.length);
    __lastMatchResult = null;

    return __lastSuccess;
  }


  /**
   * Determines if a string exactly matches a given pattern.  If
   * there is an exact match, a MatchResult instance
   * representing the match is made accesible via
   * {@link #getMatch()}.  The pattern must be
   * a Perl5Pattern instance, otherwise a ClassCastException will
   * be thrown.  You are not required to, and indeed should NOT try to
   * (for performance reasons), catch a ClassCastException because it
   * will never be thrown as long as you use a Perl5Pattern as the pattern
   * parameter.
   * <p>
   * <b>Note:</b> matches() is not the same as sticking a ^ in front of
   * your expression and a $ at the end of your expression in Perl5
   * and using the =~ operator, even though in many cases it will be
   * equivalent.  matches() literally looks for an exact match according
   * to the rules of Perl5 expression matching.  Therefore, if you have
   * a pattern <em>foo|foot</em> and are matching the input <em>foot</em>
   * it will not produce an exact match.  But <em>foot|foo</em> will
   * produce an exact match for either <em>foot</em> or <em>foo</em>.
   * Remember, Perl5 regular expressions do not match the longest
   * possible match.  From the perlre manpage:
   * <blockquote>
   *   Alternatives are tried from left to right, so the first
   *   alternative found for which the entire expression matches,
   *   is the one that is chosen. This means that alternatives
   *   are not necessarily greedy. For example: when matching
   *   foo|foot against "barefoot", only the "foo" part will
   *   match, as that is the first alternative tried, and it
   *   successfully matches the target string.
   * </blockquote>
   * <p>
   * @param input  The String to test for an exact match.
   * @param pattern  The Perl5Pattern to be matched.
   * @return True if input matches pattern, false otherwise.
   * @exception ClassCastException If a Pattern instance other than a
   *         Perl5Pattern is passed as the pattern parameter.
   */
  public boolean matches(String input, Pattern pattern) {
    return matches(input.toCharArray(), pattern);
  }


  /**
   * Determines if the contents of a PatternMatcherInput instance
   * exactly matches a given pattern.  If
   * there is an exact match, a MatchResult instance
   * representing the match is made accesible via
   * {@link #getMatch()}.  Unlike the
   * {@link #contains(PatternMatcherInput, Pattern)}
   * method, the current offset of the PatternMatcherInput argument
   * is not updated.  You should remember that the region between
   * the begin (NOT the current) and end offsets of the PatternMatcherInput
   * will be tested for an exact match.
   * <p>
   * The pattern must be a Perl5Pattern instance, otherwise a
   * ClassCastException will be thrown.  You are not required to, and 
   * indeed should NOT try to (for performance reasons), catch a
   * ClassCastException because it will never be thrown as long as you use
   * a Perl5Pattern as the pattern parameter.
   * <p>
   * <b>Note:</b> matches() is not the same as sticking a ^ in front of
   * your expression and a $ at the end of your expression in Perl5
   * and using the =~ operator, even though in many cases it will be
   * equivalent.  matches() literally looks for an exact match according
   * to the rules of Perl5 expression matching.  Therefore, if you have
   * a pattern <em>foo|foot</em> and are matching the input <em>foot</em>
   * it will not produce an exact match.  But <em>foot|foo</em> will
   * produce an exact match for either <em>foot</em> or <em>foo</em>.
   * Remember, Perl5 regular expressions do not match the longest
   * possible match.  From the perlre manpage:
   * <blockquote>
   *   Alternatives are tried from left to right, so the first
   *   alternative found for which the entire expression matches,
   *   is the one that is chosen. This means that alternatives
   *   are not necessarily greedy. For example: when matching
   *   foo|foot against "barefoot", only the "foo" part will
   *   match, as that is the first alternative tried, and it
   *   successfully matches the target string.
   * </blockquote>
   * <p>
   * @param input  The PatternMatcherInput to test for a match.
   * @param pattern  The Perl5Pattern to be matched.
   * @return True if input matches pattern, false otherwise.
   * @exception ClassCastException If a Pattern instance other than a
   *         Perl5Pattern is passed as the pattern parameter.
   */
  public boolean matches(PatternMatcherInput input, Pattern pattern) {
    char[] inp;
    Perl5Pattern expression;

    expression = (Perl5Pattern)pattern;

    __originalInput = input._originalBuffer;
    if(expression._isCaseInsensitive) {
      if(input._toLowerBuffer == null)
	input._toLowerBuffer = _toLower(__originalInput);
      inp = input._toLowerBuffer;
    } else
      inp = __originalInput;

    __initInterpreterGlobals(expression, inp, input._beginOffset,
			     input._endOffset, input._beginOffset);

    __lastMatchResult = null;

    if(__tryExpression(input._beginOffset)) {
      if(__endMatchOffsets[0] == input._endOffset ||
	 input.length() == 0 || input._beginOffset == input._endOffset) {
	__lastSuccess = true;
	return true;
      }
    }

    __lastSuccess = false;

    return false;
  }



  /**
   * Determines if a string contains a pattern.  If the pattern is
   * matched by some substring of the input, a MatchResult instance
   * representing the <b> first </b> such match is made acessible via 
   * {@link #getMatch()}.  If you want to access
   * subsequent matches you should either use a PatternMatcherInput object
   * or use the offset information in the MatchResult to create a substring
   * representing the remaining input.  Using the MatchResult offset 
   * information is the recommended method of obtaining the parts of the
   * string preceeding the match and following the match.
   * <p>
   * The pattern must be a Perl5Pattern instance, otherwise a
   * ClassCastException will be thrown.  You are not required to, and 
   * indeed should NOT try to (for performance reasons), catch a
   * ClassCastException because it will never be thrown as long as you use
   * a Perl5Pattern as the pattern parameter.
   * <p>
   * @param input  The String to test for a match.
   * @param pattern  The Perl5Pattern to be matched.
   * @return True if the input contains a pattern match, false otherwise.
   * @exception ClassCastException If a Pattern instance other than a
   *         Perl5Pattern is passed as the pattern parameter.
   */
  public boolean contains(String input, Pattern pattern) {
    return contains(input.toCharArray(), pattern);
  }


  /**
   * Determines if a string (represented as a char[]) contains a pattern.
   * If the pattern is
   * matched by some substring of the input, a MatchResult instance
   * representing the <b> first </b> such match is made acessible via 
   * {@link #getMatch()}.  If you want to access
   * subsequent matches you should either use a PatternMatcherInput object
   * or use the offset information in the MatchResult to create a substring
   * representing the remaining input.  Using the MatchResult offset 
   * information is the recommended method of obtaining the parts of the
   * string preceeding the match and following the match.
   * <p>
   * The pattern must be a Perl5Pattern instance, otherwise a
   * ClassCastException will be thrown.  You are not required to, and 
   * indeed should NOT try to (for performance reasons), catch a
   * ClassCastException because it will never be thrown as long as you use
   * a Perl5Pattern as the pattern parameter.
   * <p>
   * @param input  The char[] to test for a match.
   * @param pattern  The Perl5Pattern to be matched.
   * @return True if the input contains a pattern match, false otherwise.
   * @exception ClassCastException If a Pattern instance other than a
   *         Perl5Pattern is passed as the pattern parameter.
   */
  public boolean contains(char[] input, Pattern pattern) {
    Perl5Pattern expression;

    expression = (Perl5Pattern)pattern;

    __originalInput = input;

    if(expression._isCaseInsensitive)
      input = _toLower(input);

    return __interpret(expression, input, 0, input.length, 0);
  }


  private static final int __DEFAULT_LAST_MATCH_END_OFFSET = -100;
  private int __lastMatchInputEndOffset = __DEFAULT_LAST_MATCH_END_OFFSET;
  /**
   * Determines if the contents of a PatternMatcherInput, starting from the
   * current offset of the input contains a pattern.
   * If a pattern match is found, a MatchResult
   * instance representing the <b>first</b> such match is made acessible via 
   * {@link #getMatch()}.  The current offset of the
   * PatternMatcherInput is set to the offset corresponding to the end
   * of the match, so that a subsequent call to this method will continue
   * searching where the last call left off.  You should remember that the
   * region between the begin and end offsets of the PatternMatcherInput are
   * considered the input to be searched, and that the current offset
   * of the PatternMatcherInput reflects where a search will start from.
   * Matches extending beyond the end offset of the PatternMatcherInput
   * will not be matched.  In other words, a match must occur entirely
   * between the begin and end offsets of the input.  See
   * {@link PatternMatcherInput} for more details.
   * <p>
   * As a side effect, if a match is found, the PatternMatcherInput match
   * offset information is updated.  See the
   * {@link PatternMatcherInput#setMatchOffsets(int, int)}
   * method for more details.
   * <p>
   * The pattern must be a Perl5Pattern instance, otherwise a
   * ClassCastException will be thrown.  You are not required to, and 
   * indeed should NOT try to (for performance reasons), catch a
   * ClassCastException because it will never be thrown as long as you use
   * a Perl5Pattern as the pattern parameter.
   * <p>
   * This method is usually used in a loop as follows:
   * <blockquote><pre>
   * PatternMatcher matcher;
   * PatternCompiler compiler;
   * Pattern pattern;
   * PatternMatcherInput input;
   * MatchResult result;
   *
   * compiler = new Perl5Compiler();
   * matcher  = new Perl5Matcher();
   *
   * try {
   *   pattern = compiler.compile(somePatternString);
   * } catch(MalformedPatternException e) {
   *   System.err.println("Bad pattern.");
   *   System.err.println(e.getMessage());
   *   return;
   * }
   *
   * input   = new PatternMatcherInput(someStringInput);
   *
   * while(matcher.contains(input, pattern)) {
   *   result = matcher.getMatch();  
   *   // Perform whatever processing on the result you want.
   * }
   *
   * </pre></blockquote>
   * <p>
   * @param input  The PatternMatcherInput to test for a match.
   * @param pattern  The Pattern to be matched.
   * @return True if the input contains a pattern match, false otherwise.
   * @exception ClassCastException If a Pattern instance other than a
   *         Perl5Pattern is passed as the pattern parameter.
   */
  public boolean contains(PatternMatcherInput input, Pattern pattern) {
    char[] inp;
    Perl5Pattern expression;
    boolean matchFound;


    //if(input.length() > 0) {
    // We want to allow a null string to match at the end of the input
    // which is why we don't check endOfInput.  Not sure if this is a
    // safe thing to do or not.
    if(input._currentOffset > input._endOffset)
      return false;
    //}
    /* else 
      if(input._endOfInput())
	return false;
	*/
    expression = (Perl5Pattern)pattern;
    __originalInput = input._originalBuffer;

    // Todo:
    // Really should only reduce to lowercase that part of the
    // input that is necessary, instead of the whole thing.
    // Adjust MatchResult offsets accordingly.  Actually, pass an adjustment
    // value to __interpret.
    __originalInput = input._originalBuffer;
    if(expression._isCaseInsensitive) {
      if(input._toLowerBuffer == null)
	input._toLowerBuffer = _toLower(__originalInput);
      inp = input._toLowerBuffer;
    } else
      inp = __originalInput;

    __lastMatchInputEndOffset = input.getMatchEndOffset();

    matchFound =
      __interpret(expression, inp, input._beginOffset, input._endOffset,
		  input._currentOffset);

    if(matchFound) {
      input.setCurrentOffset(__endMatchOffsets[0]);
      input.setMatchOffsets(__beginMatchOffsets[0], __endMatchOffsets[0]);
    } else {
      input.setCurrentOffset(input._endOffset + 1);
    }

    // Restore so it doesn't interfere with other unrelated matches.
    __lastMatchInputEndOffset = __DEFAULT_LAST_MATCH_END_OFFSET;

    return matchFound;
  }


  /**
   * Fetches the last match found by a call to a matches() or contains()
   * method.  If you plan on modifying the original search input, you
   * must call this method BEFORE you modify the original search input,
   * as a lazy evaluation technique is used to create the MatchResult.
   * This reduces the cost of pattern matching when you don't care about
   * the actual match and only care if the pattern occurs in the input.
   * Otherwise, a MatchResult would be created for every match found,
   * whether or not the MatchResult was later used by a call to getMatch().
   * <p>
   * @return A MatchResult instance containing the pattern match found
   *         by the last call to any one of the matches() or contains()
   *         methods.  If no match was found by the last call, returns
   *         null. 
   */
  public MatchResult getMatch() { 
    if(!__lastSuccess)
      return null;

    if(__lastMatchResult == null)
      __setLastMatchResult();

    return __lastMatchResult;
  }

}