regexp.java

Grammatica是一个C#和Java的语法分析程序生成器(编译器的编译器)。它可以用LL(k)语法创建可读的和带有注释的源代码。它也支持创建一个运行时语法分析器

        return new RepeatElement(elem, min, max, type);
    }

    /**
     * Parses a regular expression character set. This method handles 
     * the contents of the '[...]' construct in a regular expression.
     * 
     * @return the element representing this character set
     * 
     * @throws RegExpException if an error was encountered in the 
     *             pattern string
     */
    private Element parseCharSet() throws RegExpException {
        CharacterSetElement  charset;
        Element              elem;
        boolean              repeat = true;
        char                 start;
        char                 end;
        
        if (peekChar(0) == '^') {
            readChar('^');
            charset = new CharacterSetElement(true);
        } else {
            charset = new CharacterSetElement(false);
        }
        
        while (peekChar(0) > 0 && repeat) {
            start = (char) peekChar(0);
            switch (start) {
            case ']':
                repeat = false;
                break;
            case '\\':
                elem = parseEscapeChar();
                if (elem instanceof StringElement) {
                    charset.addCharacters((StringElement) elem);
                } else {
                    charset.addCharacterSet((CharacterSetElement) elem);
                }
                break;
            default:
                readChar(start);
                if (peekChar(0) == '-'
                 && peekChar(1) > 0 
                 && peekChar(1) != ']') {

                    readChar('-');
                    end = readChar();
                    charset.addRange(start, end);
                } else {
                    charset.addCharacter(start);
                }
            }
        }
        
        return charset;
    }

    /**
     * Parses a regular expression character. This method handles 
     * a single normal character in a regular expression.
     * 
     * @return the element representing this character
     * 
     * @throws RegExpException if an error was encountered in the 
     *             pattern string
     */
    private Element parseChar() throws RegExpException {
        switch (peekChar(0)) {
        case '\\':
            return parseEscapeChar();
        case '^':
        case '$':
            throw new RegExpException(
                RegExpException.UNSUPPORTED_SPECIAL_CHARACTER,
                pos,
                pattern);
        default:
            return new StringElement(readChar());
        }
    }

    /**
     * Parses a regular expression character escape. This method 
     * handles a single character escape in a regular expression.
     * 
     * @return the element representing this character escape
     * 
     * @throws RegExpException if an error was encountered in the 
     *             pattern string
     */
    private Element parseEscapeChar() throws RegExpException {
        char    c;
        String  str;
        
        readChar('\\');
        c = readChar();
        switch (c) {
        case '0':
            c = readChar();
            if (c < '0' || c > '3') {
                throw new RegExpException(
                    RegExpException.UNSUPPORTED_ESCAPE_CHARACTER,
                    pos - 3,
                    pattern);
            }
            str = String.valueOf(c);
            c = (char) peekChar(0);
            if ('0' <= c && c <= '7') {
                str += String.valueOf(readChar());
                c = (char) peekChar(0);
                if ('0' <= c && c <= '7') {
                    str += String.valueOf(readChar());
                }
            }
            try {
                c = (char) Integer.parseInt(str, 8);
                return new StringElement(c);
            } catch (NumberFormatException e) {
                throw new RegExpException(
                    RegExpException.UNSUPPORTED_ESCAPE_CHARACTER,
                    pos - str.length() - 2,
                    pattern);
            }
        case 'x':
            str = String.valueOf(readChar()) + 
                  String.valueOf(readChar());
            try {
                c = (char) Integer.parseInt(str, 16);
                return new StringElement(c);
            } catch (NumberFormatException e) {
                throw new RegExpException(
                    RegExpException.UNSUPPORTED_ESCAPE_CHARACTER,
                    pos - str.length() - 2,
                    pattern);
            }
        case 'u':
            str = String.valueOf(readChar()) + 
                  String.valueOf(readChar()) +
                  String.valueOf(readChar()) +
                  String.valueOf(readChar());
            try {
                c = (char) Integer.parseInt(str, 16);
                return new StringElement(c);
            } catch (NumberFormatException e) {
                throw new RegExpException(
                    RegExpException.UNSUPPORTED_ESCAPE_CHARACTER,
                    pos - str.length() - 2,
                    pattern);
            }
        case 't':
            return new StringElement('\t');
        case 'n':
            return new StringElement('\n');
        case 'r':
            return new StringElement('\r');
        case 'f':
            return new StringElement('\f');
        case 'a':
            return new StringElement('\u0007');
        case 'e':
            return new StringElement('\u001B');
        case 'd':
            return CharacterSetElement.DIGIT;
        case 'D':
            return CharacterSetElement.NON_DIGIT;
        case 's':
            return CharacterSetElement.WHITESPACE;
        case 'S':
            return CharacterSetElement.NON_WHITESPACE;
        case 'w':
            return CharacterSetElement.WORD;
        case 'W':
            return CharacterSetElement.NON_WORD;
        default:
            if (('A' <= c && c <= 'Z') || ('a' <= c && c <= 'z')) {
                throw new RegExpException(
                    RegExpException.UNSUPPORTED_ESCAPE_CHARACTER,
                    pos - 2,
                    pattern);             
            }
            return new StringElement(c);
        }
    }

    /**
     * Reads a number from the pattern. If the next character isn't a
     * numeric character, an exception is thrown. This method reads
     * several consecutive numeric characters. 
     * 
     * @return the numeric value read
     * 
     * @throws RegExpException if an error was encountered in the 
     *             pattern string
     */
    private int readNumber() throws RegExpException {
        StringBuffer  buf = new StringBuffer();
        int           c;
        
        c = peekChar(0);
        while ('0' <= c && c <= '9') {
            buf.append(readChar());
            c = peekChar(0);
        }
        if (buf.length() <= 0) {
            throw new RegExpException(
                RegExpException.UNEXPECTED_CHARACTER,
                pos,
                pattern);
        }
        return Integer.parseInt(buf.toString());
    }

    /**
     * Reads the next character in the pattern. If no next character
     * exists, an exception is thrown.
     * 
     * @return the character read 
     * 
     * @throws RegExpException if no next character was available in  
     *             the pattern string
     */
    private char readChar() throws RegExpException {
        int  c = peekChar(0);
        
        if (c < 0) {
            throw new RegExpException(
                RegExpException.UNTERMINATED_PATTERN, 
                pos,
                pattern);
        } else {
            pos++;
            return (char) c;
        }
    }

    /**
     * Reads the next character in the pattern. If the character 
     * wasn't the specified one, an exception is thrown.
     * 
     * @param c              the character to read
     * 
     * @return the character read 
     * 
     * @throws RegExpException if the character read didn't match the
     *             specified one, or if no next character was 
     *             available in the pattern string
     */
    private char readChar(char c) throws RegExpException {
        if (c != readChar()) {
            throw new RegExpException(
                RegExpException.UNEXPECTED_CHARACTER, 
                pos - 1,
                pattern);
        }
        return c;
    }

    /**
     * Returns a character that has not yet been read from the 
     * pattern. If the requested position is beyond the end of the 
     * pattern string, -1 is returned.
     * 
     * @param count          the preview position, from zero (0)
     * 
     * @return the character found, or
     *         -1 if beyond the end of the pattern string
     */
    private int peekChar(int count) {
        if (pos + count < pattern.length()) {
            return pattern.charAt(pos + count);
        } else {
            return -1;
        }
    }
    
    /**
     * Combines a list of elements. This method takes care to always 
     * concatenate adjacent string elements into a single string 
     * element.  
     * 
     * @param list           the list with elements
     * 
     * @return the combined element
     */
    private Element combineElements(ArrayList list) {
        Element  prev;
        Element  elem;
        String   str;
        int      i;

        // Concatenate string elements
        prev = (Element) list.get(0);
        for (i = 1; i < list.size(); i++) {
            elem = (Element) list.get(i); 
            if (prev instanceof StringElement 
             && elem instanceof StringElement) {

                str = ((StringElement) prev).getString() +
                      ((StringElement) elem).getString();
                elem = new StringElement(str);
                list.remove(i);
                list.set(i - 1, elem);
                i--;
            }
            prev = elem;
        }

        // Combine all remaining elements
        elem = (Element) list.get(list.size() - 1);
        for (i = list.size() - 2; i >= 0; i--) {
            prev = (Element) list.get(i);
            elem = new CombineElement(prev, elem);
        }
        return elem;  
    }
}