regexparser.java

java1.6众多例子参考

              case 'g': return this.processBacksolidus_g();
              case 'X': return this.processBacksolidus_X();
              case '1':  case '2':  case '3':  case '4':
              case '5':  case '6':  case '7':  case '8':  case '9':
                return this.processBackreference();

              case 'P':
              case 'p':
                int pstart = this.offset;
                tok = processBacksolidus_pP(this.chardata);
                if (tok == null)  throw this.ex("parser.atom.5", pstart);
                break;

              default:
                tok = Token.createChar(this.chardata);
            }
            this.next();
            break;

          case T_CHAR:
            if (this.chardata == ']' || this.chardata == '{' || this.chardata == '}')
                throw this.ex("parser.atom.4", this.offset-1);
            tok = Token.createChar(this.chardata);
            int high = this.chardata;
            this.next();
            if (REUtil.isHighSurrogate(high)
                && this.read() == T_CHAR && REUtil.isLowSurrogate(this.chardata)) {
                char[] sur = new char[2];
                sur[0] = (char)high;
                sur[1] = (char)this.chardata;
                tok = Token.createParen(Token.createString(new String(sur)), 0);
                this.next();
            }
            break;

          default:
            throw this.ex("parser.atom.4", this.offset-1);
        }
        return tok;
    }

    protected RangeToken processBacksolidus_pP(int c) throws ParseException {

        this.next();
        if (this.read() != T_CHAR || this.chardata != '{')
            throw this.ex("parser.atom.2", this.offset-1);

        // handle category escape
        boolean positive = c == 'p';
        int namestart = this.offset;
        int nameend = this.regex.indexOf('}', namestart);

        if (nameend < 0)
            throw this.ex("parser.atom.3", this.offset);

        String pname = this.regex.substring(namestart, nameend);
        this.offset = nameend+1;

        return Token.getRange(pname, positive, this.isSet(RegularExpression.XMLSCHEMA_MODE));
    }

    int processCIinCharacterClass(RangeToken tok, int c) {
        return this.decodeEscaped();
    }

    /**
     * char-class ::= '[' ( '^'? range ','?)+ ']'
     * range ::= '\d' | '\w' | '\s' | category-block | range-char
     *           | range-char '-' range-char
     * range-char ::= '\[' | '\]' | '\\' | '\' [,-efnrtv] | bmp-code | character-2
     * bmp-code ::= '\' 'u' [0-9a-fA-F] [0-9a-fA-F] [0-9a-fA-F] [0-9a-fA-F]
     */
    protected RangeToken parseCharacterClass(boolean useNrange) throws ParseException {
        this.setContext(S_INBRACKETS);
        this.next();                            // '['
        boolean nrange = false;
        RangeToken base = null;
        RangeToken tok;
        if (this.read() == T_CHAR && this.chardata == '^') {
            nrange = true;
            this.next();                        // '^'
            if (useNrange) {
                tok = Token.createNRange();
            } else {
                base = Token.createRange();
                base.addRange(0, Token.UTF16_MAX);
                tok = Token.createRange();
            }
        } else {
            tok = Token.createRange();
        }
        int type;
        boolean firstloop = true;
        while ((type = this.read()) != T_EOF) {
            if (type == T_CHAR && this.chardata == ']' && !firstloop)
                break;
            firstloop = false;
            int c = this.chardata;
            boolean end = false;
            if (type == T_BACKSOLIDUS) {
                switch (c) {
                  case 'd':  case 'D':
                  case 'w':  case 'W':
                  case 's':  case 'S':
                    tok.mergeRanges(this.getTokenForShorthand(c));
                    end = true;
                    break;

                  case 'i':  case 'I':
                  case 'c':  case 'C':
                    c = this.processCIinCharacterClass(tok, c);
                    if (c < 0)  end = true;
                    break;
                    
                  case 'p':
                  case 'P':
                    int pstart = this.offset;
                    RangeToken tok2 = this.processBacksolidus_pP(c);
                    if (tok2 == null)  throw this.ex("parser.atom.5", pstart);
                    tok.mergeRanges(tok2);
                    end = true;
                    break;

                  default:
                    c = this.decodeEscaped();
                } // \ + c
            } // backsolidus
                                                // POSIX Character class such as [:alnum:]
            else if (type == T_POSIX_CHARCLASS_START) {
                int nameend = this.regex.indexOf(':', this.offset);
                if (nameend < 0) throw this.ex("parser.cc.1", this.offset);
                boolean positive = true;
                if (this.regex.charAt(this.offset) == '^') {
                    this.offset ++;
                    positive = false;
                }
                String name = this.regex.substring(this.offset, nameend);
                RangeToken range = Token.getRange(name, positive,
                                                  this.isSet(RegularExpression.XMLSCHEMA_MODE));
                if (range == null)  throw this.ex("parser.cc.3", this.offset);
                tok.mergeRanges(range);
                end = true;
                if (nameend+1 >= this.regexlen || this.regex.charAt(nameend+1) != ']')
                    throw this.ex("parser.cc.1", nameend);
                this.offset = nameend+2;
            }
            this.next();
            if (!end) {                         // if not shorthands...
                if (this.read() != T_CHAR || this.chardata != '-') { // Here is no '-'.
                    tok.addRange(c, c);
                } else {
                    this.next(); // Skips '-'
                    if ((type = this.read()) == T_EOF)  throw this.ex("parser.cc.2", this.offset);
                    if (type == T_CHAR && this.chardata == ']') {
                        tok.addRange(c, c);
                        tok.addRange('-', '-');
                    } else {
                        int rangeend = this.chardata;
                        if (type == T_BACKSOLIDUS)
                            rangeend = this.decodeEscaped();
                        this.next();
                        tok.addRange(c, rangeend);
                    }
                }
            }
            if (this.isSet(RegularExpression.SPECIAL_COMMA)
                && this.read() == T_CHAR && this.chardata == ',')
                this.next();
        }
        if (this.read() == T_EOF)
            throw this.ex("parser.cc.2", this.offset);
        if (!useNrange && nrange) {
            base.subtractRanges(tok);
            tok = base;
        }
        tok.sortRanges();
        tok.compactRanges();
        //tok.dumpRanges();
        /*
        if (this.isSet(RegularExpression.IGNORE_CASE))
            tok = RangeToken.createCaseInsensitiveToken(tok);
        */
        this.setContext(S_NORMAL);
        this.next();                    // Skips ']'

        return tok;
    }

    /**
     * '(?[' ... ']' (('-' | '+' | '&') '[' ... ']')? ')'
     */
    protected RangeToken parseSetOperations() throws ParseException {
        RangeToken tok = this.parseCharacterClass(false);
        int type;
        while ((type = this.read()) != T_RPAREN) {
            int ch = this.chardata;
            if (type == T_CHAR && (ch == '-' || ch == '&')
                || type == T_PLUS) {
                this.next();
                if (this.read() != T_LBRACKET) throw ex("parser.ope.1", this.offset-1);
                RangeToken t2 = this.parseCharacterClass(false);
                if (type == T_PLUS)
                    tok.mergeRanges(t2);
                else if (ch == '-')
                    tok.subtractRanges(t2);
                else if (ch == '&')
                    tok.intersectRanges(t2);
                else
                    throw new RuntimeException("ASSERT");
            } else {
                throw ex("parser.ope.2", this.offset-1);
            }
        }
        this.next();
        return tok;
    }

    Token getTokenForShorthand(int ch) {
        Token tok;
        switch (ch) {
          case 'd':
            tok = this.isSet(RegularExpression.USE_UNICODE_CATEGORY)
                ? Token.getRange("Nd", true) : Token.token_0to9;
            break;
          case 'D':
            tok = this.isSet(RegularExpression.USE_UNICODE_CATEGORY)
                ? Token.getRange("Nd", false) : Token.token_not_0to9;
            break;
          case 'w':
            tok = this.isSet(RegularExpression.USE_UNICODE_CATEGORY)
                ? Token.getRange("IsWord", true) : Token.token_wordchars;
            break;
          case 'W':
            tok = this.isSet(RegularExpression.USE_UNICODE_CATEGORY)
                ? Token.getRange("IsWord", false) : Token.token_not_wordchars;
            break;
          case 's':
            tok = this.isSet(RegularExpression.USE_UNICODE_CATEGORY)
                ? Token.getRange("IsSpace", true) : Token.token_spaces;
            break;
          case 'S':
            tok = this.isSet(RegularExpression.USE_UNICODE_CATEGORY)
                ? Token.getRange("IsSpace", false) : Token.token_not_spaces;
            break;

          default:
            throw new RuntimeException("Internal Error: shorthands: \\u"+Integer.toString(ch, 16));
        }
        return tok;
    }

    /**
     */
    int decodeEscaped() throws ParseException {
        if (this.read() != T_BACKSOLIDUS)  throw ex("parser.next.1", this.offset-1);
        int c = this.chardata;
        switch (c) {
          case 'e':  c = 0x1b;  break; // ESCAPE U+001B
          case 'f':  c = '\f';  break; // FORM FEED U+000C
          case 'n':  c = '\n';  break; // LINE FEED U+000A
          case 'r':  c = '\r';  break; // CRRIAGE RETURN U+000D
          case 't':  c = '\t';  break; // HORIZONTAL TABULATION U+0009
          //case 'v':  c = 0x0b;  break; // VERTICAL TABULATION U+000B
          case 'x':
            this.next();
            if (this.read() != T_CHAR)  throw ex("parser.descape.1", this.offset-1);
            if (this.chardata == '{') {
                int v1 = 0;
                int uv = 0;
                do {
                    this.next();
                    if (this.read() != T_CHAR)  throw ex("parser.descape.1", this.offset-1);
                    if ((v1 = hexChar(this.chardata)) < 0)
                        break;
                    if (uv > uv*16) throw ex("parser.descape.2", this.offset-1);
                    uv = uv*16+v1;
                } while (true);
                if (this.chardata != '}')  throw ex("parser.descape.3", this.offset-1);
                if (uv > Token.UTF16_MAX)  throw ex("parser.descape.4", this.offset-1);
                c = uv;
            } else {
                int v1 = 0;
                if (this.read() != T_CHAR || (v1 = hexChar(this.chardata)) < 0)
                    throw ex("parser.descape.1", this.offset-1);
                int uv = v1;
                this.next();
                if (this.read() != T_CHAR || (v1 = hexChar(this.chardata)) < 0)
                    throw ex("parser.descape.1", this.offset-1);
                uv = uv*16+v1;
                c = uv;
            }
            break;

          case 'u':
            int v1 = 0;
            this.next();
            if (this.read() != T_CHAR || (v1 = hexChar(this.chardata)) < 0)
                throw ex("parser.descape.1", this.offset-1);
            int uv = v1;
            this.next();
            if (this.read() != T_CHAR || (v1 = hexChar(this.chardata)) < 0)
                throw ex("parser.descape.1", this.offset-1);
            uv = uv*16+v1;
            this.next();
            if (this.read() != T_CHAR || (v1 = hexChar(this.chardata)) < 0)
                throw ex("parser.descape.1", this.offset-1);
            uv = uv*16+v1;
            this.next();
            if (this.read() != T_CHAR || (v1 = hexChar(this.chardata)) < 0)
                throw ex("parser.descape.1", this.offset-1);
            uv = uv*16+v1;
            c = uv;
            break;

          case 'v':
            this.next();
            if (this.read() != T_CHAR || (v1 = hexChar(this.chardata)) < 0)
                throw ex("parser.descape.1", this.offset-1);
            uv = v1;
            this.next();
            if (this.read() != T_CHAR || (v1 = hexChar(this.chardata)) < 0)
                throw ex("parser.descape.1", this.offset-1);
            uv = uv*16+v1;
            this.next();
            if (this.read() != T_CHAR || (v1 = hexChar(this.chardata)) < 0)
                throw ex("parser.descape.1", this.offset-1);
            uv = uv*16+v1;
            this.next();
            if (this.read() != T_CHAR || (v1 = hexChar(this.chardata)) < 0)
                throw ex("parser.descape.1", this.offset-1);
            uv = uv*16+v1;
            this.next();
            if (this.read() != T_CHAR || (v1 = hexChar(this.chardata)) < 0)
                throw ex("parser.descape.1", this.offset-1);
            uv = uv*16+v1;
            this.next();
            if (this.read() != T_CHAR || (v1 = hexChar(this.chardata)) < 0)
                throw ex("parser.descape.1", this.offset-1);
            uv = uv*16+v1;
            if (uv > Token.UTF16_MAX)  throw ex("parser.descappe.4", this.offset-1);
            c = uv;
            break;
          case 'A':
          case 'Z':
          case 'z':
            throw ex("parser.descape.5", this.offset-2);
          default:
        }
        return c;
    }

    static private final int hexChar(int ch) {
        if (ch < '0')  return -1;
        if (ch > 'f')  return -1;
        if (ch <= '9')  return ch-'0';
        if (ch < 'A')  return -1;
        if (ch <= 'F')  return ch-'A'+10;
        if (ch < 'a')  return -1;
        return ch-'a'+10;
    }
}