regularexpression.java

java1.6众多例子参考

                ret = Op.createCapture(-tok.getParenNumber(), next);
            } else {
                next = Op.createCapture(-tok.getParenNumber(), next);
                next = compile(tok.getChild(0), next, reverse);
                ret = Op.createCapture(tok.getParenNumber(), next);
            }
            break;

        case Token.LOOKAHEAD:
            ret = Op.createLook(Op.LOOKAHEAD, next, compile(tok.getChild(0), null, false));
            break;
        case Token.NEGATIVELOOKAHEAD:
            ret = Op.createLook(Op.NEGATIVELOOKAHEAD, next, compile(tok.getChild(0), null, false));
            break;
        case Token.LOOKBEHIND:
            ret = Op.createLook(Op.LOOKBEHIND, next, compile(tok.getChild(0), null, true));
            break;
        case Token.NEGATIVELOOKBEHIND:
            ret = Op.createLook(Op.NEGATIVELOOKBEHIND, next, compile(tok.getChild(0), null, true));
            break;

        case Token.INDEPENDENT:
            ret = Op.createIndependent(next, compile(tok.getChild(0), null, reverse));
            break;

        case Token.MODIFIERGROUP:
            ret = Op.createModifier(next, compile(tok.getChild(0), null, reverse),
                                    ((Token.ModifierToken)tok).getOptions(),
                                    ((Token.ModifierToken)tok).getOptionsMask());
            break;

        case Token.CONDITION:
            Token.ConditionToken ctok = (Token.ConditionToken)tok;
            int ref = ctok.refNumber;
            Op condition = ctok.condition == null ? null : compile(ctok.condition, null, reverse);
            Op yes = compile(ctok.yes, next, reverse);
            Op no = ctok.no == null ? null : compile(ctok.no, next, reverse);
            ret = Op.createCondition(next, ref, condition, yes, no);
            break;

        default:
            throw new RuntimeException("Unknown token type: "+tok.type);
        } // switch (tok.type)
        return ret;
    }


//Public

    /**
     * Checks whether the <var>target</var> text <strong>contains</strong> this pattern or not.
     *
     * @return true if the target is matched to this regular expression.
     */
    public boolean matches(char[]  target) {
        return this.matches(target, 0,  target .length , (Match)null);
    }

    /**
     * Checks whether the <var>target</var> text <strong>contains</strong> this pattern
     * in specified range or not.
     *
     * @param start Start offset of the range.
     * @param end  End offset +1 of the range.
     * @return true if the target is matched to this regular expression.
     */
    public boolean matches(char[]  target, int start, int end) {
        return this.matches(target, start, end, (Match)null);
    }

    /**
     * Checks whether the <var>target</var> text <strong>contains</strong> this pattern or not.
     *
     * @param match A Match instance for storing matching result.
     * @return Offset of the start position in <VAR>target</VAR>; or -1 if not match.
     */
    public boolean matches(char[]  target, Match match) {
        return this.matches(target, 0,  target .length , match);
    }


    /**
     * Checks whether the <var>target</var> text <strong>contains</strong> this pattern
     * in specified range or not.
     *
     * @param start Start offset of the range.
     * @param end  End offset +1 of the range.
     * @param match A Match instance for storing matching result.
     * @return Offset of the start position in <VAR>target</VAR>; or -1 if not match.
     */
    public boolean matches(char[]  target, int start, int end, Match match) {

        synchronized (this) {
            if (this.operations == null)
                this.prepare();
            if (this.context == null)
                this.context = new Context();
        }
        Context con = null;
        synchronized (this.context) {
            con = this.context.inuse ? new Context() : this.context;
            con.reset(target, start, end, this.numberOfClosures);
        }
        if (match != null) {
            match.setNumberOfGroups(this.nofparen);
            match.setSource(target);
        } else if (this.hasBackReferences) {
            match = new Match();
            match.setNumberOfGroups(this.nofparen);
            // Need not to call setSource() because
            // a caller can not access this match instance.
        }
        con.match = match;

        if (RegularExpression.isSet(this.options, XMLSCHEMA_MODE)) {
            int matchEnd = this. matchCharArray (con, this.operations, con.start, 1, this.options);
            //System.err.println("DEBUG: matchEnd="+matchEnd);
            if (matchEnd == con.limit) {
                if (con.match != null) {
                    con.match.setBeginning(0, con.start);
                    con.match.setEnd(0, matchEnd);
                }
                con.inuse = false;
                return true;
            }
            return false;
        }

        /*
         * The pattern has only fixed string.
         * The engine uses Boyer-Moore.
         */
        if (this.fixedStringOnly) {
            //System.err.println("DEBUG: fixed-only: "+this.fixedString);
            int o = this.fixedStringTable.matches(target, con.start, con.limit);
            if (o >= 0) {
                if (con.match != null) {
                    con.match.setBeginning(0, o);
                    con.match.setEnd(0, o+this.fixedString.length());
                }
                con.inuse = false;
                return true;
            }
            con.inuse = false;
            return false;
        }

        /*
         * The pattern contains a fixed string.
         * The engine checks with Boyer-Moore whether the text contains the fixed string or not.
         * If not, it return with false.
         */
        if (this.fixedString != null) {
            int o = this.fixedStringTable.matches(target, con.start, con.limit);
            if (o < 0) {
                //System.err.println("Non-match in fixed-string search.");
                con.inuse = false;
                return false;
            }
        }

        int limit = con.limit-this.minlength;
        int matchStart;
        int matchEnd = -1;

        /*
         * Checks whether the expression starts with ".*".
         */
        if (this.operations != null
            && this.operations.type == Op.CLOSURE && this.operations.getChild().type == Op.DOT) {
            if (isSet(this.options, SINGLE_LINE)) {
                matchStart = con.start;
                matchEnd = this. matchCharArray (con, this.operations, con.start, 1, this.options);
            } else {
                boolean previousIsEOL = true;
                for (matchStart = con.start;  matchStart <= limit;  matchStart ++) {
                    int ch =  target [  matchStart ] ;
                    if (isEOLChar(ch)) {
                        previousIsEOL = true;
                    } else {
                        if (previousIsEOL) {
                            if (0 <= (matchEnd = this. matchCharArray (con, this.operations,
                                                                       matchStart, 1, this.options)))
                                break;
                        }
                        previousIsEOL = false;
                    }
                }
            }
        }

        /*
         * Optimization against the first character.
         */
        else if (this.firstChar != null) {
            //System.err.println("DEBUG: with firstchar-matching: "+this.firstChar);
            RangeToken range = this.firstChar;
            if (RegularExpression.isSet(this.options, IGNORE_CASE)) {
                range = this.firstChar.getCaseInsensitiveToken();
                for (matchStart = con.start;  matchStart <= limit;  matchStart ++) {
                    int ch =  target [  matchStart ] ;
                    if (REUtil.isHighSurrogate(ch) && matchStart+1 < con.limit) {
                        ch = REUtil.composeFromSurrogates(ch,  target [  matchStart+1 ] );
                        if (!range.match(ch))  continue;
                    } else {
                        if (!range.match(ch)) {
                            char ch1 = Character.toUpperCase((char)ch);
                            if (!range.match(ch1))
                                if (!range.match(Character.toLowerCase(ch1)))
                                    continue;
                        }
                    }
                    if (0 <= (matchEnd = this. matchCharArray (con, this.operations,
                                                               matchStart, 1, this.options)))
                        break;
                }
            } else {
                for (matchStart = con.start;  matchStart <= limit;  matchStart ++) {
                    int ch =  target [  matchStart ] ;
                    if (REUtil.isHighSurrogate(ch) && matchStart+1 < con.limit)
                        ch = REUtil.composeFromSurrogates(ch,  target [  matchStart+1 ] );
                    if (!range.match(ch))  continue;
                    if (0 <= (matchEnd = this. matchCharArray (con, this.operations,
                                                               matchStart, 1, this.options)))
                        break;
                }
            }
        }

        /*
         * Straightforward matching.
         */
        else {
            for (matchStart = con.start;  matchStart <= limit;  matchStart ++) {
                if (0 <= (matchEnd = this. matchCharArray (con, this.operations, matchStart, 1, this.options)))
                    break;
            }
        }

        if (matchEnd >= 0) {
            if (con.match != null) {
                con.match.setBeginning(0, matchStart);
                con.match.setEnd(0, matchEnd);
            }
            con.inuse = false;
            return true;
        } else {
            con.inuse = false;
            return false;
        }
    }

/**
 * @return -1 when not match; offset of the end of matched string when match.
 */
    private int matchCharArray (Context con, Op op, int offset, int dx, int opts) {

        char[] target = con.charTarget;


        while (true) {
            if (op == null)
                return isSet(opts, XMLSCHEMA_MODE) && offset != con.limit ? -1 : offset;
            if (offset > con.limit || offset < con.start)
                return -1;
            switch (op.type) {
            case Op.CHAR:
                if (isSet(opts, IGNORE_CASE)) {
                    int ch = op.getData();
                    if (dx > 0) {
                        if (offset >= con.limit || !matchIgnoreCase(ch,  target [  offset ] ))
                            return -1;
                        offset ++;
                    } else {
                        int o1 = offset-1;
                        if (o1 >= con.limit || o1 < 0 || !matchIgnoreCase(ch,  target [  o1 ] ))
                            return -1;
                        offset = o1;
                    }
                } else {
                    int ch = op.getData();
                    if (dx > 0) {
                        if (offset >= con.limit || ch !=  target [  offset ] )
                            return -1;
                        offset ++;
                    } else {
                        int o1 = offset-1;
                        if (o1 >= con.limit || o1 < 0 || ch !=  target [  o1 ] )
                            return -1;
                        offset = o1;
                    }
                }
                op = op.next;
                break;

            case Op.DOT:
                if (dx > 0) {
                    if (offset >= con.limit)
                        return -1;
                    int ch =  target [  offset ] ;
                    if (isSet(opts, SINGLE_LINE)) {
                        if (REUtil.isHighSurrogate(ch) && offset+1 < con.limit)
                            offset ++;
                    } else {
                        if (REUtil.isHighSurrogate(ch) && offset+1 < con.limit)