prtokeniser.java

iText是一个能够快速产生PDF文件的java类库。iText的java类对于那些要产生包含文本

        switch (ch) {
            case '[':
                type = TK_START_ARRAY;
                break;
            case ']':
                type = TK_END_ARRAY;
                break;
            case '/':
            {
                outBuf = new StringBuffer();
                type = TK_NAME;
                while (true) {
                    ch = file.read();
                    if (delims[ch + 1])
                        break;
                    if (ch == '#') {
                        ch = (getHex(file.read()) << 4) + getHex(file.read());
                    }
                    outBuf.append((char)ch);
                }
                backOnePosition(ch);
                break;
            }
            case '>':
                ch = file.read();
                if (ch != '>')
                    throwError("'>' not expected");
                type = TK_END_DIC;
                break;
            case '<':
            {
                int v1 = file.read();
                if (v1 == '<') {
                    type = TK_START_DIC;
                    break;
                }
                outBuf = new StringBuffer();
                type = TK_STRING;
                hexString = true;
                int v2 = 0;
                while (true) {
                    while (isWhitespace(v1))
                        v1 = file.read();
                    if (v1 == '>')
                        break;
                    v1 = getHex(v1);
                    if (v1 < 0)
                        break;
                    v2 = file.read();
                    while (isWhitespace(v2))
                        v2 = file.read();
                    if (v2 == '>') {
                        ch = v1 << 4;
                        outBuf.append((char)ch);
                        break;
                    }
                    v2 = getHex(v2);
                    if (v2 < 0)
                        break;
                    ch = (v1 << 4) + v2;
                    outBuf.append((char)ch);
                    v1 = file.read();
                }
                if (v1 < 0 || v2 < 0)
                    throwError("Error reading string");
                break;
            }
            case '%':
                type = TK_COMMENT;
                do {
                    ch = file.read();
                } while (ch != -1 && ch != '\r' && ch != '\n');
                break;
            case '(':
            {
                outBuf = new StringBuffer();
                type = TK_STRING;
                hexString = false;
                int nesting = 0;
                while (true) {
                    ch = file.read();
                    if (ch == -1)
                        break;
                    if (ch == '(') {
                        ++nesting;
                    }
                    else if (ch == ')') {
                        --nesting;
                    }
                    else if (ch == '\\') {
                        boolean lineBreak = false;
                        ch = file.read();
                        switch (ch) {
                            case 'n':
                                ch = '\n';
                                break;
                            case 'r':
                                ch = '\r';
                                break;
                            case 't':
                                ch = '\t';
                                break;
                            case 'b':
                                ch = '\b';
                                break;
                            case 'f':
                                ch = '\f';
                                break;
                            case '(':
                            case ')':
                            case '\\':
                                break;
                            case '\r':
                                lineBreak = true;
                                ch = file.read();
                                if (ch != '\n')
                                    backOnePosition(ch);
                                break;
                            case '\n':
                                lineBreak = true;
                                break;
                            default:
                            {
                                if (ch < '0' || ch > '7') {
                                    break;
                                }
                                int octal = ch - '0';
                                ch = file.read();
                                if (ch < '0' || ch > '7') {
                                    backOnePosition(ch);
                                    ch = octal;
                                    break;
                                }
                                octal = (octal << 3) + ch - '0';
                                ch = file.read();
                                if (ch < '0' || ch > '7') {
                                    backOnePosition(ch);
                                    ch = octal;
                                    break;
                                }
                                octal = (octal << 3) + ch - '0';
                                ch = octal & 0xff;
                                break;
                            }
                        }
                        if (lineBreak)
                            continue;
                        if (ch < 0)
                            break;
                    }
                    else if (ch == '\r') {
                        ch = file.read();
                        if (ch < 0)
                            break;
                        if (ch != '\n') {
                            backOnePosition(ch);
                            ch = '\n';
                        }
                    }
                    if (nesting == -1)
                        break;
                    outBuf.append((char)ch);
                }
                if (ch == -1)
                    throwError("Error reading string");
                break;
            }
            default:
            {
                outBuf = new StringBuffer();
                if (ch == '-' || ch == '+' || ch == '.' || (ch >= '0' && ch <= '9')) {
                    type = TK_NUMBER;
                    do {
                        outBuf.append((char)ch);
                        ch = file.read();
                    } while (ch != -1 && ((ch >= '0' && ch <= '9') || ch == '.'));
                }
                else {
                    type = TK_OTHER;
                    do {
                        outBuf.append((char)ch);
                        ch = file.read();
                    } while (!delims[ch + 1]);
                }
                backOnePosition(ch);
                break;
            }
        }
        if (outBuf != null)
            stringValue = outBuf.toString();
        return true;
    }
    
    public int intValue() {
        return Integer.parseInt(stringValue);
    }
    
    public boolean readLineSegment(byte input[]) throws IOException {
        int c = -1;
        boolean eol = false;
        int ptr = 0;
        int len = input.length;
	// ssteward, pdftk-1.10, 040922: 
	// skip initial whitespace; added this because PdfReader.rebuildXref()
	// assumes that line provided by readLineSegment does not have init. whitespace;
	if ( ptr < len ) {
	    while ( isWhitespace( (c = read()) ) );
	}
	while ( !eol && ptr < len ) {
	    switch (c) {
                case -1:
                case '\n':
                    eol = true;
                    break;
                case '\r':
                    eol = true;
                    int cur = getFilePointer();
                    if ((read()) != '\n') {
                        seek(cur);
                    }
                    break;
                default:
                    input[ptr++] = (byte)c;
                    break;
            }

	    // break loop? do it before we read() again
	    if( eol || len <= ptr ) {
		break;
	    }
	    else {
		c = read();
	    }
        }
        if (ptr >= len) {
            eol = false;
            while (!eol) {
                switch (c = read()) {
                    case -1:
                    case '\n':
                        eol = true;
                        break;
                    case '\r':
                        eol = true;
                        int cur = getFilePointer();
                        if ((read()) != '\n') {
                            seek(cur);
                        }
                        break;
                }
            }
        }
        
        if ((c == -1) && (ptr == 0)) {
            return false;
        }
        if (ptr + 2 <= len) {
            input[ptr++] = (byte)' ';
            input[ptr] = (byte)'X';
        }
        return true;
    }
    
    public static int[] checkObjectStart(byte line[]) {
        try {
            PRTokeniser tk = new PRTokeniser(line);
            int num = 0;
            int gen = 0;
            if (!tk.nextToken() || tk.getTokenType() != TK_NUMBER)
                return null;
            num = tk.intValue();
            if (!tk.nextToken() || tk.getTokenType() != TK_NUMBER)
                return null;
            gen = tk.intValue();
            if (!tk.nextToken())
                return null;
            if (!tk.getStringValue().equals("obj"))
                return null;
            return new int[]{num, gen};
        }
        catch (Exception ioe) {
            // empty on purpose
        }
        return null;
    }
    
    public boolean isHexString() {
        return this.hexString;
    }
    
}