memory.php

来自「PHP 知识管理系统（基于树结构的知识管理系统）, 英文原版的PHP源码。」· PHP 代码 · 共 556 行 · 第 1/2 页

        $this->_data .= chr($value>>24 & 0xFF) .
                        chr($value>>16 & 0xFF) .
                        chr($value>>8  & 0xFF) .
                        chr($value     & 0xFF);

        $this->_position = strlen($this->_data);
    }


    /**
     * Returns a long integer from the current position in the file
     * and advances the file pointer.
     *
     * @return integer
     * @throws Zend_Search_Lucene_Exception
     */
    public function readLong()
    {
        $str = substr($this->_data, $this->_position, 8);
        $this->_position += 8;

        /**
         * Check, that we work in 64-bit mode.
         * fseek() uses long for offset. Thus, largest index segment file size in 32bit mode is 2Gb
         */
        if (PHP_INT_SIZE > 4) {
            return  ord($str{0}) << 56  |
                    ord($str{1}) << 48  |
                    ord($str{2}) << 40  |
                    ord($str{3}) << 32  |
                    ord($str{4}) << 24  |
                    ord($str{5}) << 16  |
                    ord($str{6}) << 8   |
                    ord($str{7});
        } else {
            if ((ord($str{0})          != 0) ||
                (ord($str{1})          != 0) ||
                (ord($str{2})          != 0) ||
                (ord($str{3})          != 0) ||
                ((ord($str{0}) & 0x80) != 0)) {
                     throw new Zend_Search_Lucene_Exception('Largest supported segment size (for 32-bit mode) is 2Gb');
                 }

            return  ord($str{4}) << 24  |
                    ord($str{5}) << 16  |
                    ord($str{6}) << 8   |
                    ord($str{7});
        }
    }

    /**
     * Writes long integer to the end of file
     *
     * @param integer $value
     * @throws Zend_Search_Lucene_Exception
     */
    public function writeLong($value)
    {
        // We do not need to check if file position points to the end of "file".
        // Only append operation is supported now

        /**
         * Check, that we work in 64-bit mode.
         * fseek() and ftell() use long for offset. Thus, largest index segment file size in 32bit mode is 2Gb
         */
        if (PHP_INT_SIZE > 4) {
            settype($value, 'integer');
            $this->_data .= chr($value>>56 & 0xFF) .
                            chr($value>>48 & 0xFF) .
                            chr($value>>40 & 0xFF) .
                            chr($value>>32 & 0xFF) .
                            chr($value>>24 & 0xFF) .
                            chr($value>>16 & 0xFF) .
                            chr($value>>8  & 0xFF) .
                            chr($value     & 0xFF);
        } else {
            if ($value > 0x7FFFFFFF) {
                throw new Zend_Search_Lucene_Exception('Largest supported segment size (for 32-bit mode) is 2Gb');
            }

            $this->_data .= chr(0) . chr(0) . chr(0) . chr(0) .
                            chr($value>>24 & 0xFF) .
                            chr($value>>16 & 0xFF) .
                            chr($value>>8  & 0xFF) .
                            chr($value     & 0xFF);
        }

        $this->_position = strlen($this->_data);
    }



    /**
     * Returns a variable-length integer from the current
     * position in the file and advances the file pointer.
     *
     * @return integer
     */
    public function readVInt()
    {
        $nextByte = ord($this->_data[$this->_position++]);
        $val = $nextByte & 0x7F;

        for ($shift=7; ($nextByte & 0x80) != 0; $shift += 7) {
            $nextByte = ord($this->_data[$this->_position++]);
            $val |= ($nextByte & 0x7F) << $shift;
        }
        return $val;
    }

    /**
     * Writes a variable-length integer to the end of file.
     *
     * @param integer $value
     */
    public function writeVInt($value)
    {
        // We do not need to check if file position points to the end of "file".
        // Only append operation is supported now

        settype($value, 'integer');
        while ($value > 0x7F) {
            $this->_data .= chr( ($value & 0x7F)|0x80 );
            $value >>= 7;
        }
        $this->_data .= chr($value);

        $this->_position = strlen($this->_data);
    }


    /**
     * Reads a string from the current position in the file
     * and advances the file pointer.
     *
     * @return string
     */
    public function readString()
    {
        $strlen = $this->readVInt();
        if ($strlen == 0) {
            return '';
        } else {
            /**
             * This implementation supports only Basic Multilingual Plane
             * (BMP) characters (from 0x0000 to 0xFFFF) and doesn't support
             * "supplementary characters" (characters whose code points are
             * greater than 0xFFFF)
             * Java 2 represents these characters as a pair of char (16-bit)
             * values, the first from the high-surrogates range (0xD800-0xDBFF),
             * the second from the low-surrogates range (0xDC00-0xDFFF). Then
             * they are encoded as usual UTF-8 characters in six bytes.
             * Standard UTF-8 representation uses four bytes for supplementary
             * characters.
             */

            $str_val = substr($this->_data, $this->_position, $strlen);
            $this->_position += $strlen;

            for ($count = 0; $count < $strlen; $count++ ) {
                if (( ord($str_val{$count}) & 0xC0 ) == 0xC0) {
                    $addBytes = 1;
                    if (ord($str_val{$count}) & 0x20 ) {
                        $addBytes++;

                        // Never used. Java2 doesn't encode strings in four bytes
                        if (ord($str_val{$count}) & 0x10 ) {
                            $addBytes++;
                        }
                    }
                    $str_val .= substr($this->_data, $this->_position, $addBytes);
                    $this->_position += $addBytes;
                    $strlen          += $addBytes;

                    // Check for null character. Java2 encodes null character
                    // in two bytes.
                    if (ord($str_val{$count})   == 0xC0 &&
                        ord($str_val{$count+1}) == 0x80   ) {
                        $str_val{$count} = 0;
                        $str_val = substr($str_val,0,$count+1)
                                 . substr($str_val,$count+2);
                    }
                    $count += $addBytes;
                }
            }

            return $str_val;
        }
    }

    /**
     * Writes a string to the end of file.
     *
     * @param string $str
     * @throws Zend_Search_Lucene_Exception
     */
    public function writeString($str)
    {
        /**
         * This implementation supports only Basic Multilingual Plane
         * (BMP) characters (from 0x0000 to 0xFFFF) and doesn't support
         * "supplementary characters" (characters whose code points are
         * greater than 0xFFFF)
         * Java 2 represents these characters as a pair of char (16-bit)
         * values, the first from the high-surrogates range (0xD800-0xDBFF),
         * the second from the low-surrogates range (0xDC00-0xDFFF). Then
         * they are encoded as usual UTF-8 characters in six bytes.
         * Standard UTF-8 representation uses four bytes for supplementary
         * characters.
         */

        // We do not need to check if file position points to the end of "file".
        // Only append operation is supported now

        // convert input to a string before iterating string characters
        settype($str, 'string');

        $chars = $strlen = strlen($str);
        $containNullChars = false;

        for ($count = 0; $count < $strlen; $count++ ) {
            /**
             * String is already in Java 2 representation.
             * We should only calculate actual string length and replace
             * \x00 by \xC0\x80
             */
            if ((ord($str{$count}) & 0xC0) == 0xC0) {
                $addBytes = 1;
                if (ord($str{$count}) & 0x20 ) {
                    $addBytes++;

                    // Never used. Java2 doesn't encode strings in four bytes
                    // and we dont't support non-BMP characters
                    if (ord($str{$count}) & 0x10 ) {
                        $addBytes++;
                    }
                }
                $chars -= $addBytes;

                if (ord($str{$count}) == 0 ) {
                    $containNullChars = true;
                }
                $count += $addBytes;
            }
        }

        if ($chars < 0) {
            throw new Zend_Search_Lucene_Exception('Invalid UTF-8 string');
        }

        $this->writeVInt($chars);
        if ($containNullChars) {
            $this->_data .= str_replace($str, "\x00", "\xC0\x80");

        } else {
            $this->_data .= $str;
        }

        $this->_position = strlen($this->_data);
    }


    /**
     * Reads binary data from the current position in the file
     * and advances the file pointer.
     *
     * @return string
     */
    public function readBinary()
    {
        $length = $this->readVInt();
        $returnValue = substr($this->_data, $this->_position, $length);
        $this->_position += $length;
        return $returnValue;
    }
}