segmentinfo.php

很棒的在线教学系统

     * Result array structure: array(docId => array(pos1, pos2, ...), ...)
     *
     * @param Zend_Search_Lucene_Index_Term $term
     * @param integer $shift
     * @return Zend_Search_Lucene_Index_TermInfo
     */
    public function termPositions(Zend_Search_Lucene_Index_Term $term, $shift = 0)
    {
        $termInfo = $this->getTermInfo($term);

        if (!$termInfo instanceof Zend_Search_Lucene_Index_TermInfo) {
            return array();
        }

        $frqFile = $this->openCompoundFile('.frq');
        $frqFile->seek($termInfo->freqPointer,SEEK_CUR);
        $freqs = array();
        $docId = 0;

        for ($count = 0; $count < $termInfo->docFreq; $count++) {
            $docDelta = $frqFile->readVInt();
            if ($docDelta % 2 == 1) {
                $docId += ($docDelta-1)/2;
                $freqs[$docId] = 1;
            } else {
                $docId += $docDelta/2;
                $freqs[$docId] = $frqFile->readVInt();
            }
        }

        $result = array();
        $prxFile = $this->openCompoundFile('.prx');
        $prxFile->seek($termInfo->proxPointer, SEEK_CUR);
        foreach ($freqs as $docId => $freq) {
            $termPosition = 0;
            $positions = array();

            for ($count = 0; $count < $freq; $count++ ) {
                $termPosition += $prxFile->readVInt();
                $positions[] = $termPosition;
            }

            $result[$shift + $docId] = $positions;
        }

        return $result;
    }

    /**
     * Load normalizatin factors from an index file
     *
     * @param integer $fieldNum
     * @throws Zend_Search_Lucene_Exception
     */
    private function _loadNorm($fieldNum)
    {
        if ($this->_hasSingleNormFile) {
            $normfFile = $this->openCompoundFile('.nrm');

            $header              = $normfFile->readBytes(3);
            $headerFormatVersion = $normfFile->readByte();

            if ($header != 'NRM'  ||  $headerFormatVersion != (int)0xFF) {
                throw new  Zend_Search_Lucene_Exception('Wrong norms file format.');
            }

            foreach ($this->_fields as $fieldNum => $fieldInfo) {
                if ($fieldInfo->isIndexed) {
                    $this->_norms[$fieldNum] = $normfFile->readBytes($this->_docCount);
                }
            }
        } else {
            $fFile = $this->openCompoundFile('.f' . $fieldNum);
            $this->_norms[$fieldNum] = $fFile->readBytes($this->_docCount);
        }
    }

    /**
     * Returns normalization factor for specified documents
     *
     * @param integer $id
     * @param string $fieldName
     * @return float
     */
    public function norm($id, $fieldName)
    {
        $fieldNum = $this->getFieldNum($fieldName);

        if ( !($this->_fields[$fieldNum]->isIndexed) ) {
            return null;
        }

        if (!isset($this->_norms[$fieldNum])) {
            $this->_loadNorm($fieldNum);
        }

        return Zend_Search_Lucene_Search_Similarity::decodeNorm( ord($this->_norms[$fieldNum]{$id}) );
    }

    /**
     * Returns norm vector, encoded in a byte string
     *
     * @param string $fieldName
     * @return string
     */
    public function normVector($fieldName)
    {
        $fieldNum = $this->getFieldNum($fieldName);

        if ($fieldNum == -1  ||  !($this->_fields[$fieldNum]->isIndexed)) {
            $similarity = Zend_Search_Lucene_Search_Similarity::getDefault();

            return str_repeat(chr($similarity->encodeNorm( $similarity->lengthNorm($fieldName, 0) )),
                              $this->_docCount);
        }

        if (!isset($this->_norms[$fieldNum])) {
            $this->_loadNorm($fieldNum);
        }

        return $this->_norms[$fieldNum];
    }


    /**
     * Returns true if any documents have been deleted from this index segment.
     *
     * @return boolean
     */
    public function hasDeletions()
    {
        return $this->_deleted !== null;
    }


    /**
     * Returns true if segment has single norms file.
     *
     * @return boolean
     */
    public function hasSingleNormFile()
    {
        return $this->_hasSingleNormFile ? 1 : 0;
    }

    /**
     * Returns true if segment is stored using compound segment file.
     *
     * @return boolean
     */
    public function isCompound()
    {
        return $this->_isCompound ? 1 : 0;
    }

    /**
     * Deletes a document from the index segment.
     * $id is an internal document id
     *
     * @param integer
     */
    public function delete($id)
    {
        $this->_deletedDirty = true;

        if (extension_loaded('bitset')) {
            if ($this->_deleted === null) {
                $this->_deleted = bitset_empty($id);
            }
            bitset_incl($this->_deleted, $id);
        } else {
            if ($this->_deleted === null) {
                $this->_deleted = array();
            }

            $this->_deleted[$id] = 1;
        }
    }

    /**
     * Checks, that document is deleted
     *
     * @param integer
     * @return boolean
     */
    public function isDeleted($id)
    {
        if ($this->_deleted === null) {
            return false;
        }

        if (extension_loaded('bitset')) {
            return bitset_in($this->_deleted, $id);
        } else {
            return isset($this->_deleted[$id]);
        }
    }


    /**
     * Write changes if it's necessary.
     */
    public function writeChanges()
    {
        if (!$this->_deletedDirty) {
            return;
        }

        if (extension_loaded('bitset')) {
            $delBytes = $this->_deleted;
            $bitCount = count(bitset_to_array($delBytes));
        } else {
            $byteCount = floor($this->_docCount/8)+1;
            $delBytes = str_repeat(chr(0), $byteCount);
            for ($count = 0; $count < $byteCount; $count++) {
                $byte = 0;
                for ($bit = 0; $bit < 8; $bit++) {
                    if (isset($this->_deleted[$count*8 + $bit])) {
                        $byte |= (1<<$bit);
                    }
                }
                $delBytes{$count} = chr($byte);
            }
            $bitCount = count($this->_deleted);
        }


        // Get new generation number
        Zend_Search_Lucene_LockManager::obtainWriteLock($this->_directory);

        $delFileList = array();
        foreach ($this->_directory->fileList() as $file) {
        	if ($file == $this->_name . '.del') {
        		// Matches <segment_name>.del file name
        		$delFileList[] = 0;
        	} else if (preg_match('/^' . $this->_name . '_([a-zA-Z0-9]+)\.del$/i', $file, $matches)) {
        		// Matches <segment_name>_NNN.del file names
                $delFileList[] = (int)base_convert($matches[1], 36, 10);
            }
        }

        if (count($delFileList) == 0) {
        	// There is no deletions file for current segment in the directory
        	// Set detetions file generation number to 1
        	$this->_delGen = 1;
        } else {
        	// There are some deletions files for current segment in the directory
        	// Set detetions file generation number to the highest + 1
        	$this->_delGen = max($delFileList) + 1;
        }

        $delFile = $this->_directory->createFile($this->_name . '_' . base_convert($this->_delGen, 10, 36) . '.del');

        Zend_Search_Lucene_LockManager::releaseWriteLock($this->_directory);


        $delFile->writeInt($this->_docCount);
        $delFile->writeInt($bitCount);
        $delFile->writeBytes($delBytes);

        $this->_deletedDirty = false;
    }



    /**
     * Term Dictionary File object for stream like terms reading
     *
     * @var Zend_Search_Lucene_Storage_File
     */
    private $_tisFile = null;

    /**
     * Actual offset of the .tis file data
     *
     * @var integer
     */
    private $_tisFileOffset;

    /**
     * Frequencies File object for stream like terms reading
     *
     * @var Zend_Search_Lucene_Storage_File
     */
    private $_frqFile = null;

    /**
     * Actual offset of the .frq file data
     *
     * @var integer
     */
    private $_frqFileOffset;

    /**
     * Positions File object for stream like terms reading
     *
     * @var Zend_Search_Lucene_Storage_File
     */
    private $_prxFile = null;

    /**
     * Actual offset of the .prx file in the compound file
     *
     * @var integer
     */
    private $_prxFileOffset;


    /**
     * Actual number of terms in term stream
     *
     * @var integer
     */
    private $_termCount = 0;

    /**
     * Overall number of terms in term stream
     *
     * @var integer
     */
    private $_termNum = 0;

    /**
     * Segment index interval
     *
     * @var integer
     */
    private $_indexInterval;

    /**
     * Segment skip interval
     *
     * @var integer
     */
    private $_skipInterval;

    /**
     * Last TermInfo in a terms stream
     *
     * @var Zend_Search_Lucene_Index_TermInfo
     */
    private $_lastTermInfo = null;

    /**
     * Last Term in a terms stream
     *
     * @var Zend_Search_Lucene_Index_Term
     */
    private $_lastTerm = null;

    /**
     * Map of the document IDs
     * Used to get new docID after removing deleted documents.
     * It's not very effective from memory usage point of view,
     * but much more faster, then other methods
     *
     * @var array|null
     */
    private $_docMap = null;

    /**
     * An array of all term positions in the documents.
     * Array structure: array( docId => array( pos1, pos2, ...), ...)
     *
     * Is set to null if term positions loading has to be skipped
     *
     * @var array|null
     */
    private $_lastTermPositions;


    /**
     * Terms scan mode