1 <?php

     2 /**

     3  * Zend Framework

     4  *

     5  * LICENSE

     6  *

     7  * This source file is subject to the new BSD license that is bundled

     8  * with this package in the file LICENSE.txt.

     9  * It is also available through the world-wide-web at this URL:

    10  * http://framework.zend.com/license/new-bsd

    11  * If you did not receive a copy of the license and are unable to

    12  * obtain it through the world-wide-web, please send an email

    13  * to license@zend.com so we can send you a copy immediately.

    14  *

    15  * @category   Zend

    16  * @package    Zend_Search_Lucene

    17  * @subpackage Document

    18  * @copyright  Copyright (c) 2005-2010 Zend Technologies USA Inc. (http://www.zend.com)

    19  * @license    http://framework.zend.com/license/new-bsd     New BSD License

    20  * @version    $Id: OpenXml.php 20096 2010-01-06 02:05:09Z bkarwin $

    21  */

    22 

    23 

    24 /** Zend_Search_Lucene_Document */

    25 require_once 'Zend/Search/Lucene/Document.php';

    26 

    27 

    28 /**

    29  * OpenXML document.

    30  *

    31  * @category   Zend

    32  * @package    Zend_Search_Lucene

    33  * @subpackage Document

    34  * @copyright  Copyright (c) 2005-2010 Zend Technologies USA Inc. (http://www.zend.com)

    35  * @license    http://framework.zend.com/license/new-bsd     New BSD License

    36  */

    37 abstract class Zend_Search_Lucene_Document_OpenXml extends Zend_Search_Lucene_Document

    38 {

    39     /**

    40      * Xml Schema - Relationships

    41      *

    42      * @var string

    43      */

    44     const SCHEMA_RELATIONSHIP = 'http://schemas.openxmlformats.org/package/2006/relationships';

    45 

    46     /**

    47      * Xml Schema - Office document

    48      *

    49      * @var string

    50      */

    51     const SCHEMA_OFFICEDOCUMENT = 'http://schemas.openxmlformats.org/officeDocument/2006/relationships/officeDocument';

    52 

    53     /**

    54      * Xml Schema - Core properties

    55      *

    56      * @var string

    57      */

    58     const SCHEMA_COREPROPERTIES = 'http://schemas.openxmlformats.org/package/2006/relationships/metadata/core-properties';

    59 

    60     /**

    61      * Xml Schema - Dublin Core

    62      *

    63      * @var string

    64      */

    65     const SCHEMA_DUBLINCORE = 'http://purl.org/dc/elements/1.1/';

    66 

    67     /**

    68      * Xml Schema - Dublin Core Terms

    69      *

    70      * @var string

    71      */

    72     const SCHEMA_DUBLINCORETERMS = 'http://purl.org/dc/terms/';

    73 

    74     /**

    75      * Extract metadata from document

    76      *

    77      * @param ZipArchive $package    ZipArchive OpenXML package

    78      * @return array    Key-value pairs containing document meta data

    79      */

    80     protected function extractMetaData(ZipArchive $package)

    81     {

    82         // Data holders

    83         $coreProperties = array();

    84 

    85         // Read relations and search for core properties

    86         $relations = simplexml_load_string($package->getFromName("_rels/.rels"));

    87         foreach ($relations->Relationship as $rel) {

    88             if ($rel["Type"] == Zend_Search_Lucene_Document_OpenXml::SCHEMA_COREPROPERTIES) {

    89                 // Found core properties! Read in contents...

    90                 $contents = simplexml_load_string(

    91                     $package->getFromName(dirname($rel["Target"]) . "/" . basename($rel["Target"]))

    92                 );

    93 

    94                 foreach ($contents->children(Zend_Search_Lucene_Document_OpenXml::SCHEMA_DUBLINCORE) as $child) {

    95                     $coreProperties[$child->getName()] = (string)$child;

    96                 }

    97                 foreach ($contents->children(Zend_Search_Lucene_Document_OpenXml::SCHEMA_COREPROPERTIES) as $child) {

    98                     $coreProperties[$child->getName()] = (string)$child;

    99                 }

   100                 foreach ($contents->children(Zend_Search_Lucene_Document_OpenXml::SCHEMA_DUBLINCORETERMS) as $child) {

   101                     $coreProperties[$child->getName()] = (string)$child;

   102                 }

   103             }

   104         }

   105 

   106         return $coreProperties;

   107     }

   108 

   109     /**

   110      * Determine absolute zip path

   111      *

   112      * @param string $path

   113      * @return string

   114      */

   115     protected function absoluteZipPath($path) {

   116         $path = str_replace(array('/', '\\'), DIRECTORY_SEPARATOR, $path);

   117         $parts = array_filter(explode(DIRECTORY_SEPARATOR, $path), 'strlen');

   118         $absolutes = array();

   119         foreach ($parts as $part) {

   120             if ('.' == $part) continue;

   121             if ('..' == $part) {

   122                 array_pop($absolutes);

   123             } else {

   124                 $absolutes[] = $part;

   125             }

   126         }

   127         return implode('/', $absolutes);

   128     }

   129 }