6543/CommunityID
1
0
Fork 0
Code Issues 3 Pull Requests Releases Wiki Activity

import v1.0.0-RC4 | 2009-05-20

Browse Source
This commit is contained in:
6543
2019-07-17 22:08:50 +02:00
commit b484e522e8
2459 changed files with 1038434 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

119
libs/Zend/Search/Lucene/Document/Docx.php Normal file
View File

@ -0,0 +1,119 @@
<?php
/**
* Zend Framework
*
* LICENSE
*
* This source file is subject to the new BSD license that is bundled
* with this package in the file LICENSE.txt.
* It is also available through the world-wide-web at this URL:
* http://framework.zend.com/license/new-bsd
* If you did not receive a copy of the license and are unable to
* obtain it through the world-wide-web, please send an email
* to license@zend.com so we can send you a copy immediately.
*
* @category Zend
* @package Zend_Search_Lucene
* @subpackage Document
* @copyright Copyright (c) 2005-2008 Zend Technologies USA Inc. (http://www.zend.com)
* @license http://framework.zend.com/license/new-bsd New BSD License
*/
/** Zend_Search_Lucene_Document_OpenXml */
require_once 'Zend/Search/Lucene/Document/OpenXml.php';
if (class_exists ( 'ZipArchive' )) {
/**
* Docx document.
*
* @category Zend
* @package Zend_Search_Lucene
* @subpackage Document
* @copyright Copyright (c) 2005-2008 Zend Technologies USA Inc. (http://www.zend.com)
* @license http://framework.zend.com/license/new-bsd New BSD License
*/
class Zend_Search_Lucene_Document_Docx extends Zend_Search_Lucene_Document_OpenXml {
/**
* Xml Schema - WordprocessingML
*
* @var string
*/
const SCHEMA_WORDPROCESSINGML = 'http://schemas.openxmlformats.org/wordprocessingml/2006/main';
/**
* Object constructor
*
* @param string $fileName
* @param boolean $storeContent
*/
private function __construct($fileName, $storeContent) {
// Document data holders
$documentBody = array ( );
$coreProperties = array ( );
// Open OpenXML package
$package = new ZipArchive ( );
$package->open ( $fileName );
// Read relations and search for officeDocument
$relations = simplexml_load_string ( $package->getFromName ( "_rels/.rels" ) );
foreach ( $relations->Relationship as $rel ) {
if ($rel ["Type"] == Zend_Search_Lucene_Document_OpenXml::SCHEMA_OFFICEDOCUMENT) {
// Found office document! Read in contents...
$contents = simplexml_load_string ( $package->getFromName ( $this->absoluteZipPath ( dirname ( $rel ["Target"] ) . "/" . basename ( $rel ["Target"] ) ) ) );
$contents->registerXPathNamespace ( "w", Zend_Search_Lucene_Document_Docx::SCHEMA_WORDPROCESSINGML );
$paragraphs = $contents->xpath ( '//w:body/w:p' );
foreach ( $paragraphs as $paragraph ) {
$runs = $paragraph->xpath ( '//w:r/w:t' );
foreach ( $runs as $run ) {
$documentBody [] = ( string ) $run;
}
}
break;
}
}
// Read core properties
$coreProperties = $this->extractMetaData ( $package );
// Close file
$package->close ();
// Store filename
$this->addField ( Zend_Search_Lucene_Field::Text ( 'filename', $fileName ) );
// Store contents
if ($storeContent) {
$this->addField ( Zend_Search_Lucene_Field::Text ( 'body', implode ( ' ', $documentBody ) ) );
} else {
$this->addField ( Zend_Search_Lucene_Field::UnStored ( 'body', implode ( ' ', $documentBody ) ) );
}
// Store meta data properties
foreach ( $coreProperties as $key => $value ) {
$this->addField ( Zend_Search_Lucene_Field::Text ( $key, $value ) );
}
// Store title (if not present in meta data)
if (! isset ( $coreProperties ['title'] )) {
$this->addField ( Zend_Search_Lucene_Field::Text ( 'title', $fileName ) );
}
}
/**
* Load Docx document from a file
*
* @param string $fileName
* @param boolean $storeContent
* @return Zend_Search_Lucene_Document_Docx
*/
public static function loadDocxFile($fileName, $storeContent = false) {
return new Zend_Search_Lucene_Document_Docx ( $fileName, $storeContent );
}
}
} // end if (class_exists('ZipArchive'))

343
libs/Zend/Search/Lucene/Document/Html.php Normal file
View File

@ -0,0 +1,343 @@
<?php
/**
* Zend Framework
*
* LICENSE
*
* This source file is subject to the new BSD license that is bundled
* with this package in the file LICENSE.txt.
* It is also available through the world-wide-web at this URL:
* http://framework.zend.com/license/new-bsd
* If you did not receive a copy of the license and are unable to
* obtain it through the world-wide-web, please send an email
* to license@zend.com so we can send you a copy immediately.
*
* @category Zend
* @package Zend_Search_Lucene
* @subpackage Document
* @copyright Copyright (c) 2005-2008 Zend Technologies USA Inc. (http://www.zend.com)
* @license http://framework.zend.com/license/new-bsd New BSD License
*/
/** Zend_Search_Lucene_Document */
require_once 'Zend/Search/Lucene/Document.php';
/**
* HTML document.
*
* @category Zend
* @package Zend_Search_Lucene
* @subpackage Document
* @copyright Copyright (c) 2005-2008 Zend Technologies USA Inc. (http://www.zend.com)
* @license http://framework.zend.com/license/new-bsd New BSD License
*/
class Zend_Search_Lucene_Document_Html extends Zend_Search_Lucene_Document
{
/**
* List of document links
*
* @var array
*/
private $_links = array();
/**
* List of document header links
*
* @var array
*/
private $_headerLinks = array();
/**
* Stored DOM representation
*
* @var DOMDocument
*/
private $_doc;
/**
* Exclud nofollow links flag
*
* If true then links with rel='nofollow' attribute are not included into
* document links.
*
* @var boolean
*/
private static $_excludeNoFollowLinks = false;
/**
* Object constructor
*
* @param string $data
* @param boolean $isFile
* @param boolean $storeContent
*/
private function __construct($data, $isFile, $storeContent)
{
$this->_doc = new DOMDocument();
$this->_doc->substituteEntities = true;
if ($isFile) {
$htmlData = file_get_contents($data);
} else {
$htmlData = $data;
}
@$this->_doc->loadHTML($htmlData);
$xpath = new DOMXPath($this->_doc);
$docTitle = '';
$titleNodes = $xpath->query('/html/head/title');
foreach ($titleNodes as $titleNode) {
// title should always have only one entry, but we process all nodeset entries
$docTitle .= $titleNode->nodeValue . ' ';
}
$this->addField(Zend_Search_Lucene_Field::Text('title', $docTitle, $this->_doc->actualEncoding));
$metaNodes = $xpath->query('/html/head/meta[@name]');
foreach ($metaNodes as $metaNode) {
$this->addField(Zend_Search_Lucene_Field::Text($metaNode->getAttribute('name'),
$metaNode->getAttribute('content'),
$this->_doc->actualEncoding));
}
$docBody = '';
$bodyNodes = $xpath->query('/html/body');
foreach ($bodyNodes as $bodyNode) {
// body should always have only one entry, but we process all nodeset entries
$this->_retrieveNodeText($bodyNode, $docBody);
}
if ($storeContent) {
$this->addField(Zend_Search_Lucene_Field::Text('body', $docBody, $this->_doc->actualEncoding));
} else {
$this->addField(Zend_Search_Lucene_Field::UnStored('body', $docBody, $this->_doc->actualEncoding));
}
$linkNodes = $this->_doc->getElementsByTagName('a');
foreach ($linkNodes as $linkNode) {
if (($href = $linkNode->getAttribute('href')) != '' &&
(!self::$_excludeNoFollowLinks || strtolower($linkNode->getAttribute('rel')) != 'nofollow' )
) {
$this->_links[] = $href;
}
}
$this->_links = array_unique($this->_links);
$linkNodes = $xpath->query('/html/head/link');
foreach ($linkNodes as $linkNode) {
if (($href = $linkNode->getAttribute('href')) != '') {
$this->_headerLinks[] = $href;
}
}
$this->_headerLinks = array_unique($this->_headerLinks);
}
/**
* Set exclude nofollow links flag
*
* @param boolean $newValue
*/
public static function setExcludeNoFollowLinks($newValue)
{
self::$_excludeNoFollowLinks = $newValue;
}
/**
* Get exclude nofollow links flag
*
* @return boolean
*/
public static function getExcludeNoFollowLinks()
{
return self::$_excludeNoFollowLinks;
}
/**
* Get node text
*
* We should exclude scripts, which may be not included into comment tags, CDATA sections,
*
* @param DOMNode $node
* @param string &$text
*/
private function _retrieveNodeText(DOMNode $node, &$text)
{
if ($node->nodeType == XML_TEXT_NODE) {
$text .= $node->nodeValue ;
$text .= ' ';
} else if ($node->nodeType == XML_ELEMENT_NODE && $node->nodeName != 'script') {
foreach ($node->childNodes as $childNode) {
$this->_retrieveNodeText($childNode, $text);
}
}
}
/**
* Get document HREF links
*
* @return array
*/
public function getLinks()
{
return $this->_links;
}
/**
* Get document header links
*
* @return array
*/
public function getHeaderLinks()
{
return $this->_headerLinks;
}
/**
* Load HTML document from a string
*
* @param string $data
* @param boolean $storeContent
* @return Zend_Search_Lucene_Document_Html
*/
public static function loadHTML($data, $storeContent = false)
{
return new Zend_Search_Lucene_Document_Html($data, false, $storeContent);
}
/**
* Load HTML document from a file
*
* @param string $file
* @param boolean $storeContent
* @return Zend_Search_Lucene_Document_Html
*/
public static function loadHTMLFile($file, $storeContent = false)
{
return new Zend_Search_Lucene_Document_Html($file, true, $storeContent);
}
/**
* Highlight text in text node
*
* @param DOMText $node
* @param array $wordsToHighlight
* @param string $color
*/
public function _highlightTextNode(DOMText $node, $wordsToHighlight, $color)
{
$analyzer = Zend_Search_Lucene_Analysis_Analyzer::getDefault();
$analyzer->setInput($node->nodeValue, $this->_doc->encoding);
$matchedTokens = array();
while (($token = $analyzer->nextToken()) !== null) {
if (isset($wordsToHighlight[$token->getTermText()])) {
$matchedTokens[] = $token;
}
}
if (count($matchedTokens) == 0) {
return;
}
$matchedTokens = array_reverse($matchedTokens);
foreach ($matchedTokens as $token) {
// Cut text after matched token
$node->splitText($token->getEndOffset());
// Cut matched node
$matchedWordNode = $node->splitText($token->getStartOffset());
$highlightedNode = $this->_doc->createElement('b', $matchedWordNode->nodeValue);
$highlightedNode->setAttribute('style', 'color:black;background-color:' . $color);
$node->parentNode->replaceChild($highlightedNode, $matchedWordNode);
}
}
/**
* highlight words in content of the specified node
*
* @param DOMNode $contextNode
* @param array $wordsToHighlight
* @param string $color
*/
public function _highlightNode(DOMNode $contextNode, $wordsToHighlight, $color)
{
$textNodes = array();
if (!$contextNode->hasChildNodes()) {
return;
}
foreach ($contextNode->childNodes as $childNode) {
if ($childNode->nodeType == XML_TEXT_NODE) {
// process node later to leave childNodes structure untouched
$textNodes[] = $childNode;
} else {
// Skip script nodes
if ($childNode->nodeName != 'script') {
$this->_highlightNode($childNode, $wordsToHighlight, $color);
}
}
}
foreach ($textNodes as $textNode) {
$this->_highlightTextNode($textNode, $wordsToHighlight, $color);
}
}
/**
* Highlight text with specified color
*
* @param string|array $words
* @param string $color
* @return string
*/
public function highlight($words, $color = '#66ffff')
{
if (!is_array($words)) {
$words = array($words);
}
$wordsToHighlight = array();
$analyzer = Zend_Search_Lucene_Analysis_Analyzer::getDefault();
foreach ($words as $wordString) {
$wordsToHighlight = array_merge($wordsToHighlight, $analyzer->tokenize($wordString));
}
if (count($wordsToHighlight) == 0) {
return $this->_doc->saveHTML();
}
$wordsToHighlightFlipped = array();
foreach ($wordsToHighlight as $id => $token) {
$wordsToHighlightFlipped[$token->getTermText()] = $id;
}
$xpath = new DOMXPath($this->_doc);
$matchedNodes = $xpath->query("/html/body");
foreach ($matchedNodes as $matchedNode) {
$this->_highlightNode($matchedNode, $wordsToHighlightFlipped, $color);
}
}
/**
* Get HTML
*
* @return string
*/
public function getHTML()
{
return $this->_doc->saveHTML();
}
}

131
libs/Zend/Search/Lucene/Document/OpenXml.php Normal file
View File

@ -0,0 +1,131 @@
<?php
/**
* Zend Framework
*
* LICENSE
*
* This source file is subject to the new BSD license that is bundled
* with this package in the file LICENSE.txt.
* It is also available through the world-wide-web at this URL:
* http://framework.zend.com/license/new-bsd
* If you did not receive a copy of the license and are unable to
* obtain it through the world-wide-web, please send an email
* to license@zend.com so we can send you a copy immediately.
*
* @category Zend
* @package Zend_Search_Lucene
* @subpackage Document
* @copyright Copyright (c) 2005-2008 Zend Technologies USA Inc. (http://www.zend.com)
* @license http://framework.zend.com/license/new-bsd New BSD License
*/
/** Zend_Search_Lucene_Document */
require_once 'Zend/Search/Lucene/Document.php';
if (class_exists('ZipArchive')) {
/**
* OpenXML document.
*
* @category Zend
* @package Zend_Search_Lucene
* @subpackage Document
* @copyright Copyright (c) 2005-2008 Zend Technologies USA Inc. (http://www.zend.com)
* @license http://framework.zend.com/license/new-bsd New BSD License
*/
abstract class Zend_Search_Lucene_Document_OpenXml extends Zend_Search_Lucene_Document
{
/**
* Xml Schema - Relationships
*
* @var string
*/
const SCHEMA_RELATIONSHIP = 'http://schemas.openxmlformats.org/package/2006/relationships';
/**
* Xml Schema - Office document
*
* @var string
*/
const SCHEMA_OFFICEDOCUMENT = 'http://schemas.openxmlformats.org/officeDocument/2006/relationships/officeDocument';
/**
* Xml Schema - Core properties
*
* @var string
*/
const SCHEMA_COREPROPERTIES = 'http://schemas.openxmlformats.org/package/2006/relationships/metadata/core-properties';
/**
* Xml Schema - Dublin Core
*
* @var string
*/
const SCHEMA_DUBLINCORE = 'http://purl.org/dc/elements/1.1/';
/**
* Xml Schema - Dublin Core Terms
*
* @var string
*/
const SCHEMA_DUBLINCORETERMS = 'http://purl.org/dc/terms/';
/**
* Extract metadata from document
*
* @param ZipArchive $package ZipArchive OpenXML package
* @return array Key-value pairs containing document meta data
*/
protected function extractMetaData(ZipArchive $package)
{
// Data holders
$coreProperties = array();
// Read relations and search for core properties
$relations = simplexml_load_string($package->getFromName("_rels/.rels"));
foreach ($relations->Relationship as $rel) {
if ($rel["Type"] == Zend_Search_Lucene_Document_OpenXml::SCHEMA_COREPROPERTIES) {
// Found core properties! Read in contents...
$contents = simplexml_load_string(
$package->getFromName(dirname($rel["Target"]) . "/" . basename($rel["Target"]))
);
foreach ($contents->children(Zend_Search_Lucene_Document_OpenXml::SCHEMA_DUBLINCORE) as $child) {
$coreProperties[$child->getName()] = (string)$child;
}
foreach ($contents->children(Zend_Search_Lucene_Document_OpenXml::SCHEMA_COREPROPERTIES) as $child) {
$coreProperties[$child->getName()] = (string)$child;
}
foreach ($contents->children(Zend_Search_Lucene_Document_OpenXml::SCHEMA_DUBLINCORETERMS) as $child) {
$coreProperties[$child->getName()] = (string)$child;
}
}
}
return $coreProperties;
}
/**
* Determine absolute zip path
*
* @param string $path
* @return string
*/
protected function absoluteZipPath($path) {
$path = str_replace(array('/', '\\'), DIRECTORY_SEPARATOR, $path);
$parts = array_filter(explode(DIRECTORY_SEPARATOR, $path), 'strlen');
$absolutes = array();
foreach ($parts as $part) {
if ('.' == $part) continue;
if ('..' == $part) {
array_pop($absolutes);
} else {
$absolutes[] = $part;
}
}
return implode('/', $absolutes);
}
}
} // end if (class_exists('ZipArchive'))

192
libs/Zend/Search/Lucene/Document/Pptx.php Normal file
View File

@ -0,0 +1,192 @@
<?php
/**
* Zend Framework
*
* LICENSE
*
* This source file is subject to the new BSD license that is bundled
* with this package in the file LICENSE.txt.
* It is also available through the world-wide-web at this URL:
* http://framework.zend.com/license/new-bsd
* If you did not receive a copy of the license and are unable to
* obtain it through the world-wide-web, please send an email
* to license@zend.com so we can send you a copy immediately.
*
* @category Zend
* @package Zend_Search_Lucene
* @subpackage Document
* @copyright Copyright (c) 2005-2008 Zend Technologies USA Inc. (http://www.zend.com)
* @license http://framework.zend.com/license/new-bsd New BSD License
*/
/** Zend_Search_Lucene_Document_OpenXml */
require_once 'Zend/Search/Lucene/Document/OpenXml.php';
if (class_exists('ZipArchive')) {
/**
* Pptx document.
*
* @category Zend
* @package Zend_Search_Lucene
* @subpackage Document
* @copyright Copyright (c) 2005-2008 Zend Technologies USA Inc. (http://www.zend.com)
* @license http://framework.zend.com/license/new-bsd New BSD License
*/
class Zend_Search_Lucene_Document_Pptx extends Zend_Search_Lucene_Document_OpenXml
{
/**
* Xml Schema - PresentationML
*
* @var string
*/
const SCHEMA_PRESENTATIONML = 'http://schemas.openxmlformats.org/presentationml/2006/main';
/**
* Xml Schema - DrawingML
*
* @var string
*/
const SCHEMA_DRAWINGML = 'http://schemas.openxmlformats.org/drawingml/2006/main';
/**
* Xml Schema - Slide relation
*
* @var string
*/
const SCHEMA_SLIDERELATION = 'http://schemas.openxmlformats.org/officeDocument/2006/relationships/slide';
/**
* Xml Schema - Slide notes relation
*
* @var string
*/
const SCHEMA_SLIDENOTESRELATION = 'http://schemas.openxmlformats.org/officeDocument/2006/relationships/notesSlide';
/**
* Object constructor
*
* @param string $fileName
* @param boolean $storeContent
*/
private function __construct($fileName, $storeContent)
{
// Document data holders
$slides = array();
$slideNotes = array();
$documentBody = array();
$coreProperties = array();
// Open OpenXML package
$package = new ZipArchive();
$package->open($fileName);
// Read relations and search for officeDocument
$relations = simplexml_load_string($package->getFromName("_rels/.rels"));
foreach ($relations->Relationship as $rel) {
if ($rel["Type"] == Zend_Search_Lucene_Document_OpenXml::SCHEMA_OFFICEDOCUMENT) {
// Found office document! Search for slides...
$slideRelations = simplexml_load_string($package->getFromName( $this->absoluteZipPath(dirname($rel["Target"]) . "/_rels/" . basename($rel["Target"]) . ".rels")) );
foreach ($slideRelations->Relationship as $slideRel) {
if ($slideRel["Type"] == Zend_Search_Lucene_Document_Pptx::SCHEMA_SLIDERELATION) {
// Found slide!
$slides[ str_replace( 'rId', '', (string)$slideRel["Id"] ) ] = simplexml_load_string(
$package->getFromName( $this->absoluteZipPath(dirname($rel["Target"]) . "/" . dirname($slideRel["Target"]) . "/" . basename($slideRel["Target"])) )
);
// Search for slide notes
$slideNotesRelations = simplexml_load_string($package->getFromName( $this->absoluteZipPath(dirname($rel["Target"]) . "/" . dirname($slideRel["Target"]) . "/_rels/" . basename($slideRel["Target"]) . ".rels")) );
foreach ($slideNotesRelations->Relationship as $slideNoteRel) {
if ($slideNoteRel["Type"] == Zend_Search_Lucene_Document_Pptx::SCHEMA_SLIDENOTESRELATION) {
// Found slide notes!
$slideNotes[ str_replace( 'rId', '', (string)$slideRel["Id"] ) ] = simplexml_load_string(
$package->getFromName( $this->absoluteZipPath(dirname($rel["Target"]) . "/" . dirname($slideRel["Target"]) . "/" . dirname($slideNoteRel["Target"]) . "/" . basename($slideNoteRel["Target"])) )
);
break;
}
}
}
}
break;
}
}
// Sort slides
ksort($slides);
ksort($slideNotes);
// Extract contents from slides
foreach ($slides as $slideKey => $slide) {
// Register namespaces
$slide->registerXPathNamespace("p", Zend_Search_Lucene_Document_Pptx::SCHEMA_PRESENTATIONML);
$slide->registerXPathNamespace("a", Zend_Search_Lucene_Document_Pptx::SCHEMA_DRAWINGML);
// Fetch all text
$textElements = $slide->xpath('//a:t');
foreach ($textElements as $textElement) {
$documentBody[] = (string)$textElement;
}
// Extract contents from slide notes
if (isset($slideNotes[$slideKey])) {
// Fetch slide note
$slideNote = $slideNotes[$slideKey];
// Register namespaces
$slideNote->registerXPathNamespace("p", Zend_Search_Lucene_Document_Pptx::SCHEMA_PRESENTATIONML);
$slideNote->registerXPathNamespace("a", Zend_Search_Lucene_Document_Pptx::SCHEMA_DRAWINGML);
// Fetch all text
$textElements = $slideNote->xpath('//a:t');
foreach ($textElements as $textElement) {
$documentBody[] = (string)$textElement;
}
}
}
// Read core properties
$coreProperties = $this->extractMetaData($package);
// Close file
$package->close();
// Store filename
$this->addField(Zend_Search_Lucene_Field::Text('filename', $fileName));
// Store contents
if ($storeContent) {
$this->addField(Zend_Search_Lucene_Field::Text('body', implode(' ', $documentBody)));
} else {
$this->addField(Zend_Search_Lucene_Field::UnStored('body', implode(' ', $documentBody)));
}
// Store meta data properties
foreach ($coreProperties as $key => $value)
{
$this->addField(Zend_Search_Lucene_Field::Text($key, $value));
}
// Store title (if not present in meta data)
if (!isset($coreProperties['title']))
{
$this->addField(Zend_Search_Lucene_Field::Text('title', $fileName));
}
}
/**
* Load Pptx document from a file
*
* @param string $fileName
* @param boolean $storeContent
* @return Zend_Search_Lucene_Document_Pptx
*/
public static function loadPptxFile($fileName, $storeContent = false)
{
return new Zend_Search_Lucene_Document_Pptx($fileName, $storeContent);
}
}
} // end if (class_exists('ZipArchive'))

255
libs/Zend/Search/Lucene/Document/Xlsx.php Normal file
View File

@ -0,0 +1,255 @@
<?php
/**
* Zend Framework
*
* LICENSE
*
* This source file is subject to the new BSD license that is bundled
* with this package in the file LICENSE.txt.
* It is also available through the world-wide-web at this URL:
* http://framework.zend.com/license/new-bsd
* If you did not receive a copy of the license and are unable to
* obtain it through the world-wide-web, please send an email
* to license@zend.com so we can send you a copy immediately.
*
* @category Zend
* @package Zend_Search_Lucene
* @subpackage Document
* @copyright Copyright (c) 2005-2008 Zend Technologies USA Inc. (http://www.zend.com)
* @license http://framework.zend.com/license/new-bsd New BSD License
*/
/** Zend_Search_Lucene_Document_OpenXml */
require_once 'Zend/Search/Lucene/Document/OpenXml.php';
if (class_exists('ZipArchive')) {
/**
* Xlsx document.
*
* @category Zend
* @package Zend_Search_Lucene
* @subpackage Document
* @copyright Copyright (c) 2005-2008 Zend Technologies USA Inc. (http://www.zend.com)
* @license http://framework.zend.com/license/new-bsd New BSD License
*/
class Zend_Search_Lucene_Document_Xlsx extends Zend_Search_Lucene_Document_OpenXml
{
/**
* Xml Schema - SpreadsheetML
*
* @var string
*/
const SCHEMA_SPREADSHEETML = 'http://schemas.openxmlformats.org/spreadsheetml/2006/main';
/**
* Xml Schema - DrawingML
*
* @var string
*/
const SCHEMA_DRAWINGML = 'http://schemas.openxmlformats.org/drawingml/2006/main';
/**
* Xml Schema - Shared Strings
*
* @var string
*/
const SCHEMA_SHAREDSTRINGS = 'http://schemas.openxmlformats.org/officeDocument/2006/relationships/sharedStrings';
/**
* Xml Schema - Worksheet relation
*
* @var string
*/
const SCHEMA_WORKSHEETRELATION = 'http://schemas.openxmlformats.org/officeDocument/2006/relationships/worksheet';
/**
* Xml Schema - Slide notes relation
*
* @var string
*/
const SCHEMA_SLIDENOTESRELATION = 'http://schemas.openxmlformats.org/officeDocument/2006/relationships/notesSlide';
/**
* Object constructor
*
* @param string $fileName
* @param boolean $storeContent
*/
private function __construct($fileName, $storeContent)
{
// Document data holders
$sharedStrings = array();
$worksheets = array();
$documentBody = array();
$coreProperties = array();
// Open OpenXML package
$package = new ZipArchive();
$package->open($fileName);
// Read relations and search for officeDocument
$relations = simplexml_load_string($package->getFromName("_rels/.rels"));
foreach ($relations->Relationship as $rel) {
if ($rel["Type"] == Zend_Search_Lucene_Document_OpenXml::SCHEMA_OFFICEDOCUMENT) {
// Found office document! Read relations for workbook...
$workbookRelations = simplexml_load_string($package->getFromName( $this->absoluteZipPath(dirname($rel["Target"]) . "/_rels/" . basename($rel["Target"]) . ".rels")) );
$workbookRelations->registerXPathNamespace("rel", Zend_Search_Lucene_Document_OpenXml::SCHEMA_RELATIONSHIP);
// Read shared strings
$sharedStringsPath = $workbookRelations->xpath("rel:Relationship[@Type='" . Zend_Search_Lucene_Document_Xlsx::SCHEMA_SHAREDSTRINGS . "']");
$sharedStringsPath = (string)$sharedStringsPath[0]['Target'];
$xmlStrings = simplexml_load_string($package->getFromName( $this->absoluteZipPath(dirname($rel["Target"]) . "/" . $sharedStringsPath)) );
if (isset($xmlStrings) && isset($xmlStrings->si)) {
foreach ($xmlStrings->si as $val) {
if (isset($val->t)) {
$sharedStrings[] = (string)$val->t;
} elseif (isset($val->r)) {
$sharedStrings[] = $this->_parseRichText($val);
}
}
}
// Loop relations for workbook and extract worksheets...
foreach ($workbookRelations->Relationship as $workbookRelation) {
if ($workbookRelation["Type"] == Zend_Search_Lucene_Document_Xlsx::SCHEMA_WORKSHEETRELATION) {
$worksheets[ str_replace( 'rId', '', (string)$workbookRelation["Id"]) ] = simplexml_load_string(
$package->getFromName( $this->absoluteZipPath(dirname($rel["Target"]) . "/" . dirname($workbookRelation["Target"]) . "/" . basename($workbookRelation["Target"])) )
);
}
}
break;
}
}
// Sort worksheets
ksort($worksheets);
// Extract contents from worksheets
foreach ($worksheets as $sheetKey => $worksheet) {
foreach ($worksheet->sheetData->row as $row) {
foreach ($row->c as $c) {
// Determine data type
$dataType = (string)$c["t"];
switch ($dataType) {
case "s":
// Value is a shared string
if ((string)$c->v != '') {
$value = $sharedStrings[intval($c->v)];
} else {
$value = '';
}
break;
case "b":
// Value is boolean
$value = (string)$c->v;
if ($value == '0') {
$value = false;
} else if ($value == '1') {
$value = true;
} else {
$value = (bool)$c->v;
}
break;
case "inlineStr":
// Value is rich text inline
$value = $this->_parseRichText($c->is);
break;
case "e":
// Value is an error message
if ((string)$c->v != '') {
$value = (string)$c->v;
} else {
$value = '';
}
break;
default:
// Value is a string
$value = (string)$c->v;
// Check for numeric values
if (is_numeric($value) && $dataType != 's') {
if ($value == (int)$value) $value = (int)$value;
elseif ($value == (float)$value) $value = (float)$value;
elseif ($value == (double)$value) $value = (double)$value;
}
}
$documentBody[] = $value;
}
}
}
// Read core properties
$coreProperties = $this->extractMetaData($package);
// Close file
$package->close();
// Store filename
$this->addField(Zend_Search_Lucene_Field::Text('filename', $fileName));
// Store contents
if ($storeContent) {
$this->addField(Zend_Search_Lucene_Field::Text('body', implode(' ', $documentBody)));
} else {
$this->addField(Zend_Search_Lucene_Field::UnStored('body', implode(' ', $documentBody)));
}
// Store meta data properties
foreach ($coreProperties as $key => $value)
{
$this->addField(Zend_Search_Lucene_Field::Text($key, $value));
}
// Store title (if not present in meta data)
if (!isset($coreProperties['title']))
{
$this->addField(Zend_Search_Lucene_Field::Text('title', $fileName));
}
}
/**
* Parse rich text XML
*
* @param SimpleXMLElement $is
* @return string
*/
private function _parseRichText($is = null) {
$value = array();
if (isset($is->t)) {
$value[] = (string)$is->t;
} else {
foreach ($is->r as $run) {
$value[] = (string)$run->t;
}
}
return implode('', $value);
}
/**
* Load Xlsx document from a file
*
* @param string $fileName
* @param boolean $storeContent
* @return Zend_Search_Lucene_Document_Xlsx
*/
public static function loadXlsxFile($fileName, $storeContent = false)
{
return new Zend_Search_Lucene_Document_Xlsx($fileName, $storeContent);
}
}
} // end if (class_exists('ZipArchive'))