*/ class EPub { public $xml; //FIXME change to protected, later protected $xpath; protected $file; protected $meta; protected $namespaces; protected $imagetoadd=''; /** * Constructor * * @param string $file path to epub file to work on * @throws Exception if metadata could not be loaded */ public function __construct($file){ // open file $this->file = $file; $zip = new ZipArchive(); if(!@$zip->open($this->file)){ throw new Exception('Failed to read epub file'); } // read container data $data = $zip->getFromName('META-INF/container.xml'); if($data == false){ throw new Exception('Failed to access epub container data'); } $xml = new DOMDocument(); $xml->registerNodeClass('DOMElement','EPubDOMElement'); $xml->loadXML($data); $xpath = new EPubDOMXPath($xml); $nodes = $xpath->query('//n:rootfiles/n:rootfile[@media-type="application/oebps-package+xml"]'); $this->meta = $nodes->item(0)->attr('full-path'); // load metadata $data = $zip->getFromName($this->meta); if(!$data){ throw new Exception('Failed to access epub metadata'); } $this->xml = new DOMDocument(); $this->xml->registerNodeClass('DOMElement','EPubDOMElement'); $this->xml->loadXML($data); $this->xml->formatOutput = true; $this->xpath = new EPubDOMXPath($this->xml); $zip->close(); } /** * file name getter */ public function file(){ return $this->file; } /** * Writes back all meta data changes */ public function save(){ $zip = new ZipArchive(); $res = @$zip->open($this->file, ZipArchive::CREATE); if($res === false){ throw new Exception('Failed to write back metadata'); } $zip->addFromString($this->meta,$this->xml->saveXML()); // add the cover image if($this->imagetoadd){ $path = dirname('/'.$this->meta).'/php-epub-meta-cover.img'; // image path is relative to meta file $path = ltrim($path,'/'); $zip->addFromString($path,file_get_contents($this->imagetoadd)); $this->imagetoadd=''; } $zip->close(); } /** * Get or set the book author(s) * * Authors should be given with a "file-as" and a real name. The file as * is used for sorting in e-readers. * * Example: * * array( * 'Pratchett, Terry' => 'Terry Pratchett', * 'Simpson, Jacqeline' => 'Jacqueline Simpson', * ) * * @params array $authors */ public function Authors($authors=false){ // set new data if($authors !== false){ // Author where given as a comma separated list if(is_string($authors)){ if($authors == ''){ $authors = array(); }else{ $authors = explode(',',$authors); $authors = array_map('trim',$authors); } } // delete existing nodes $nodes = $this->xpath->query('//opf:metadata/dc:creator[@opf:role="aut"]'); foreach($nodes as $node) $node->delete(); // add new nodes $parent = $this->xpath->query('//opf:metadata')->item(0); foreach($authors as $as => $name){ if(is_int($as)) $as = $name; //numeric array given $node = $parent->newChild('dc:creator',$name); $node->attr('opf:role', 'aut'); $node->attr('opf:file-as', $as); } $this->reparse(); } // read current data $rolefix = false; $authors = array(); $nodes = $this->xpath->query('//opf:metadata/dc:creator[@opf:role="aut"]'); if($nodes->length == 0){ // no nodes where found, let's try again without role $nodes = $this->xpath->query('//opf:metadata/dc:creator'); $rolefix = true; } foreach($nodes as $node){ $name = $node->nodeValue; $as = $node->attr('opf:file-as'); if(!$as){ $as = $name; $node->attr('opf:file-as',$as); } if($rolefix){ $node->attr('opf:role','aut'); } $authors[$as] = $name; } return $authors; } /** * Set or get the book title * * @param string $title */ public function Title($title=false){ return $this->getset('dc:title',$title); } /** * Set or get the book's language * * @param string $lang */ public function Language($lang=false){ return $this->getset('dc:language',$lang); } /** * Set or get the book' publisher info * * @param string $publisher */ public function Publisher($publisher=false){ return $this->getset('dc:publisher',$publisher); } /** * Set or get the book's copyright info * * @param string $rights */ public function Copyright($rights=false){ return $this->getset('dc:rights',$rights); } /** * Set or get the book's description * * @param string $description */ public function Description($description=false){ return $this->getset('dc:description',$description); } /** * Set or get the book's ISBN number * * @param string $isbn */ public function ISBN($isbn=false){ return $this->getset('dc:identifier',$isbn,'opf:scheme','ISBN'); } /** * Set or get the Google Books ID * * @param string $google */ public function Google($google=false){ return $this->getset('dc:identifier',$google,'opf:scheme','GOOGLE'); } /** * Set or get the Amazon ID of the book * * @param string $amazon */ public function Amazon($amazon=false){ return $this->getset('dc:identifier',$amazon,'opf:scheme','AMAZON'); } /** * Set or get the book's subjects (aka. tags) * * Subject should be given as array, but a comma separated string will also * be accepted. * * @param array $subjects */ public function Subjects($subjects=false){ // setter if($subjects !== false){ if(is_string($subjects)){ if($subjects === ''){ $subjects = array(); }else{ $subjects = explode(',',$subjects); $subjects = array_map('trim',$subjects); } } // delete previous $nodes = $this->xpath->query('//opf:metadata/dc:subject'); foreach($nodes as $node){ $node->delete(); } // add new ones $parent = $this->xpath->query('//opf:metadata')->item(0); foreach($subjects as $subj){ $node = $this->xml->createElement('dc:subject',htmlspecialchars($subj)); $node = $parent->appendChild($node); } $this->reparse(); } //getter $subjects = array(); $nodes = $this->xpath->query('//opf:metadata/dc:subject'); foreach($nodes as $node){ $subjects[] = $node->nodeValue; } return $subjects; } /** * Read the cover data * * Returns an associative array with the following keys: * * mime - filetype (usually image/jpeg) * data - the binary image data * found - the internal path, or false if no image is set in epub * * When no image is set in the epub file, the binary data for a transparent * GIF pixel is returned. * * When adding a new image this function return no or old data because the * image contents are not in the epub file, yet. The image will be added when * the save() method is called. * * @param string $path local filesystem path to a new cover image * @param string $mime mime type of the given file * @return array */ public function Cover($path=false, $mime=false){ // set cover if($path !== false){ // remove current pointer $nodes = $this->xpath->query('//opf:metadata/opf:meta[@name="cover"]'); foreach($nodes as $node) $node->delete(); // remove previous manifest entries if they where made by us $nodes = $this->xpath->query('//opf:manifest/opf:item[@id="php-epub-meta-cover"]'); foreach($nodes as $node) $node->delete(); if($path){ // add pointer $parent = $this->xpath->query('//opf:metadata')->item(0); $node = $parent->newChild('opf:meta'); $node->attr('opf:name','cover'); $node->attr('opf:content','php-epub-meta-cover'); // add manifest $parent = $this->xpath->query('//opf:manifest')->item(0); $node = $parent->newChild('opf:item'); $node->attr('id','php-epub-meta-cover'); $node->attr('opf:href','php-epub-meta-cover.img'); $node->attr('opf:media-type',$mime); // remember path for save action $this->imagetoadd = $path; } $this->reparse(); } // load cover $nodes = $this->xpath->query('//opf:metadata/opf:meta[@name="cover"]'); if(!$nodes->length) return $this->no_cover(); $coverid = (String) $nodes->item(0)->attr('opf:content'); if(!$coverid) return $this->no_cover(); $nodes = $this->xpath->query('//opf:manifest/opf:item[@id="'.$coverid.'"]'); if(!$nodes->length) return $this->no_cover(); $mime = $nodes->item(0)->attr('opf:media-type'); $path = $nodes->item(0)->attr('opf:href'); $path = dirname('/'.$this->meta).'/'.$path; // image path is relative to meta file $path = ltrim($path,'/'); $zip = new ZipArchive(); if(!@$zip->open($this->file)){ throw new Exception('Failed to read epub file'); } $data = $zip->getFromName($path); return array( 'mime' => $mime, 'data' => $data, 'found' => $path ); } /** * A simple getter/setter for simple meta attributes * * It should only be used for attributes that are expected to be unique * * @param string $item XML node to set/get * @param string $value New node value * @param string $att Attribute name * @param string $aval Attribute value */ protected function getset($item,$value=false,$att=false,$aval=false){ // construct xpath $xpath = '//opf:metadata/'.$item; if($att){ $xpath .= "[@$att=\"$aval\"]"; } // set value if($value !== false){ $value = htmlspecialchars($value); $nodes = $this->xpath->query($xpath); if($nodes->length == 1 ){ if($value === ''){ // the user want's to empty this value -> delete the node $nodes->item(0)->delete(); }else{ // replace value $nodes->item(0)->nodeValue = $value; } }else{ // if there are multiple matching nodes for some reason delete // them. we'll replace them all with our own single one foreach($nodes as $n) $n->delete(); // readd them if($value){ $parent = $this->xpath->query('//opf:metadata')->item(0); $node = $this->xml->createElement($item,$value); $node = $parent->appendChild($node); if($att) $node->attr($att,$aval); } } $this->reparse(); } // get value $nodes = $this->xpath->query($xpath); if($nodes->length){ return $nodes->item(0)->nodeValue; }else{ return ''; } } /** * Return a not found response for Cover() */ protected function no_cover(){ return array( 'data' => base64_decode('R0lGODlhAQABAIAAAAAAAP///yH5BAEAAAEALAAAAAABAAEAAAIBTAA7'), 'mime' => 'image/gif', 'found' => false ); } /** * Reparse the DOM tree * * I had to rely on this because otherwise xpath failed to find the newly * added nodes */ protected function reparse() { $this->xml->loadXML($this->xml->saveXML()); $this->xpath = new EPubDOMXPath($this->xml); } } class EPubDOMXPath extends DOMXPath { public function __construct(DOMDocument $doc){ parent::__construct($doc); if(is_a($doc->documentElement, 'EPubDOMElement')){ foreach($doc->documentElement->namespaces as $ns => $url){ $this->registerNamespace($ns,$url); } } } } class EPubDOMElement extends DOMElement { public $namespaces = array( 'n' => 'urn:oasis:names:tc:opendocument:xmlns:container', 'opf' => 'http://www.idpf.org/2007/opf', 'dc' => 'http://purl.org/dc/elements/1.1/' ); public function __construct($name, $value='', $namespaceURI=''){ list($ns,$name) = $this->splitns($name); $value = htmlspecialchars($value); if(!$namespaceURI && $ns){ $namespaceURI = $this->namespaces[$ns]; } parent::__construct($name, $value, $namespaceURI); } /** * Create and append a new child * * Works with our epub namespaces and omits default namespaces */ public function newChild($name, $value=''){ list($ns,$local) = $this->splitns($name); if($ns){ $nsuri = $this->namespaces[$ns]; if($this->isDefaultNamespace($nsuri)){ $name = $local; $nsuri = ''; } } // this doesn't call the construcor: $node = $this->ownerDocument->createElement($name,$value); $node = new EPubDOMElement($name,$value,$nsuri); return $this->appendChild($node); } /** * Split given name in namespace prefix and local part * * @param string $name * @return array (namespace, name) */ public function splitns($name){ $list = explode(':',$name,2); if(count($list) < 2) array_unshift($list,''); return $list; } /** * Simple EPub namespace aware attribute accessor */ public function attr($attr,$value=null){ list($ns,$attr) = $this->splitns($attr); $nsuri = ''; if($ns){ $nsuri = $this->namespaces[$ns]; if(!$this->namespaceURI){ if($this->isDefaultNamespace($nsuri)){ $nsuri = ''; } }elseif($this->namespaceURI == $nsuri){ $nsuri = ''; } } if(!is_null($value)){ if($value === false){ // delete if false was given if($nsuri){ $this->removeAttributeNS($nsuri,$attr); }else{ $this->removeAttribute($attr); } }else{ // modify if value was given if($nsuri){ $this->setAttributeNS($nsuri,$attr,$value); }else{ $this->setAttribute($attr,$value); } } }else{ // return value if none was given if($nsuri){ return $this->getAttributeNS($nsuri,$attr); }else{ return $this->getAttribute($attr); } } } /** * Remove this node from the DOM */ public function delete(){ $this->parentNode->removeChild($this); } }