/**
* Method: getStartAndEndNodes
* =========================================================================
* Searches the xml with the given blockname and returns
* the corresponding start and end nodes.
*
* Parameters:
* -------------------------------------------------------------------------
* - $xml: An instance of ```SimpleXMLElement```
* - $blockname: The name of the block to find.
*
* Returns:
* -------------------------------------------------------------------------
* array
*/
protected function getStartAndEndNodes($xml, $blockname)
{
// Assume the nodes don't exist
$startNode = false;
$endNode = false;
// Search for the block start and end tags
foreach ($xml->xpath('//w:t') as $node) {
if (Str::contains($node, $this->normaliseStartTag($blockname))) {
$startNode = $node;
continue;
}
if (Str::contains($node, $this->normaliseEndTag($blockname))) {
$endNode = $node;
break;
}
}
// Bail out if we couldn't find anything
if ($startNode === false || $endNode === false) {
return false;
}
// Find the parent <w:p> node for the start tag
$node = $startNode;
$startNode = null;
while (is_null($startNode)) {
$node = $node->xpath('..')[0];
if ($node->getName() == 'p') {
$startNode = $node;
}
}
// Find the parent <w:p> node for the end tag
$node = $endNode;
$endNode = null;
while (is_null($endNode)) {
$node = $node->xpath('..')[0];
if ($node->getName() == 'p') {
$endNode = $node;
}
}
// Return the start and end node
return [$startNode, $endNode];
}
/**
* Performs some intial Setup.
*
* @param string $document This is either a filepath to a docx or html file.
* Or it may be a HTML string. The HTML string must
* contain a valid DOCTYPE.
*
* @param array $config Further configuration for the di container.
*
* @throws RuntimeException When not of the correct document type.
*/
public function __construct($document, $config = [])
{
// Configure the container
parent::__construct($config);
// Is the document a file
if (is_file($document)) {
// So that the save method can save the PDF in the same folder as
// the original source document we need a refrence to it.
$this->originalDocument = $this->file($document);
// Grab the files extension
$ext = $this->originalDocument->getExtension();
if ($ext !== 'docx' && $ext !== 'html') {
throw new RuntimeException('Must be a DOCX or HTML file.');
}
$this->documentType = $ext;
// Save the document to a new temp file
// In the case of DOCX files we may make changes to the document
// before converting to PDF so to keep the API consitent lets create
// a the temp file now.
$this->document = $this->tempFile(file_get_contents($document), $ext);
} elseif (Str::contains($document, 'DOCTYPE')) {
// Again lets save a temp file
$this->document = $this->tempFile($document, 'html');
$this->documentType = 'html';
} else {
throw new RuntimeException('Unrecognised document type!');
}
// Now create a new backend
$class = '\\Gears\\Pdf\\' . ucfirst($this->documentType) . '\\Backend';
$this->backend = new $class($this->document, $config);
}
/**
* Method: convertDoc
* =========================================================================
* This is where we actually do some converting of docx to pdf.
* We use the command line utility unoconv. Which is basically a slightly
* fancier way of using OpenOffice/LibreOffice Headless.
*
* See: http://dag.wiee.rs/home-made/unoconv/
*
* Parameters:
* -------------------------------------------------------------------------
* - $docx: This must be an instance of ```SplFileInfo```
* pointing to the document to convert.
*
* Returns:
* -------------------------------------------------------------------------
* void
*/
public function convertDoc(TempFile $docx)
{
if (!is_executable($this->binary)) {
throw new RuntimeException('The unoconv command ("' . $this->binary . '") ' . 'was not found or is not executable by the current user! ');
}
// Check to see if the profile dir exists and is writeable
if (is_dir($this->profile) && !is_writable($this->profile)) {
throw new RuntimeException('If unoconv does not have permissions to the User ' . 'Profile directory ("' . $this->profile . '") the conversion ' . 'will fail!');
}
// Build the unoconv cmd
$cmd = 'export HOME=' . $this->profile . ' && ' . $this->binary . ' ' . '--stdout ' . '-f pdf ' . '"' . $docx->getPathname() . '"';
// Run the command
$process = $this->process($cmd);
$process->run();
// Check for errors
$error = null;
if (!$process->isSuccessful()) {
$error = $process->getErrorOutput();
// NOTE: For some really odd reason the first time the command runs
// it does not complete successfully. The second time around it
// works fine. It has something to do with the homedir setup...
if (Str::contains($error, 'Error: Unable to connect')) {
$process->run();
if (!$process->isSuccessful()) {
$error = $process->getErrorOutput();
} else {
$error = null;
}
}
if (!is_null($error)) {
throw new RuntimeException($error);
}
}
// Clean up after ourselves
exec('rm -rf ' . $this->profile);
// Return the pdf data
return $process->getOutput();
}
