/**
* Reads the content of an external file being indexed.
*
* @param string $ext File extension, eg. "pdf", "doc" etc.
* @param string $absFile Absolute filename of file (must exist and be validated OK before calling function)
* @param string $cPKey Pointer to section (zero for all other than PDF which will have an indication of pages into which the document should be split.)
* @return array Standard content array (title, description, keywords, body keys)
*/
public function readFileContent($ext, $absFile, $cPKey)
{
$contentArr = null;
// Return immediately if initialization didn't set support up:
if (!$this->supportedExtensions[$ext]) {
return false;
}
// Switch by file extension
switch ($ext) {
case 'pdf':
if ($this->app['pdfinfo']) {
$this->setLocaleForServerFileSystem();
// Getting pdf-info:
$cmd = $this->app['pdfinfo'] . ' ' . escapeshellarg($absFile);
CommandUtility::exec($cmd, $res);
$pdfInfo = $this->splitPdfInfo($res);
unset($res);
if ((int) $pdfInfo['pages']) {
list($low, $high) = explode('-', $cPKey);
// Get pdf content:
$tempFileName = GeneralUtility::tempnam('Typo3_indexer');
// Create temporary name
@unlink($tempFileName);
// Delete if exists, just to be safe.
$cmd = $this->app['pdftotext'] . ' -f ' . $low . ' -l ' . $high . ' -enc UTF-8 -q ' . escapeshellarg($absFile) . ' ' . $tempFileName;
CommandUtility::exec($cmd);
if (@is_file($tempFileName)) {
$content = GeneralUtility::getUrl($tempFileName);
unlink($tempFileName);
} else {
$content = '';
$this->pObj->log_setTSlogMessage(sprintf($this->sL('LLL:EXT:indexed_search/Resources/Private/Language/locallang_main.xlf:pdfToolsFailed'), $absFile), 2);
}
if ((string) $content !== '') {
$contentArr = $this->pObj->splitRegularContent($this->removeEndJunk($content));
}
}
if (!empty($pdfInfo['title'])) {
$contentArr['title'] = $pdfInfo['title'];
}
$this->setLocaleForServerFileSystem(true);
}
break;
case 'doc':
if ($this->app['catdoc']) {
$this->setLocaleForServerFileSystem();
$cmd = $this->app['catdoc'] . ' -d utf-8 ' . escapeshellarg($absFile);
CommandUtility::exec($cmd, $res);
$content = implode(LF, $res);
unset($res);
$contentArr = $this->pObj->splitRegularContent($this->removeEndJunk($content));
$this->setLocaleForServerFileSystem(true);
}
break;
case 'pps':
case 'ppt':
if ($this->app['ppthtml']) {
$this->setLocaleForServerFileSystem();
$cmd = $this->app['ppthtml'] . ' ' . escapeshellarg($absFile);
CommandUtility::exec($cmd, $res);
$content = implode(LF, $res);
unset($res);
$content = $this->pObj->convertHTMLToUtf8($content);
$contentArr = $this->pObj->splitHTMLContent($this->removeEndJunk($content));
$contentArr['title'] = basename($absFile);
$this->setLocaleForServerFileSystem(true);
}
break;
case 'xls':
if ($this->app['xlhtml']) {
$this->setLocaleForServerFileSystem();
$cmd = $this->app['xlhtml'] . ' -nc -te ' . escapeshellarg($absFile);
CommandUtility::exec($cmd, $res);
$content = implode(LF, $res);
unset($res);
$content = $this->pObj->convertHTMLToUtf8($content);
$contentArr = $this->pObj->splitHTMLContent($this->removeEndJunk($content));
$contentArr['title'] = basename($absFile);
$this->setLocaleForServerFileSystem(true);
}
break;
case 'docx':
case 'dotx':
case 'pptx':
case 'ppsx':
case 'potx':
case 'xlsx':
case 'xltx':
if ($this->app['unzip']) {
$this->setLocaleForServerFileSystem();
switch ($ext) {
case 'docx':
case 'dotx':
// Read document.xml:
$cmd = $this->app['unzip'] . ' -p ' . escapeshellarg($absFile) . ' word/document.xml';
break;
case 'ppsx':
case 'pptx':
case 'potx':
// Read slide1.xml:
$cmd = $this->app['unzip'] . ' -p ' . escapeshellarg($absFile) . ' ppt/slides/slide1.xml';
break;
case 'xlsx':
case 'xltx':
// Read sheet1.xml:
$cmd = $this->app['unzip'] . ' -p ' . escapeshellarg($absFile) . ' xl/worksheets/sheet1.xml';
break;
}
CommandUtility::exec($cmd, $res);
$content_xml = implode(LF, $res);
unset($res);
$utf8_content = trim(strip_tags(str_replace('<', ' <', $content_xml)));
$contentArr = $this->pObj->splitRegularContent($utf8_content);
// Make sure the title doesn't expose the absolute path!
$contentArr['title'] = basename($absFile);
// Meta information
$cmd = $this->app['unzip'] . ' -p ' . escapeshellarg($absFile) . ' docProps/core.xml';
CommandUtility::exec($cmd, $res);
$meta_xml = implode(LF, $res);
unset($res);
$metaContent = GeneralUtility::xml2tree($meta_xml);
if (is_array($metaContent)) {
$contentArr['title'] .= ' ' . $metaContent['cp:coreProperties'][0]['ch']['dc:title'][0]['values'][0];
$contentArr['description'] = $metaContent['cp:coreProperties'][0]['ch']['dc:subject'][0]['values'][0];
$contentArr['description'] .= ' ' . $metaContent['cp:coreProperties'][0]['ch']['dc:description'][0]['values'][0];
$contentArr['keywords'] = $metaContent['cp:coreProperties'][0]['ch']['cp:keywords'][0]['values'][0];
}
$this->setLocaleForServerFileSystem(true);
}
break;
case 'sxi':
case 'sxc':
case 'sxw':
case 'ods':
case 'odp':
case 'odt':
if ($this->app['unzip']) {
$this->setLocaleForServerFileSystem();
// Read content.xml:
$cmd = $this->app['unzip'] . ' -p ' . escapeshellarg($absFile) . ' content.xml';
CommandUtility::exec($cmd, $res);
$content_xml = implode(LF, $res);
unset($res);
// Read meta.xml:
$cmd = $this->app['unzip'] . ' -p ' . escapeshellarg($absFile) . ' meta.xml';
CommandUtility::exec($cmd, $res);
$meta_xml = implode(LF, $res);
unset($res);
$utf8_content = trim(strip_tags(str_replace('<', ' <', $content_xml)));
$contentArr = $this->pObj->splitRegularContent($utf8_content);
$contentArr['title'] = basename($absFile);
// Make sure the title doesn't expose the absolute path!
// Meta information
$metaContent = GeneralUtility::xml2tree($meta_xml);
$metaContent = $metaContent['office:document-meta'][0]['ch']['office:meta'][0]['ch'];
if (is_array($metaContent)) {
$contentArr['title'] = $metaContent['dc:title'][0]['values'][0] ? $metaContent['dc:title'][0]['values'][0] : $contentArr['title'];
$contentArr['description'] = $metaContent['dc:subject'][0]['values'][0] . ' ' . $metaContent['dc:description'][0]['values'][0];
// Keywords collected:
if (is_array($metaContent['meta:keywords'][0]['ch']['meta:keyword'])) {
foreach ($metaContent['meta:keywords'][0]['ch']['meta:keyword'] as $kwDat) {
$contentArr['keywords'] .= $kwDat['values'][0] . ' ';
}
}
}
$this->setLocaleForServerFileSystem(true);
}
break;
case 'rtf':
if ($this->app['unrtf']) {
$this->setLocaleForServerFileSystem();
$cmd = $this->app['unrtf'] . ' ' . escapeshellarg($absFile);
CommandUtility::exec($cmd, $res);
$fileContent = implode(LF, $res);
unset($res);
$fileContent = $this->pObj->convertHTMLToUtf8($fileContent);
$contentArr = $this->pObj->splitHTMLContent($fileContent);
$this->setLocaleForServerFileSystem(true);
}
break;
case 'txt':
case 'csv':
$this->setLocaleForServerFileSystem();
// Raw text
$content = GeneralUtility::getUrl($absFile);
// @todo Implement auto detection of charset (currently assuming utf-8)
$contentCharset = 'utf-8';
$content = $this->pObj->convertHTMLToUtf8($content, $contentCharset);
$contentArr = $this->pObj->splitRegularContent($content);
$contentArr['title'] = basename($absFile);
// Make sure the title doesn't expose the absolute path!
$this->setLocaleForServerFileSystem(true);
break;
case 'html':
case 'htm':
$fileContent = GeneralUtility::getUrl($absFile);
$fileContent = $this->pObj->convertHTMLToUtf8($fileContent);
$contentArr = $this->pObj->splitHTMLContent($fileContent);
break;
case 'xml':
$this->setLocaleForServerFileSystem();
// PHP strip-tags()
$fileContent = GeneralUtility::getUrl($absFile);
// Finding charset:
preg_match('/^[[:space:]]*<\\?xml[^>]+encoding[[:space:]]*=[[:space:]]*["\'][[:space:]]*([[:alnum:]_-]+)[[:space:]]*["\']/i', substr($fileContent, 0, 200), $reg);
$charset = $reg[1] ? $this->pObj->csObj->parse_charset($reg[1]) : 'utf-8';
// Converting content:
$fileContent = $this->pObj->convertHTMLToUtf8(strip_tags(str_replace('<', ' <', $fileContent)), $charset);
$contentArr = $this->pObj->splitRegularContent($fileContent);
$contentArr['title'] = basename($absFile);
// Make sure the title doesn't expose the absolute path!
$this->setLocaleForServerFileSystem(true);
break;
case 'jpg':
case 'jpeg':
case 'tif':
$this->setLocaleForServerFileSystem();
// PHP EXIF
if (function_exists('exif_read_data')) {
$exif = @exif_read_data($absFile, 'IFD0');
} else {
$exif = false;
}
if ($exif) {
$comment = trim($exif['COMMENT'][0] . ' ' . $exif['ImageDescription']);
} else {
$comment = '';
}
$contentArr = $this->pObj->splitRegularContent($comment);
$contentArr['title'] = basename($absFile);
// Make sure the title doesn't expose the absolute path!
$this->setLocaleForServerFileSystem(true);
break;
default:
return false;
}
// If no title (and why should there be...) then the file-name is set as title. This will raise the hits considerably if the search matches the document name.
if (is_array($contentArr) && !$contentArr['title']) {
// Substituting "_" for " " because many filenames may have this instead of a space char.
$contentArr['title'] = str_replace('_', ' ', basename($absFile));
}
return $contentArr;
}
