//first we have to figure out how many pages
//are in the file. this is a rough method.
//we have gs kick up an error after it opens
//the file and sees how many pages there are
$numpages = `{$gs} -dNODISPLAY "{$filepath}" -c quit`;
$pos1 = strpos($numpages, 'through');
$numpages = substr($numpages, $pos1);
$pos2 = strpos($numpages, ".");
$numpages = trim(substr($numpages, 8, $pos2 - 8));
for ($page = 1; $page <= $numpages; $page++) {
//gs the page and return as a string
$tempstring = `{$gs} -q -dNODISPLAY -dNOBIND -dWRITESYSTEMDICT -dSIMPLE -dFirstPage={$page} -dLastPage={$page} -c save -f ps2ascii.ps "{$filepath}" -c quit`;
//strip out all the trash from the string
//$tempstring = string_clean($tempstring,$preventIndex,$keepIndex);
$filetext = strtolower($tempstring) . "\n" . $filename;
doindexfile($db, $filetext, $rfiles->fields['id'], $rindextable->fields['associated_table'], $rfiles->fields['ftableid'], $page);
}
$db->Execute("UPDATE files SET indexed=1 WHERE id=" . $rfiles->fields['id']);
$pdffilecounter++;
}
}
}
echo 'Indexed file: ' . $rfiles->fields['filename'] . '.<br>';
$rfiles->MoveNext();
}
// The rest just serves to report some statistics..
if (!$textfilecounter) {
$textfilecounter = 0;
}
if (!$pdffilecounter) {
$pdffilecounter = 0;
/**
* Indexes the content of the given file
*
* The file is converted to a text file (pdfs with ghost script,
* word files were already converted to html,html characters are stripped),
* all words are lowercased, it is checked whether an entry in the table words
* already exists, if not, it is added. A relation to the word is made in
* the table associated with the given column
*/
function indexfile($db, $tableinfo, $indextable, $recordid, $fileid, $htmlfileid)
{
return false;
if (!$indextable) {
return false;
}
// if the html file exists, we'll work with that one
if ($htmlfileid) {
$fp = fopen(file_path($db, $htmlfileid), "r");
if ($fp) {
while (!feof($fp)) {
$filetext .= fgetss($fp, 64000);
}
fclose($fp);
}
$filetext = strtolower($filetext);
doindexfile($db, $filetext, $htmlfileid, $indextable, $recordid, $pagenr);
}
}
