Ümit Tunç
36 lines
1.0 KiB
PHP
36 lines
1.0 KiB
PHP
<?php
|
|
function docx_to_html($docx, $html_dir) {
|
|
// Ensure the LANG environment variable is set
|
|
putenv('LANG=ru_RU.UTF-8');
|
|
|
|
// Ensure the output directory exists
|
|
if (!is_dir($html_dir)) {
|
|
mkdir($html_dir, 0777, true);
|
|
}
|
|
|
|
// Construct the command
|
|
$command = "libreoffice --headless --convert-to html --outdir " . escapeshellarg($html_dir) . " " . escapeshellarg($docx);
|
|
// Execute the command
|
|
$output = shell_exec($command);
|
|
$output = extract_between_markers_docx($output);
|
|
|
|
update_src_paths($output, $html_dir);
|
|
|
|
return $output;
|
|
}
|
|
|
|
function extract_between_markers_docx($input_str) {
|
|
// Define the regular expression pattern to match the content between the markers
|
|
$pattern = '/->\s*(.*?)\s*using filter : HTML \(StarWriter\)/';
|
|
|
|
// Perform the regex match
|
|
if (preg_match($pattern, $input_str, $matches)) {
|
|
// Return the matched content
|
|
return $matches[1];
|
|
}
|
|
|
|
// Return null if no match is found
|
|
return null;
|
|
}
|
|
|
|
?>
|