MS Word Files corrupt when downloading from different PHP scripts

Question

I'm facing a strange issue with a web app I'm developing, when I try to download the same file from two different PHP scripts one file is fine and the other is detected as corrupt by MS Word.

Both scripts contain the same exact code, one is a literal copy of the other.

ob_start();

header('Content-Description: File Transfer');
header("Content-Type: $contentType");
header('Content-Disposition: attachment; filename='.basename($file));
header('Content-Length: ' . filesize($file));
ob_clean();

readfile($file);
exit;

I'm using a windows server 2012 running Apache 2.4 and PHP7. I use two different folders for production and development /app/ and /appBeta/ respectively.

When I request the file download from the download script in the production folder I get the corrupted file.

When I request the same file from the download script in the development folder it is fine. It may be important to note that the file is located inside of the production folder, both production and development are in document root.

Here are the request/response headers for both downloads:

Production (corrupt)

Development (not corrupt)

Another thing I noticed is a discrepancy between the "Origin" data included with files. (Right Click File->Properties->Details)

Production (corrupt)

Development(not corrupt)

I'm at a loss for what could be causing this, I suspect something with Apache or permissions...anyone have any idea what's going on here?

EDIT: The file is an existing static file in the production folder. It is not a generated file.

Answer 1

I just had a similar problem with a word doc generator. It turned out the problem was mixed encodings. Managed to solve the issue by feeding everything through the ForceUTF8 library to standardize the encoding issues:

use \ForceUTF8\Encoding;

$output = Encoding::toUTF8($raw_input); //where $raw_input is your unsanitized source data

The resulting $output then rendered into a functional word doc. In your case, if the file is a preexisting static file (not mentioned in the original question), you may be having an encoding issue defined by the response headers also.

You will need to include the ForceUTF8 library in your project to use this, see their github page for setup instructions here: https://github.com/neitanod/forceutf8