Question

I’m using PDFKit to generate a PDF in the browser. I’m then posting the PDF to a tornado webserver, which I want to return to the browser as a download. Here is the request handler:

class PDF(RequestHandler):
    def post(self):
        logger.info("Received PDF.")
        self.set_header('Content-Type', 'application/pdf')
        self.set_header('Content-Disposition', 'attachment; filename="export.pdf"')
        self.write(self.request.body)
        logger.info("Returning PDF.")

However, the downloaded PDF appears to be URL-encoded (slashes appear as %2F, etc.). What am I doing wrong here?

Answer 1

The problem was actually a combination of base64 decoding and URL unescaping. The following works now:

class PDF(RequestHandler):
    def post(self):
        self.set_header('Content-Type', 'application/pdf')
        self.set_header('Content-Disposition', 'attachment; filename="export.pdf"')
        self.write(b64decode(escape.url_unescape(self.request.body.split("=")[1])))