Is it possible to generate PDF with StreamingHttpResponse as it's possible to do so with CSV for large dataset?

Question

I have a large dataset that I have to generate CSV and PDF for. With CSV, I use this guide: https://docs.djangoproject.com/en/3.1/howto/outputting-csv/

import csv

from django.http import StreamingHttpResponse

class Echo:
    """An object that implements just the write method of the file-like
    interface.
    """
    def write(self, value):
        """Write the value by returning it, instead of storing in a buffer."""
        return value

def some_streaming_csv_view(request):
    """A view that streams a large CSV file."""
    # Generate a sequence of rows. The range is based on the maximum number of
    # rows that can be handled by a single sheet in most spreadsheet
    # applications.
    rows = (["Row {}".format(idx), str(idx)] for idx in range(65536))
    pseudo_buffer = Echo()
    writer = csv.writer(pseudo_buffer)
    response = StreamingHttpResponse((writer.writerow(row) for row in rows),
                                     content_type="text/csv")
    response['Content-Disposition'] = 'attachment; filename="somefilename.csv"'
    return response

It works great. However, I can't find anything that can be done for PDF. Can it? I use render_to_pdf as well as I use a template for PDF.

Answer 1

I had a similar situation where I am able to "generate and stream download" files of csv, json and xml types and I want to do the same with Excel - xlsx file.

Unfortunately, I couldn't do that. But, during that time I found a few things

1. The files , CSV, JSON and XML are text files with a proper representation. But, when comes to PDF or Excel (or similar files), these files are built with a proper formatting and proper metadata.

2. The binary data of PDF and similar docs are written to the io buffer only when we call some specific methods. [ showPage() and save() methods of reportlab. (source- Django Doc) ]

3. If we inspect the file stream, PDF and Excel require sophisticated special applications (eg: PDF reader, Bowsers etc) to view/read the data whereas, with CSV and JSON, we need only a simple text editor.

So, I conclude that the process of "on the fly generation of file with stream download" (not sure what is the correct technical term I should use) is not possible for all file types, but only possible for a few text-oriented files

Note: This is my limited experience, which may be wrong.

Answer 2

Have you tried FileResponse?

Something like this should work, it is basically what you can find in the Django doc:

import io
from django.http import FileResponse
from reportlab.pdfgen import canvas

def stream_pdf(request):
    buffer = io.BytesIO()
    p = canvas.Canvas(buffer)
    p.drawString(10, 10, "Hello world.")
    p.showPage()
    p.save()
    buffer.seek(io.SEEK_SET)
    return FileResponse(buffer, as_attachment=True, filename='helloworld.pdf')

Answer 3

Looking at the link you provided it does provide a link to a page on creating and sending pdf files dynamically using reportlab.

import io
from django.http import FileResponse
from reportlab.pdfgen import canvas

def some_view(request):
    # Create a file-like buffer to receive PDF data.
    buffer = io.BytesIO()

    # Create the PDF object, using the buffer as its "file."
    p = canvas.Canvas(buffer)

    # Draw things on the PDF. Here's where the PDF generation happens.
    # See the ReportLab documentation for the full list of functionality.
    p.drawString(100, 100, "Hello world.")

    # Close the PDF object cleanly, and we're done.
    p.showPage()
    p.save()

    # FileResponse sets the Content-Disposition header so that browsers
    # present the option to save the file.
    buffer.seek(0)
    return FileResponse(buffer, as_attachment=True, filename='hello.pdf')

Here's a link to the reportlab api documentation. Its kinda lengthy and stored in a annoying to navigate single page pdf, but it should get you up and running and able to nicely format the PDFs as you want.

Answer 4

Think of CSV as a fruit salad. You can slice bananas in a big pot, add some grapefruits, some pineapple, ... and then split the whole into individual portions that you bring together to the table (this is: you generate your CSV file, and then you send it to the client). But you could also make individual portions directly: Cut some slices of a banana in a small bowl, add some grapefruits, some pineapple, ... bring this small bowl to the table, and repeat the process for other individual portions (this is: you generate your CSV file and send it part by part to the client as you generate it).

Well if CSV is a fruit salad, then PDF is a cake. You have to mix all your ingredients and put it in the oven. This means you can't bring a slice of the cake to the table until you have baked the whole cake. Likewise, you can't start sending your PDF file to the client until it's entirely generated.

So, to answer your question, this (response = StreamingHttpResponse((writer.writerow(row) for row in rows), content_type="text/csv")) can't be done for PDF.

However, once your file is generated, you can stream it to the client using FileResponse as mentioned in other answers.

If your issue is that the generation of the PDF takes too much time (and might trigger a timeout error for instance), here are some things to consider:

1. Try to optimize the speed of your generation algorithm
2. Generate the file in the background before the client requests it and store it in your storage system. You might want to use a cronjob or celery to trigger the generation of the PDF without blocking the HTTP request.
3. Use websockets to send the file to the client as soon as it is ready to be downloaded (see django-channels)