如何使用 python-docx 从模板流式传输文件

Question

Ok so right now I have a function in my django app that creates a word document like so:

def form_view(request):

    if request.method == 'POST':
       #do a bunch of things
       context = { 
        'model_1' : model_1,
        }
        in_template = "Forms/mytemplate.docx"

        doc = DocxTemplate(in_template)
        doc.render(context)
        out_filename = "outfile.docx"
        http_word_response = HttpResponse(content_type='application/vnd.openxmlformats-officedocument.wordprocessingml.document')
        http_word_response['Content-Disposition'] = 'attachment; filename=%s' % out_filename
        doc.save(http_word_response)
        return http_word_response

     else:
       return render( request, 'mysite/form.html', context)

So right now, the file is served up as the http_word_response variable and it works great

I want to change this so that the file is actually created and saved into S3, but I'd like to not save the file locally first and just use the stream function of the python-docx package: https://python-docx.readthedocs.io/en/latest/user/documents.html#opening-a-file-like-document

Here's the sample it gives:

with open('foobar.docx', 'rb') as f:
    source_stream = StringIO(f.read())
document = Document(source_stream)
source_stream.close()
...
target_stream = StringIO()
document.save(target_stream)

I'm not sure how to translate that sample into reading in my template file and outputting a file/object I can then send to S3 and do something with.

My best guess would start like this:

        with open(in_template, 'rb') as f:
            source_stream = StringIO(f.read())
        doc = DocxTemplate(source_stream)
        source_stream.close()

        target_stream = StringIO()
        document.save(target_stream)

But then I get confused on where my doc.render(context) part goes and how I can change the name of the target_stream to a filename I want.

Any help on getting started with that is appreciated.

Answer 1

Late to the party here, I assume you already found the answer, but anyway...

source_stream and target_stream are StringIO objects, text streams that use an in-memory text buffer. Think of them as in-memory files, without any info about their paths.

Your guess is good, and the doc.render(context) goes in the empty line, right after creating the DocxTemplate and closing the source_stream (since you don't need it anymore).

To post the target_stream to s3, you can use the smart open library, with something like this:

from smart_open import open

with open('s3://bucket/key.txt', 'wb', transport_params=transport_params) as fout:
    fout.write(target_stream)

See smart open documentation for info about the parameters.