允许用户下载在AWS上生成的zip文件

Question

Here is how I do to generate ZIP and download it from a server, it works well in local development.

import zipfile

doc = get_object_or_404(Document,id=id_obj)

filepath = doc.file.path
filename = os.path.basename(doc.file.name)
directory = os.path.dirname(filepath)

xzip = zipfile.ZipFile(os.path.join(directory,"%s.zip" % filename), "w")
xzip.write(filepath,filename)
xzip.close()

zip_file = open(xzip.filename, 'rb')
response = HttpResponse(zip_file, content_type='application/zip')
response['Content-Disposition'] = 'attachment; filename="%s.zip"' % 
                                   os.path.splitext(filename)[0]
return response

All my static & media files are uploaded to AWS in production. So I change a little bit

 # filepath becomes
 filepath = settings.MEDIA_ROOT + "/" + doc.file.name

But When I try to download it, it gives me [Errno 2] No such file or directory with the link:

https://bucket_name.s3.amazonaws.com/media/public/files/file.pdf.zip

the settings.MEDIA_ROOT is:

AWS_ACCESS_KEY_ID = config('AWS_ACCESS_KEY_ID')
AWS_SECRET_ACCESS_KEY = config('AWS_SECRET_ACCESS_KEY')

AWS_STORAGE_BUCKET_NAME = 'bucket_name'
AWS_S3_CUSTOM_DOMAIN = '%s.s3.amazonaws.com' % AWS_STORAGE_BUCKET_NAME

AWS_PUBLIC_MEDIA_LOCATION = 'media/public'

doc.file.path gives me the error: 'This backend doesn't support absolute paths' , that's why I changed to MEDIA_ROOT + doc.file.name

How to do that to download from AWS the zip file generated?

Answer 1

The file exists on S3, not the local file system. When you call those os.path.* functions the code is trying to find the file on the local file system. It's giving you that error because that S3 URL you are giving it as the path can't be mapped to anything on the local file system.

Why don't you allow S3 to serve the file directly to the end-user's browser by simply return a redirect response with the URL of the S3 file instead of trying to read the file and return the contents in the response?