DeliciousPie：是否可以在批量请求中发布多个文件

Question

I would like to make a bulk post. The problem is that each item requires an image (or, maybe even a few). Is it possible to do this via bulk request?

The model:

class CollageItem(models.Model):
  url = models.URLField(null = True)
  image = models.FileField(upload_to = 'i')
  thumbnail = models.FileField(upload_to = 't')

And the TastyPie object:

class CollageItemResource(ModelResource):
  image = fields.FileField(attribute = 'image', null = True, blank = true)
  thumbnail = fields.FileField(attribute = 'thumbnail', null = True, blank = true)
  class Meta:
    queryset = CollageItem.objects.all(
    resource_name = "collage_item"

Can I post multiple images using bulk request or should I revert to individual posts?

Answer 1

Ofcourse you can! Depending on the image size, you'll have to decide whether it takes too long to upload them or not, but it is possible. According to tastypie docs, bulk creation and updation are possible via the Patch option.

Docs here

and in detail here

Answer 2

I went with custom serializer road:

class FormPostSerializer(Serializer):
    formats = ['form']
    content_types = {
        'form': 'multipart/form-data',
    }

    def from_form(self, content):
        try:
            dict = cgi.parse_multipart(StringIO(content), self.form_boundary)
        except Exception, e:
            raise e
        for key, value in dict.iteritems():
            dict[key] = value[0] if len(value) > 0 else None
        return dict

And base class for all the resources that require posting multiple files:

class FormResource(ModelResource):
    class Meta:
        serializer = FormPostSerializer()

    def dispatch(self, request_type, request, **kwargs):
        cth = request.META.get('CONTENT_TYPE') or \
            request.META.get('Content-type') or \
            self._meta.serializer.content_types['json']
        self.Meta.serializer.form_boundary = self.parse_content_type_header(cth)
        return super(FormResource, self).dispatch(request_type, request, **kwargs)

    def parse_content_type_header(self, content_type_header):
        parts = cgi.parse_header(content_type_header)
        rv = {}
        for p in parts:
            if isinstance(p, dict):
                rv = dict(rv.items() + p.items())
        return rv

Of course, the serializer requires some additional processing (UTF8 fields, for example) I omitted those from the answer.