Any plan to support m2m fields? about django-queryset-csv HOT 2 CLOSED

Thanks for your interest, I hope the project continues to be useful for you.

This library only really supports exports where the schema of the CSV closely aligns to a django model. When we have more elaborate CSV needs, we just roll them ourselves.

Many-to-many relationships are hard to express in a CSV. Not only would you have to choose a particular field or string serialization technique for the related model, you also have to stuff it into a collection or reduce to a single value.

For example:

class Person(models.model):
    name = models.CharField()
    places = models.ManyToManyField('Place')

class Place(models.model):
    description = models.CharField()
    location = models.PointField()

could lead to a csv like:

name,places\n
rust,"[(0,0),(1,1),(2,2)]"\n
marty,"[(0,1),(1,0),(3,3)]"\n

Where you'd have chosen the location field on Place, as well as an array as the way to collapse the multiple values.

There's no plan to support this, but I'll leave this issue open for a time to keep the discussion open for things I've failed to consider.

from django-queryset-csv.

Any way of specifying number of items only? Like in above example one could choose "Kathmandu" and "Pokhara" (cities from Nepal), and then the result could be "2"? That might be easier? If so, how?

from django-queryset-csv.