Importing External Blogs

Mezzanine has the ability to import blog posts from other blogging platforms using a Django management command. These are the currently supported formats and their commands:

Each command takes a Mezzanine username to assign the blog posts to as well as certain arguments specific to the blog platform. For example to import an existing Wordpress blog:

$ python manage.py import_wordpress --mezzanine-user=username [options]

Use the --help argument to learn more about the arguments specific to each blog platform’s command. For example you can see all options for Wordpress by running:

$ python manage.py import_wordpress --help

Considerations

There are some known issues with HTML formatting loss - specifically where a heading tag is followed by a paragraph tag or another block HTML element that is not typically enclosed with a <p> tag is followed by a paragraph. This depends heavily on the originating platform and how it encodes the blog post’s copy. The import processor gets this about 90% correct but you may need to do some quick clean up afterwards.

Generally speaking you shouldn’t be able to import your data twice. There is a check in place to either create or update for both comments and posts as they are processed, so even if you run the importer multiple times you should only end up with data imported once. However if you have changed any data this will be overwritten.

Importing from Wordpress

Dependencies

The first step is to export your Wordpress data. Login to Wordpress and go to Tools -> Export. Here you can select your filters, otherwise only published posts will be exported. Once you have saved your export file make a note of the location you saved it to.

Note

It is faster to import directly from your filesystem if you can, especially if you have a large blog with lots of comments.

The next step is to run the import_wordpress command where the url argument contains the path or URL to your export file:

$ python manage.py import_wordpress --mezzanine-user=.. --url=[path|URL]

Importing from Blogger

The Blogger import currently has one known limitation which is a maximum of 500 blogs or 500 comments per blog that can be imported. If you have more than this the import will still work but end up being truncated.

Dependencies

  • Google’s gdata Library

The first step is to obtain your Blogger ID. Login to Blogger and go to Settings. You’ll see that the address in your browser end with BlogID=XXX where XXX is your Blogger ID. Make a note of this and while you’re in settings, go to Site Feed then set Allow Blog Feeds to be Full - this will give you all your data when you run the import.

The next step is to run the import_blogger command where the blogger-id argument contains the Blogger ID you retrieved:

$ python manage.py import_blogger --mezzanine-user=.. --blogger-id=XXX

Importing from Tumblr

Simply run the import_tumblr command where the tumblr-user argument contains your Tumblr username:

$ python manage.py import_blogger --mezzanine-user=.. --tumblr-user=username

Importing RSS

Dependencies

Simply run the import_rss command where the rss-url argument contains the URL for your RSS feed:

$ python manage.py import_rss --mezzanine-user=.. --rss-url=url

Importing from Posterous

Dependencies

Simply run import_posterous command with the right params. You need to get your API key from the Posterous API Reference:

$ python manage.py import_posterous --mezzanine-user=.. --api-token=.. --posterous-user=your_posterous_login --posterous-pass=your_posterous_password

If you have more than one blog on your posterous account check out the -posterous-host option. Be aware that like the tumblr importer, this leaves your media assets on the Posterous servers. If you’re worried about posterous being shut down you may want want to have a closer look at the API to actually export your media.

Importing BlogML

Dependencies

Simply run the import_blogml command where the blogxmlfname argument contains local blogml file name:

$ python manage.py import_blogml --blogxmlfname=filename

Importer API - Adding New Importers

The importer system has been designed to be extensible so that import commands can easily be added for other blogging platforms.

Each importer’s management command is located in the mezzanine.blog.management.commands package, and should have its module named import_type where type represents the type of import the command is for. This module will then contain a class named Command which subclasses mezzanine.blog.management.base.BaseImporterCommand.

The first step is to define any custom arguments the command will require using Python’s argparse handling.

The main responsbility of the Command class is then to implement a handle_import() method which handles retrieving blog posts and comments from the particular blogging platform. The handle_import() method is passed a dictionary of options for the command. The add_post() and add_comment() methods should be called inside the handle_import() method, adding posts and comments respectively. The add_post() method returns a post to be used with the add_comment() method. For example:

from optparse import make_option
from django.core.management.base import CommandError
from mezzanine.blog.management.base import BaseImporterCommand

class Command(BaseImporterCommand):

    def add_arguments(self, parser):
        super(Command, self).add_arguments(parser)
        parser.add_argument(
            "-s", "--some-arg-name", dest="some_arg_var",
            help="Description of some-arg-name")

    def handle_import(self, options):
        # Perform the tasks that need to occur to retrieve blog posts.
        # We'll use an imaginary "posts" variable that contains a list of
        # post dicts with keys: title, author, pub_date, tags and content.
        # In this example we have access to the command line argument
        # "some-arg-name" via "options["some_arg_var"]".
        for retrieved_post in posts:
            added_post = self.add_post(**retrieved_post)
            # Another imaginary variable to demo the API.
            for retrieved_comment in comments:
                self.add_comment(post=added_post, **retrieved_comment)