Introduction

transmogrify.wordpress provides a collective.transmogrifier pipeline for importing a Wordpress blog into Plone.

In particular, it is useful if you have a Wordpress export in WXR format, and want to end up with a Plone-based blog using Scrawl and plone.app.discussion. (The complementary use of collective.blog.star for additional blog functionality is recommended, but not required.)

A WXR export of a Wordpress.com blog can be obtained via Tools -> Export in the blog's dashboard.

Usage

A sample pipeline, wordpress.cfg, is provided. You can use it by editing it in place to set a few variables at the top of the file, and then importing the transmogrify.wordpress profile via portal_setup.

The following settings must be configured in the options part of wordpress.cfg:

filename

Full path to the Wordpress WXR export file.

path

Path, relative to the site root, of the folder where blog entries should be created.

type

The desired portal_type for blog entries. Must be an Archetypes-based type with a 'text' field, and must already exist in the site. The default, 'Blog Entry', may be obtained by installing Scrawl.

entry-selector

CSS selector to find the body of the post within its original HTML page. (The body is fetched this way instead of using the one in the WXR, because the WXR contains special Wordpress markup.)

Other prerequisites:

• The importing of comments requires that plone.app.discussion is installed.

Note

The pipeline will take some time to run, and you'll probably get a timeout error unless you are accessing Zope directly with no proxy or web server in front. Even if this happens the import will continue to run. Progress is logged to the Zope event log at the INFO level, so you can watch the log to see what is happening.

Alternatively, you can copy the wordpress.cfg pipeline into your own package, and edit and run it as part of your own GenericSetup profile by naming it in a transmogrifier.txt file placed in the profile directory (see the transmogrify.wordpress profile for an example of how this is done).

Caveats

This importer is currently in an alpha state. Don't expect it to work perfectly without a little coaxing. Contributions to improve it are welcome.

• For now, the import can only create Archetypes-based content types. That could probably be changed with a little work.
• No effort is made to map between differing sets of usernames in Wordpress and Plone.
• Wordpress' "categories" and "tags" are lumped together as Plone "tags".
• Image sources are currently updated with path-based links even if the site is configured to link by UID.
• Height and width specified on image tags is left intact. No effort is made to choose the most reasonable scale size to download for the specified height and width.
• No effort has been made to make the pipeline do something reasonable if it is run more than once. In particular, comments will be re-added, and images will be re-fetched, even if they are already present in the site.
• No effort is made to make sure that the old Wordpress URLs will continue to work in the new site if the blog's domain is pointed at the new site. This can be handled via rewrites in something in front of Zope, such as Apache or nginx.
• Comments that are unapproved or marked as spam are not imported.
• Pingbacks are imported, but don't actually contain a link to the referring page, as plone.app.discussion doesn't support HTML in comments at this time.

Included Blueprints

Note

This section contains advanced information on the components of the pipeline.

transmogrify.wordpress makes use of several custom transmogrifier blueprints. Some of these would probably be useful in non-Wordpress-related pipelines and should probably get factored out into separate packages.

transmogrify.wordpress.blueprints.wxrsource

Parses a WXR-format export file and injects an item into the pipeline for each blog post and comment.

Settings:

filename

Path to the WXR file on disk.

path

Path relative to the site root where blog posts should be imported.

type

portal_type that should be used for imported blog posts.

import-comments

you can choose to import wordpress comments. defaults to false

Sets the following pipeline keys:

portal_type

Value of the type setting for blog posts. plone.Comment for comments.

_orig_url

The URL of the original blog post.

_path

Path of the new item, based on the path setting and the Wordpress post_name. For comments, path of the item to which the comment should be added.

title

Post title

description

Post description

creation_date, effectiveDate, modification_date

Wordpress creation date

creators

Wordpress post author

subject

A list of Wordpress categories and tags, sorted alphabetically

text

The text of the blog post or comment. Unmodified from Wordpress' markup (so uses newlines instead of P and BR tags, among other things).

_transitions

Set to 'publish' to control the workflow state later in the pipeline.

_comment_id

For comments only, the unique Wordpress ID of the comment (used for threading).

_in_reply_to

For comments only, the Wordpress ID of the comment's parent (used for threading).

author_name

For comments only, name of the comment's author.

author_email

For comments only, email of the comment's author.

created

For comments only, date of the comment.

transmogrify.wordpress.blueprints.text_cleanup

Cleans up Wordpress markup into more standard HTML. In particular, it will:

• Encode the text if necessary
• Normalize carriage returns to newlines
• Replace double newlines with P tags
• Replace single newlines with BR tags

These operations are performed on the pipeline key named in the blueprint's key setting (defaults to "text").

transmogrify.wordpress.blueprints.fetch_html

Fetches the content of an HTML page and selects a portion of it, which it places (UTF8-encoded) into a pipeline key.

Settings:

url_key

Name of the pipeline key which gives the URL to be fetched. (Default: _orig_url.)

selector

CSS selector specifying which portion of the retrieved page to retain.

target_key

Name of the pipeline key where the fetched HTML should be stored.

If the url_key is not found for the current item, or the page cannot be retrieved, no change will be made to the target key.

transmogrify.wordpress.blueprints.html_image_source

This blueprint parses HTML for images, fetches the images from their current remote location, and injects new items into the pipeline so that those images will be added to the Plone site. It also updates the src attribute of the image tags so that they will refer to the new local images.

If an IMG tag is inside an A tag whose href points at another image, it is assumed that the latter image is a larger version, and it is fetched instead of the one referred to by the IMG's src attribute.

Settings:

key

Name of the pipeline key which should be parsed for images, and updated. (Default: "text")

path

Path relative to the site root of the container to which images should be added. (Default: "images")

transmogrify.wordpress.blueprints.set_mimetype

Wraps text in a Zope File object with a particular mimetype, so that the mimetype can be correctly transferred when the value is set on an item via an Archetypes File or Text field mutator.

Settings:

key

Name of the pipeline key which contains the text, and which should be replaced with the File object.

mimetype

Mimetype that should be set. (Default: "text/html")

transmogrify.wordpress.blueprints.comment_constructor

Constructs a plone.app.discussion comment, and adds it to the conversation for a particular item.

Threading of comments is handled as long as the injection of comments into the pipeline is ordered such that a comment's parent has always been already created.

Uses the following keys from the item in the pipeline:

portal_type

Must be plone.Comment or the item will be skipped.

path

Path to the item which is being commented on.

text

Text of the comment (should be plain text).

author_name

Name of the commenter.

author_email

Email of the commenter.

created

Date the comment was made.

_comment_id

Source system's unique ID of this comment.

_in_reply_to

Source system's unique ID of the parent of this comment. That comment must have already been added or things will fail. Defaults to '0', which means a top-level comment.

Credits

transmogrify.wordpress was created by David Glick for Groundwire.