These tools allow simple roundtrippig of data between a WordPress content database and git-controlled files on the local machine.

HTML stored in a WordPress database may have a variety of problems (typos, out-of-date '™' versus '®' references, outdated company names, and so forth. It's much easier to fix such things using text processing tools, but hard to turn such tools loose against the database. Furthermore, a text processing flow allows the use of git, which in turn allows careful step-by-step review and collaboration.

Although it's technically possible to edit a WordPress XML archive directly, it's quite unpleasant.

This package therefore has two tools. The first tool expands an XML archive into a directory containg the archive contents, split up by page, post or other item type.

The second tool reverses this process, creating an XML archive that can be imported into WordPress.

The tools use Python's lxml library, and therefore are aware of the structure of what they're processing.

This Python3 script scans a WordPress archive, and splits it up into files, one file per item. It will optionally remove some Divi-specific tags as it creates the files.

It is intended for use in conjunction with a tool like Flywheel local, when you want to make massive regular changes to the content of a WordPress site, remove Divi tags, etc.

Documntation is here

This python3 script takes a collection of XML files (typically as generated by extract-wp-items, and puts them together into an archive that can subsequently be imported.

Documentation is here.

To use this, you must make two changes in your WordPress set up.

1. You must install the WordPress Importer.

2. You must add wordpress-import-update.php to your WordPress installation's app/public/wp-content/mu-plugins directory. (Without this, duplicate pages are skipped).

   Note: If you're using Flywheel local, you may need to re-install this after pulling an image from the main site.

1. Using Flywheel's local, clone your site to a Ubuntu system. (We have not tested on Mac, Windows, or other distributions. We tested on Ubuntu 18.04 LTS.)

2. Copy your local copy of wordpress-import-update.php to ~/Local Files/$SITE/app/public/wp-content/mu-plugins.

3. Create a directory you'll use for working on the site.

4. Change to that directory, and git init .

1. Using WordPress wp-admin>Tools>Export, create a full archive from the Flywheel local image. (You can also do a subset, but my normal workflow is the full site.)

2. Extract to your working git repository directory.

   cd {path-to-repo}
   python3 {path-to-script}/extract-wordpress-items.py . --strip-divi

3. Capture any changes.

   git add .
   git commit -m "Import from WordPress"

4. Make your local changes.

5. Create an archive.

   python3 {path-to-script}/compose-wordpress-items.py . /tmp/archive.xml

6. Use WordPress Tools>Import to update your Flywheel local image of the repo.

Terry Moore, MCCI Corporation

2021-01-23: These tools are very much works in progress. However, I successfully used them on MCCI's website to make a number of site-wide changes, check them with Flywheel local, and then updated the MCCI site with the results. This was much less painful than using the on-line WordPress editor -- I could use Visual Studio Code, command-line tools like grep and aspell, etc. Since I've only used it in my use case, it's possible that there are ugly mistakes that I've not yet encountered.

It is clear that it would be even more convenient to further split each page and post item so that the <content:encoded> CDATA is placed into a separate .html file, parallel to the item's .xml file. This would allow Visual Studio Code (or other editor) to apply syntax analysis to the HTML body.

Tested with python3 v3.6.9 and python3-lxml version 4.2.1.

Released under MIT license.