(But it tries, at least on first appearances.)
As you may know, I’m working on consolidating my various blogging-ish sites into a single, combined, Github Pages site. This involves moving some 200 posts from my current blog and my old, nearly-defunct personal blog to GH Pages (which is based on Jekyll).
Fortunately, Blogger/Google give you a proof-against-most-idiots process to export everything about your blog — posts, comments, templates, and several kitchen sinks, all rolled up into one ginormous XML file. (My raw export data was ~5,260 lines; pretty-formatted (as in “legibly formatted”), it’s 16,239 lines long. Well, I did start blogging in 2003.
I am not hand-splitting that up into individual _posts files.. That’s what software is for. Right?
Step 1. Google convert blogger to github pages. That led me to this Gist which will pull your posts’ RSS feed and spit out appropriately-formatted Jekyll/Markdown files, suitable for publishing in GH Pages.
For your 25 most recent posts. Archimedes’ Lever has roughly seven times that. Oops.
Step 2: More Google-fu. Learn that we can solve the ‘25 posts’ problem by adding ?max-posts=500 to the feed URL. Run the script and see all the pretty files, almost ready to publish on Github Pages.
The operative word there is “almost”. There are two problems immediately apparent:
- The files contain the original HTML markup; no Markdown conversion has been applied. This should have been expected, and really shouldn’t be a problem, but still. More important is
- Those tags-turned-categories are not separated in any way; the original tags “best practices”, “competence”, “development”, “ruby”, “ruby on rails”, “startup”, “teams” get written to the
categoriesfield as “best practices competence development ruby ruby on rails startup teams”. Definitely not what we want.
But we’re closer. Now all I need is a program to either read a list of all tags from a different file, and then search each post’s categories string for each match, longest match first; or read the Blogger/Google export file to build a list of categories for each post (since they’re properly quoted there), figure out which file in _posts matches each post in the export file, and blast in the categories as a proper YAML array.
We’re a lot closer. Stay tuned.
