Friday, February 8, 2008, 2:34pm
I had to do something similar recently - Winderz to LAMP Drupal. Here's what I did...
1. I pulled the entire site using wget --mirror
2. I ran each page through HTML tidy... uh... find . | grep html | xargs -n1 tidy -ibmq (I think that was it.)
3. I examined a few pages and found that all content was stored insides of , so I sed'ed out all everything inside of there.
4. Cut and paste into new pages, doing my best to preserve paths.
5. Everything that was not HTML was uploaded to new site in old path location.
6. Thundercats were go.
Neat how some commandline magic just cuts to the chase and cuts the task down to size. Sure, a database migration script would be even more elegant, but this one is hacked together nicely.