Script or utility to export from ScrewTurn Wiki into Confluence?

Question

Has anyone tried migrating from ScrewTurn to Confluence? I'm hoping that I can export the ScrewTurn database to xml and then use a utility to put it in a format that Confluence can understand - perhaps the format used by the Universal Wiki Converter.

Has anyone used such a utility? Is there anything I should know before I try to write one myself?

Update: I've installed the Confluence trial and I took a look at the backup. There's an entities.xml file that looks like the data store. The root tag is <hibernate-generic>, which leads me to believe that they use Hibernate. I'm not familiar with Hibernate, but this should be useful.

Phloem

Answer 1

1. Implement a Wiki Converter for ScrewTurn. I used MediaWiki's Syntax Converter as a base since the basic wiki syntax is very similar. I also implemented a few Converter classes, UserDateConverter (requires the Confluence UDMF plugin), PagenameConverter, AttachmentsConverter, MetaDataCleaner (to remove the first three lines in ScrewTurn page files that include page name, date, and ##PAGE##).
2. In ScrewTurn, change the page storage provider to Local Pages Provider (if its using a different provider such as SQL).
3. Run the customized UWC implemented in step 1 and convert one namespace at a time.

You can download the source code from GitHub.

Phloem

I've used your implementation with success. Note to others, at this time the Confluence UWC (Universal Wiki Converter) only supports Confluence 3.5 and lower so you can't first install Confluence 4.0 and import your screwturn data. Also note that I had to use a ver 6 JDK instead of the ver 7 that I had installed to compile UWC with Ant. Thanks John. – Photoactinic