Login
You're viewing the functional.cafe public feed.
  • Apr 26, 2026, 11:45 PM

    Yesterday, I decided to take a crack at recovering someone's story from the format used by a popular, if old, macOS writing tool (writing.stackexchange.com/a/72). I've concluded that it's possible, albeit fiddly: the file format isn't great, but it's shelf-stable enough for archival (due to appropriate choice of self-documenting, generic container formats). This work's a bit fiddly to do on the command-line. If only I had a tool specifically designed to do this task!

    Oh, wait! I do!

    💬 1🔄 0⭐ 0

Replies

  • Apr 26, 2026, 11:50 PM

    Of course, 3½ years on, I've decided that everything about the program is bad and needs to be replaced. The first version was too fixated on a single-pass source-target transformation model, when really it needs to have a multi-stage transformation model. (I identified this problem at the time, and not knowing how to fix it is a major reason I lost interest in the project.)

    I didn't realise that I'd got it to a working (albeit hacky) prototype, though! I'm really pleased about that achievement.

    💬 0🔄 0⭐ 0
  • Apr 26, 2026, 11:59 PM

    One of the key components of this generic file transformer (named GFTrans) is being able to create schema description languages, which can be built from one file, and used to extract data from other files.

    I haven't committed to finishing it, but hopefully I'll get something good enough to export stories from this file format. (What I've done so far is very nearly enough to do a manual copy-paste of the main story text, but I want something with a shiny user interface and batch processing.)

    💬 0🔄 1⭐ 0