The lexicon app

I remember when, two years ago, Philip and I attended our “Advanced language workshop” where we started analysing Kovol texts. We had each transcribed 2-minute Kovol stories, and during the workshop, both my texts and Philip’s were up on the projector for discussion.

A hurdle we encountered was that both of us had difficulty reading each other’s texts because the spelling was so different. There were spoken abbreviations we had chosen to represent differently, different vowels in words, and it was enough to break the visual recognition of a word. We had work to do to “standardise” within our team on how we were writing.

What we needed was a spellcheck we could both use to help us move towards each other. The problem was, of course, that there was no such thing; there was no Kovol language spellcheck. I had already written a lexicon app that our team were using to work on our lexicon; now it just needed spellcheck and integration into other tools. I whipped something up, and we’ve been enjoying the benefits. Our team’s spelling is now (on the computer at least!) consistent. I don’t think our team’s spelling is always 100% accurate to what the Kovol people may choose once they start their literacy journey, but it is consistent.

As we’ve been using it and others have seen it, I’ve had requests to make it available for other languages. The problem was that it was a cobbled-together bunch of features written as we needed them, specifically for the Kovol language. There was a bit of work needed to generalise the app and make it suitable for other languages, and a level of polish needed that’s required when you make something that’s intended for others, and not just for yourself to use.

Over the last 2 months, I’ve been able to spend the mornings “on the code face” chipping away at the problem and I’ve got to the point where I can say “it’s ready”. Ready doesn’t mean of course, bug-free, with all the bells and whistles but it’s ready for others to test.

https://lexicon.reachkovol.com

The app lets you:

• Manage lexicon entries in a centralised place as a team
• Create paradigm grids to display verb conjugations
• Write rules to automatically conjugate word + affix combinations (can be hundreds per word!)
• Export a LibreOffice extension that adds the language and spellcheck
• Update spell check through LibreOffice itself
• Export spellcheck data to use in Paratext Bible translation software
• Export a .dic file to use for spellcheck in Word

As version 2 of the system our team has already been using to manage our spellcheck, we’ll be moving over to it. The new code is actually unit tested, which makes me feel much happier about its reliability.

As we don’t want the whole internet to have permission to mess with our language data, there is a sign-up and permissions system, but it’s free to sign up, and I’m willing to help interested people get started.

I really, really enjoy programming. It feels like a guilty pleasure – it’s far too easy to spend time on it when I should be spending time on the Kovol literacy program itself. It’s been a balance between how much time to put into refining the tool I’m using and how much time I should be using the tool to do work. It’s been really nice to have some time away from Kovol to push and get the app finished and ready for others. Will anyone else use it? We’ll have to see 🙂