A small project/application/library intended for generating conlang (artificial language) resources, such as basic phonologies and word lists, however one may wish to use it.
This started primarily as a Python study project, but gradually extended to include studies in other topics and languages. I then badly borked some work I had done both in Python and Rust on February 2022, because of not adequately commiting stuff and then profoundly messing up git commands (at least, I learned a bit).
As a result, I restarted from scratch in Rust and vanilla JS, and later a very glued together frontend in Svelte. This has been an enormously important project for me to learn dev practices, Rust, JS, Python, git and overall building slightly more complex applications. This is still a super ambitious project, and I don't know when it's going to be usable, but it remains as it began - a study project.
There's no predefined set of features and capabilities I want to add to this project, just some overall ideas such as generating (broadly) phoneme inventories, phonologies and pseudotext. Some stuff I might work on at some point are listed below.
- IPA symbol picker, for easier input
- Parse X-SAMPA to IPA and vice-versa
- Analyze word patterns for setup extraction
- Generate all anagrams for a given word
- Generate all possible words for a given setup
- Generate filtered anagrams for a given word
- Generate random words with weighted/filtered probabilities
- Generate words with static affixes (e.g. baCV)
- Generate text with weighted word probabilities
- Generate pseudotext (text that is not just a bunch of fully random words without punctuation, but could be mistaken for actual text)
- Support syllable and word-level rules
- Support sandhi rules
Later on, if things go right, I might try adding real language concepts to the generator, supporting phological, morphosyntactical and other kinds of representation layers. Words would be generated based on real-world phonetic and phonemic rules per setup, and we could support variations such as noun and verb declension for each generated word.