There is no good reason to compile to WebAssembly and run under Node.js aside from testing, and doing so makes everything more complicated and prevents us from easily packaging SoundSwallower as an ES6 module (though maybe that bug is fixed: https://github.com/emscripten-core/emscripten/pull/17915)
Still, there isn't any good reason to do it. SoundSwallower is always used via Python for offline, command-line, etc, and this makes sense since all the code stuff for manipulating speech data and models is there anyway.
The problem is that we need to test the JavaScript code, and Node.js is pretty useful for that. There are I think like 15 different solutions for providing an automated test environment that looks like the browser, all of them have cutsey names and probably icons too, and I have no clue.
There is no good reason to compile to WebAssembly and run under Node.js aside from testing, and doing so makes everything more complicated and prevents us from easily packaging SoundSwallower as an ES6 module (though maybe that bug is fixed: https://github.com/emscripten-core/emscripten/pull/17915)
Still, there isn't any good reason to do it. SoundSwallower is always used via Python for offline, command-line, etc, and this makes sense since all the code stuff for manipulating speech data and models is there anyway.
The problem is that we need to test the JavaScript code, and Node.js is pretty useful for that. There are I think like 15 different solutions for providing an automated test environment that looks like the browser, all of them have cutsey names and probably icons too, and I have no clue.
So probably I should figure this out.