You won't find much information, especially not if you specifically include a console as your destination device -- generally anybody that knows how to do it will know how to do it and how much effort it will be, then likely not bother. Some do anyway (see several entries on
https://osgameclones.com/ ) but it usually takes some determined coders a number of years to pull off, and more for extremely popular (or popular among high end coders -- heroes of might and magic and transport tycoon are there for a reason).
Is there not an android port these days? Or possibly versions on systems which are emulated?
Anyway three main approaches
1) You have the source code because it was open source, because the devs released it or it leaked in some form. This is where 99% of homebrew ports of games start, so much so that most just assume if there is no source code then there is no point in trying.
2) You don't have the source code and have to generate some, either by recreating the effects yourself or actual reverse engineering.
3) You use the assets and just port the engine around, this is even easier if there is an engine common to a lot of games (think what happens when people port doom, quake and all that to various consoles and get all the games based on them almost for free, if not actually for free as the user made tweaks to the engines have adaptations to encounter whatever the other game devs did over the baseline). SCUMMVM is probably the most famous of these and most of the Japanese based stuff I have seen is for the smaller indy stuff.
Visual novels run the gamut from essentially a slideshow of BMP images (stuck around for quite a while in the more indy scene in Japan but has largely died off) done as a choose your own adventure right through to full bore game like any other if minigames and whatnot get involved (or just some of the big boy animations and such). If it is the slideshow of BMP then that is one of the easier things you might do both in general coding options and if you decide to view it and recreate it -- you notice it is a simple fade to black between scenes, easy to make that yourself rather than looking at existing code and figuring out what it is/does.
There are hybrids of all of those approaches/methods. Reverse engineering code fully for a game is among the harder things to do in computing, though something you can learn happily enough. If you are going to go for PC efforts it also changes things a bit more as things can be a lot harder (anti piracy, hacker protection methods and the like), or at least a lot more variable than what we see on consoles (knowledge of SQL has likely never helped a console hacker do a ROM hack, those making World of Warcraft custom servers on the other hand benefit extensively from it). There are some nice tools that were not available in years prior (decompilation of C is a real thing and while C++ is unlikely to fall in the same way any time soon it does not have nothing, with those you have a lot of games fall before you)
You also then get to write an engine for the Switch afterwards which is not the easiest/nicest compared to writing it for the PC.
So things to pick up
PC coding, probably something in the C and C++ region as well as some others. Spend extra time learning about file formats in general.
Game coding, especially game coding for the region and time period you are in. Not as necessary but makes things 10 times easier if you know what was used to make it and can predict what was done very having to go from step 1 every time.
Encryption and anti reverse engineering methods, including how to defeat them.
Coding for the Switch. Being visual novels they are often quite simple, and the Switch allows some reasonably high level languages so you might not have to learn the hardcore low level programming for the Switch.
All of those are pretty intense areas of computing that people spend decades playing in before they might consider themselves good. If you are not so interested in being tip top in today's games and instead are content to play with older you can get up to speed in a reasonable amount of time (the Switch might even still be a current console -- we are 3 years in at this point after all and they tend to only last about 5 or 6) as a lot of things have already been done/happened and documented. As such most people that ask a question like this fail to grasp the enormity of the task and then run into the wall of Turing's halting problem at almost light speed and that seldom ends pretty.
