I've no experience with code, but do have a bit of experience with trying to export/convert files from one format to another. You often end up with so much work that you wonder if it would have been better to recreate the file from scratch.
Surely with code, you will need to be able to code anyway, in order to correct all the minor things that dont convert well (?)
Yeah, I'm with Tomos - I'm no coder, but I did used to do file/data conversion as a key part of my old job. One program wants to add extra header lines to stuff, so then if you dump it your line count is hopelessly ruined. Another program wants to fiddle with pagination so then Adobe's Acrobat X comparison tool bombs with false positives because the entire document is off by a paragraph, creating False Positives.
Or, to be funny, look at the "advertising" graphic on that site:https://varycode.com...version_conveyor.png
The first two rows of a Rubik's Cube are a snap. "Anyone with twelve brain cells" (shout out to Ren!) can do the first two rows. It's the last row that's a killer. I'm no SpeedSolver, so the method I use is sloppy and slow, but it all boils down to two cubes at the very end. Wait for it ...
It takes a 36 move sequence to do it!
That's like (File) conversion: It looks good, yay this is a snap, ... until that last 20% chews up 80% of your time!
Code, to do actual stuff, has to be worse. Data is just data. Code has to actually Do Stuff. I will absolutely say I don't trust any program created solely by a "converter". I'm fine if a programmer wants to use tools, but code is for ... uh ...