wasm2ps VS resvg

It's come up in discussions I've been in before, mostly with the PostScript document language, since you can run programs on your printer with PS, albeit incredibly slowly. I've written a roundabout C-to-printer compiler of sorts which compiles WASM code to PS, and it is admittedly even slower because it doesn't do much on the stack. But stack languages and linear logic have a pretty neat correspondence.

I'm not a fan of the usual uses of S-expressions in data formats, but for quite the opposite reasons. People seem to have a thing for making up new syntax, so you can't just throw a normal S-expression parser at the problem, and be done with it. For example, the wasm2ps reader [1] has to know how to read a comment in the WebAssembly text format. I decided to leave the project at handling integer instructions and control flow only, as there are other oddities like using a single atom for alignment and offsets in load and store instructions [2]. Some people I've talked to wished that people would stop inventing ad-hoc syntaxes, and just use S-expressions or something. Using S-expressions but adding more ad-hoc syntax gives you the worst of both options.

While admittedly less of a problem for implementation, it is also annoying to see lists with dangling parens on their own lines, and symbols with underscores or camelCase in the names, once you are used to the normal way of formatting Lisp code.

[1] https://github.com/no-defun-allowed/wasm2ps/blob/master/Code...

I've seen resvg as a potential pick, but it feels huge and seems to be importing skia, which itself is a whole rendering engine. Furthermore, I have no idea if I can pass my own 2D buffer to resvg and let it draw to it.

For example, when converting vector SVG images to raster PNG images with resvg, most of the time is spent compressing the PNG image. This is a lot of wasted work if we just want to read the image instead of transferring it over the network! The fast compression mode eliminates all this wasted work, resulting in huge performance and efficiency gains.

Just as a point of reference, I have a ~75KLOC project (includes dependencies) called resvg which takes just 4s in the debug mode and 8s in the release mode to build on M1 Pro.

resvg is very fast, although the performance depends on the exact SVG you feed it - sometimes faster than librsvg, sometimes slower (although librsvg is also written in Rust now, it does use unsafe while resvg doesn't)

So I am absolutely clueless about Rust and just installed it an hour ago to use this tool called "usvg" https://github.com/RazrFalcon/resvg/tree/master/usvg

So, all you gotta do is rewrite the parsers. Funny you mention librsvg because there is a library called resvg that has a thumbnailer implementation for Windows Explorer. https://github.com/RazrFalcon/resvg