I’m enjoying writing literate programming-esque posts. I have written two of them in the last months, and I’ll definitely write more in the future.
I’m new to the idea of literate programming, so I hope you’ll excuse my lack of knowledge on the matter.
Those two posts were written by taking the sources of an already written program and splitting them into pieces, adding commentary text here and there.
I liked, but I have a greater plan that I’ll try to follow. This Great Plan requires various steps, many of which I don’t really know how to do, but I did an initial one that I think I’d like to share: tangling literate text/gemini into a source file.
If I’m not mistaking, in the scope of literate programming the term “tangle” means “putting the chunk of codes together into a source file.”
I wrote a quick program to do this, and it’s called tango. (no, it’s not written in go)Tango git repositoryGitHub mirror
It’s a C program that parses text/gemini from standard input (or from given files) and tangle the code blocks into a single stream (standard output or a given file). It’s able to conditionally export block of codes using a given regexp that will match on the alt text of the preformatted block.
Even if I’m posting it after, it predates my AWK script to convert text/gemini to HTML. In retrospect, I could have probably saved some lines of code by writing it in AWK.
At the moment tango is quite simplistic: it outputs the block in the order it finds them. Other, and more sophisticate, literate programming tools allow for more control over this. For instance, in org-mode you can name a block of code and then “include” it into another block using the syntax:
Honestly, I don’t like this. I mean, I would like to have something like this in tango, but I heavily dislike the syntax for inclusion. Ideally, the information of “put that thing here” should be outside the code block. Or maybe it’s a silly idea.
Anyway, this is all I have this time, hope you enjoy the idea.