I already have a draft of a post about GESS #stenography system for my website. It’s kinda useless, because it’s an inherently #Russian system, so explaining it in #English is somewhat hard and counterproductive. Still, I’m trying!

And, as a nice bonus, I’m planning to translate it to #Interslavic, because that’s where it’d be most useful, as a system designed for a Slavic language! I haven’t seen any shorthand system for Interslavic, so this might be my contribution to the community!

Playing #visualNovel and #interactiveFiction games makes me want to write. I guess this is the reading-writing brain pathway (metaphorical, unscientific) I’ve been hearing about? Like, reading good stuff makes you make stuff yourself too? Took me quite long to get to that stage. I even had a draft of a post debunking this phenomenon as not working for me. Matter of language proficiency?

Now I want to make a game.

But first finish my Making Sense of Y series, maybe? A good exercise in committing to longer form medium.

And finish that #Interslavic project too.

And write some #Tarot stories too. Would be useful as a fiction practice.

God do I have lots of writing projects all of a sudden.

Turns out, meta-programming is faster.

So I had this #ed(1) #Interslavic dictionary search script. Taking a word from file name, concatenating it to every line in the dictionary, and matching all the lines with 2+ occurrences of the word. Easy, right?

And slow. It took, like, two seconds for the script to search for a word in a 20K word dictionary (Interslavic is quite frugal!) which is beyond acceptable. I tried to filter out the non-matching lines before concatenation (the most resource-intensive and allocation-heavy part) to no avail.

And then it dawned at me: what if all this runtime pattern search was compiled into a simple g-lobal search and ran through another instance of ed(1)? So I did:

!# Meta trick to read current file name (word to search) into the buffer
0r !echo %
!# Turn into a meta script to feed into ed: enable help, search for word, output, quit
s;\(.*\);H\
g/^[^ ]* \\([^ ]*\\) .* \\([^ ]*\\<\1\\>[^ ]*\\) .* \\([^ ]*\\) \\([^ ]*\\) [^ ]*$/s//\\1: \\2 (\\4)/p\
Q;
-2,.w ! ed -s ~/Documents/isv.tsv

Looks scary, but only because I have to manage tabs in a TSV file. But, otherwise, it's a nice solution that lowered the lookup time to cognitively imperceptible numbers. I'm happy with it!

I kinda finished the #Interslavic dictionary lookup tool implemented in #ed(1) regex. It's still slow, but it's bearable, especially for a tool with that ugly regexen. Find the script (and running instructions) at:

https://codeberg.org/aartaka/aed/src/branch/main/isv.ed

Yes, it's part of aed(1), my wrapper for ed(1), because I want aed(1) to be the most comfortable text editor with a built-in Interslavic dictionary.

Another editing session in #ed(1) on a bunch of files (#Lisp, #HTML, ed scripts, including a serious #Interslavic dictionary processing one.) Not much to report, but I seem to be getting into the flow.

(Interslavic dictionary lookup done, I just need to optimize it a bit…)

The most noticeable hack is finding some anchor sequence of chars near the necessary editing point / “cursor” and s-ubstituting this anchor with the new content. And, well, copying things from terminal with a mouse and pasting in in insert mode 🙈

Ололо сайт меджуслава с российских ip не открывается
«Славянское братство блаблабла»
https://interslavic.fun/ru/

#interslavic #междуславянский #межславянский

Кто там за #interslavic / #меджуслав топил? Давайте переведём
https://translate.codeberg.org/projects/readeck/-/isv/

#междуславянский #межславянский