mastodon.online is one of the many independent Mastodon servers you can use to participate in the fediverse.
A newer server operated by the Mastodon gGmbH non-profit

Server stats:

11K
active users

#parser

0 posts0 participants0 posts today

Парсинг с помощью LLM: зачем, как и сколько стоит?

Во всю идет 2025 год, и нейросети перестают быть чем‑то фантастическим. Они уже повсюду в нашей жизни: от умных колонок в квартирах до сложнейших систем, управляющих логистикой и финансами. Вместе с ними стремительно меняется подход к работе с данными. В этой статье мы поговорим о том, как современные LLM помогают автоматизировать сбор данных с веб‑сайтов и сводят к минимуму рутинную настройку и «подкручивание» парсеров. Что еще вы найдете в этой статье?

habr.com/ru/articles/892954/

ХабрПарсинг с помощью LLM: зачем, как и сколько стоит?Во всю идет 2025 год, и нейросети перестают быть чем-то фантастическим. Они уже повсюду в нашей жизни: от умных колонок в квартирах до сложнейших систем, управляющих логистикой и финансами. Вместе с...

To the #Rust #rustlang community: I once started to write a #compiler / #parser with #nom #crate. However I had some struggles with it of how to provide **multiple** errors with line+column indicators for a parsed context.
Now I read a tiny bit about #syn and #chumsky 👀
Are they the right crates for me? Are there others?
I do not parse rust-code, but a completely custom language (similar to pugjs)

(Boost for reach 🙏)

Saviez-vous qu'en JavaScript `parseInt('2️⃣`)` retourne 2 ?

Il y a une raison logique à cela.

C'est une combinaison de la composition de cet emoji et du fonctionnement de `parseInt` et cet article l'explique très bien.

🔗 aleksandrhovhannisyan.com/blog

www.aleksandrhovhannisyan.comTo Parse an Int | Aleksandr HovhannisyanWhy does JavaScript's parseInt treat keycap emoji as integers? Is it a bug, a feature, or perhaps both?

Trying to improve my Tree-sitter #parser for my new #esolang Bespoke, so that it recognizes things like commands and comments.

Trouble is, I can't figure out how best to handle block comments. And when I tried a workaround...it crashed my computer.

This is gonna be tough.

In Bespoke, words are first turned into digits, then the digits are turned into commands.

Block comments start and end with 0...0, where ... is an arbitrary digit sequence.

I need the digits before I can detect the commands, so I'm unsure how to handle this with Tree-sitter.

輸入:
@foo
{@foo|@bar|12\|}

輸出:
Any[Main.uahgi.Parsing.ID("foo"), "\n", 
Main.uahgi.Parsing.SEQ(Any[Main.uahgi.Parsing.ELE(Any[Main.uahgi.Parsing.ID("foo")]), 
Main.uahgi.Parsing.ELE(Any[Main.uahgi.Parsing.ID("bar")]), Main.uahgi.Parsing.ELE(Any["1", "2", 
Main.uahgi.Parsing.ESC_CHAR("|")])])]```

Parsec這樣還是很不錯的。(參考:
https://repo.kianting.info/yoxem/uahgi/src/branch/main/src/parsing.jl#Julia #Parser #ParserCombinator
Tan Kian-ting's Repos: 個人的程式碼倉庫uahgi/src/parsing.jl at mainuahgi - another experimential typesetting tool

(Towards):
A #sidescrolling #textAdventure #game as the host of a Pratt Grammar #Parser:
lispy-gopher-show.itch.io/lisp
#itch #itch_io #gameDev #lisp #commonLisp #repl

Sharpsign lengthy and wandering. Sharpsign happy 2025

Basically, I was having trouble imagining parsing a grammar at all. So I imagined input tokens as a sidescroller level I could cdr-code speedrun passes of. And you know, when lisp people imagine things it happens in real life too.

I just released pyparsing 3.2.1 with a very cool addition to the railroad diagrams that pyparsing generates for its parsers - non-terminal diagram elements now link to their definitions elsewhere in the diagram! This is pretty handy when the diagram is for a complex parser, as in this demo of a parser/evaluator for time references like "3 weeks from today" or "0800 two days from now" or "in exactly twenty-four hours". #pyparsing #python #parser #railroaddiagram

Currently dallying with a simple embedded, statically typed scripting language. The magic is an extensible lexer and parser, with the hosting application able to do something like

VM_addCommand(vm, KEYWORD(“open”), KEYWORD(“file”), TYPE(STRING), KEYWORD(“as”), REFERENCE, END);

to allow something like

open file “foo.txt” as myFile

It’s obviously been done before, but it’s fun to play with. I haven’t done enough #compiler or #parser work lately.