wydm "non three link version of the link" and "try could be calling"

The old version that only used two links, and*they... Stupid auto-correct

no that doesn't break anything, it just allows you to golf your code more, no?

Refer to this: chat.stackexchange.com/transcript/message/43202144#43202144

Huh. (dyad)(nilad)(monad)$ still works (ignore the first line)

NVM then

My phone doesn't want to lead TIO

@cairdcoinheringaahing What do you mean? Did you expect that to break?

Oh wait, so did $ change?

@DJMcMayhem no

well I mean the source code changed

from >1 to >=2

so in other words, functionality did not change

it's just that there are more golfy ways of writing code in some cases where $ used to be used

AFAIK Dennis is pretty good with making Jelly backwards-compatible so old answers never break when new additions are made

So why rip dyad nilad monad?

@DJMcMayhem I've already made a suggestion in the Jelly chat room above your message

@DJMcMayhem nvm

So no breaks

?

No break (AFAWCT)

Long live Jelly backwards compatibility

CMC: Given an array of integers and an integer n, return whether all elements in the array aren't divisible by n. Example: [1, 2, 4, 5], 3 => 1 and [1, 2, 3, 4], 3 => 0

Add++, 5 bytes: L,ª%

@cairdcoinheringaahing Jelly 2 bytes

@Mr.Xcoder Divides, then product? Just a guess, as my phone doesn't want to load TIO

@cairdcoinheringaahing Canvas, 6 bytes

@Zacharý Modulo, then all

@cairdcoinheringaahing so I was somewhat close?

Bah I am sick of literature.

Yeah, kinda

@Mr.Xcoder Pourquoi

I have an exam simulation tomorrow and had to learn a few types of compositions :|

(Sorry for the french, I found out that my phone can predict text for languages other than English. Including Espéranto, Interlingua, and Latin)

Seriously? Espéranto?

Jes.

But it lacks a polytonic Greek keyboard

@cairdcoinheringaahing autocorrect

@Mr.Xcoder Alternatively, %P

Ninja'd sorta

@DJMcMayhem you have been ninjad by me, I think

except that he hasn't

his version is modulo and product, not divides and product

the latter is invalid

I'm on a phone without access to the wiki... Just imagine a @ and move on

not just the @, but also a Ṇ

Why would I need that?

Q: If a major change to a language is made, and it's current version is 3.3, should the new version become 4.0 or 3.4?

Who gave Dennis the idea to use the hooked letters as things other than nilads?

@cairdcoinheringaahing how major

This major

@Zacharý I'm guessing he thought of it himself

4.0

@Zacharý because the point is to use single-byte tokens, and we are running out of those

at least another reason to do that

@EriktheOutgolfer ... I have to ask why <= and >= were never one-byters...

I think I wanna post this challenge, any feedback or objections?

I found my problem with my modgrammar stuff earlier: Left Recursion doesn't seem to exist there.

Anyone here know how to use modgrammar?

@Zacharý I mostly know how to use it. Send code?

I don't know how to do left-recursion either but I think it can be done using a better method anyway. Depends on what you're doing though.

Two different problems, I figured out how to "do" left-recursion (by not doing it).

from modgrammar import *
from functools import *
from math import *

#grammar_whitespace_mode = 'optional'

NAMESPACE = {}

class NonZeroDigit(Grammar):
	grammar = OR("1","2","3","4","5","6","7","8","9")

class Digit(Grammar):
	grammar = NonZeroDigit | "0"

class Integer(Grammar):
	grammar = (NonZeroDigit, ZERO_OR_MORE(Digit))
	def grammar_elem_init(self, sd):
		self.value = int(self.string)

class FloatingPoint(Grammar):
	grammar = OR((Integer, '.', ZERO_OR_MORE(Digit)) , ('.', ONE_OR_MORE(Digit)))

I'm guessing the commented line is the one where you needed to figure out how to avoid recursion errors?

You could just do grammar = LIST_OF(Value, sep = "+")

How do you dereference a REF (for lack of a better term)

I'm not getting recursion errors

I've overcome THAT problem

o

@Zacharý When you're done with all of your statements you can do Expression.grammar_resolve_refs()

Now I need to know what I do to get user-defined properties of REF'd Grammars.

@mudkip201 how would you check if two arrays are exactly equivalent in pyt?

@HyperNeutrino That doesn't fix anything, I think

@Zacharý what exactly is the problem

AttributeError: '<GRAMMAR>' object has no attribute 'value'

dereferencing a REF doesn't fix that either

the object returned by the parser just doesn't have a "value"

if you want to see the text returned by the parser (which is just the input text), just print it

value is my user-defined property

oh

uh

Expression.parser.parse_string(...) won't return an Expression object...

Well, I've tried it on other objects with a value property ... it worked...

example?

It only borks when I add parens

> if self[0] != None: self.value = self[0].value

self[0] in this case is the (...

I don't think you quite understand how OR constructs work

I'm a noob and just hacking my way along...

it doesn't return a list of all of its matches, it just returns the first one that matches

Ah ........... that might fix everything.

class Value(Grammar):
	grammar = Integer | FloatingPoint | Var | (L('('), REF('Expression'), ')')
	def grammar_elem_init(self, sd):
		if self[0].grammar_name in ["Integer", "FloatingPoint", "Var"]: self.value = self[0].value
		else: self.value = self[0][1].value

replace your Value class with that. works

Merci!

It's one less character than english...

or you could just say "ty"

@tfbninja I probably should make a built-in for that, but as of right now, the best I can do is 6 characters: =ĐŁ⇹Ʃ=

Or you could use regex :P

@cairdcoinheringaahing Yeah, but for what I'm doing ... that'd involve PHP's named regexes ... BLECH, php

Wait, you're using PHP? Why?

that'd, as in that WOULD. I WOULD have to use PHP if I used regexes

wat

how does that make sense

If you're using Python at the moment, why can't you use Python regex?

because Python regex isn't good enough (I think)

Better than PHP's

I'm just testing the waters for modgrammar.

The regex module supports a lot of stuff that re doesn't, like recursive references, I think

modgrammar is way better than regex for something like this

As in once I get the hang of it, I might program a language that ISN'T a hack

You should declare a separate class for addition like Add which is LIST_OF(Value, sep = "+") because that way the number of chained +s doesn't increase recursion depth

@HyperNeutrino Best would be a custom grammar... but I have no clue how to do this, and honestly don't want to learn right yet.

I always prefer using regex for parsing

Expression should be the third-top-most thing

@Zacharý I mean, modgrammar is pretty good and building your own parser using it isn't bad

@cairdcoinheringaahing I use it for tokenizing all the time, but with modgrammar I don't even need to tokenize it first because modgrammar parses text directly. However, I use regex for all golfing languages (because tokenization and minimal parsing)

Screw style conventions.

¯\_(ツ)_/¯ I'm still trying to create a (non-brainfuck) language in Jelly/Add++ :P

eval

and transpile some stuff.

done

No, that's not as fun :P

AppDelegate.swift:11:1: error: 'UIApplicationMain' attribute cannot be used in a module that contains top-level code
@UIApplicationMain
^
AppDelegate.swift:1:1: note: top-level code defined in this source file
//
^

yes swift, that's some complex top-level code called "comments" -_-

@cairdcoinheringaahing that sounds like a fun idea I should do that

but I kind of also want to make a completely functional language (which will probably just be a variant of Proton except a lot easier to implement and also probably a lot better)

basically just Proton but now functional as in both the programming paradigm and as in it will actually work (:P)

Anyone know a good way to slurp a file into a char * in C? I've spent over half an hour googling and I've only found C++ solutions.

I have three language designs I eventually want to implement: made in Jelly, made in Add++ and declarative

I should design an anti-functional language (as in everything must modify the global state to do anything) and call it a dysfunctional language

nahhhh... seems too easy to break

@HyperNeutrino So, Assembly?

o lol

@Pavel to be fair assembly has stack

You have to implement the stack yourself, and it's still global.

also OS manages context so not everything is shared will all other things

at least in terms of what you can do without doing a segfault

ok maybe these messages should be moved to a new room @Mego

also why does f# in cyrillic characters, probably russian, always show up as a possible duplicate?

thx

@HyperNeutrino 2d language which edits the grid?

so basically 2d smbf?

not exactly

what if changing the stuff is the only way arbitrary control flow can be achieved

ooh interesting

i think that there was a functional language where the only way you could achieve turing completeness was using the grid as storage

i believe it was called flobnar

though control flow still worked without it i think

@Adám, is there an EBNF-like document for APL?

@Zacharý I think the ISO standard has one.

NOT behind a paywall...

I found one barred by horrible mojibake: pmichaud.com/2006/pres/yapc-apl/slide13.html

@Zacharý Too bad. Looks like an interesting presentation.

@Zacharý Turns out it doesn't. It uses railroad diagrams.

@Adám Ah. Well, dissecting ngn's code is a possible option, but the trains wouldn't work...

@Zacharý If you want to know how (Dyalog) APL "works", this may help (or confuse) you.

Thanks.

I found out a useful advantage of RTL over LTR: cumulative reduction over subtraction can be used for alternating series.

Ah

@Zacharý fold(r/l), flip?

2.6 second now that I tried again. SymPy got a lot faster.

@Οurous As in: 1-2-3-4 resulting in 1-(2-(3-4))... I DID find a hacky (?) way to do it in MY's source code, IIRC