Vyxal

A username

12:40 AM

Yes

see the ascii-art thing reference implementation

lyxal

3:41 AM

Me when I'm going to majorily change the parser specs later

hyper-neutrino

3:52 AM

thinkies

lyxal

Switch from dictionaries to lists

hyper-neutrino

i'll let you handle the parser if that's fine with you cuz i think we had quite different ideas for the lexer-parser setup (besides like reviewing your code or modifying small things or smth)

lyxal

4:21 AM

Cool cool

That might make it easier for me to actually write parser specs

Because I can get a MVP working with little documentation and then formalise it after that

hyper-neutrino

4:48 AM

good point

getting a start on things with just one person working on it is not a bad idea

then once things get sort of going people can add contributions

lyxal

5:03 AM

@hyper-neutrino I'm not quite sure the lexer needs to actually produce name tokens

Except for variables and function references, having name tokens doesn't add much

Because once you pop the name token, there might still be stuff needed to be grouped together

hyper-neutrino

1. what about variable names ending with k 2. what do you mean there still might be stuff needing grouping

lyxal

(abc:def|...)

And @name:params|...

@hyper-neutrino obviously it's useful for variables, just not so much for for loops and functions

I just don't know if it's actually useful or if it's something like Jira

I'll design the first version of the new parser around name tokens and see how it turns out

hyper-neutrino

oh i see

either way i still feel like we're going to run into issues with variable names ending with k but

¯\_(ツ)_/¯

lyxal

2 mins ago, by lyxal

@hyper-neutrino obviously it's useful for variables, just not so much for for loops and functions

The solution here would be to have a token just for variables

Ez

hyper-neutrino

true

which are trivial anyway, fortunately

lyxal

5:10 AM

Correct

Also, I'll be making a structure class

And it'll be like tokens for structures

hyper-neutrino

good idea

lyxal

Here's the parser so far:

My idea is to also have different functions for each structure type that process branches if need be

So for example, there's be a function that turns the name branch of a function into its name and parameters

hyper-neutrino

i feel like a structure object can just have an instance method that assembles its meaning from its branches and compiles it

so an if statement (at its base level) might take statements a and b and wrap in in

if pop():
    {a.compile()}
else:
    {b.compile()}

(not exactly like that, but just for demo)

and that would be .compile() on an if statement

lyxal

5:26 AM

You mean like this?

hyper-neutrino

something like that is what i was thinking, yeah

it should be fine for you to just focus on the parser for now and as we get the structure classes i can see if i can get this idea working

lyxal

Sure

lyxal

7:13 AM

\          /   \    /   \    /        /\         |
 \        /     \  /     \  /        /  \        |
  \      /       \/       \/        /    \       |
   \    /        /        /\       /      \      |
    \  /        /        /  \     /        \     |
     \/        /        /    \   /          \    | ______

made using the reference implementation from the @Ausername's ascii art challenge

Input: '\ / \/ \/ /\ |\n \/ / /\ / \ |__', 3

I love that moment when black doesn't actually stick to the 80 char limit

Vyxal Bot

Lyxal pushed a commit to Vyxal/fresh-beginnings: Update README.md

hyper-neutrino

i don't think it can break apart long strings

lyxal

it isn't a string

        tokens.append(Token(TokenType.VARIABLE_SET, contextual_token_value))

it can easily be split over two lines

        tokens.append(Token(TokenType.VARIABLE_SET,
                            contextual_token_value))

but black keeps changing it back to the first option

hyper-neutrino

7:28 AM

wat

sounds like a configuration issue? ¯\_(ツ)_/¯

lyxal

nope

using the black playground doesn't work either

> You probably noticed the peculiar default line length. Black defaults to 88 characters per line, which happens to be 10% over 80.

from the black docs

aha

I see now

welp, time to go update the code style guide

Vyxal Bot

Lyxal pushed a commit to Vyxal/fresh-beginnings: Update code guide.md

lyxal

dang, forgot to change the revision number

hyper-neutrino

i just set it to 80 in my atom package config lul

Vyxal Bot

Lyxal pushed a commit to Vyxal/fresh-beginnings: Update code guide.md

lyxal

7:38 AM

I like how the style guide has gotten 13 revisions

anyhow

back to the parser

Vyxal Bot

Lyxal pushed a commit to Vyxal/fb-parser: Create the parser

Lyxal created branch Vyxal/fb-parser

lyxal

there we go

AviFS

Check TNB @lyxal

I think you'll be happy ; )

Congrats on starting on the fresh start after a fresh while of freshly not starting

lyxal

7:56 AM

Thank you

We'll get there eventually

lyxal

8:34 AM

@AviFS but we are making it a priority to take things a bit slower

Even if that means periods of stagnation

Besides, it hasn't even been 2 weeks since I actually created the rewrite branch

hyper-neutrino

the thing is vyxal right now is (AFAIK) in a pretty stable state and is fine to just stay as-is - but developing more features and bug-fixing got really difficult

the point is to do things right this time, so since the current copy on the interpreter is fine, we can just leave it as that and take our time

and then once our rewrite is done properly, future development will be much easier, faster, more reliable, and more maintainable

lyxal

That's precisely it

We're not aiming for quantity, we're aiming for quality

And for 13 days of beginning fresh, I think our pace is good

hyper-neutrino

especially since all of CGCC has felt rather quiet lately, lyxal's uni classes just started up again, and i (and a few others i believe, at least aaron) have a job so

lyxal

Heck, I'm marginally concerned it might be a tad bit too fast lol

@hyper-neutrino correct

A username

Me when writing code on paper

lyxal

8:41 AM

> hopefully sometime in the future we'll find some way to gracefully handle multi-threading :-)

~ pythonanywhere

They asked for feedback, so I gave them feedback

And that's what they said back

A username

When you log on and see this

It's beautiful

And I'm glad I didn't offer a 500-point bounty

lyxal

9:00 AM

!!/flowey quote

Vyxal Bot

@lyxal Down here, LOVE is shared through little white... 'friendliness pellets'

lyxal

10:56 AM

Can I just say that the pattern matching syntax of python 3.10 would make type overloading way easier for vyxal

Like you wouldn't have to wrap everything in a dictionary with lambdas

It'd be really cool

A username

11:07 AM

Agreed

If you wait two months...

PyGamer0

11:25 AM

@lyxal same for 51AC8

cause i will have a default

A username

12:01 PM

!!/flowey quote

Vyxal Bot

@Ausername So you were able to play by your own rules. You spared the life of a single person. Hee hee hee... I bet you feel really great.

A username

!!/'night

Vyxal Bot

@Ausername o/

lyxal

1:09 PM

So are we alphabetically sorting functions or no?

PyGamer0

@lyxal for what

lyxal

when writing cod

e

do the functions go in alphabetica

l order or

do they go in any order

?

(half of that was intentional)

PyGamer0

idea: they should go in the order of the code page

lyxal

this is also non-element functions

like say in the parser

PyGamer0

hmmm random order?

Vyxal Bot

1:17 PM

Lyxal pushed a commit to Vyxal/fb-parser: add stub code

PyGamer0

@lyxal ok αβical order

user

@lyxal Methinks we ahould group them by what they do

Vyxal Bot

Lyxal pushed a commit to Vyxal/fb-parser: Add more stub

user

So a category for mathy functiona, one for ranges, one for strings, one for stack ops, etc.

PyGamer0

@user yes this is a good idea

user

1:24 PM

@lyxal yes

When will puthonanywhere get 3.10 btw?

PyGamer0

each sub category functions will be sorted in αβical order

@user when python 3.10 is released officially

user

Oh

Vyxal Bot

Lyxal pushed a commit to Vyxal/fb-parser: Add even more stub

lyxal

very cool

I think I've done enough parsing for one night

sleep is my next file to work on

sleep.irl

@VyxalBot goodnight

Vyxal Bot

@lyxal o/

PyGamer0

1:39 PM

@lyxal idea: maybe make a separate repo for a custom vyxal keyboard layout

like APL

user

2:11 PM

Or we could ask Adam to add Vyxal to his language bar thingy

PyGamer0

3:41 PM

@user vyxal already has a keyboard in the tìo

hyper-neutrino

having a layout or language bar lets you type in chat and other places too

though the usefulness of that is debatable

user

9:13 PM

It's useful to have that in this room, at least

hyper-neutrino

that's true

user

That was a smart point

Vyxal Bot

@user I second that

A username

That was a smart point

?

Oh I see

!!/run `Nutritious!

Vyxal Bot

9:29 PM

@Ausername Nutritious!

hyper-neutrino

@VyxalBot What's 1+1?

Vyxal Bot

@hyper-neutrino 3

user

@hyper-neutrino lol

A username

@VyxalBot Who is Lyxal?

Vyxal Bot

@Ausername Joe mama

Vyxal Bot

10:19 PM

hyper-neutrino pushed a commit to VyxalBotSE/master: update chatbot.py

hyper-neutrino pushed a commit to VyxalBotSE/master: frontend send message endpoint returns message ID

A username

@VyxalBot Why is hyper so overprotective of you?

Vyxal Bot

@Ausername Because I'm special

user

10:35 PM

You're not special

Vyxal Bot

@user ಠ_ಠ

A username

10:46 PM

@VyxalBot Do you think I should be an admin?

Vyxal Bot

@Ausername @hyper-neutrino I think that A username should be an admin

lyxal

11:40 PM

@VyxalBot are you a sussy baka?

Transcript for