Why aren't you writing it in Haskell, Jow Forums?

Regex is inherently "impure", as Husklelltards say. So no.

Attached: advanced_language.jpg (889x218, 55K)

anything other than c is homosexual

How is it impure? You just take a string and return a string

>we won't explain precisely
Because they can't. Haskell is unrigorous garbage.

Lol retard. You have no clue what pure means.

then out with the garbage it goes

Yes
hackage.haskell.org/package/regex-compat-0.95.1/docs/Text-Regex.html

ah ok well I take it back

I will be soon again. Currently reading up on what Haskell is being compiled to, the Core (Fancy Lambda Calculus + few extras) so that I can translate Haskell into another, potentially more efficient language. Currently I wonder how one can deal with the lazy semantics of Haskell in a strict target language that is turing-complete, but comes with severe restrictions: How the fuck would you translate [1 .. ] into a lambda calculus without Y-combinator or other forms of unbounded parallelism?

Actually I am looking forward to writing more practical Haskell soon, haven't been in some time. This language is so fucking cool.

Of course it does. But check out parser combinators, for more complex tasks they can be much more readable than a regex. I mean you shouldn't write complex regexes anyway, but parser combinators are just neat.

because I have no ivy league phd to apply to a place where they pay for writing Haskell

I can write anything but Java or Javascript, but there is demand only for $5/h JS coding niggers on Upwork

Because it's restrictive.

Attached: Lisplogo.png (811x805, 29K)

$ is totally fucking pointless if it's left associative, it's the same as putting everything after it in parenthesis, it's one of the ways that Haskell manages to not look like lisp. You don't need a left associative version because it loses the entire purpose of the fucking function/operator

I agree on do notation though,it's really just a way for people to write code as if Haskell is another language, and in most cases it makes you lose some of Haskell value. At worst, it gets abused as a way to make Haskell into a procedural language

lol you have no idea what you are babbling about lol

$ has lover precedence, so it could be used to simplify long chains of composed functions, and the possibility to define a function just like that is absolutely amazing, you fucking imbecile

no one is forcing you to use it even once

>Use parens
Are you really saying the first is easier to read than the second? ($) is great.
f (g (h x)
f $ g $ h x

not him, but yes

(f . g . h) x

Nigger.

You're confusing associativity with precedence. ($) could still have very low precedence, but it ought to be left-associative.

a ? b ? c -- given this:
(a ? b) ? c -- left associative
a ? (b ? c) -- right associative

f (g (h x)
f $ g $ h $ x -- 1
f . g . h $ x -- 2

1 and 2 are equivalent. ($)'s right associativity is not useful. If it were left associative, you could express this without parens:

f (g x) y
f $ g x $ y

Have you seen the people that are using Haskell?
Urgh, I'm not going to be part of that crowd.

Haskell is a blast to program in and might be my favorite language of all time but the tooling is just god awful. Haskell is pitched as a cross platform Lang but windows support is a total joke. I recently wrote a command line program to emulate terry Davis' god talk program only to find that there's a bug in the standard library on windows that doesn't let you shut off echoing on the command line. It's been there for years and still no one's bothered to fix it.

Also, ghc uses libgmp for their integer implementation. Libgmp is licensed under lgpl. On windows, dynamic linking to libraries is broken and has been for the past decade. I've been following it on their git repo and it gets mentioned every few years and then ignored by the developers. What this means is that a company is essentially locked out of creating non-open source products on windows since lgpl requires dynamic linking to the lgpl library.

All of this stuff on top of lack of any real ide has caused me to switch over to using f# mainly. Less powerful of a language with no typeclasses, monads or lazy evaluation, but it's still a blast to use and the support actually exists.

can you paredit $?

eta-lang.org/ is trying to make Haskell industry-friendly

For what reason?
It's the most useless niche shit around, period.

You've never even heard of the most niche language.

Obviously.
But also I have, because I made it fuch you.

Compilation takes forever, especially if Template Haskell is involved. Cabal and Stack are ugly and Nix spreads to your projects like kudzu. However, the main problem is that open source libraries written in Haskell suffer from poor performance and documentation, and a general lack of production testing. Comparable libraries in a worse language like Java are better quality.

Yes. Haskell even has good PCRE bindings.

>Currently I wonder how one can deal with the lazy semantics of Haskell in a strict target language that is turing-complete
gitlab.haskell.org/ghc/ghc/wikis/commentary/compiler/strictness-analysis
>Y-combinator or other forms of unbounded parallelism?
>unbounded parallelism
What?

>he writes his shit in some academia meme programming paradigm

Attached: resume-trash1-300x171.jpg (300x171, 8K)

Use Lisp already.

Agree 100%.
The language itself is great, the tooling unfortunately not so much.

haskell is hard and not very useful

>he unironically uses c

>t. abstractFactoryBeanPOO

>implying Java isn't the same but worse.