The Nineteenth Byte

Challenger5

8:02 AM

@LeakyNun It's not.

Leaky Nun

@Mr.Xcoder just dissect my pyth answer to that challenge

Mr. Xcoder

@LeakyNun What Pyth answer?

Leaky Nun

@Challenger5 you can't expand an array right? so the size must be stored somewhere?

@Mr.Xcoder the partition challenge on main

Mr. Xcoder

Ok

3

A: How many partitions contain only perfect squares?

Pyth, 16 bytes lf!f!sI@sjkY2T./ Test suite.

This one?

Challenger5

@LeakyNun It's known at compile time, but not runtime

Accessing out of bounds is UB

Leaky Nun

8:06 AM

@Challenger5 oh, interesting!

so malloc is done at compile time?

@Mr.Xcoder yes

Challenger5

@LeakyNun Arrays are stored on the stack. The size of anything on the stack must be known at compile time, otherwise you can't get the return address, etc.

Leaky Nun

@Challenger5 quite interesting...

how does sizeof work then?

also, is the array pointer constant, i.e. cannot be changed?

Challenger5

@LeakyNun You should consider every different array length as its own type. sizeof expressions are replaced at compile time.

Leaky Nun

@Challenger5 is there anything done at runtime?

is 1+1 done at runtime?

Challenger5

@LeakyNun Any sort of local-variable stuff is handled on the stack for efficiency.

@LeakyNun No, it's a constant expression so it will be replaced.

Leaky Nun

8:11 AM

@Challenger5 can you name anything done at runtime?

Challenger5

(This is all assuming a standard compiler; the C spec does not guarantee it.)

Leaky Nun

let's say i have int a[99] and int b[9], and i do sizeof(expression?a:b), then what would happen?

Challenger5

@LeakyNun Calls to printf and I/O functions are done at runtime, of course. Function bodies will sometimes be inlined if they are sufficiently small, but generally they are called at runtime as well. C optimization in most compilers is pretty aggressive.

Leaky Nun

let's say the expression is done at runtime

Challenger5

@LeakyNun I'm pretty sure that wouldn't be allowed. Let me test it real quick

Leaky Nun

8:13 AM

@Challenger5 thanks

if it works, then probably it would be compiled to expression?396:36, i guess

Challenger5

@LeakyNun Also, you can't allocate an array with a size that's not known at compile time

Leaky Nun

@Challenger5 so nothing like int a[i]?

if i is determined at runtime?

maybe I should grab a book about C

Challenger5

@LeakyNun Nope.

You can fill it with constexprs like a[48 + 65535 + 1] though.

Leaky Nun

have you tested the 99 or 9 one?

Challenger5

Yeah, but I have absolutely no idea what it's doing.

Leaky Nun

8:19 AM

@Challenger5 wild guess: it's the size of that expression somehow

Challenger5

Yeah, that's what I was thinking

Leaky Nun

try disassembly?

Challenger5

OK, apparently actually you can have variable-length arrays, but I think they would have to be allocated on the heap.

Leaky Nun

I'm on mobile now so you might be more convenient

Challenger5

Wikipedia says, "although the feature is considered an optional implementation in later versions of the C standard"

Mr. Xcoder

8:21 AM

Strangely, if you assign a variable k to 0 ? a : b, you get another result.

Challenger5

Which basically means that it you do it your program is no longer compatible and compilers will optimize it into smithereens.

@LeakyNun Oh, I know - it's the size of the pointer.

Leaky Nun

@Mr.Xcoder someone forgot to press save

@Challenger5 but sizeof(b) returns 36?

@Challenger5 UB?

Mr. Xcoder

@LeakyNun Fixed.

Challenger5

@LeakyNun No, but it's "optional implementation" which means that your code is no longer portable across compilers/versions.

Not that C was portable anyways.

@LeakyNun Because "[sizeof] is one of the few exceptions to the rule that the name of an array is converted to a pointer to the first element of the array".

So if you use the array variables in an expression, they just mean pointers.

But if you use the names in sizeof, it means something different.

Leaky Nun

@Challenger5 but pointers have size 4?

Challenger5

8:24 AM

On 32 bit systems they do.

On 64 bit they're size 8.

Leaky Nun

if *k is a pointer, then what is k?

can you test TIO is 32 or 64?

Challenger5

sizeof(long) is 8, which makes me think it's 64.

But I think Dennis would be a more trustworthy source.

Leaky Nun

can you test a pointer?

@Challenger5 sizeof long is 8 even on 32...

Challenger5

OK. I just tested, and long* is also 8.

Mr. Xcoder

That's the reason I avoid C ^

Liad

8:28 AM

im a big confused about $(0+1) \ ^ {*}$ meaning. is $010 $ is in that language ?

(is this the right chat to ask this question ? :P )

Leaky Nun

@Liad dont use latex for regex

Challenger5

What language?

Leaky Nun

(0+1)*

no, it doesn't match 010.

Liad

alright

Challenger5

Because (0+1) matches one or more zeroes, followed by a one.

Liad

8:29 AM

i read somewhere that the language of (0+1)*bb(0+1)* is all the words that contains bb

Challenger5

If you want a character class, you can use [0+1].

Leaky Nun

@Challenger5 no it doesn't

Challenger5

Oh, whoops. I forgot about +.

Leaky Nun

@Liad then it is wrong

Liad

they used union instead of +

Leaky Nun

8:31 AM

what?

Liad

im learning computability theory and it is a just shown that regex are equivalent to regular languages

Leaky Nun

[01]* matches 010.

Challenger5

If bb is a bitstring, then .*bb.* matches any string that contains bb.

Leaky Nun

if + is union then in regex it is [01]*.

Liad

and they gave as an example that $(a \cup b) \ ^{ *} bb(a \cup b) \ ^{ *}$ is the language of all words that contains $bb$

Leaky Nun

8:32 AM

sorry for the confusion

@Liad that is true

Liad

So how come $aba \in (a \cup b )\ ^ {*} $

Challenger5

By the way, have you read Regular Expression Matching can be Simple and Fast? The first part of it was very instructive to me.

Liad

isn't it a* union b* ?

Challenger5

The second part describes a C algorithm for matching regex, which isn't quite as useful theoretically

Leaky Nun

@Liad no

Liad

8:34 AM

can you explain ?

i could have guessed it is not :P

Challenger5

In regular expression notation, a* means "zero or more of a".

Leaky Nun

@Liad a in [ab], b in [ab], a in [ab]

Mr. Xcoder

@Challenger5 Quantifier, in other words.

Leaky Nun

@Liad I'm on mobile. I'm typing.

Challenger5

@Mr.Xcoder Yes.

Leaky Nun

8:35 AM

@Liad where [ab] is a union b

so aba in [ab]*

Mr. Xcoder

Have we changed the subject from C to regex?

Challenger5

I think so.

Leaky Nun

in each * you can choose a different choice

I'm done @Liad

Liad

got it , thanks @LeakyNun

Leaky Nun

a* union b* is (a*|b*) in regex

Mr. Xcoder

8:36 AM

Can we now changed the subject to Pyth training? :P

Leaky Nun

@Mr.Xcoder just ask

Mr. Xcoder

I don't have anything to ask :p

But will have once I give myself a task

Leaky Nun

@Mr.Xcoder given [3,1,4,1], return [[1,4,1],[3,4,1],[3,1,1],[3,1,4]]

Mr. Xcoder

@LeakyNun All contiguous permutations of length len(list-1)?

Leaky Nun

not contiguous at all

Mr. Xcoder

8:40 AM

I don't get it

I think I got it

Leaky Nun

remove the first element

remove the second element

remove the third element

remove the fourth element

Mr. Xcoder

Yeah, I got it

Mayube

@LeakyNun CMC it

Liad

@LeakyNun just realized , $(0\cup 1) \ ^ {}$ is just $\{ 0,1 \} \ ^ {} $ in regex, right?

Leaky Nun

@Liad in regex it is (0|1)*.

Mr. Xcoder

8:42 AM

@LeakyNun For a list of any length, right?

Leaky Nun

@Mr.Xcoder yes

Liad

@LeakyNun the left one is in regex, the right in "regular languages" form, that is what i meant

@LeakyNun never mind

Leaky Nun

regex doesn't use union either

but I get what you mean

and I'll answer yes @Liad

Liad

@LeakyNun alright :P

Mr. Xcoder

@LeakyNun Golfing hint?

Leaky Nun

8:47 AM

@Mr.Xcoder no hint

Mr. Xcoder

@LeakyNun Good

Found one but does not work if the list has duplicate elements :( too Naïve

@LeakyNun (this one works for duplicate elements) 13 bytes: m+<Qd>QhdtMSl

Leaky Nun

try harder

Mr. Xcoder

@LeakyNun To golf it?

Sorry, forgot to press permalink

Leaky Nun

yes

Mr. Xcoder

Good link version: m+<Qd>QhdtMSl (To be golfed)

Mayube

8:55 AM

oh dang how did I not know that MySQL can do regex

Mr. Xcoder

I have an idea, I just have to search through the built-ins. May take bit

Leaky Nun

@Mayube do my challenge in braingolf?

@Mr.Xcoder if you haven't scrolled through the list of built-ins, then there is some problem

Mayube

@LeakyNun which one? The remove each element one?

Leaky Nun

yes

Mayube

I'll give it a shot

isaacg

8:58 AM

@LeakyNun That's a nice one, I have it in 5 bytes

Mr. Xcoder

@LeakyNun I know it's not yet good, but I have 11 bytes: m+<Qd>QhdUl

isaacg

Sure, sorry.

Mr. Xcoder

Then, 10 bytes: m+<Qd>QhdU

@isaacg Let me solve it myself pls

isaacg

Ul is the same as U, by the way.

Ugh, chat is unintuitive

Mr. Xcoder

And sorry for the crappy issue I raised on Github yesterday

I closed it in the meanwhile

isaacg

9:01 AM

I saw. Don't worry about it - there's no harm in such things.

Mr. Xcoder

@isaacg Great

isaacg

:) It's nice to have people who appreciate the language and want to improve it.

Mr. Xcoder

I found something that might be of great use: .( - Is it relevant here?

Leaky Nun

@isaacg 5 bytes... can you post a spoiler?

isaacg

>!Test spoiler

Apparently no, I can't post a spoiler.

Leaky Nun

9:03 AM

@isaacg link

isaacg

Oh, sure.

Leaky Nun

[spoiler](link to that)

Mr. Xcoder

@LeakyNun Not even a small hint to golf it?

What's your byte count, then?

isaacg

Spoiler gone.

Leaky Nun

I haven't tried it

Mr. Xcoder

9:04 AM

Now I am struggling to golf it and struggling to not look at the spoiler either, thanks :)

Leaky Nun

@isaacg you can just link to the compiler...

isaacg

I could. But that would put the code in the link URL.

Leaky Nun

you can use [spoiler](...) to cover the link

isaacg

Both ways work.

Leaky Nun

interestingly you used a built-in that I don't know

Mr. Xcoder

9:07 AM

@LeakyNun if you haven't scrolled through the list of built-ins, then there is some problem

isaacg

Ok, here's another way to do it with no weird builtins: spoiler, same length.

Leaky Nun

@Mr.Xcoder yes, I have some problem.

@isaacg that's clever

isaacg

:) Wrong order though, it would make it longer to fix it.

Mr. Xcoder

@LeakyNun Does the order matter?

NVM

Leaky Nun

yes

Mr. Xcoder

9:11 AM

@LeakyNun *Faceplam*

@LeakyNun 6 bytes: m.DQdU

isaacg

Nice!

Mr. Xcoder

Took me a long time to find .D

> .D <seq> <num> Returns A with the element at index B removed.

I was searching for "ommited","index" but I found it when I searched for "removed"...

@LeakyNun Next? (or should I golf further?)

Leaky Nun

use L

Mr. Xcoder

@LeakyNun Ok....

> L <pfn> <n-1:any> <col> Left map. n>=2. Map A(B, ..., _) over C, where _ is the lambda variable. Map uses m underneath, and B may refer to its lambda variable.

This one ^?

Leaky Nun

yes

Mr. Xcoder

9:17 AM

Ok, I first gotta play around with L and see how it works

@isaacg I would give you a bounty for the beautiful language you created... I am upset that I can't.

isaacg

@Mr.Xcoder I appreciate the thought. And I have plenty of rep as it is.

Mr. Xcoder

BTW What's <pfn>?

I know <pf1> means I should put sth in front, but n?

isaacg

Click on the link at the top to see the key.

n is used again in the text. It says n >= 2.

Mr. Xcoder

Preceding function of arity n. *F, for instance.

So I must put at least two things in front.

isaacg

No, you must put one function of arity at least 2 in front.

Mr. Xcoder

9:21 AM

@isaacg Oh, I must put a dyad or something with more parameters in front of it

isaacg

Exactly.

Mr. Xcoder

Should I still use .D?

isaacg

Yep

CMC: Reduce by ternary. Given an odd length sequence of 0s and 1s, replace the first 3 numbers by the second one if the first is 1, and the third one if the first is 0. Repeat until there is one number, then return that number. For instance, [0, 0, 1, 1, 0] -> [1, 1, 0] -> 0.

Mr. Xcoder

@isaacg Wait for me to finish this plsss

isaacg

It's not going anywhere

Mr. Xcoder

9:25 AM

What's the variable for L? I think It is something like .DQ<something>L

Oh, I'm completely wrong.

I have to put the dyad in front and I have to loop on something, it should start with .DL, right?

Mayube

@isaacg Shouldn't that be [0,0,1,1,0] -> [1,1,0] -> 1?

isaacg

@Mayube Yes, you're right.

Mr. Xcoder

6 bytes again: .DLQUl

Oh, no, Ul is U

5 bytes: .DLQU

Mayube

@isaacg Braingolf, 22 bytes: &,1>[?>$_<:$_|<$_l1->]

Mr. Xcoder

Good ^^?

isaacg

9:30 AM

Yes! Correct.

Mr. Xcoder

I shall move on to your CMC, shall I?

isaacg

Sounds good. I'll try to solve it too, now.

Mr. Xcoder

@isaacg I don't get you CMC: I have [0, 0, 1, 1, 0]. The first 3 numbers are 0,0,1. The first is 0, so I replace them all with the third: 1,1,1. I get [1,1,1,1,0].

Ahh, I got It.

The first and the second after the first three.

Mayube

@Mr.Xcoder not 1,1,1, just 1

Mr. Xcoder

@Mayube I misunderstood formerly

How does [1, 1, 0] become 0?

isaacg

9:36 AM

That was a typo, sorry.

It should become 1

Mr. Xcoder

@isaacg But how, anyway?

isaacg

First is 1, so we take the second number, which is 1.

Leaky Nun

@isaacg 10 bytes

isaacg

Same length as what I had, and essentially the same code. Nice!

Leaky Nun

thanks

Mr. Xcoder

9:45 AM

Sorry for my absence, I had a netfall

Another CMC?

isaacg

Feel free to come up with one, if you want. Those are just as fun to solve.

Leaky Nun

still isaac's one

Mr. Xcoder

@LeakyNun Nah, I personally wouldn't like solving that one

And I've looked at your solution anyway, so...

CMC: Given two lists of numbers of the same length, return the product of their corresponding elements

[1,2,3,4], [5,6,7,8] -> [5,12,21,32]

Leaky Nun

three bytes in pyth

Mr. Xcoder

Okay

@isaacg Does Pyth have a zip?

Leaky Nun

9:51 AM

C

isaacg

It has C which is n-wise zip, and it has V, which is 2 wise zip as a pfn

Mr. Xcoder

> zip argument #1 must support iteration

isaacg

@LeakyNun I can do 3 bytes three different ways

@Mr.Xcoder What's your code?

Mr. Xcoder

@isaacg NVM

Leaky Nun

@isaacg do you want a cookie?

isaacg

9:55 AM

Was that a "well done" or a "calm down, you're getting too competitive"? I'm not good at reading tone of voice.

Mr. Xcoder

Do you have Q and E in your code?

Leaky Nun

that was a question

isaacg

I like cookies, especially the kind I can eat.

Mr. Xcoder

> Do you have Q and E in your code?

isaacg

@Mr.Xcoder In one version, I have one of those.

Mr. Xcoder

9:56 AM

And how do you read two lists, then?

isaacg

In the other versions, I take a tuple of two lists, like you originally framed it

[1,2,3,4], [5,6,7,8]

In the version using E, they're newline separated.

Mr. Xcoder

Is there a built-in I don't know?

isaacg

All of them use only common characters, no strange builtins

Mr. Xcoder

@isaacg Then I guess I can solve it

IDk

@isaacg How should I use C to get (1,5),(2,6)...? Or V?

isaacg

If you start with [1,2,3,4], [5,6,7,8] and use C, it's very easy.

Mr. Xcoder

10:06 AM

@isaacg ~~That didn't really help~~ That did help

isaacg

The code is C.

Mr. Xcoder

@isaacg Figured that in the meanwhile. Thanks!

isaacg

Ah

You're welcome.

Mr. Xcoder

*MC

isaacg

Perfect!

Mr. Xcoder

10:08 AM

I should really scroll more through that list of built-ins

isaacg

My favorite way to solve this is *VF

Mr. Xcoder

@isaacg What's the strong point of Pyth? I mean, in what type of challenges does it excel?

I see it's good with list manipulation

And String manipulation

isaacg

In my opinion, its strong point is understandability. It's much easier to write a long (e.g. 80 byte) program in Pyth than other golfing languages

But to be honest, Jelly and 05AB1E are typically a bit shorter at just about everything.

And that's even after compensating for the larger range of characters.

It's natural for that to happen over time - Pyth is basically only beaten than newer languages than it.

Mr. Xcoder

That issue I opened yesterday was a matter of winning to Leaky's Jelly solution by 1 byte and losing to JonathanAllan's Jelly solution by 1 byte

isaacg

It only would have saved 1 byte per use. Were you using it twice?

Mr. Xcoder

10:13 AM

Now I am tied with Leaky's Jelly one and 2 bytes behind Jonathan

@isaacg Wait...

Destructible Lemon

hi

Mr. Xcoder

4

A: Randomly select a character, plinko-style

Pyth, 10 8 bytes O+<Q2*2t Uses the exact same algorithm as is @JonathanAllan's Python answer. Try it online! (I have the same test case multiple times such that you can see it's random) Explanation O - Takes a random element of the String made by appending (with +): <Q2 - The first tw...

No, just once.

isaacg

So if *2 was a 1 byte builtin, it would be 7 bytes long.

Mr. Xcoder

The jelly solutions were 8 and 6 bytes respectively. And I had 8 too

@isaacg Yep, Not really useful though

isaacg

I get it, I misunderstood the first time.

@DestructibleLemon Hi

Mr. Xcoder

10:15 AM

@DestructibleLemon Hi

Destructible Lemon

@Mr.Xcoder ninja'd :P

Mr. Xcoder

@DestructibleLemon Can't multiple people say "HI"?

Now we have greeting rules too?

Destructible Lemon

hence the ":P"

Mr. Xcoder

@isaacg Pyth assignment?

isaacg

I was just working on golfing your plinko code. It can be shortened by at least one byte.

Mr. Xcoder

10:18 AM

@isaacg Really?

Thanks BTW

isaacg

Well, the code looks completely different, so I'm not sure it's really golfing your code, but: @Q|O8 1

Mr. Xcoder

What's |?

isaacg

Logical or

Mr. Xcoder

Logical or?

How is that random?

isaacg

The O

Mr. Xcoder

10:20 AM

I don't understand your code >_>

I actually do

But I doubt that's valid

isaacg

Why not?

Mr. Xcoder

@isaacg You must pick a random character from a string that has the first char repeated once, the second repeated thrice and the other two repeated twice each

How does your code do that?

isaacg

No, that's not a requirement of the challenge. It just specifies the probabilities, not how you select the character.

Mr. Xcoder

@isaacg ?

How does the second character have a 3/8 probability?

With you code?

isaacg

If O8 outputs 0, 1 or 5, the second character will be picked.

Because |O8 1 will be 1, 1 or 5 respectively.

Mr. Xcoder

10:24 AM

@isaacg How does the first have 1/8?

isaacg

It is only output if 4 is generated by O8.

Mr. Xcoder

Oh, I didn't see the <space>1

isaacg

Yeah, that's crucial.

Mr. Xcoder

I thought your code was @Q|O8, sorry, my mistake

Cool

Can I use @isaacg ?

isaacg

I think it should probably be a separate answer, to be honest.

There's no overlap with your code.

Mr. Xcoder

10:26 AM

@isaacg Go ahead and post it, then

isaacg

Will do.

Mr. Xcoder

@isaacg Consider making it a test suite with the same string multiple times, such that people can see it's random. And because it's a brilliant idea, consider adding a good explanation to it.

Ping me when you post it

isaacg

@Mr.Xcoder pyth.herokuapp.com/…

1

A: Randomly select a character, plinko-style

Pyth, 7 bytes @z|O8 1 Test suite O8 generates a random number from 0 to 7. | ... 1 applies a logical or with 1, converting the 0 to a 1 and leaving everything else the same. The number at this stage is 1 2/8th of the time, and 2, 3, 4, 5, 6, 7 or 8 1/8 of the time. @z indexes into the input ...

Mr. Xcoder

@isaacg Already upvoted before you first comment

isaacg

And now, I need to sleep. Good night, everyone.

Mr. Xcoder

10:34 AM

GN

isaacg

:)

Mayube

10:45 AM

Huh, surprised we don't have a catalog tag

programmer5000

11:20 AM

What's the gofiest way to generate a string / array containing all lowercase letters in the alphabet in JS ES6?

dzaima

there was a main post about that

14

Q: Generating the alphabet in JavaScript

I'm pretty sure there's not a better way to do this but figured it couldn't hurt to ask. I'm tired of typing out a='abcdefghijklmnopqrstuvwxyz'. Cool languages have Range('a'..'z') or similar What can we come up with with JS that's as short as possible?? for(i=97,a='';i<123;){a+=String.fromCh...

Mr. Xcoder

@programmer5000 "abcdefghijklmnopqrstuvwxyz"

programmer5000

:( I wish there was something shorter.

Mr. Xcoder

@programmer5000 There ain't

totallyhuman

Python has a nice short way :3

map(chr,range(65,91))

Mr. Xcoder

11:30 AM

@totallyhuman That's uppercase

totallyhuman

@ASCII-only I don't know how it's supposed to look, but grey text on green is very bork to me

@Mr.Xcoder map(chr,range(97,123))

Mr. Xcoder

@totallyhuman Pyth has a nice, short way:3

G

Oh, a new ascii art challenge

New Main Posts

1

Q: ASCII 2D game-map

Problem Given input a where a is a grid of characters in any input format as long as it has only one element for each 'block' of the output. And input b where b is a grid of numbers the same size as input a. There are two types of road, a 1 represents a stone road marked by @ and a 2 represent...

code-golf ascii-art

New Sandboxed Posts

0

A: Sandbox for Proposed Challenges

The answer to life, the universe, and ASCII-art Simple challenge: try to output the following text in as few bytes as you can: the answer toli fetheuniv ersea nde ver ything the ans wer tol ife the uni ver sean dev ery...