Stack Exchange
log in

Grimmy

 Mathematical Regexes

Grimmy
Jul 12, 2022 18:15
store.steampowered.com/app/247080/Crypt_of_the_NecroDancer
Grimmy
Jul 12, 2022 18:10
@Deadcode sure, you 100% have my permission to post that! i’m busy with game dev at the moment so i don’t have time to write up an explanation for it (plus i haven’t done mathematical regex in a while, so i’m not sure i’d still understand how it works)
Grimmy
Jul 12, 2022 17:20
@Deadcode oh no that was years ago, i don’t know if i still have it anywhere
Grimy
May 6, 2019 14:32
^(((x+)\3+)(?=\3\3\2$)x((?=\3+(x*))(?=\5+$)(?=((x+)(?=\5\7*$)x)\6*$)\3+(?=\5\7$)x)*\3x|x?)x$ is an equivalent 92
Grimy
May 6, 2019 14:32
That's not where the golf is, though
Grimy
May 6, 2019 14:26
92: ^(((x+)\3+)(?=\3\3\2$)(x(?=\3+(x*))(?=\5+$)(?=((x+)(?=\5\7*$)x)\6*$)\3+(?=\5\7$))*\3xx|x?)x$
Grimy
May 6, 2019 12:50
And of course you can get a slow 94 from it
Grimy
May 6, 2019 12:50
Here's a counting downwards 95: ^((x+)\2+)(?=\2\2\1$)x((?=\2+(x*))(?=\4(?=\4*$)((x*)(?=\4\6*$)x)\5*$)\2+(?=\4\5$)x)*\2xx$|^xx?$
Grimy
May 6, 2019 12:09
down to 100:
^((x*)(\2x)*)(?=xx\1$)((?=\2+(x*))(?=\5(?=(\5xx)*$)((x*)(?=\6\8*$)x)\7*$)\2\3*(?=\6\8$))*\2\2$|^xx?$
Grimy
May 6, 2019 11:03
sub-100 should still be possible
Grimy
May 6, 2019 11:03
an easy golf is
`^((x*)(\2x)*)(?=xx\1$)((?=\2+(x*))\5(?=(\5xx)*$)(?=((x*)(?=\6(\8*$))x)\7*$)\9(?=\3+\5x$))*\2\2$|^xx?$`
Grimy
May 6, 2019 11:03
oh thanks for noticing that
Grimy
May 5, 2019 14:46
that’s what i get for only testing up to 10000
Grimy
May 5, 2019 14:46
grr
Grimy
May 5, 2019 14:41
down to 93, nice
Grimy
May 5, 2019 14:36
so 96
Grimy
May 5, 2019 14:36
the ? after the (\2\3*) is no longer needed
Grimy
May 5, 2019 14:36
oh wait i missed an easy -1
Grimy
May 5, 2019 14:36
i changed how i count a bit, but it’s still in the same direction
 

 CG&CC-gaming

For CG&CC users to discuss and play video games and board game...
1
Grimmy
Nov 6, 2019 23:39
Oh, makes sense! Didn’t realize it was you
Grimmy
Nov 6, 2019 23:29
Hmm are you on the Zachtronics Discord?
Grimmy
Nov 6, 2019 23:23
Nice to see Molek-Syntez talk here (: If you’re interested in advanced optimization, feel free to add me on Steam: steamcommunity.com/profiles/76561198052453103
 

 05AB1E - Oasis

For discussing and golfing in 05AB1E and Oasis
Grimy
Sep 24, 2019 16:37
Still room for optimization though, I think sub-100 is possible.
Grimy
Sep 24, 2019 16:34
I'm pretty happy with how short codegolf.stackexchange.com/a/193481/6484 turned out, when I first looked at the challenge I thought it was gonna be impossible in 05AB1E.
Grimy
Sep 9, 2019 16:19
@Adnan bug with sentence capitalization (.ª): if the last sentence is one character long (as in "hello. i".ª), nothing gets capitalized at all.
Grimy
Aug 29, 2019 11:51
1;· isn't always truthy. 1;·i42 gives 42, but 3Lʒ1;· gives an empty list. Floating-point shenanigans? @Adnan
Grimy
Jun 5, 2019 15:30
à seems to be borked for non-numbers
Grimy
Jun 5, 2019 15:30
why is à not the same thing as {θ T_T
Grimy
Jun 4, 2019 10:28
Golfed odd-size greco-latin squares: tio.run/##ASQA2/9vc2FiaWX//0xE4oCaSUdEzrXDgE5pw4BdKeKCrM62/8K7/…
Grimy
May 29, 2019 13:22
0L returns [1, 0], it would be more convenient/consistent if it returned [] instead.
Grimy
May 28, 2019 11:42
I guess it makes sense, though
Grimy
May 28, 2019 11:41
Oh! Q vectorizes in the list-number case, but not in the list-list case. This is so weird.
Grimy
May 28, 2019 09:00
Also, how do I do a boolean or? Best I could find is ΘsΘ~, but there’s got to be a better way, right?
Grimy
May 28, 2019 08:58
So what’s the difference between Q and .Q? info.txt says .Q is non-vectorizing, but afaict Q doesn’t vectorize either (tio.run/##yy9OTMpM/f/f0Mg4WMHQDEgE/v8PAA).
Grimy
May 24, 2019 14:54
Maybe I should make an 05AB1E layout
Grimy
May 24, 2019 14:54
Yeah my keyboard doesn't have a ‹
Grimy
May 24, 2019 14:42
Ooh yeah I didn't think about ¤‹
Grimy
May 24, 2019 14:40
Oh wait āZÊ> also gives 22
Grimy
May 24, 2019 14:16
It would be 22 with my suggested change to ß
Grimy
May 24, 2019 14:16
Here's an alternative 23-bytes Walk the Words to demonstrate ÅL: tio.run/…
Grimy
May 24, 2019 12:28
ÅL does vectorize, otherwise it'd be completely redundant with ‚ß
Grimy
May 24, 2019 12:28
Nah it doesn't, just tested it
Grimy
May 24, 2019 12:27
Or does it?
Grimy
May 24, 2019 12:27
Doesn't vectorize though
Grimy
May 24, 2019 12:22
As an alternative, a two-byte max(a, b) would also be cool (we already have a two-byte min(a,b), it's ÅL).
Grimy
May 24, 2019 12:21
Yeah, the change to à/ß would not be retro-compatible. I haven't seen any answer that would break because of it, but it's a good idea to be careful.
Grimy
May 24, 2019 12:02
Alright, here's a grab-bag of feature suggestions for @Adnan:

* The big one: change `õ` to a variable that defaults to `""`, and have a single-byte command (`ñ`?) to append to it. This way, we'd have incremental variables for all types: `¾`/`¼` for numbers, `¯`/`ˆ`for arrays, `õ`/`ñ` for strings. (Probably needs two-byte commands to clear or directly assign to it as well).

* I've never wanted suffixes in the order returned by `.s`. A two-byte shortcut for `.sR` (`.η`?) would be nice.

* As far as I can tell, `O` and `P` on lists that contain both numbers and non-numbers do nothing. Ignori…
Grimy
May 23, 2019 18:43
Found it! lSAÁ0ª7₃Ÿžq÷9š‡þO
Grimy
May 23, 2019 17:13
I'm pretty happy with codegolf.stackexchange.com/a/186035/6484, but I think there's still one byte waiting to be saved
Grimy
May 23, 2019 17:13
I figured I should hang around here (=