🦋 Welcome to the MAIN() IRC channel of the Raku Programming Language (raku.org). Log available at irclogs.raku.org/raku/live.html . If you're a beginner, you can also check out the #raku-beginner channel!
Set by lizmat on 6 September 2022.
00:10 kjp left 00:11 kjp joined 00:12 kjp left, kjp joined 01:20 jgaz left 01:28 hulk joined, kylese left 02:15 hulk left, kylese joined
coleman rebooting docs and main website. 02:55
elcaro perryprog: There's DateTime::Parse, but it's not super documented. Looking at the tests, it looks like it supports a few formats, but I'm not sure if you can specify your own à la strptime 03:10
perryprog there is?? 03:11
elcaro parsing 'YYYYmmddHHMMSS' is not too difficult. One way: DateTime.new: |'20250910131152'.match(/(....)(..)(..)(..)(..)(..)/)
perryprog ... that is better than what I went with 03:12
DateTime.new("{.[0..3].join}-{.[4..5].join}-{.[6..7].join}T{.[8..9].join}:{.[10..11].join}:{.[12..13].join}Z") given $time.comb
elcaro Or if your on linux, here's a test module I wrote for strftime/strptime: gist.github.com/0racle/67c4e9e8f51...1ac2162f3c
I never released it, or tested/used it much... but it seems to work ok 03:13
perryprog interesting 👀
this is just for a one off so it's not too big a deal, but it is definitely one of those things that's a big nice to have in a stdlib
elcaro Here's another way to do what you did: '20250910130537'.comb.rotor(4,2,2,2,2,2)».join 03:15
jubilatious1_98524 Maybe DateTime::Parse will be of interest: raku.land/github:sergot/DateTime::Parse
unix.stackexchange.com/a/709253 03:17
perryprog elcaro: in my defense it's been a while since I've done Raku🫣
that is really elegant, though
thanks for the pointer jubilatious! 03:19
man rotor has got to be one of the most weirdly powerful functions that I always forget about 03:22
elcaro Yeah, even tho I don't use it heaps, I always appreciate it when I need it. 03:30
Note also that if what you need is `.rotor(n, :partial)`, that's just `.batch(n)`
Voldenet m: grammar MyDateFormat { regex TOP { <year=d4> <month=d2> <day=d2> <hour=d2> <minute=d2> <second=d2> { make DateTime.new(|$/) } }; regex d2 { \d**2 }; regex d4 { \d**4 } }; say MyDateFormat.parse("20250910130537").made 03:31
camelia 2025-09-10T13:05:37Z
Voldenet here, have some sledgehammer solution 03:32
though jokes aside, it's very easy to adjust to any other format 03:34
m: grammar MyDateFormat { regex TOP { <year=.d4> <month=.d2> <day=.d2> <hour=.d2> <minute=.d2> <second=.d2> { make DateTime.new(|$/) } }; regex d2 { \d**2 }; regex d4 { \d**4 } }; say MyDateFormat.parse("20250910130537").made # don't pass those d2 to datetime's constructor 03:36
camelia 2025-09-10T13:05:37Z
elcaro FWIW, the first .match one I posted benches faster than the .rotor one or the grammar one. 03:47
03:51 arkiuat left
elcaro Using .substr would be even faster, but annoying to write 03:53
Like so: DateTime.new: |(.substr(0,4),.substr(4,2),.substr(6,2),.substr(8,2),.substr(10,2),.substr(12,2)) given '20250910131152' 03:55
~250% faster than the .match version
Actually don't need the `|` on that last one 03:58
Voldenet the grammar one is actually going to be pretty slow without tons of optimizations 04:02
because it's using the hash to initialize new and grammars are not horribly fast either
m: grammar MyDateFormat { regex TOP { <day=.d2> <year=.d4> <month=.d2> <hour=.d2> <minute=.d2> <second=.d2> { make DateTime.new(|$/) } }; regex d2 { \d**2 }; regex d4 { \d**4 } }; say MyDateFormat.parse("20250910130537").made # but you can do that easily 04:03
camelia 2509-10-20T13:05:37Z
04:03 arkiuat joined 04:31 tbrowder left 04:34 tbrowder joined 05:32 Sgeo left 05:37 Aedil joined 05:50 arkiuat left 05:52 abraxxa joined 06:04 arkiuat joined 07:07 arkiuat left 07:21 arkiuat joined 07:26 arkiuat left 07:31 lichtkind joined 07:49 arkiuat joined 07:54 arkiuat left 08:09 dakkar joined 08:16 Aedil left 08:23 arkiuat joined 08:28 arkiuat left 08:31 melezhik joined 08:48 Guest31 joined 08:50 arkiuat joined 08:54 apac joined 09:06 itaipu left 09:17 Guest31 left 09:21 arkiuat left 09:34 arkiuat joined 09:38 arkiuat left 10:06 arkiuat joined 10:11 arkiuat left 10:16 apac left 10:29 crnlskn joined 10:32 arkiuat joined 10:39 arkiuat left 10:50 melezhik left 11:09 arkiuat joined 11:14 arkiuat left 11:20 arkiuat joined 11:24 arkiuat left 11:49 Aedil joined 11:52 arkiuat joined 11:57 arkiuat left
antononcube @jubilatious1_98524 "DateTime::Grammar" is better: raku.land/zef:antononcube/DateTime::Grammar 12:04
12:16 arkiuat joined 12:21 arkiuat left 12:34 simcop2387 left 12:36 simcop2387 joined 12:45 arkiuat joined 12:48 simcop2387 left 12:50 arkiuat left 13:17 arkiuat joined, itaipu joined 13:22 arkiuat left 13:25 oodani left 13:27 itaipu left 13:30 oodani joined 13:31 xinming left 13:32 xinming joined 13:34 simcop2387 joined 13:38 arkiuat joined 13:43 itaipu joined 13:44 arkiuat left 14:15 arkiuat joined 14:20 arkiuat left
perryprog elcaro, wait HOW IS .MATCH FASTER 14:20
*than rotor 14:21
14:29 crnlskn left 14:39 arkiuat joined 14:43 arkiuat left 15:11 arkiuat joined 15:16 arkiuat left 15:19 apac joined
ab5tract m: "20250910131152" ==> -> $ts { [[0,4],[4,2],[6,2],[8,2],[10,2],[12,2]].map: ->[ \f,\l] { $ts.substr(f,l)} }() ==> { DateTime.new(|$_) }() ==> dd() 15:20
camelia DateTime.new(2025,9,10,13,11,52)
ab5tract was inspired to pull out the old feed operator 15:22
15:40 arkiuat joined 15:44 arkiuat left 16:09 apac left 16:13 arkiuat joined 16:17 arkiuat left 16:39 dakkar left 16:41 arkiuat joined 16:45 arkiuat left 16:46 arkiuat joined 16:51 arkiuat left 17:20 arkiuat joined 17:24 arkiuat left 17:42 arkiuat joined 17:47 arkiuat left 17:49 itaipu left 18:02 itaipu joined 18:16 arkiuat joined 18:21 arkiuat left 18:34 arkiuat joined 18:35 abraxxa left 18:39 arkiuat left 18:45 arkiuat joined 18:49 arkiuat left 19:18 arkiuat joined 19:24 arkiuat left 19:27 apac joined 19:44 arkiuat joined 19:48 Aedil left, arkiuat left 20:11 arkiuat joined 20:18 arkiuat left 20:36 arkiuat joined
librasteve Voldenet: et al ... how about github.com/Raku/problem-solving/issues/494 .FatRatStr 21:14
21:55 ACfromTX left
Voldenet ah, so it would never become Num in the first place 22:07
22:08 ACfromTX joined
Voldenet m: say FatRat.new(16605402, 10**34).nude; say 1.6605402e-27.FatRat(1e-100) # those are equal, but str can be directly interpreted as FatRat 22:10
camelia (8302701 5000000000000000000000000000000000)
0.00000000000000000000000000166054
Voldenet m: say FatRat.new(16605402, 10**34); say 1.6605402e-27.FatRat(1e-100)
camelia 0.0000000000000000000000000016605402
0.00000000000000000000000000166054
22:11 human-blip left 22:13 human-blip joined
Voldenet m: sub z(Str() $_) { S:g/('0'+)/{ "\{0x{$0.chars}}" }/ }; say z FatRat.new(16605402, 10**34); say z 1.6605402e-27.FatRat(1e-100) # more readable 22:18
camelia {0x1}.{0x26}166{0x1}54{0x1}2
{0x1}.{0x26}166{0x1}54
Voldenet m: sub z(Str() $_) { S:g/('0'+)/{ $0.chars > 3 ?? "\{0x{$0.chars}}" !! $0 }/ }; say z FatRat.new(16605402, 10**34); say z 1.6605402e-27.FatRat(1e-100) 22:20
camelia 0.{0x26}16605402
0.{0x26}166054
Voldenet I'd say that special casing NumStr -> FatRat should be enough 22:22
> glot.io/snippets/hazwbo9b0m 22:31
in fact, NumStr -> Rat would be more than enough to handle this, after all 16605402 is not a large number 22:33
(the above is llm generated, because I hoped that it'd cover all possible cases, I didn't test it on more than one example) 22:39
23:09 itaipu left 23:17 Sgeo joined 23:24 itaipu joined 23:25 vasko left 23:27 vasko joined 23:46 itaipu left

Powered by Cro and the Raku® Programming Language.

Please report any issues / comments / feature requests as an issue on App::Raku::Log.

Thank you!