🦋 Welcome to Raku! raku.org/ | evalbot usage: 'p6: say 3;' or /msg camelia p6: ... | irclog: colabti.org/irclogger/irclogger_log/raku
Set by ChanServ on 14 October 2019.
xinming Is there a way to pause a Supply.interval(1) supply? 03:34
What I mean is something like, my Channel $c .= new; react whenever Supply.interval(1) { "pause interval here".say; my $t = $c.receive; "Process here"; "resume interval here again".say; } 03:35
I know we have throttle already, But I wish to do it myself to understand how these kind of problems to be solved correctly. 03:36
Hmm, BTW, I know I can use loop { .. sleep 1; .. } thing, But I wish to try it in Supply 03:37
codesections xinming: I'm not sure if this does quite what you're looking for (I haven't done a whole lot with Supplies), but how about: 04:05
my $s = Supply.interval(1).share; my $t0 = $s.tap; $t0.close; my $t1 = $s.tap 04:08
xinming codesections: hmm, not about tap, I mean pause the Supply source itself. :-) 05:16
raku-bridge <EsperLily> I can't figure out why (0, 0), (* Z+ (3, 1)) ... * produces ((0 0) (5) (4) (4) (4) (4) (4) (4) (4) ...) 05:25
<EsperLily> huh $(0, 0) Z+ (3, 1) produces (5). wtf? 05:29
guifa2 EsperLily: when using the whatever code, the * is getting treated in scalar contexts, and +(0,0) --> 2, and 2 Z+ (3,1) --> 5 as the 1 is seen as extra and is ignored
compare
raku-bridge <EsperLily> ahhh I missed that +(0, 0) == 2. I had just come to the realization that * was being treated as scalar though 05:30
guifa2 my @a = (0, 0), {$^i Z+ (3, 1)} ... *; say @a[0..20]; # explicit scalar
evalable6 ((0 0) (5) (4) (4) (4) (4) (4) (4) (4) (4) (4) (4) (4) (4) (4) (4) (4) (4) (4) (4) (4))
guifa2 my @a = (0, 0), {@^i Z+ (3, 1)} ... *; say @a[0..20]; # explicit array
evalable6 ((0 0) (3 1) (6 2) (9 3) (12 4) (15 5) (18 6) (21 7) (24 8) (27 9) (30 10) (33 11) (36 12) (39 13) (42 14) (45 15) (48 16) (51 17) (54 18) (57 19) (60 20))
guifa2 my @a = (0, 0), {*<> Z+ (3, 1)} ... *; say @a[0..20]; # I wonder if zen operator works
err
raku-bridge <EsperLily> yeah it does 05:31
guifa2 m: my @a = (0, 0), *<> Z+ (3, 1) ... *; say @a[0..20];
camelia 5===SORRY!5=== Error while compiling <tmp>
Only identical operators may be list associative; since 'Z+' and '...' differ, they are non-associative and you need to clarify with parentheses
at <tmp>:1
------> 3my @a = (0, 0), *<> Z+ (3, 1)7…
raku-bridge <EsperLily> oh you need parens
guifa2 my @a = (0, 0), (*<> Z+ (3, 1)) ... *; say @a[0..20];
m: my @a = (0, 0), (*<> Z+ (3, 1)) ... *; say @a[0..20];
camelia The iterator of this Seq is already in use/consumed by another Seq
(you might solve this by adding .cache on usages of the Seq, or
by assigning the Seq into an array)
in block <unit> at <tmp> line 1
guifa2 m: my @a = (0, 0), (|* Z+ (3, 1)) ... *; say @a[0..20];
camelia The iterator of this Seq is already in use/consumed by another Seq
(you might solve this by adding .cache on usages of the Seq, or
by assigning the Seq into an array)
in block <unit> at <tmp> line 1
guifa2 m: my @a = (0, 0), (*.list Z+ (3, 1)) ... *; say @a[0..20]; 05:32
camelia The iterator of this Seq is already in use/consumed by another Seq
(you might solve this by adding .cache on usages of the Seq, or
by assigning the Seq into an array)
in block <unit> at <tmp> line 1
guifa2 jeez
haha
raku-bridge <EsperLily> I was testing with .head(10) myself 05:33
<EsperLily> hmm I can't seem to figure out how to iterate a sequence and stop once an element fails a predicate 05:36
<EsperLily> (in a functional manner I mean) 05:37
guifa2 something probably like 05:47
.head( .first( not *, :k) ) 05:48
but obviously in a way that it'll compiile
lol
raku-bridge <EsperLily> The use here is for terminating an infinite sequence. Scanning it with first would presumably consume it so that's a no-go 06:02
guifa2 my @a = 1,2,3,4,5,Any,6,7,8,9,10; @a.map({ $^i // IterationEnd }).say 06:06
evalable6 (1 2 3 4 5)
guifa2 Granted, that consumes it, but it also gets you the values. But you can do other things obviously inside the map and just terminate by returning IterationEnd whenever you hit an undefined or whatever sentinel value you want 06:07
raku-bridge <EsperLily> Hmm 06:20
<EsperLily> Here's a weird one: (1, 2; 3, 4)[1; 1] returns 4 but (1, 2; 3, 4)[1; 1 % *] returns (4). What the heck? 06:21
elcaro m: say ((0, 0), (*[] Z+ (3, 1)) ...^ *[0] > 20); 06:24
camelia ((0 0) (3 1) (6 2) (9 3) (12 4) (15 5) (18 6))
elcaro or could use toggle 06:25
m: ((0, 0), (*[] Z+ (3, 1)) ... *).toggle(*[0] < 20);
m: say ((0, 0), (*[] Z+ (3, 1)) ... *).toggle(*[0] < 20);
camelia (timeout)
((0 0) (3 1) (6 2) (9 3) (12 4) (15 5) (18 6))
elcaro if you've already got a (maybe infinite) sequence you want to stop when it fails (or passes) a predicate, use toggle 06:26
it's a generalization of takewhile/dropwhile/takeuntil/dropuntil... depending on how you use it 06:27
guifa2 EsperLily: Makes sense to me 06:29
[1; * % 1] is equivalent to [1; .elems % 1]
Which is to say, [1; 1] 06:30
elcaro yep, and elems is 2, so `1 % 2` is still 1, so there's no difference in the output of both expressions
guifa2 just finished getting the <units> element prepped for CLDR. Hooray. 06:31
Although it's killed loading time, dang. I need to figure something out. .28s per language is too slow 06:32
raku-bridge <EsperLily> ok toggle looks neat, oddly hard to find 06:37
raku-bridge <EsperLily> guifa2: [1; 1 % *] should be the same as [1; 1] in this case, but the output is different! 06:37
<EsperLily> Subscripting with [1; 1] gives me a single element. Subscripting with [1; 1 % *] gives me a list of one element
guifa2 Yeah 06:38
Because you can also do
Well
since 1 % * returns a single value, it shouldn't go into slicing mode
raku-bridge <EsperLily> it only seems to do this in the multidimensional subscript (or whatever you call the [a; b] form) 06:39
guifa2 Yeah, I think it's being overzealous perhaps with slices 06:40
m: say (1, 2; 3, 4)[*; 1] # that I expect to return a list
camelia (2 4)
raku-bridge <EsperLily> 3[0; 0 % *] returns (3)
guifa2 say (1, 2; 3, 4, 5, 6)[*; * % 2] 06:41
evalable6 (1 3)
raku-bridge <EsperLily> also curiously, tacking additional ; 0s on the end doesn't unwrap the list. I have to use a separate subscript entirely
guifa2 oh there you go
say (1, 2, 3, 4; 5, 6, 7, 8)[*; * % 2] 06:42
evalable6 (1 5)
guifa2 hmm I dunno, weird
raku-bridge <EsperLily> This feels like a bug
<EsperLily> I'd just use multiple subscripts but AFIAK there's no way to say "subscript the invocant only if it's non-nil", and someNilValue[1 % *] gives a divide-by-zero error 06:43
<EsperLily> oh, regarding toggle, it doesn't exist as a standalone sub so I can't use it with ==> :'( 06:44
raku-bridge <EsperLily> well, I could use ==> { @^a.toggle(*.defined) }() but that just feels ugly 06:44
guifa2 Actually wait. The reason is probably that whatever code could still potentially return a list 06:48
There's no reason to distinguish whatever code from any other callable
guifa2 Except a callable returning a list still gets evaluated as a number. Hm.. okay, I'm definitely agreeing with there's something weird going on here 06:55
bisectable: (1, 2; 3, 4)[1; 1 % *] 07:03
bisectable6 guifa2, Will bisect the whole range automagically because no endpoints were provided, hang tight
guifa2, ¦6c (49 commits): «»
guifa2, Nothing to bisect!
guifa2 so been that way for basically ever
guifa2 is afk 07:09
coldpress Hi, what's the idiomatic way of grouping and then summing contiguous 1s in the array [1, 1, 0, 1, 0, 1, 1, 1]? The result should be [2, 1, 3] 07:41
moritz coldpress: I don't understand how the grouping is supposed to work 07:44
moritz split on 0s? 07:44
coldpress yes, split on 0s 07:49
moritz m: say [1, 1, 0, 1, 0, 1, 1, 1].join('').split(0).map(*.comb.sum) 07:54
camelia (2 1 3)
moritz but that's cheating, assuming that each entry is just one digit 07:55
m: say [1, 1, 0, 1, 0, 1, 1, 1].grep(:k, 0) 07:56
camelia (2 4)
moritz this gives you the indexes to split on
moritz too distracted with $work to come up with something clever 07:57
I guess you could always just do a boring iteration
m: say gather { my $sum = 0; for [1, 1, 0, 1, 0, 1, 1, 1] -> $x { if ( $x == 0 ) { take $sum; $sum = 0 } else { $sum += $x } } 07:58
camelia 5===SORRY!5=== Error while compiling <tmp>
Missing block
at <tmp>:1
------> 3e $sum; $sum = 0 } else { $sum += $x } }7⏏5<EOL>
moritz m: say gather { my $sum = 0; for [1, 1, 0, 1, 0, 1, 1, 1] -> $x { if $x == 0 { take $sum; $sum = 0 } else { $sum += $x } } 07:59
camelia 5===SORRY!5=== Error while compiling <tmp>
Missing block
at <tmp>:1
------> 3e $sum; $sum = 0 } else { $sum += $x } }7⏏5<EOL>
moritz m: say gather { my $sum = 0; for [1, 1, 0, 1, 0, 1, 1, 1] -> $x { if $x == 0 { take $sum; $sum = 0 } else { $sum += $x } } }
camelia (2 1)
coldpress moritz: yeah, assuming one digit is cheating
moritz m: say gather { my $sum = 0; for [1, 1, 0, 1, 0, 1, 1, 1] -> $x { if $x == 0 { take $sum; $sum = 0 } else { $sum += $x } }; take $sum }
camelia (2 1 3)
coldpress moritz: I am doing iteration, and I don't like it because it's too verbose. .grep(:k, 0) looks like a good solution
moritz a split + reduce approach would be more satisfying, I guess
samebchase- there should be a way to split on arrays like Str.split 08:00
coldpress indeed
samebchase- which was take arbitary predicates
s/was/can/ 08:01
coldpress what's the Raku way to scan (as in parallel scan, as in hackage.haskell.org/package/foldl-...ml#v:scan) 08:13
(not necessarily parallel)
moritz sorry, I don't follow that terminology 08:42
frost-lab Hi, everybody. How can I make sure that an array contains a specific part? For example, to determine whether [3,1,2,0,0,8,7] contains the part 2,0,0.
moritz coldpress: what do you want to achieve? 08:42
m: say [3,1,2,0,0,8,7] ~~ [*, 2, 0, 0, *] 08:43
camelia False
frost-lab I just know transform the array to a string and use string method to match the specific part, but is there any other way? 08:44
moritz m: say any([3,1,2,0,0,8,7].rotor(4 => -3) ~~ [2, 0, 0] 08:46
camelia 5===SORRY!5=== Error while compiling <tmp>
Unable to parse expression in argument list; couldn't find final ')' (corresponding starter was at line 1)
at <tmp>:1
------> 031,2,0,0,8,7].rotor(4 => -3) ~~ [2, 0, 0]7⏏5<EOL>
moritz m: say any([3,1,2,0,0,8,7].rotor(4 => -3)) ~~ [2, 0, 0]
camelia any(False, False, False, False)
frost-lab :m so [3,1,2,0,0,8,7].Str ~~ /'2 0 0'/
m:so [3,1,2,0,0,8,7].Str ~~ /'2 0 0'/
evalable6 WARNINGS for /tmp/FYwaBQQD9e:
Useless use of "so " in expression "so [3,1,2,0,0,8,7].Str ~~" in sink context (line 1)
moritz m: say any([3,1,2,0,0,8,7].rotor(3 => -2)) ~~ [2, 0, 0]
camelia any(False, False, False, False, False)
frost-lab p6:so [3,1,2,0,0,8,7].Str ~~ /'2 0 0'/ 08:47
evalable6 WARNINGS for /tmp/2EYhR7PIFX:
Useless use of "so " in expression "so [3,1,2,0,0,8,7].Str ~~" in sink context (line 1)
moritz m: say any([3,1,2,0,0,8,7].rotor(3 => -2)) ~~ (2, 0, 0)
camelia any(False, False, False, False, False)
moritz my: say so [3,1,2,0,0,8,7].Str ~~ /'2 0 0'/
m: say so [3,1,2,0,0,8,7].Str ~~ /'2 0 0'/
camelia True
frost-lab yep
moritz frost-lab: nothing clever comes to mind
frost-lab moritz thanks for your answer. 08:49
elcaro frost-lab: To compare lists, you need to use `eqv`... so to fix moritz's example 09:46
m: say so any([3,1,2,0,0,8,7].rotor(3 => -2)) eqv (2,0,0);
camelia True
frost-lab Aha, interesting! 09:51
frost-lab m: (6,7,8) ~~ (6,7,8) 10:18
camelia ( no output )
frost-lab m: say (6,7,8) ~~ (6,7,8)
camelia True
frost-lab m: say ((6,7,8) | (6,8,7)) ~~ (6,7,8) 10:19
camelia any(False, False)
frost-lab Weird, why (6,7,8) ~~ (6,7,8) is true but ((6,7,8) | (6,8,7)) ~~ (6,7,8) gets two fales? 10:20
moritz the matcher is always on the right-hand side 10:21
m: say (6,7,8) ~~ (6,7,8) | (6, 8, 7)
camelia True
frost-lab Oh, I forget the characteristic of the ~~, thanks. 10:22
elcaro you can also use `cmp` 10:34
m: say (1,1,1) cmp (1,1,1)
camelia Same
elcaro m: say (1,1,1) cmp (1,2,1)
camelia Less
holyghost DWIM ? 10:38
There's several equalitiy operators as you see 10:40
eq? equal?, eqv? in Scheme for example
moritz stackoverflow.com/questions/176343...erators-eq 10:56
frost-lab m:say ((1,2)|(2,1,9)).WAHT, ((1,2)|(2,1,9)).Seq.WHAT 11:57
evalable6 (exit code 1) No such method 'WAHT' for invocant of type 'List'. Did you mean any of
these: 'Rat', 'WALK', 'WHAT', 'WHO'?
in block <unit> at /tmp/SzIDdH4LYN line 1
frost-lab m:say ((1,2)|(2,1,9)).WHAT, ((1,2)|(2,1,9)).Seq.WHAT
evalable6 (Junction)(Junction)
frost-lab m: say ((1,2)|(2,1,9)).Seq eqv (1,2), ((1,2)|(2,1,9)) eqv (1,2)
camelia any(False, False)any(True, False)
frost-lab m: say so ((1,2)|(2,1,9)).Seq eqv (1,2), so ((1,2)|(2,1,9)) eqv (1,2) 11:58
camelia FalseTrue
frost-lab What is the difference between ((1,2)|(2,1,9)).Seq and ((1,2)|(2,1,9)) ? 11:59
There are still things that confuse me X) . 12:01
moritz Seq is an iterator type 12:42
with it, you can you produce potentially inifinte streams that are never held fully in memory
perfect for things like a file iterator
or a UNIX pipe 12:43
notandinus m: my @t = 0; push @t, 1..3; say @t.raku; 12:48
camelia [0, 1..3]
notandinus m: my @t = 0; my @k = 1 .. 3; push @t, @k; say @t.raku;
camelia [0, [1, 2, 3]]
notandinus how do i do the above but make it push each element one by one instead of pushing the whole array? 12:49
m: my @t = 0; my @k = 1 .. 3; push @t, $_ for @k; say @t.raku;
camelia [0, 1, 2, 3]
notandinus like ^, for loop works fine, are there more ways?
Altai-man m: my @t = 0; my @k = 1 .. 3; push @t, |@k; say @t.raku;
camelia [0, 1, 2, 3]
Altai-man m: my @t = 0; my @k = 1 .. 3; @t.append(@k); say @t.raku; 12:50
camelia [0, 1, 2, 3]
notandinus i see, thanks. 12:51
Altai-man notandinus, use `append` or `|` "slip" operator.
notandinus also, is unshift more costly compared to push? cost being cpu time.
dakkar ah, we had the same question yesterday ☺ 12:52
I ran some benchmarks, and couldn't see a difference
notandinus yeah, looks like there is not much difference when testing against large arrays
i found push to be faster on small arrays.
Altai-man github.com/MoarVM/MoarVM/pull/1392 ?
push is faster in cases of large inputs, but in new release this will be fixed. :] 12:53
notandinus oh their raku initializes in less than 200ms 12:54
mine takes about 1s
Altai-man notandinus, what rakudo version do you use? 12:55
notandinus ah yeah, i was talking about unshifing a single item vs initializing with that item and pushing the rest. looks like both are equally as fast. 12:56
Altai-man: It's Rakudo v2020.10
Altai-man I see, that's pretty fresh. :)
notandinus, for small inputs it won't really matter, I'd say.
notandinus yeah i compiled it myself, can i do anything to improve this initialization time? 12:57
Altai-man notandinus, what do you mean by "initialization time"? Repl start? Particular script start? 12:58
notandinus Altai-man: time taken to run "raku -e ''" 12:59
Altai-man `Executed in 93,26 millis` for me.
I don't think there is anything for the user to improve particularly that. 13:00
notandinus i see
ggoebel raku: (|(1, 2..3)) 14:08
evalable6
ggoebel raku: say (|(1, 2..3))
evalable6 (1 2..3)
ggoebel raku: say (1, 2..3).flat
evalable6 (1 2 3)
ggoebel wondering why these are different?
tbrowder .tell jmerelo i have NOT scheduled it 14:41
tellable6 tbrowder, I'll pass your message to jjmerelo
guifa2 ggoebel: a slip doesn't flatten 15:10
ggoebel thx 15:13
guifa2++
notandinus what will be the equivalent of following perl code: 16:46
print sprintf('%02d ', $_) for(0...10) 16:48
lizmat m: print sprintf('%02d ', $_) for ^11 16:48
camelia 00 01 02 03 04 05 06 07 08 09 10
lizmat m: printf('%02d ', $_) for ^11 16:49
camelia 00 01 02 03 04 05 06 07 08 09 10
notandinus oh
thanks, i was trying something with say, map, sprintf
lizmat you don't have to use .fmt if you don't want to :-)
notandinus i don't know what fmt is, i'll check it otu 16:50
m: say (0 ... 10).fmt("%02d") 16:51
camelia 00 01 02 03 04 05 06 07 08 09 10
notandinus thanks, that makes it easy
lizmat yw
notandinus is there a better way to get last index other than @_.elems - ? 16:53
@_.elems -1
lizmat m: my @a = ^10; say @a[*-1] 16:55
camelia 9
lizmat m: my @a = ^10; say @a[*-1,*-2, *-3]
camelia (9 8 7)
lizmat notandinus: is that what you're looking for?
m: my @a = ^10; say @a.end
camelia 9
lizmat if you just want .elems - 1 16:56
notandinus no i was looking for the index
yeah .end is what i was looking for :)
MasterDuke huh, i didn't know about .end 16:57
lizmat
.oO( it's not the end :-)
 16:58
notandinus yeah raku always has a method you dont know about
dakkar imagines a raku repl with tab-completion of method names via the mop… 17:01
tony-o that'd not be difficult
tellable6 2020-12-04T10:20:46Z #raku <patrickb> tony-o If there is any think I can help you with wrt. a new ecosystem please do contact me!
dakkar in the meantime, `.^methods` helps 17:02
tony-o .tell patrickb i kind of gave up on it, there didn't seem to be a lot of discussion about it in raku community. if you think this is not the case please lmk and i'll resume testing and playing around with it
tellable6 tony-o, I'll pass your message to patrickb
dakkar m: my @a;say @a.^methods
camelia (iterator from-iterator from-list new STORE reification-target Method+{is-nodal}.new Method+{is-nodal}.new shape Method+{is-nodal}.new Method+{is-nodal}.new grab name of default dynamic clone Method+{is-nodal}.new Method+{is-nodal}.new Method+{is-noda…
jmerelo dakkar: that would be pretty cool. With colors, too. 17:10
tellable6 2020-12-10T14:41:34Z #raku <tbrowder> jmerelo i have NOT scheduled it
jmerelo .tell tbrowder OK, I'm checking it and scheduling it then. 17:10
tellable6 jmerelo, I'll pass your message to tbrowder
tbrowder thanks, jj, wordpress and i don't get along very well :-( 17:12
notandinus .tell tellable6 hi
tellable6 notandinus, Thanks for the message
dakkar oh nice, REPL already has completion hooks (only for Linenoise, which doesn't install for me, but shouldn't be too hard to add them to readline as well) 17:14
jmerelo tbrowder: does not seem to get along with anyone, up to and including my cable provider. I have to log via private navigation every time.
tbrowder is there any way to get Pod::To::HTML to get better styling for wordpress by creating a mustache template for it?
jmerelo Doesn't latest version have Mustache baked in?
tbrowder i really like using pod6 for writing this year! 17:15
tbrowder yeah, baked in i think, but it looks pretty bland when i view it locally with firefox, so i'm not sure. was i supposed to use partials or something like that? 17:17
btw, lizmat seems to have wordpress tamed pretty well, at least style-wise 17:18
lizmat it did take some getting used to
and every now and then, it fights me still 17:19
tbrowder lizmat: i assume you use a script to collect the wkly info. do you create the html outside of wp and then copy/paste? 17:21
lizmat only for the new / update modules
tbrowder hm. 17:22
lizmat which gfldex kindly created a script for
tbrowder so the styling is from wp?
lizmat yup, for that style that I selected well over a year ago now
it's only disadvantage is really is that Preview doesn't work 17:23
OTOH, the editor is close enough so that doesn't really feel like a problem most of the time
tbrowder thnx 17:24
notandinus how do you match something like "<this is a name>"in raku grammar? 17:25
i'm doing this: token name { '<' \w+ '>'}
but it only matches when there are no space characters inside of '<>' 17:26
dakkar yes, because `\w` does not match spaces
guifa Easiest way is probably ‘<‘ ~ ‘>’ .*?
MasterDuke { '<' [^>]+ '>' } is the normal regex way
dakkar m: 'foo <bar baz> moo' ~~ /'<' [\w|\s]+? '>'/ 17:27
camelia ( no output )
MasterDuke and ^^^ is the raku way
dakkar m: say 'foo <bar baz> moo' ~~ /'<' [\w|\s]+? '>'/
camelia ｢<bar baz>｣
dakkar change that alternation to match what you mean
tony-o btw dakkar, Bench showed a pretty decent difference in shift/pop gist.github.com/tony-o/22bb89f23c4...7a76da4721
notandinus i see
dakkar (I fear that guifa's suggestion may capture more than you want) 17:28
notandinus also why doesn't '<' .* '>' work?
is it because token doesn't backtrack?
dakkar tony-o: yes, but for me it was not stable, different runs would produce very different results
guifa notandinus because ‘>’ is a ., and tokens don’t backtrack
dakkar m: say 'foo <bar baz> moo' ~~ /'<' .*? '>'/
camelia ｢<bar baz>｣
guifa but if you make it frugal, it works
tony-o i'm skeptical of 2m ops with that taking less than .5s 17:29
guifa I remember jnhtn saying that there can be a slight memory efficiency in using native strs (especially if in an array) 17:30
notandinus guifa: i see, thanks
dakkar: that last one works
guifa Is there much of a penalty for unboxing them?
dakkar tony-o: yes, me to
MasterDuke highly recommend www.amazon.com/Mastering-Regular-E...596528124/ one of only two programming books i've read cover to cover (the other being Programming Perl)
tony-o do you mind trying the one liner with Bench the way i wrote it?
notandinus '<' ~ '>' .* doesn't seem to work with that
MasterDuke also moritz++ has a good Raku grammar book
guifa moritz is the Raku grammar guru 17:31
MasterDuke www.amazon.com/Parsing-Perl-Regexe...484232275/
dakkar tony-o: (running) 17:32
notandinus i see, i'll check those out
btw i've never ever completed any programming book
very bad at reading
i lose focus after a while 17:33
tony-o dakkar: i'm asking because if Bench is broken then i'd like to fix it - it might've just evaluated Block rather than running it
dakkar tony-o: Benchy was the one that returned very fast, not Bench
lizmat fwiw, Benchy is up for adoption: github.com/raku-community-modules/Benchy 17:34
notandinus so i wrote this grammar to parse a log file (included), anything i could improve there? paste.debian.net/hidden/a7a6275f/
lizmat other modules up for adoption: github.com/Raku/ecosystem/blob/mas...DOPT-ME.md 17:35
dakkar github.com/raku-community-modules/...hy.pm6#L24 would that actually call &old?
guifa notandinus: you might try jmerelo’s Recipes for Raku. It’s a bit more task-oriented but in bite-size pieces 17:36
jmerelo guifa: thanks :-)
guifa . o O ( you know my OCD is bad when I rename constants from FOO-OFFSET and BAR-A-OFFSEST to OFFSET-FOO and OFFSET-BAR-A so they vertically align ) 17:37
lizmat guifa: that's just doing your future self a favour!
jmerelo .tell tbrowder I think I've seen what you wanted to do; I've switched back the old version to draft, promoted the new version to be published in ~ 7 hours.
tellable6 jmerelo, I'll pass your message to tbrowder
guifa lizmat: I’ve taken vertical alignment to an extreme I think 17:38
notandinus guifa: i see, can you link me to it?
was that the first book you mentioned?
guifa but it’s so good for catching copy paste bugs
MasterDuke’s recommendation is from moritz, this one is from jmerelo 17:39
www.apress.com/gp/book/9781484262573
tony-o m: sub b (&o) { my $i=-1; nqp::until(nqp::islt_i(1, $i = nqp::add_i($i, 1)), o, :nohandler); }; b({ 'hi'.say; });
camelia 5===SORRY!5=== Error while compiling <tmp>
Could not find nqp::add_i, did you forget 'use nqp;' ?
at <tmp>:1
------> 3il(nqp::islt_i(1, $i = nqp::add_i($i, 1)7⏏5), o, :nohandler); }; b({ 'hi'.say; });
guifa They are both very good books
tony-o m: use nqp; sub b (&o) { my $i=-1; nqp::until(nqp::islt_i(1, $i = nqp::add_i($i, 1)), o, :nohandler); }; b({ 'hi'.say; });
camelia hi
hi
tony-o dakkar: looks like it should
notandinus thanks, i'll check it out 17:40
dakkar tony-o: then I don't see why benchy was suspiciously fast for me 17:41
tony-o what version of raku are you on?
you might try that one nqp one liner and see if it outputs 'hi\nhi\n' 17:42
dakkar it does
`raku -e 'use Benchy; my $a=0;my $b=0;my @front = ^100; my @back = ^100; b 2_000_000, { @front.unshift(1); @front.pop(); ++$a}, { @back.push(1); @back.shift(); ++$b };say $a;say $b'` prints 2000002 at the end 17:43
so stuff gets called 17:44
ok, Bench has finished, pop: 72.302 shift: 73.779
anyway, end of workday for me, see I'll be back on monday 17:45
notandinus so i wrote a script to find out the most active hour over here 18:02
paste.debian.net/hidden/c73a0889/
looks like it's 12:00 (UTC-5) 18:03
tony-o what were the other values dakkar? 18:05
my benchmark was on 2020.11 18:06
tbrowder jmerelo: thanks! 18:07
tellable6 2020-12-10T17:37:50Z #raku <jmerelo> tbrowder I think I've seen what you wanted to do; I've switched back the old version to draft, promoted the new version to be published in ~ 7 hours.
ggoebel I'm not sure whether to be proud or ashamed of my 66 character solution to part 1 of today's AoC puzzle... 18:15
say lines».Int.sort.&{((|@$_,.max+3) Z-(0,|@$_)).Bag.&{.{1}*.{3}}}
lizmat doesn't like .&{ ... } very much 18:16
feels like we should have better syntactic sugar for that 18:17
guifa It’s ugly but it’s also internally consistent I suppose 18:19
tbrowder: re DateTime:: prefix, I’m actually thinking now about using an entirely different one, a User:: prefix for all types of “get information about current system/user preferences”, so User::Timezone, User::Locale, etc 18:21
lizmat guifa: yeah, it's internally consistent :-) 18:22
bit it also looks like linenoise :-)
at least to me
guifa I’m going to assume it’d not be visibly distinguished for most people but maybe 18:23
lizmat when the RBP arrives (Raku Best Practices) book arrives...
guifa maybe use the · for call-sub/block-as-method 18:24
foo·bar; foo·{bar}
eh…nah too small
jmerelo We're waiting for your talks @ the FOSDEM perl & Raku devroom news.perlfoundation.org/post/fosde...for-papers you can send short, long, or longer talks, it's up to you! 18:27
guifa jmerelo: I’ve got an idea for one 18:28
patrickb tony-o: Are you around for a little longer? 18:29
tellable6 2020-12-10T17:02:34Z #raku <tony-o> patrickb i kind of gave up on it, there didn't seem to be a lot of discussion about it in raku community. if you think this is not the case please lmk and i'll resume testing and playing around with it
patrickb tony-o: I kicked off a rather lenghty discussion about fixing our ecosystems in github.com/Raku/ecosystem/pull/512 18:31
jmerelo BTW, check out also other tracks and tell them things about Raku. We want to expand the community as much as possible, and not get a blank stare and a "Rakuwhat?" every time we talk about Raku
patrickb tony-o: I was basically at the point of trying to do something about CPAN being a real pain to use. (I got lots of "you'll have little chance of succeeding, touching PAUSE is near to impossible" comments.) 18:33
lizmat patrickb: you have no idea how long it took to get CPAN where it is now
then again, maybe things have changed
fwiw, I am the least likely person to pursue this and be successful 18:34
patrickb tony-o: People tend to either lean on the side of "Let's do it like Go does and only have a very simple git repo approach" (aka keep p6c) and the other crew that says, p6c is difficult to get right and we should go for a proper solution.
lizmat patrickb: the world is different from when CPAN started 18:35
patrickb tony-o: Then I read about your proposal and was very happy that wouldn't have to try persuing a goal everyone warned me about.
lizmat so an ecosystem that depends on one ore more cloud based content stores, should be workable nowadays
patrickb tony-o: Then your proposal got rejected, and now I am in misery again. I still firmly believe we need to do something about the ecosystem, it's really really broken. Discussions tend to not flourish action as there are naysayers for all possible approaches I know of. So I think someone just has to push for a solution. 18:38
tony-o: You are in a good position to come up with a good approach because of your knowledge of zef and knowing ugexe well. (ugexe being one of the few people who really understands the problems with our current ecosystems). 18:39
tony-o: So yes, you do have my full support to push further with your ecosystem rewrite. I am willing to actually help. 18:40
I hope tony-o actually notices the above... 18:41
patrickb is away for a bit. He does backlog. 18:42
guifa jmerelo: I already have my title slide ready 18:43
jmerelo guifa: which is...
guifa Well, first slide after the title slide haha
one sec
guifa imgur.com/WlzhL0e 18:48
Title is going to be “Surprisingly unsurprising: Making things work for end users” 18:49
Or maybe something like “The joy in making things work for end users”
jmerelo guifa: that's mirandés?
Préstame is not lend me?
guifa asturiano
prestar = gustar 18:50
jmerelo guifa: love it :-)
guifa I’m thinking about starting to do a poetic epigraph for different files in my modules 18:51
When I was writing the advent post I remembered how cool it was to go through the tz database and its code: full of references to literature, waxing poetic on historians, fanciful and enlightening — if totally superfluous — musings 18:52
jmerelo guifa++
Grinnz feel free to ask me or mst about feasibility of any particular PAUSE or CPAN changes, or tag us in to liaise 18:59
guifa notes to self. access penalty using method { .return with $!foo; $!foo = load-foo } over loading it at BUILD is 0.0002s 19:07
perryprog gasps 19:08
guifa perryprog: does it feel abnormally high or low to you? 19:09
perryprog That seems very low to me
tony-o patrickb: yea i'm around
perryprog I would have to guess it's within experimental error 19:10
tony-o patrickb: my code was at the point where it needs a zef plugin to use but was working in an alpha sort of way 19:11
guifa It’s quite consistent in my tests. I do `say CLDR<en>.dates.calendars.gregorian.months.stand-alone.wide[11]` with 5 languages. the .dates element is what I was testing on lazy load penalty 19:12
all else is pre-cached
tony-o go's ecosystem is horrific. so is pythons. haskell's is fine but is way too manual
tony-o lizmat: my code uses cloudflare and s3 and was pretty pluggable with other similar serices 19:13
guifa whoa 19:17
actually it’s even less
m: say my $lazy-load-avg = ([+] 0.0003798, 0.0004655, 0.00043866, 0.000378, 0.0003691, 0.00038276, 0.000360024, 0.0003852, 0.00038173, 0.00043363) / 10;
camelia 0.0003974404
guifa say my $eager-load-avg = ([+] 0.0003076, 0.000297, 0.00032552, 0.00032676, 0.000297, 0.0002976, 0.0003199, 0.0003535, 0.0003071, 0.000342) / 10; say "Difference: ", $lazy-load-avg - $eager-load-avg
guifa m: say my $lazy-load-avg = ([+] 0.0003798, 0.0004655, 0.00043866, 0.000378, 0.0003691, 0.00038276, 0.000360024, 0.0003852, 0.00038173, 0.00043363) / 10; say my $eager-load-avg = ([+] 0.0003076, 0.000297, 0.00032552, 0.00032676, 0.000297, 0.0002976, 0.0003199, 0.0003535, 0.0003071, 0.000342) / 10; say "Difference: ", $lazy-load-avg - $eager-load-avg 19:17
camelia 0.0003974404
0.000317398
Difference: 0.0000800424
guifa m: class Foo is Associative { method AT-KEY ($) { 42 } }; say Foo<a>; 20:05
camelia 42
guifa class Foo is Associative { method AT-KEY ($) { 42 } }; say Foo{'a'}
m: class Foo is Associative { method AT-KEY ($) { 42 } }; say Foo{'a'}
camelia 5===SORRY!5=== Error while compiling <tmp>
Autovivifying object closures not yet implemented. Sorry.
at <tmp>:1
------> 3method AT-KEY ($) { 42 } }; say Foo{'a'}7⏏5<EOL>
guifa Anyway to get around that compile error? 20:06
patrickb tony-o: We are in different timezones. Communication tends to be a bit difficult. How could I get involved? Is there any code I could dig into? Do you have a list of open tasks? 20:08
stoned75 guifa: F{'a'}. really ? Are you sure you do not mean Foo.new{'a'} (Foo.new<a> btw) ? 20:10
s/F/Foo/
guifa stoned75: yeah. It’s a singleton object, so I’ve no need for it to be instantiated 20:11
stoned75 I see
Map<a> Map{'a'} behave the same
moritz jsut be careful, < > splits on whitespace 20:12
os %h<a b c > is the same as %h{'a', 'b', 'c'}, not %h{'a b c '}
s/os/so/
[Coke] back in review saw this question, and an almost answer, but this works: 20:13
m: my @a = 1,2,3,4,2,0,0,7,8,9; dd @a.rotor(3 => -2).grep(* ~~ (2,0,0)) 20:14
camelia ((2, 0, 0),).Seq
guifa moritz: Yeah. In this use case, I don’t think anyone would ever call it with spaces it (it’s to call up language codes)
I mean I guess I can give the class a different name and export an instantiated copy but … feels like a lot of work ^_- 20:15
tony-o patrickb: i can start a channel or pm is fine, i need to stand up the infrastructure again and we can work on a zef plugin 20:18
tony-o nick had a few notes too, need to see if i can find them or if i lost them 20:19
Geth doc/more-catch-index: 19741db102 | (Stoned Elipot)++ | doc/Language/exceptions.pod6
Add more CATCH index locations
 20:31
doc: stoned++ created pull request #3729:
Add more CATCH index locations
raku-bridge <EsperLily> Why does this not work? "abc" ~~ /:r [ "a" || "ab" ] "c" / 20:51
<EsperLily> || is documented as trying alternatives even in non-backtracking contexts. Using | works but || doesn't
moritz in ["a" || "ab"], the first alternative is prefered, so it matches "a". The :r (ratchet) tells the regex engine not to try a different path when the following token doesn't match 20:57
moritz in [a|ab]c, all parts of the regex are part of a "regular" language, and so are compiled into an NFA that matches the whole string without backtracking 20:59
|| is not regular, it's explicitly sequential
raku-bridge <EsperLily> docs.raku.org/language/regexes#Alternation:_|| says "Even in non-backtracking contexts, the alternation operator || tries all the branches in order until the first one matches." 21:23
moritz yes, abut it doesn't try the second one, if the first one already matched 21:26
raku-bridge <EsperLily> since it's mentioning non-backtracking contexts, I interpret "matched" as meaning the entire branch matches 21:27
<EsperLily> bleh, /:r [ a | {} ab ] c/ won't match either 21:32
<EsperLily> ooh wait, I just realized my confusion with [ a || ab ]. The whole group matches on the first branch, then fails after the group, so it won't backtrack 21:37
<EsperLily> and [ a | ab ] works because it's matching the longest declarative prefix, which ends up being "ab" 21:38
<EsperLily> [ a | {} ab ] doesn't because the longest declarative prefix matches "a" instead
guifa EsperLily: exactly 21:48
coldpress moritz │ coldpress: what do you want to achieve? 22:46
moritz: calculate prefix sum of an array of numbers en.wikipedia.org/wiki/Prefix_sum 22:48
guifa coldpress: ([\+] @numbers).tail 23:06
guifa m: say [\+] ^10 23:06
camelia (0 1 3 6 10 15 21 28 36 45)
guifa m: say ([\+] ^10).tail 23:07
camelia 45
coldpress guifa: thanks, it's not easy to search for this in Raku docs 23:09
guifa The [\+] is in the metaoperators section. Basically, you can put any operator inside of [ ] and it will turn it into a reduce function for you 23:10
If you use [\ ] (in Raku terminology, a triangle reduction becase [\ looks like a triangle lol) it does the euqivalent of what other languages call a scan 23:11
guifa (even better, it’s lazy, so you can use it on infinite lists) 23:13
guifa my @a = [\+] ^Inf; say @a; say @a[10] 23:14
evalable6 [...]
55
melezhik .tell @patrickb at some point I was trying to help tony-o as well, he is aware, with new eco system stuff . For example - sparrowhub.io/search?q=function 23:20
tellable6 melezhik, I cannot recognize this command. See wiki for some examples: github.com/Raku/whateverable/wiki/Tellable
melezhik .tell @patrickb at some point I was trying to help tony-o as well, he is aware, with new eco system stuff . For example - sparrowhub.io/search?q=function 23:21
tellable6 melezhik, I cannot recognize this command. See wiki for some examples: github.com/Raku/whateverable/wiki/Tellable
melezhik .tell patrickb: at some point I was trying to help tony-o as well, he is aware, with new eco system stuff . For example - sparrowhub.io/search?q=function
tellable6 melezhik, I'll pass your message to patrickb
tbrowder .tell jmerelo the latest source is here: github.com/tbrowder/advent2020/blo...dvent.html 23:23
tellable6 tbrowder, I'll pass your message to jmerelo
tbrowder .tell jmerelo same title: Santa Claus TWEAKs with a Class 23:25
tellable6 tbrowder, I'll pass your message to jmerelo

Powered by Cro and the Raku® Programming Language.

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

Thank you!