Home freenode/#haskell: Logs Calendar

Logs: freenode/#haskell

←Prev  Next→
Page 1 .. 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 .. 5022
— 502,152 events total
2020-11-20 21:38:49 → mputz joins (~Thunderbi@dslb-084-058-211-084.084.058.pools.vodafone-ip.de)
2020-11-20 21:41:03 <jcd> So, I see that it is a valid choice! I don't know if it's the one I _should_ choose though.
2020-11-20 21:41:03 → conal joins (~conal@172.255.125.150)
2020-11-20 21:42:21 × nyaomi quits (~naomi@2604:6000:1509:c86e:c878:29ff:fedf:ce89) (Quit: meow)
2020-11-20 21:42:22 × geekosaur quits (ac3a56ed@172.58.86.237) (Remote host closed the connection)
2020-11-20 21:42:28 <dminuoso> That's up to you. In this case, the idea is "parseRequest" could fail. Rather than forcing the function to use say `Maybe` or `IO` to communicate failure in, it defers the choice to you.
2020-11-20 21:42:51 <sm[m]> interactively inspect Haskell values at runtime ... that's interesting
2020-11-20 21:42:51 <dminuoso> It says "Hey, you pick something of type `m`, as long as I can throw an exception with it", for some value of "exception"
2020-11-20 21:43:23 <sm[m]> doc: https://hackage.haskell.org/package/heap-console-0.1.0.0/docs/Heap-Console.html
2020-11-20 21:45:08 <dminuoso> Internally the package will use `throwM (InvalidUrlException s "Invalid URL")` if it failed to parse your input. What this ends up doing depends on your choice, if you pick Maybe, you get a Nothing back.
2020-11-20 21:45:12 → justanotheruser joins (~justanoth@unaffiliated/justanotheruser)
2020-11-20 21:45:19 <dminuoso> If you pick IO, you get an IO exception
2020-11-20 21:45:27 <dminuoso> And so forth
2020-11-20 21:45:30 hackage predicate-typed 0.7.4.5 - Predicates, Refinement types and Dsl https://hackage.haskell.org/package/predicate-typed-0.7.4.5 (gbwey)
2020-11-20 21:46:09 <jcd> Okay! Got it! How would I properly wrap it? Is there an approriate place to show a 3-line code example?
2020-11-20 21:46:20 <dminuoso> What do you mean with "wrap it"?
2020-11-20 21:46:25 → nyaomi joins (~naomi@cpe-74-75-6-125.maine.res.rr.com)
2020-11-20 21:46:27 <dminuoso> @where paste
2020-11-20 21:46:27 <lambdabot> Help us help you: please paste full code, input and/or output at eg https://paste.tomsmeding.com
2020-11-20 21:47:10 <jcd> https://paste.tomsmeding.com/5trqeFMa
2020-11-20 21:47:35 <jcd> Would it be better to say 'assign'?
2020-11-20 21:47:59 <dminuoso> jcd: I see. So since httpLbs puts you into IO already, you could chose IO for parseRequest too
2020-11-20 21:48:38 <dminuoso> Also note that this use of `Maybe` is not valid
2020-11-20 21:48:50 <dminuoso> Sadly, there's not a very explicit way to make the choice, it is inferred on usage
2020-11-20 21:48:52 <dminuoso> Say, if you wrote:
2020-11-20 21:49:12 → johnw_ joins (~johnw@haskell/developer/johnw)
2020-11-20 21:49:14 → minimario joins (2fe3e53b@047-227-229-059.res.spectrum.com)
2020-11-20 21:49:18 → hlisp joins (~hlisp@114.246.35.11)
2020-11-20 21:49:22 <dminuoso> let f :: Maybe Request; f = parseRequest "https://google.de" in ...
2020-11-20 21:49:38 <dminuoso> Then the choice is made in the type signature. Equivalently if you wrote:
2020-11-20 21:49:49 <dminuoso> case parseRequest "https://google.de" of
2020-11-20 21:49:51 <dminuoso> Nothing -> ...
2020-11-20 21:49:54 <dminuoso> Just x -> ...
2020-11-20 21:49:57 → gxt joins (~gxt@gateway/tor-sasl/gxt)
2020-11-20 21:50:19 × heatsink quits (~heatsink@107-136-5-69.lightspeed.sntcca.sbcglobal.net) (Remote host closed the connection)
2020-11-20 21:50:21 <dminuoso> Then the choice is implicit because GHC can infer Maybe from the data constructors in the pattern matching.
2020-11-20 21:51:07 <dminuoso> For your case, IO is probably a fine choice. Then you can just write: https://paste.tomsmeding.com/USndzrdo
2020-11-20 21:51:20 <dminuoso> (The choice of IO is inferred based on the context automatically)
2020-11-20 21:51:56 × ubert quits (~Thunderbi@p200300ecdf1e5361e6b318fffe838f33.dip0.t-ipconnect.de) (Remote host closed the connection)
2020-11-20 21:52:31 hackage phonetic-languages-simplified-common 0.2.1.0 - A simplified version of the phonetic-languages-functionality https://hackage.haskell.org/package/phonetic-languages-simplified-common-0.2.1.0 (OleksandrZhabenko)
2020-11-20 21:52:43 <jcd> Because of the 'do'!
2020-11-20 21:52:50 <dminuoso> no, not the do
2020-11-20 21:52:54 <jcd> Oh!
2020-11-20 21:52:59 <jcd> the '<-'
2020-11-20 21:53:04 <dminuoso> Also, no.
2020-11-20 21:53:12 <dminuoso> If you keep guessing, you'll eventually figure it out.
2020-11-20 21:53:14 <dminuoso> :)
2020-11-20 21:53:25 <dminuoso> Well, the `do` notation is related, but not quite the point
2020-11-20 21:53:33 <minimario> quick question: what's the logic behind making andM return Just False instead of Nothing when we call it on andM (Just False) Nothing?
2020-11-20 21:53:45 × hlisp quits (~hlisp@114.246.35.11) (Ping timeout: 260 seconds)
2020-11-20 21:54:16 → elliott__ joins (~elliott@pool-108-51-141-12.washdc.fios.verizon.net)
2020-11-20 21:54:50 <dminuoso> jcd: https://paste.tomsmeding.com/Xx49wa4s here there's two points of interest.
2020-11-20 21:55:04 ski . o O ( <https://hackage.haskell.org/package/hood>,<https://hackage.haskell.org/package/GHood>,<https://wiki.haskell.org/Debugging#Printf_and_friends> )
2020-11-20 21:55:40 <dminuoso> jcd: First there's the type signature telling us the do-notation uses IO. You left it out, which gives us another point of inference
2020-11-20 21:55:44 <dminuoso> The type of httpLbs
2020-11-20 21:56:15 <jcd> Since the httpLbs is the expression that gets returned?
2020-11-20 21:56:27 <dminuoso> Not because it's returned, just its presence.
2020-11-20 21:56:33 <dminuoso> It could be on any line in the do block, in fact.
2020-11-20 21:56:50 <dminuoso> Roughly, do-notation is just a syntax desugarer
2020-11-20 21:57:15 <dminuoso> It desugars uses of line breaks and <- into (>>) and (>>=)
2020-11-20 21:57:32 <dminuoso> % :t (>>)
2020-11-20 21:57:33 <yahb> dminuoso: Monad m => m a -> m b -> m b
2020-11-20 21:57:34 <dminuoso> % :t (>>+)
2020-11-20 21:57:35 <yahb> dminuoso: ; <interactive>:1:1: error:; * Variable not in scope: >>+; * Perhaps you meant one of these: `>>=' (imported from Prelude), `>>' (imported from Prelude), `>>>' (imported from Control.Arrow)
2020-11-20 21:57:37 <dminuoso> % :t (>>=)
2020-11-20 21:57:37 <yahb> dminuoso: Monad m => m a -> (a -> m b) -> m b
2020-11-20 21:57:41 <ski> minimario : it's short-circuiting. since the first action (successfully) executes to `False', we abort and don't try the second
2020-11-20 21:57:56 <dminuoso> jcd: Do you see in each type, there's multiple occurences of the *same* m?
2020-11-20 21:58:35 → rprije joins (~rprije@124.148.131.132)
2020-11-20 21:58:42 <dminuoso> jcd: Recall, same story as above! This is a polymorphic value, *you* decide what you want for each of those type variables. So if you make the choice `IO` for the first `m` in say (>>=), then the other `m` in (>>=) become IO too.
2020-11-20 21:58:54 <ski> minimario : similarly, if you did `andM readLn readLn', and you typed in `False' for the first one, it would abort the `andM', and not try executing the second `readLn'
2020-11-20 21:59:45 <dminuoso> jcd: It's roughly similar to if you had `a + b + c + d`, and if `a` was of type Int, then `b`, `c` and `d` have to be of type Int.
2020-11-20 21:59:51 <dminuoso> % :t (+)
2020-11-20 21:59:51 <yahb> dminuoso: Num a => a -> a -> a
2020-11-20 21:59:52 <ski> minimario : if you don't want short-circuiting, you can use `liftM2 (&&)'/`liftA2 (&&)'. but the point here was exactly when you want short-circuiting
2020-11-20 22:00:18 <jcd> dminuoso: Yeah, I do! That explains the errors my REPL was dumping out.
2020-11-20 22:00:45 × takuan quits (~takuan@178-116-218-225.access.telenet.be) (Ping timeout: 240 seconds)
2020-11-20 22:01:08 <jcd> dminuoso: Thank you so much for the help! Recommend any literature so I don't have cry anymore.
2020-11-20 22:01:27 <minimario> ski: ah ok that makes sense
2020-11-20 22:01:40 <ski> @hoogle Monad m => (a -> m Bool) -> [a] -> m Bool
2020-11-20 22:01:41 <lambdabot> Control.Monad.Extra anyM :: Monad m => (a -> m Bool) -> [a] -> m Bool
2020-11-20 22:01:41 <lambdabot> Control.Monad.Extra allM :: Monad m => (a -> m Bool) -> [a] -> m Bool
2020-11-20 22:01:41 <lambdabot> Extra anyM :: Monad m => (a -> m Bool) -> [a] -> m Bool
2020-11-20 22:01:50 <ski> @hoogle Monad m => [m Bool] -> m Bool
2020-11-20 22:01:51 <lambdabot> Control.Monad.Extra orM :: Monad m => [m Bool] -> m Bool
2020-11-20 22:01:51 <lambdabot> Control.Monad.Extra andM :: Monad m => [m Bool] -> m Bool
2020-11-20 22:01:51 <lambdabot> Extra orM :: Monad m => [m Bool] -> m Bool
2020-11-20 22:01:56 <ski> similar thing, in those
2020-11-20 22:02:17 × johnw_ quits (~johnw@haskell/developer/johnw) (Quit: ZNC - http://znc.in)
2020-11-20 22:02:23 <ski> minimario : ooc, which module were you looking in ?
2020-11-20 22:02:27 <dminuoso> jcd: Im not good on book references. I keep suggesting CIS194 because it's the only reference material I looked at in the past 12 months and it looked decent.
2020-11-20 22:02:53 <dminuoso> And it happens to be one of the few resources that's not as dated
2020-11-20 22:02:58 → borne joins (~fritjof@200116b86455d000f1dc39039d201adf.dip.versatel-1u1.de)
2020-11-20 22:03:16 × invaser quits (~Thunderbi@31.148.23.125) (Ping timeout: 272 seconds)
2020-11-20 22:03:21 <ski> @where CIS194
2020-11-20 22:03:21 <lambdabot> https://www.seas.upenn.edu/~cis194/spring13/lectures.html
2020-11-20 22:03:43 <minimario> oh lol i was just doing an intro haskell tutorial and one of the exercises was to implement andM haha
2020-11-20 22:03:44 <dminuoso> There's also a more recent version of CIS194 from Joachim Breitner around, the style is slightly different.
2020-11-20 22:03:45 → hlisp joins (~hlisp@114.246.35.11)
2020-11-20 22:03:51 <dminuoso> https://www.seas.upenn.edu/~cis194/fall16/
←Prev  Next→
Page 1 .. 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 .. 5022

All times are in UTC.